
From doug.arnold2@gmail.com  Wed Jul 31 21:24:14 2013
Return-Path: <doug.arnold2@gmail.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5E99E21E8064 for <tictoc@ietfa.amsl.com>; Wed, 31 Jul 2013 21:24:14 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.599
X-Spam-Level: 
X-Spam-Status: No, score=-2.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HTML_MESSAGE=0.001, NO_RELAYS=-0.001]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id nM0JxGsrg4HC for <tictoc@ietfa.amsl.com>; Wed, 31 Jul 2013 21:24:13 -0700 (PDT)
Received: from mail-ob0-x22b.google.com (mail-ob0-x22b.google.com [IPv6:2607:f8b0:4003:c01::22b]) by ietfa.amsl.com (Postfix) with ESMTP id 6F0D121E805F for <tictoc@ietf.org>; Wed, 31 Jul 2013 21:24:13 -0700 (PDT)
Received: by mail-ob0-f171.google.com with SMTP id tb18so2941237obb.2 for <tictoc@ietf.org>; Wed, 31 Jul 2013 21:24:13 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; bh=SL86KIP56usIHU3XECvexqmukrB8B2FN6Uecrdi4dU0=; b=uezoJV89AJvxS6lHzHqHd3jkQZfoyCZjlc+BWe9Vonw9HeCwa+ma+lxto6NsOyZRBb +wxWW/TSHn0asIZnBxh0Wl6JjKzyW3IdHNUM4pakGo8gr1vPxNXZiSS6RY//IlGnFRyE DoUPgagE6JRCZyIcN/eE3jk1H1YBx7f6fUwHlGJMwHZMPGLRTg6ySpKh0FVmZQOyXoff DjZeg/zg/N0TsqUbv4sKikWoc2FvZ5a58wCXtZo7eBcmB9CuVo+4v1vuT6UvLVGpFPs+ D9baOmfSEC2SQPqFGQ7L16AIdbUyXUtY1ccJMZf1UObox88YeTx5CN+nw9owZh0K3wv5 jw7A==
MIME-Version: 1.0
X-Received: by 10.50.126.74 with SMTP id mw10mr1054384igb.20.1375331052946; Wed, 31 Jul 2013 21:24:12 -0700 (PDT)
Received: by 10.64.225.2 with HTTP; Wed, 31 Jul 2013 21:24:12 -0700 (PDT)
In-Reply-To: <74470498B659FA4687F0B0018C19A89C01A0FA10A369@IL-MB01.marvell.com>
References: <CACQYgzEfFq0S1+aASOZ3VcODVMoaDDPhcDkM4Hi=o7ihLFhv1A@mail.gmail.com> <74470498B659FA4687F0B0018C19A89C01A0FA10A369@IL-MB01.marvell.com>
Date: Wed, 31 Jul 2013 21:24:12 -0700
Message-ID: <CACQYgzH3ROmKkUcRqZCo527ewHoPMBOJXDXLLnMR1-Hp=Q-YAg@mail.gmail.com>
From: Douglas Arnold <doug.arnold2@gmail.com>
To: Tal Mizrahi <talmi@marvell.com>
Content-Type: multipart/alternative; boundary=047d7b414176dc0ee804e2db3614
X-Mailman-Approved-At: Thu, 01 Aug 2013 00:51:40 -0700
Cc: "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] Enterprise profile comments
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 01 Aug 2013 04:24:14 -0000

--047d7b414176dc0ee804e2db3614
Content-Type: text/plain; charset=ISO-8859-1

Hello Tal,

This is an important point. I believe that the Enterprise IT world will
want to use multipath PTP when implementations become available.
 Unfortunately I believe that it will be a few years before COTS PTP
equipment has this feature.  The exact form of MPPTP is yet to be defined
in the 1588 revision which just started and I expect that most of the
vendors will wait until at least the 1588 committee is in agreement about
how MPPTP will operate in particular mappings.  At that time I would expect
to revise the Enterprise Profile.  The short version of this answer is that
I don't want to include MPPTP until I am sure how the 1588 committee is
going to treat it.

In IEEE 1588-2008 Unicast negotiation is used only for master clock message
capacity management.   This is, so far, not wanted by the network
architects in the finance industry, whose needs we are trying to meet.

Here I am only speaking my personal opinion.  I have not discussed this
with Heiko yet.

Doug


On Wed, Jul 31, 2013 at 6:13 AM, Tal Mizrahi <talmi@marvell.com> wrote:

> Hi Heiko, Doug,****
>
> ** **
>
> Thanks for putting this draft together. I believe it can be very useful to
> the industry.****
>
> ** **
>
> After going over the updated draft I have a single comment:****
>
> Please consider allowing unicast negotiation. The reason I am raising this
> is that multi-path PTP (draft-shpiner-multi-path-synchronization) would
> require Sync messages to be sent as unicast, and hence require unicast
> negotiation. I believe MPPTP can be useful in the context of this profile,
> allowing both redundancy and higher accuracy in some scenarios. Will
> appreciate if you can give this some thought.****
>
> ** **
>
> Thanks,****
>
> Tal.****
>
> ** **
>

--047d7b414176dc0ee804e2db3614
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hello Tal,<div><br></div><div>This is an important point. =
I believe that the Enterprise IT world will want to use multipath PTP when =
implementations become available. =A0Unfortunately I believe that it will b=
e a few years before COTS PTP equipment has this feature. =A0The exact form=
 of MPPTP is yet to be defined in the 1588 revision which just started and =
I expect that most of the vendors will wait until at least the 1588 committ=
ee is in agreement about how MPPTP will operate in particular mappings. =A0=
At that time I would expect to revise the Enterprise Profile. =A0The short =
version of this answer is that I don&#39;t want to include MPPTP until I am=
 sure how the 1588 committee is going to treat it. =A0</div>
<div><br></div><div>In IEEE 1588-2008 Unicast negotiation is used only for =
master clock message capacity management. =A0 This is, so far, not wanted b=
y the network architects in the finance industry, whose needs we are trying=
 to meet. =A0</div>
<div><br></div><div>Here I am only speaking my personal opinion. =A0I have =
not discussed this with Heiko yet.</div><div><br></div><div>Doug</div></div=
><div class=3D"gmail_extra"><br><br><div class=3D"gmail_quote">On Wed, Jul =
31, 2013 at 6:13 AM, Tal Mizrahi <span dir=3D"ltr">&lt;<a href=3D"mailto:ta=
lmi@marvell.com" target=3D"_blank">talmi@marvell.com</a>&gt;</span> wrote:<=
br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div lang=3D"EN-US" link=3D"blue" vlink=3D"p=
urple"><div><p class=3D"MsoNormal">Hi Heiko, Doug,<u></u><u></u></p><p clas=
s=3D"MsoNormal">
<u></u>=A0<u></u></p><p class=3D"MsoNormal">Thanks for putting this draft t=
ogether. I believe it can be very useful to the industry.<u></u><u></u></p>=
<p class=3D"MsoNormal"><u></u>=A0<u></u></p><p class=3D"MsoNormal">After go=
ing over the updated draft I have a single comment:<u></u><u></u></p>
<p class=3D"MsoNormal">Please consider allowing unicast negotiation. The re=
ason I am raising this is that multi-path PTP (draft-shpiner-multi-path-syn=
chronization) would require Sync messages to be sent as unicast, and hence =
require unicast negotiation. I believe MPPTP can be useful in the context o=
f this profile, allowing both redundancy and higher accuracy in some scenar=
ios. Will appreciate if you can give this some thought.<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=A0<u></u></p><p class=3D"MsoNormal">Thanks,<=
span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,&quot;sans-s=
erif&quot;"><u></u><u></u></span></p><p class=3D"MsoNormal"><span style=3D"=
font-size:10.0pt;font-family:&quot;Arial&quot;,&quot;sans-serif&quot;">Tal.=
<u></u><u></u></span></p>
<p class=3D"MsoNormal"><u></u>=A0<u></u></p></div></div></blockquote></div>=
<br></div>

--047d7b414176dc0ee804e2db3614--

From mills@udel.edu  Thu Aug  1 13:42:24 2013
Return-Path: <mills@udel.edu>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B41AE11E815E for <tictoc@ietfa.amsl.com>; Thu,  1 Aug 2013 13:42:24 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.598
X-Spam-Level: 
X-Spam-Status: No, score=-6.598 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id TfoaMw13374W for <tictoc@ietfa.amsl.com>; Thu,  1 Aug 2013 13:42:18 -0700 (PDT)
Received: from mail.eecis.udel.edu (louie.udel.edu [128.4.40.12]) by ietfa.amsl.com (Postfix) with ESMTP id 8382411E813F for <tictoc@ietf.org>; Thu,  1 Aug 2013 13:42:18 -0700 (PDT)
Received: from [192.168.1.2] (pool-74-109-13-78.phlapa.fios.verizon.net [74.109.13.78]) (Authenticated sender: mills@mail.eecis.udel.edu) by mail.eecis.udel.edu (Postfix) with ESMTP id 2E31966D; Thu,  1 Aug 2013 16:42:11 -0400 (EDT)
Message-ID: <51FAC820.3090401@udel.edu>
Date: Thu, 01 Aug 2013 20:42:08 +0000
From: "David L. Mills" <mills@udel.edu>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax)
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: NTP Working Group <ntpwg@lists.ntp.org>
Content-Type: multipart/alternative; boundary="------------040509030609050009080808"
Cc: "tictoc@ietf.org" <tictoc@ietf.org>
Subject: [TICTOC] security ID submitted for review
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 01 Aug 2013 20:42:24 -0000

This is a multi-part message in MIME format.
--------------040509030609050009080808
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Guys,

Here are some general comments on the ID.  I apologize if they are a 
little rough as I have to dictate this message instead of my usual 
polished wordsmithing.

format.
The ID does not specify an extension field format.  Presumably, the 
extension field format described in RFC 5906 is adequate with new 
message types as required.  However, in keeping with long Internet 
tradition, it must be possible to determine correct format without 
interpreting the header or extension fields.  In addition, it must be 
possible to verify message integrity without interpreting the extension 
fields.  In particular, the private key for the message authentication 
code (MAC) , if used, must be known beforehand.  This is where the 
"cookie snatcher" issue comes up as described in previous messages.  In 
particular, an important constraint for normal operation is that 
extension fields are not used unless required for cookie determination.

Glaring vulnerability exists when using extension fields, as the key 
value is public and must be known in advance.  This invites a 
cut-and-paste attack, as well as header changes, even if subsequently 
detected by the on-wire protocol.


certificate.
Both RFC 5906 and the ID require the server certificate.  The server 
certificate can be conveyed, either off-line, or directly or indirectly 
via a certificate trail as described in RFC 5906 or the ID.  If obtained 
via the trail, there is a hazzard due to a middle-man attack and 
masquerade.  A solution proposed in RFC 5906 uses an identity scheme 
such as IFF.  There may be an equivalent, or better schemes available now.

protocol.
Most of the discussion following the ID announcement is from protocol 
weenies, affectionately known as protomeister.  When the Autokey scheme 
was first defined in 1996,  the signature time required ranged up to one 
second on the computers of the time.  Therefore, a clogging hazzard 
existed in the form of  a replay attack.  This is why the timestamp was 
included in the extension field format to suppress duplicates.  Most of 
the extension fields were computed and signed off-line.  Nevertheless, a 
clogging vulnerability remains in the cookie request message.  This 
requires the server to do two public key incriptions and two hash 
functions. Today, this overhead may no longer  be considered  excessive; 
however, this may not be true for public time servers with thousands and 
thousands of clients.

In broadcast mode, the protocols described in RFC 5906 and the ID are 
messy and with several possible vulnerabilities.  Here is a suggestion 
that might be useful.  Define a new broadcast packet format as the NTP 
header followed by one extension field representing the MAC.  The 
extension field represents the signature of the header computed using 
the private key of the server and can be verified by the client using 
the public key in the server certificate.  Originally, this was 
considered an unacceptable dilution of accuracy due to the variability 
of the signature computing time and the length of the signature.  
However, at least in modern NTP, the interleaved broadcast mode can be 
used to correct for this overhead.  This assumes that the overhead for 
either the server or client is reduced using fast modern processors.

As mentioned previously, in NTP symmetric mode the current protocol in 
RFC 5906 should be changed to use Diffie-Hellman key agreement.  This is 
so easy to do that it can be incorporated separately in the standard.

While the interleaved mode is useful in symmetric and broadcast modes, 
it can't be used when the server does not maintain state separately for 
each client.  However, as in the orginal NTP design many years ago, the 
output delay can be computed for each transmitted server packet.  The 
output delay can be averaged and used to augment  the transmit timestamp 
or retrieve using an extension field.

Note that this message does not address the cookie snatcher issue 
discussed by our German protomeisters, who have concocted a brillient 
solution.  Many of the issues discussed are relevant to both NTP and 
PTP, although the on-wire protocol details may be different.  I hope 
this discussion is useful in either context.

Dave



--------------040509030609050009080808
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
  <title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
<font size="+2">Guys,<br>
<br>
Here are some general comments on the ID.&nbsp; I apologize if they are a
little rough as I have to dictate this message instead of my usual
polished wordsmithing.<br>
<br>
format.<br>
The ID does not specify an extension field format.&nbsp; Presumably, the
extension field format described in RFC 5906 is adequate with new
message types as required.&nbsp; However, in keeping with long Internet
tradition, it must be possible to determine correct format without
interpreting the header or extension fields.&nbsp; In addition, it must be
possible to verify message integrity without interpreting the extension
fields.&nbsp; In particular, the private key for the message authentication
code (MAC) , if used, must be known beforehand.&nbsp; This is where the
"cookie snatcher" issue comes up as described in previous messages.&nbsp; In
particular, an important constraint for normal operation is that
extension fields are not used unless required for cookie determination.<br>
<br>
Glaring vulnerability exists when using extension fields, as the key
value is public and must be known in advance.&nbsp; This invites a
cut-and-paste attack, as well as header changes, even if subsequently
detected by the on-wire protocol.<br>
<br>
<br>
certificate.<br>
Both RFC 5906 and the ID require the server certificate.&nbsp; The server
certificate can be conveyed, either off-line, or directly or indirectly
via a certificate trail as described in RFC 5906 or the ID.&nbsp; If
obtained via the trail, there is a hazzard due to a middle-man attack
and masquerade.&nbsp; A solution proposed in RFC 5906 uses an identity
scheme such as IFF.&nbsp; There may be an equivalent, or better schemes
available now.<br>
<br>
protocol.<br>
Most of the discussion following the ID announcement is from protocol
weenies, affectionately known as protomeister.&nbsp; When the Autokey scheme
was first defined in 1996,&nbsp; the signature time required ranged up to
one second on the computers of the time.&nbsp; Therefore, a clogging hazzard
existed in the form of&nbsp; a replay attack.&nbsp; This is why the timestamp was
included in the extension field format to suppress duplicates.&nbsp; Most of
the extension fields were computed and signed off-line.&nbsp; Nevertheless,
a clogging vulnerability remains in the cookie request message.&nbsp; This
requires the server to do two public key incriptions and two hash
functions. Today, this overhead may no longer&nbsp; be considered&nbsp;
excessive; however, this may not be true for public time servers with
thousands and thousands of clients.<br>
<br>
In broadcast mode, the protocols described in RFC 5906 and the ID are
messy and with several possible vulnerabilities.&nbsp; Here is a suggestion
that might be useful.&nbsp; Define a new broadcast packet format as the NTP
header followed by one extension field representing the MAC.&nbsp; The
extension field represents the signature of the header computed using
the private key of the server and can be verified by the client using
the public key in the server certificate.&nbsp; Originally, this was
considered an unacceptable dilution of accuracy due to the variability
of the signature computing time and the length of the signature.&nbsp;
However, at least in modern NTP, the interleaved broadcast mode can be
used to correct for this overhead.&nbsp; This assumes that the overhead for
either the server or client is reduced using fast modern processors.<br>
<br>
As mentioned previously, in NTP symmetric mode the current protocol in
RFC 5906 should be changed to use Diffie-Hellman key agreement.&nbsp; This
is so easy to do that it can be incorporated separately in the standard.<br>
<br>
While the interleaved mode is useful in symmetric and broadcast modes,
it can't be used when the server does not maintain state separately for
each client.&nbsp; However, as in the orginal NTP design many years ago, the
output delay can be computed for each transmitted server packet.&nbsp; The
output delay can be averaged and used to augment&nbsp; the transmit
timestamp or retrieve using an extension field.<br>
<br>
Note that this message does not address the cookie snatcher issue
discussed by our German protomeisters, who have concocted a brillient
solution.&nbsp; Many of the issues discussed are relevant to both NTP and
PTP, although the on-wire protocol details may be different.&nbsp; I hope
this discussion is useful in either context.<br>
<br>
Dave<br>
<br>
<br>
</font>
</body>
</html>

--------------040509030609050009080808--

From stbryant@cisco.com  Fri Aug  2 02:07:23 2013
Return-Path: <stbryant@cisco.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 8B50F11E833F; Fri,  2 Aug 2013 02:07:22 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -110.599
X-Spam-Level: 
X-Spam-Status: No, score=-110.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id sWU1kD89p-JK; Fri,  2 Aug 2013 02:07:01 -0700 (PDT)
Received: from ams-iport-2.cisco.com (ams-iport-2.cisco.com [144.254.224.141]) by ietfa.amsl.com (Postfix) with ESMTP id 8466511E82ED; Fri,  2 Aug 2013 02:01:50 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=58039; q=dns/txt; s=iport; t=1375434119; x=1376643719; h=message-id:date:from:reply-to:mime-version:to:cc:subject: content-transfer-encoding; bh=6RKliOZig2Q4B/WoFXscPAj7WgBtMisPEzxRgz0CBl0=; b=A8vL9iV16t5mXlY9rHbrlAeCMxJVAH45yCmgbM/tw/FalwkL428OdnXG ya+julkhqTucosppplQWTvcQZMDIxrf4h56ZImrJA4j+ly/0/ccFKSt/v UbvCdlOEG/OxCRxT1isnCNFF6tbGmH+95KItIzKMzMChNiFi9FI+6PlpA Q=;
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AkEFAKd0+1GQ/khM/2dsb2JhbABRCoMGATS/HYEbFnSCRQEdLgESLQ8MKAI4FAEJAwEFAgEBF4d1DJ8/mT6ONAIJBgWBGyIRhAIDiy6BEocHQYNXgSqFbIo4gxiBaAgX
X-IronPort-AV: E=Sophos;i="4.89,800,1367971200"; d="scan'208";a="85379318"
Received: from ams-core-3.cisco.com ([144.254.72.76]) by ams-iport-2.cisco.com with ESMTP; 02 Aug 2013 09:00:54 +0000
Received: from cisco.com (mrwint.cisco.com [64.103.70.36]) by ams-core-3.cisco.com (8.14.5/8.14.5) with ESMTP id r7290p3w026368 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 2 Aug 2013 09:00:52 GMT
Received: from [IPv6:::1] (localhost [127.0.0.1]) by cisco.com (8.14.4+Sun/8.8.8) with ESMTP id r7290oBR001361; Fri, 2 Aug 2013 10:00:50 +0100 (BST)
Message-ID: <51FB7543.801@cisco.com>
Date: Fri, 02 Aug 2013 10:00:51 +0100
From: Stewart Bryant <stbryant@cisco.com>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:17.0) Gecko/20130620 Thunderbird/17.0.7
MIME-Version: 1.0
To: "mpls-chairs@tools.ietf.org" <mpls-chairs@tools.ietf.org>, "tictoc-chairs@tools.ietf.org" <tictoc-chairs@tools.ietf.org>, int-ads@tools.ietf.org, "rtg-ads@tools.ietf.org" <rtg-ads@tools.ietf.org>, "mpls-chairs@tools.ietf.org" <mpls-chairs@tools.ietf.org>, draft-ietf-tictoc-1588overmpls@tools.ietf.org
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
Reply-To: stbryant@cisco.com
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 02 Aug 2013 09:07:34 -0000
X-List-Received-Date: Fri, 02 Aug 2013 09:07:34 -0000

SB> This draft does not seem to provide a precise definition
SB> the properties of the new LSP type that it wishes to
SB> define, in particular it does it define the PHB of those
SB> LSPs, nor the full interaction with the MPLS
SB> architecture.
SB>
SB> I have not tracked TICTOC for a while but I thought that
SB> the original plan was to define the concept of an offset
SB> into a packet to do the correction.
SB>
SB> It is disappointing that the opportunity was not taken
SB> to define a timing shim inside the timing LSP so that
SB> a time correction could be added to any packet such that
SB> the MPLS system was isolated from the details of the
SB> complexity of the particular time transfer type.

SB> I think that much more clarify is needed in terms of
SB> definition of the new LSP type, since it is unclear
SB> from this text how to implement one.
SB>
SB> There are a lot of other MPLS services such as
SB> LSP ping that need to be considered.
SB>
SB> Please see inline for more comments. However these
SB> comments are made in the context of the text as written
SB> whilst I have a fundamental concern that this approach
SB> lacks an MPLS architectural soundness that need
SB> greater thought with significant impact on the
SB> draft.

- Stewart


TICTOC Working Group                                           S. Davari
Internet-Draft                                                   A. Oren
Intended status: Standards Track                          Broadcom Corp.
Expires: December 17, 2013                                     M. Bhatia
                                                               P. Roberts
                                                           Alcatel-Lucent
                                                               L. Montini
                                                               L. Martini
                                                            Cisco Systems
                                                            June 15, 2013


             Transporting Timing messages over MPLS Networks
                    draft-ietf-tictoc-1588overmpls-05

Abstract

    This document defines the method for transporting Timing messages
    such as PTP and NTP over an MPLS network.  The method allows for the
    easy identification of these PDUs at the port level to allow for port

SB> What is a port

    level processing of these PDUs in both LERs and LSRs.

    The basic idea is to transport Timing messages inside dedicated MPLS
    LSPs.  These LSPs only carry Timing messages and possibly Control and
    Management packets, but they do not carry customer traffic.

SB> More specifically they only carry traffic associated with the
SB> timing service and its support.
SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
SB> also it gets carried in a structure that causes it to get
SB> timestamped.

    Two methods for transporting Timing messages over MPLS are defined.

SB> Perhaps the right approach is to define the new LSP type and then
SB> seperately to define  the mapping of the various timing services
SB> over that LSP type.

    The first method is to transport Timing messages directly over the
    dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
    MPLS networks.  The second method is to transport Timing messages
    inside a PW via Ethernet encapsulation.

SB> I think that we should note that there are some
SB> h/w reasons for this preference. A clean sheet approach
SB> would have been to use PTP over MPLS with no intermediate
SB> layers.

Status of this Memo

    This Internet-Draft is submitted in full conformance with the
    provisions of BCP 78 and BCP 79.

    Internet-Drafts are working documents of the Internet Engineering
    Task Force (IETF).  Note that other groups may also distribute
    working documents as Internet-Drafts.  The list of current Internet-
    Drafts is at http://datatracker.ietf.org/drafts/current/.

    Internet-Drafts are draft documents valid for a maximum of six months
    and may be updated, replaced, or obsoleted by other documents at any
    time.  It is inappropriate to use Internet-Drafts as reference
    material or to cite them other than as "work in progress."

    This Internet-Draft will expire on December 17, 2013.



Davari, et al.          Expires December 17, 2013               [Page 1]

Internet-Draft        Transporting Timing over MPLS            June 2013


Copyright Notice

    Copyright (c) 2013 IETF Trust and the persons identified as the
    document authors.  All rights reserved.

    This document is subject to BCP 78 and the IETF Trust's Legal
    Provisions Relating to IETF Documents
    (http://trustee.ietf.org/license-info) in effect on the date of
    publication of this document.  Please review these documents
    carefully, as they describe your rights and restrictions with respect
    to this document.  Code Components extracted from this document must
    include Simplified BSD License text as described in Section 4.e of
    the Trust Legal Provisions and are provided without warranty as
    described in the Simplified BSD License.


Table of Contents

    1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  5

    2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .  7

    3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .  8

    4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .  9

    5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . . 12

    6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . . 13
      6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . . 13
      6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . . 13
      6.3.  Other Timing Encapsulation methods . . . . . . . . . . . . 14

    7.  Timing message Processing  . . . . . . . . . . . . . . . . . . 15

    8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . . 16

    9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

    10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

    11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

    12. OAM, Control and Management  . . . . . . . . . . . . . . . . . 20

    13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . . 21

    14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . . 22



Davari, et al.          Expires December 17, 2013               [Page 2]

Internet-Draft        Transporting Timing over MPLS            June 2013


    15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . . 23
      15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . . 23
      15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . . 23
      15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . . 24

    16. Other considerations . . . . . . . . . . . . . . . . . . . . . 25

    17. Security Considerations  . . . . . . . . . . . . . . . . . . . 26

    18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 27

    19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 28

    20. References . . . . . . . . . . . . . . . . . . . . . . . . . . 29
      20.1. Normative References . . . . . . . . . . . . . . . . . . . 29
      20.2. Informative References . . . . . . . . . . . . . . . . . . 29

    Appendix 1.  Routing extensions for Timing-aware Routers . . . . . 32

    Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . . 33

    Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 34





























Davari, et al.          Expires December 17, 2013               [Page 3]

Internet-Draft        Transporting Timing over MPLS            June 2013


    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
    "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
    document are to be interpreted as described in RFC2119 [RFC2119].

    When used in lower case, these words convey their typical use in
    common language, and are not to be interpreted as described in
    RFC2119 [RFC2119].












































Davari, et al.          Expires December 17, 2013               [Page 4]

Internet-Draft        Transporting Timing over MPLS            June 2013


1.  Introduction

    The objective of Precision Time Protocol (PTP) and Network Timing
    Protocol (NTP) are to synchronize independent clocks running on
    separate nodes of a distributed system.

    [IEEE-1588] defines PTP messages for frequency, phase and time
    synchronization.  The PTP messages include PTP PDUs over UDP/IP
    (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F of
    [IEEE-1588]).

SB> Sure BUT it is acknowledged that IEEE is open to the definition
SB> of other PTP mappings if they provide better optimisation.

    This document defines mapping and transport of the PTP
    messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
    defines several clock types: ordinary clocks, boundary clocks, end-
    to-end transparent clocks, and peer-to-peer transparent clocks.
    Transparent clocks require intermediate nodes to update correction
    field inside PTP message that reflects the transit time in the node.

    [RFC5905] defines NTP messages for clock and time synchronization.
    The PTP messages (PDUs) are transported over UDP/IP.  This document
SB> Should that be NTP messages?
SB> It needs to be made clear as soon as you introduce NTP that
SB> they use different time representations.

    defines mapping and transport of the NTP messages defined in
    [RFC5905] over MPLS networks.

    One key attribute of all of these Timing messages is that the Time
    stamp processing should occur as close as possible to the actual
    transmission and reception at the physical port interface.  This
    targets optimal time and/or frequency recovery by avoiding variable
    delay introduced by queues internal to the clocks.

SB> As I recall NTP has no epoch point defined, and I am not sure
SB> where that point is in the case of PTP in this mapping
SB> Hopefully this will get defined in due course.

    To facilitate the fast and efficient recognition of Timing messages
    at the port level when the Timing messages are carried over MPLS
    LSPs,

SB> Over a new LSP type with time optimied characteristics

    this document defines the specific encapsulations that should
    be used.
SB> Hopefully it will also define the PHP

    In addition, it can be expected that there will exist LSR/
    LERs where only a subset of the physical ports will have the port-
    based Timing message processing capabilities.
SB> Do you need to clarify that this only works at base and not in
SB> a label heirarchy.


    In order to ensure
    that the LSPs carrying Timing packets always enter and exit ports
    with this capability, routing extensions are defined to advertise
    this capability on a port basis and to allow for the establishment of
    LSPs that only transit such ports.  While this path establishment
    restriction may be applied only at the LER Ingress and/or egress
    ports, it becomes more important when using transparent clock capable
    LSRs in the path.
SB> I do not understand the implications of the last
SB> sentences - starting ", it becomes"


    Port based Timing message processing involves Timing message
    recognition.  Once the Timing messages are recognized they can be
    modified based on the reception or transmission Time-stamp.

    This document provides two methods for transporting Timing messages
    over MPLS.  One is applicable to MPLS environment and the other one
    is applicable to MPLS/MPLS-TP environment

SB> I think the sentence is incomplete.



Davari, et al.          Expires December 17, 2013               [Page 5]

Internet-Draft        Transporting Timing over MPLS            June 2013


    The solution involves transporting Timing messages over dedicated
    LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
    carry Management and control messages, but not data plane client
    traffic.

SB> It is not clear why this restriction applies.

    Timing LSPs can be established statically or via signaling.
SB> s/statically/by provisioning/network management/

    Extensions to control plane (OSPF, ISIS, etc.) is required to enable
    routers to distribute their Timing processing capabilities over MPLS
    to other routers.  However such extensions are outside the scope of
    this document.

    When signaling is used to setup the PTP LSP, Extensions to signaling
SB> is it a PTP LSP or a Timing LSP?

    protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
    However such extensions are outside the scope of this document.

SB> for mpls-tp GMPLS is the signalling protocol

    While the techniques included herein allow for the establishment of
    paths optimized to include Time-stamping capable links, the
    performance of the Slave clocks is outside the scope of this
    document.

    At the time of publishing this specification, Transparent Clocking
    (TC) is only defined for PTP.  Therefore at this time any part of
    this specification that talks about Transparent Clocking applies only
    to PTP.





























Davari, et al.          Expires December 17, 2013               [Page 6]

Internet-Draft        Transporting Timing over MPLS            June 2013


2.  Terminology

    1588: The timing and synchronization as defined by IEEE 1588.

SB> I think that there is a more formal name for the 1588 group
SB> that needs to be used here.
SB> Also do we need to talk about 1588-200? as there is
SB> an update in progress

    NTP: The timing and synchronization protocol defined by IETF RFC-1305
    and RFC-5905.

    PTP: The timing and synchronization protocol used by 1588.
SB> need the proper name for 1588

    Master Clock: The source of 1588 timing to a set of slave clocks.

    Master Port: A port on a ordinary or boundary clock that is in Master
    state.  This is the source of timing toward slave ports.

SB> I am not sure the reader knows what a port is

    Slave Clock: A receiver of 1588 timing from a master clock.

    Slave Port: A port on a boundary clock or ordinary clock that is
    receiving timing from a master clock.

    Ordinary Clock: A device with a single PTP port.

    Transparent Clock.  A device that measures the time taken for a PTP
    event message to transit the device and then updates the
    correctionField of the message with this transit time.

    Boundary Clock: A device with more than one PTP port.  Generally
    boundary clocks will have one port in slave state to receive timing
    and then other ports in master state to re-distribute the timing.

    PTP LSP: An LSP dedicated to carry PTP messages

SB> PTP or timing?


    PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
    messages.

SB> Ah I don't think that PWE3 know what one of these is

    CW: Pseudowire Control Word

    LAG: Link Aggregation

    ECMP: Equal Cost Multipath

    CF: Correction Field, a field inside certain PTP messages (message
    type 0-3)that holds the accumulative transit time inside intermediate
    switches

    Timing messages: Timing Protocol messages that are exchanged between
    routers in order to establish a synchronized clock.

SB> A number of these definitions look like copies of IEEE1588
SB> definitions. We need to provide references and note the
SB> priority of the IEEE base reference.



Davari, et al.          Expires December 17, 2013               [Page 7]

Internet-Draft        Transporting Timing over MPLS            June 2013


3.  Problem Statement

    [IEEE-1588] has defined methods for transporting PTP messages over
    Ethernet and IP networks.  [RFC5905] has defined the method of
    transporting NTP messages over IP networks.  There is a need to
    transport Timing messages over MPLS networks while supporting the
    Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
    functionality in the LER and LSRs in the MPLS network.

    There are multiple ways of transporting Timing over MPLS.  However,
    there is a requirement to limit the possible encapsulation options to
    simplify the Timing message identification and processing required at
    the port level.

    When Timing-awareness is needed, Timing messages should not be
    transported over LSPs or PWs that are carrying customer traffic
    because LSRs perform Label switching based on the top label in the
    stack.

SB> Have you explained why?

    To detect Timing messages inside such LSPs require special
    hardware to do deep packet inspection at line rate.  Even if such
    hardware exists, the payload can't be deterministically identified by
    LSRs because the payload type is a context of the PW label, and the
    PW label and its context are only known to the Edge routers (PEs/
    LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CES,
    etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
    LSRs dont have the knowledge of whether PW Control Word (CW) is
    present or not and therefore can not deterministically identify the
    payload.

    A generic method is defined in this document that does not require
    deep packet inspection at line rate, and can deterministically
    identify Timing messages.  This method can be used to detect Timing
    Messages in both one-step and two-step clock implementations of
    ordinary, boundary and transparent clocks.

SB> Needs a ref and I am sure many MPLS specialists will not understand
SB> the msg types.


















Davari, et al.          Expires December 17, 2013               [Page 8]

Internet-Draft        Transporting Timing over MPLS            June 2013


4.  Timing over MPLS Architecture

    Timing messages are exchange between Timing ports on ordinary and

SB> Have you defined a timing port?

    boundary clocks.  Boundary clocks terminate the Timing messages and
    act as master for other boundary clocks or for slave clocks.  End-to-
    End Transparent clocks do not terminate the Timing messages but they
    do modify the contents of the Timing messages as they transit across
    the transparent clock.

    Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Clock

    (TC) could be implemented in either LERs or LSRs.

SB> LER and LSR need to be expanded

    An example is shown in Figure 1, where the LERs act as Ordinary Clock
    (OC) and are the initiating/terminating point for Timing messages.
    The ingress LER encapsulates the Timing messages in Timing LSP and
    the Egress LER terminates the Timing LSP.  The LSRs act as
    Transparent Clock (TC) and just update the Timing field in the Timing
    messages.


       +--------+     +-------+     +-------+     +-------+     +--------+
       |Switch, |     |       |     |       |     |       |     |Switch, |
       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
       |        |     |  OC   |     |  TC   |     |  OC   |     |        |
       +--------+     +-------+     +-------+     +-------+     +--------+
                      /                                 \
       +-------+     /                                   \     +-------+
       |  LER  |    /                                     \    |  LER  |
       | Master|---/                                       \---| Slave |
       | Clock |                                               | Clock |
       +-------+                                               +-------+

      Figure (1) - Deployment example 1 of timing over MPLS network

    Another example is shown in Figure2, where LERs terminate the Timing
    messages received from switch/routers that are outside of the MPLS
    network acting as OC or BC.  In this example LERs regenerate the
    clock and initiate timing messages encapsulated in Timing LSP toward
    the MPLS network, while the LSRs act as Transparent Clock (TC) and
    just update the Timing field in the Timing messages, which are
    already encapsulated in Timing LSPs.









Davari, et al.          Expires December 17, 2013               [Page 9]

Internet-Draft        Transporting Timing over MPLS            June 2013


      +--------+     +-------+     +-------+     +-------+     +--------+
      |Switch, |     |       |     |       |     |       |     |Switch, |
      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
      | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC  |
      +--------+     +-------+     +-------+     +-------+     +--------+

      Figure (2) - Deployment example 2 of timing over MPLS network


    Another example is shown in Figure 3, where LERs do not terminate the
    Timing messages received from switch/routers that are outside of the
    MPLS network acting as OC, TC or BC.  The LERs act as TC and update
    the Timing field in the Timing messages as they transit the LER,
    while encapsulating them in timing LSP.  The LSRs also act as
    Transparent Clock (TC) and just update the Timing field in the Timing
    messages which are already encapsulated in Timing LSPs.

       +--------+     +-------+     +-------+     +-------+     +--------+
       |Switch, |     |       |     |       |     |       |     |Switch, |
       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
       |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC/BC|
       +--------+     +-------+     +-------+     +-------+     +--------+

     Figure (3) - Deployment example 3 of timing over MPLS network

    Another example is shown in Figure 4, where LERs and LSRs support
    Boundary Clocks.  A single-hop LSP is created between two adjacent
    LSRs engaged in BC operation.  Other methods such as PTP transport
    over Ethernet MAY be used for transporting timing messages if the
    link between the two routers is Ethernet.

      +--------+     +-------+     +-------+     +-------+     +--------+
      |Switch, |     |       |     |       |     |       |     |Switch, |
      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
      | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC  |
      +--------+     +-------+     +-------+     +-------+     +--------+

    Figure (4) - Deployment example 3 of timing over MPLS network

    An MPLS domain MAY serve multiple customers.  In these cases the MPLS
    domain (maintained by a service provider) may provide timing services
    to multiple customers, each having their own Timing domain.

    The Timing over MPLS architecture assumes full mesh of Timing LSPs
    between all LERs supporting this specification.

SB> Note sure this is right - the salves surely do not need to
SB> exchange timing amongst themselves

    It supports
    Point-to- point (VPWS) and Multipoint (VPLS) services.

SB> What does that mean? You do not carry user data traffic?
SB> Maybe it's the ordering of the statemnets that is causing
SB> confusion.

    This means
    that a customer may purchase a Point-to-point Timing service between
    two customer sites or a Multipoint Timing service between more than



Davari, et al.          Expires December 17, 2013              [Page 10]

Internet-Draft        Transporting Timing over MPLS            June 2013


    two customer sites.

    The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
    This means that the Timing Multicast messages such as PTP Multicast
    event messages can be transported over P2MP Timing LSP or be
    replicated and transported over many P2P Timing LSPs.

SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
SB> nor a P2MP PW, although we are close.

    Timing messages, that do not require Time stamping or Correction
    Field update MAY be transported over Timing LSPs to simplify hardware
    and software.

    PTP Announce messages that determine the Timing LSP terminating point
    behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
    to simplify hardware and software.

SB> have you defined and referenced PTP announce msgs?





































Davari, et al.          Expires December 17, 2013              [Page 11]

Internet-Draft        Transporting Timing over MPLS            June 2013


5.  Dedicated LSPs for Timing messages

    Many methods have been considered for identifying the Timing messages
    when they are encapsulated in MPLS such as using GAL/G-ACH or a new
    reserved label.  These methods were not attractive since they either
    required deep packet inspection at line rate in the intermediate LSRs
    or they required use of a scarce new reserved label.  Also one of the
    goals was to reuse existing OAM mechanisms.

SB> RLs = SPLs are not so rare now. In any case needs a ref.

    The method defined in this document can be used by LER and LSRs to
    identify Timing messages in MPLS tunnels by just looking at the top
    label in the MPLS label stack, which only carry Timing messages as
    well as OAM, but not data plane client traffic.

    Compliant implementations MUST use dedicated LSPs to carry Timing
    messages over MPLS.

SB> I think that we need a definition of the properies of these LSPs

    These LSPs are herein referred to as "Timing
    LSPs" and the labels associated with these LSPs as "Timing LSP
    labels".  The Timing LSPs that runs between Ingress and Egress LERs
    MUST be co-routed.  Alternatively, a single bidirectional co-routed
    LSP can be used.

SB> I though that you said you could use M2MP LSPs - these are not
SB> bidirectional.

    Co-routing of the two directions is required to limit the difference
    in the delays in the Master clock to Slave clock direction compared
    to the Slave clock to Master clock direction.  The Timing LSP MAY be
    MPLS/MPLS-TP LSP.

    The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
    New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
    outside the scope of this document.

    The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such as
    BFD and LSP Ping but the LSP data plane client plane traffic MUST be
    Timing packets only.

SB> Why?


















Davari, et al.          Expires December 17, 2013              [Page 12]

Internet-Draft        Transporting Timing over MPLS            June 2013


6.  Timing over LSP Encapsulation

The encapsulations is not LSP is it?

    This document defines two methods for carrying Timing messages over
    MPLS.  The first method is carrying UDP/IP encapsulated Timing
    messages over Timing LSPs, and the second method, is carrying
    Ethernet encapsulated Timing messages over Ethernet PWs inside Timing
    LSPs.

6.1.  Timing over UDP/IP over MPLS Encapsulation

    The simplest method of transporting Timing messages over MPLS is to
    encapsulate Timing PDUs in UDP/IP and then encapsulate them in Timing
    LSP.  This format is shown in Figure 4.


                     +----------------------+
                     |   Timing LSP Label   |
                     +----------------------+
                     |        IPv4/6        |
                     +----------------------+
                     |         UDP          |
                     +----------------------+
                     |     Timing PDU       |
                     +----------------------+

       Figure (4) - Timing over UDP/IP over MPLS Encapsulation


    This encapsulation is very simple and is useful when the network
    between Timing Master Clock and Slave Clock is MPLS network.

SB> Simple is a judgement call

    In order for an LER/LSR to process Timing messages, the Timing LSP
    Label must be at the top label of the label stack.  The LER/LSR MUST
    know that the Timing LSP Label is used for carrying Timing messages.
    This can be accomplished via static configuration or via RSVP-TE
    signaling.

    The UDP/IP encapsulation of PTP MUST follow Annex D and E of
    [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
    [RFC5905].

6.2.  Timing over PW Encapsulation

    Another method of transporting Timing over MPLS networks is by
    encapsulating Timing PDUs in PW which in turn is transported over
    Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
    shown in Fig 5(A) MUST be used and the Ethernet encapsulation of PTP
    MUST follow Annex F of [IEEE-1588].



Davari, et al.          Expires December 17, 2013              [Page 13]

Internet-Draft        Transporting Timing over MPLS            June 2013


    The RAW mode or Tagged mode defined in [RFC4448] MAY be used and the
    Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
    Timing over PW encapsulation MUST use the Control Word (CW) as
    specified in [RFC4448] to ensure proper detection of PTP messages
    inside the MPLS packets for Timing over LSP and Timing over PW
    encapsulation.

SB> That needs explanation

    The use of Sequence Number in the CW is optional.

SB> Given that s/n are never in practice deployed, you could probably
SB> simplify things by sayig that they are not used.

    Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over PW
    (the IP PW discussed in [RFC4447]) shown in Fig 5(B).

                     +----------------+  +----------------+
                      |Timing LSP Label|  |Timing LSP Label|
                      +----------------+  +----------------+
                      |    PW Label    |  |    PW Label    |
                      +----------------+  +----------------+
                      |  Control Word  |  |      IP        |
                      +----------------+  +----------------+
                      |    Ethernet    |  |      UDP       |
                      |     Header     |  +----------------+
                      +----------------+  |   Timing PDU   |
                      |S-VLAN(Optional)|  |                |
                      +----------------+  +----------------+
                      |C-VLAN(Optional)|        (B)
                      +----------------+
                      |   Timing PDU   |
                      |                |
                      +----------------+
                             (A)

               Figure (5) - Timing over PW Encapsulations

    In order for an LSR to process PTP messages, the top label of the
    label stack (the Tunnel Label) MUST be a Timing label.

S> You said that before.

6.3.  Other Timing Encapsulation methods

    In future other timing encapsulation methods may be introduced, such
    as a new shim header after the Bottom of Stack to carry the Timing
    information.  Such new encapsulations are outside the scope of this
    document.


SB> Taking a pure MPLS pov, you can simplify a lot of the text
SB> out of the definition of the LSP








Davari, et al.          Expires December 17, 2013              [Page 14]

Internet-Draft        Transporting Timing over MPLS            June 2013


SB> I think we need a section on LSP processing

7.  Timing message Processing

    Each Timing protocol such as PTP and NTP, define their set of Timing
    messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
    FOLLOW_UP, etc messages.

    Some of the Timing messages require time stamping or correction field
    update at port level and some dont.  It is the job of the LER/LSR to
    parse the timing message and find out the type of the Timing message
    and decide whether and how to Time- stamp it (e.g., BC) or update
    correction field(e.g., TC).


SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
SB> function rather than the LER?

    For example the following PTP messages (called Event messages)
    require time-stamping or correction field update:

    o  SYNC

    o  DELAY_REQ (Delay Request)

    o  PDELAY_REQ (Peer Delay Request)

    o  PDELAY_RESP (Peer Delay Response)

    SYNC and DELAY_REQ are exchanged between Master Clock and Slave Clock
    and MUST be transported over PTP LSPs.  PDELAY_REQ and PDELAY_RESP
    are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
    Boundary, or Transparent) and SHOULD be transported over single hop
    PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
    and PDELAY_RESP_FOLLOW_UP messages MUST also be transported over the
    PTP LSPs.

    For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
    transported over two PTP LSPs that are in opposite directions.  These
    PTP LSPs, which are in opposite directions MUST be congruent and co-
    routed.  Alternatively, a single bidirectional co-routed LSP can be
    used.

    Except as indicated above for the two-step PTP clocks, Non-Event PTP
    message types do not need to be processed by intermediate routers.
    These message types MAY be carried in PTP Tunnel LSPs.

SB> Are you saying that a timing P router has to be msg type sensitive?









Davari, et al.          Expires December 17, 2013              [Page 15]

Internet-Draft        Transporting Timing over MPLS            June 2013


8.  Protection and Redundancy


SB> This is a bit of a jump - I don't know how the LSP itself works yet!

    In order to ensure continuous uninterrupted operation of slave
    clocks, usually as a general practice, slave clocks (or ports) track
    redundant master clocks.

    It is the responsibility of the network operator to ensure that
    physically disjoint Timing LSPs are established between a slave clock
    (or port) and redundant master clocks (or ports).

    When a slave clock (or port) listens to redundant master clocks or
    ports, any prolonged Timing LSP outage will trigger the slave clock
    or port to switch to a redundant master clock or port.

    LSP/PW protection such as Linear protection Switching (1:1, 1+1),
    Ring protection switching or MPLS Fast Reroute (FRR) generally switch
    alternative path that usually cause a change in delay, which if
    undetected by slave clock can reduce accuracy of the slave clock.

    Therefore protection switching MAY be used, as long as phase jumps
    upon switchover due to differences in path latency are detected and
    compensated for (such compensation not being required if BCs or peer-
    peer TCs are used throughout).

    Note that any protection or reroute mechanism that adds additional
    MPLS label to the label stack, such as Facility Backup Fast Reroute,
    MUST ensure that the pushed label is also a Timing Label to ensure
    recognition of the MPLS frame as containing Timing messages, as it
    transits the backup path.






















Davari, et al.          Expires December 17, 2013              [Page 16]

Internet-Draft        Transporting Timing over MPLS            June 2013


9.  ECMP

    To ensure the optimal operation of slave clocks and avoid error
    introduced by forward and reverse path delay asymmetry, the physical
    path for Timing messages from master clock to slave Clock and vice
    versa must be the same for all Event Timing messages listed in
    section 7.

    Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
    Multipath).









































Davari, et al.          Expires December 17, 2013              [Page 17]

Internet-Draft        Transporting Timing over MPLS            June 2013


10.  PHP

    To ensure that the label on the top of the label stack is the Timing
    LSP Label, PHP MUST not be used.















































Davari, et al.          Expires December 17, 2013              [Page 18]

Internet-Draft        Transporting Timing over MPLS            June 2013


11.  Entropy

    To ensure all Timing messages in a Timing LSP take the same path,
    Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
    Entropy Label MUST NOT be used for the PWs that are carried inside
    Timing LSP [RFC6391].

SB> This is incorrect - you mean that all msgs of the same timing
SB> flow need to have the same EL value.













































Davari, et al.          Expires December 17, 2013              [Page 19]

Internet-Draft        Transporting Timing over MPLS            June 2013


12.  OAM, Control and Management

    In order to monitor Timing LSPs and their encapsulated PWs, they MUST
    be able to carry OAM and management messages.  These management
    messages MUST be differentiated from Timing messages via already
    defined IETF methods.

    For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
    over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
    Management protocols can easily be identified by the UDP Destination
    Port number or by GAL/G-ACH respectively.

    Also BFD, LSP-Ping and other management messages MAY run over the PWs
    encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3 or
    4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is going
    to be deprecated by IETF).  In this case G-ACH, PW label (TTL=1) or
    GAL-ACH are used to identify such management messages.


































Davari, et al.          Expires December 17, 2013              [Page 20]

Internet-Draft        Transporting Timing over MPLS            June 2013


13.  QoS Considerations

    In network deployments where not every LSR/LER is Timing-aware, it is
    important to reduce the impact of the non-Timing-aware LSR/LERs on
    the timing recovery in the slave clock.  The Timing messages are time
    critical and must be treated with the highest priority.  Therefore
    Timing over MPLS messages must be treated with the highest priority
    in the routers.  This can be achieved by proper setup of Timing LSPs.

    It is recommended that the Timing LSPs are setup or configured
    properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC2697]
    for drop eligibility.







































Davari, et al.          Expires December 17, 2013              [Page 21]

Internet-Draft        Transporting Timing over MPLS            June 2013


14.  FCS and Checksum Recalculation

    When time-stamp generation and timing packet adjustment is performed
    near the physical port hardware, the process MUST include
    recalculation of the Ethernet FCS.

SB> The above is confusing - an LSR always recomputes the link layer
SB> CRC which may or may not be Ethernet.

    Also FCS retention for the
    payload Ethernet described in [RFC4720] MUST NOT be used.

    For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
    may be required as per UDP transport standards.

SB> You really need to be working on getting the IPv6 C?S computation
SB> removed from PTP msgs.

    When UDP checksum is used, each Timing-aware LER/LSR must either
    incrementally update the UDP checksum after Time stamping or
    Correction Field update or verify the UDP checksum on reception from
    upstream and recalculate the checksum completely on transmission to
    downstream node after Time stamping or Correction Field update.




































Davari, et al.          Expires December 17, 2013              [Page 22]

Internet-Draft        Transporting Timing over MPLS            June 2013


15.  Behavior of LER/LSR

    Timing-capable/aware LERs and LSRs are routers that have one or more

SB> You mean physical interfaces?

    interfaces that can perform Timing operations (OC/BC/TC) on Timing
    packets and are configured to do so.  Timing-capable/aware LERs and
    LSRs can advertise their Timing-capability per-interface via control
    plane such as OSPF or IS-IS.
SB> ISIS and OSPF are routing protocols.

   The Timing-capable/aware LERs can then
    signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timing
    capability of LER and LSRs may be configured in a centralized
    controller and the Timing LSP may be setup using manual configuration
    or other methods such as SDN.

SB> it can also be configured individually rather then through
SB> a cebtral controllwe

15.1.  Behavior of Timing-capable/aware LER

    When a Timing-capable/aware LER behaves as a Transparent clock and
    receives a Timing message from a Timing-capable/aware non-MPLS
    interface, the LER updates the Correction Field (CF) and encapsulates
    and forwards the timing message over previously established Timing
    LSP.

SB> You need to call out the details so that people properly
SB> understand the definition of the new LSP.

    Also when a Timing message is received from a Timing-capable/
    aware MPLS interface, LER updates the Correction Filed (CF) and
    decapsulates the MPLS encapsulation and forwards the timing message
    to a non-MPLS interface.

    When a Timing-capable/aware LER behaves as a Boundary clock and
    receives a Timing message from a Timing-capable/aware non MPLS
    interface, the LER Timestamps the Timing packet and sends it to the
    LERs Boundary clock processing module.  Also when a Timing message is
    received from a Timing- capable/aware MPLS interface, the LER
    Timestamps the Timing packet and sends it to the LERs Boundary clock
    processing module.

    When a Timing-capable/aware LER behaves as an Ordinary Clock toward
    the MPLS network, and receives a Timing message from a Timing-
    capable/aware MPLS interface, the LER Timestamps the Timing packet
    and sends it to the LERs Ordinary clock processing module.

15.2.  Behavior of Timing-capable/aware LSR

    When a Timing-capable/aware LSR behaves as a Transparent clock and
    receives a Timing message from a Timing-capable/aware MPLS interface,
    The LSR updates the Correction Filed (CF) and forwards the timing
    message over another MPLS interface.

    When a Timing-capable/aware LSR behaves as a Boundary clock and
    receives a Timing message from a Timing-capable/aware MPLS interface.
    The LSR performs the functions of a Boundary Clock in terminating the
    received Timing message and re-generating a new timing message over
    another (or the same) MPLS interface.



Davari, et al.          Expires December 17, 2013              [Page 23]

Internet-Draft        Transporting Timing over MPLS            June 2013


15.3.  Behavior of non-Timing-capable/aware LSR

    It is most beneficial when all LSRs in the path of a Timing LSP be
    timing-Capable/aware LSRs.  This would ensure the highest quality
    time and clock synchronization by Timing Slave Clocks.  However, this
    specification does not mandate that all LSRs in path of a Timing LSP
    be Timing- capable/aware.

    Non-Timing-capable/aware LSRs just switch the packets encapsulated in
    Timing LSPs and dont perform any Timing operation (TC or BC).
    However as explained in QoS section the Timing over MPLS packets MUST
    be still be treated with the highest priority based on their Traffic
    Class (TC) marking.






































Davari, et al.          Expires December 17, 2013              [Page 24]

Internet-Draft        Transporting Timing over MPLS            June 2013


16.  Other considerations

    [IEEE-1588] defines an optional peer-to-peer Transparent clocking
    that requires peer delay measurement between two adjacent Timing-
    capable/ aware routers/switches.  Peer delay measurement messages
    need to be time stamped and terminated by the Timing-capable/aware
    routers/ switches.  This means that two adjacent LSRs may be engaged
    in a peer delay measurement.

    For transporting such peer delay measurement messages a single-hop
    LSP SHOULD to be created between the two adjacent LSRs engaged in
    peer delay measurement to carry peer delay measurement messages.
    Other methods such as PTP transport over Ethernet MAY be used for
    transporting peer delay measurement messages if the link between the
    two routers is Ethernet.

    In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ ware
    routers/switches MUST maintain a list of all the neighbors it needs
    to send a PDelay_Req to, where each neighbor corresponds to a timing
    LSP.

    The use of Explicit Null Label (Label= 0 or 2) is acceptable as long
    as either the Explicit Null label is the bottom of stack label
    (applicable only to UDP/IP encapsulation) or the label below the
    Explicit Null label is a PTP label.


























Davari, et al.          Expires December 17, 2013              [Page 25]

Internet-Draft        Transporting Timing over MPLS            June 2013


17.  Security Considerations

    MPLS PW security considerations in general are discussed in [RFC3985]
    and [RFC4447],and those considerations also apply to this document.

    An experimental security protocol is defined in [IEEE-1588].The PTP
    security extension and protocol provides group source authentication,
    message integrity, and replay attack protection for PTP messages.

    When the MPLS network (provider network) serves multiple customers,
    it is important to maintain and process each customers clock and
    Timing messages separately from other customers to ensure there is no
    cross- customer effect.  For example if an LER BC is synchronized to
    a specific grandmaster, belonging to customer A, then the LER MUST
    use that BC clock only for customer A to ensure that customer A
    cannot attack other customers by manipulating its time.

    Timing messages MAY be encrypted or authenticated, provided that the
    LERs/LSRs that are Timing capable/aware can authenticate/ decrypt the
    timing messages.































Davari, et al.          Expires December 17, 2013              [Page 26]

Internet-Draft        Transporting Timing over MPLS            June 2013


18.  Acknowledgements

    The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizrahi,
    Stefano Ruffini, Peter Meyer, and other members of IETF for reviewing
    and providing feedback on this draft.














































Davari, et al.          Expires December 17, 2013              [Page 27]

Internet-Draft        Transporting Timing over MPLS            June 2013


19.  IANA Considerations

    There are no IANA requirements in this specification.



















































Davari, et al.          Expires December 17, 2013              [Page 28]

Internet-Draft        Transporting Timing over MPLS            June 2013


20.  References

20.1.  Normative References

    [IEEE-1588]
               IEEE 1588-2008, "IEEE Standard for a Precision Clock
               Synchronization Protocol for Networked Measurement and
               Control Systems".

    [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
               Requirement Levels", BCP 14, RFC 2119, March 1997.

    [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
               Edge (PWE3) Architecture", RFC 3985, March 2005.

    [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
               Proxies (ND Proxy)", RFC 4389, April 2006.

    [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
               Heron, "Pseudowire Setup and Maintenance Using the Label
               Distribution Protocol (LDP)", RFC 4447, April 2006.

    [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
               "Encapsulation Methods for Transport of Ethernet over MPLS
               Networks", RFC 4448, April 2006.

    [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
               Emulation Edge-to-Edge (PWE3) Frame Check Sequence
               Retention", RFC 4720, November 2006.

    [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
               Connectivity Verification (VCCV): A Control Channel for
               Pseudowires", RFC 5085, December 2007.

    [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detection
               (BFD)", RFC 5880, June 2010.

    [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
               "Bidirectional Forwarding Detection (BFD) for MPLS Label
               Switched Paths (LSPs)", RFC 5884, June 2010.

20.2.  Informative References

    [I-D.ietf-pwe3-fat-pw]
               Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
               J., and S. Amante, "Flow Aware Transport of Pseudowires
               over an MPLS Packet Switched Network",
               draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.



Davari, et al.          Expires December 17, 2013              [Page 29]

Internet-Draft        Transporting Timing over MPLS            June 2013


    [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
               system routeing information exchange protocol for use in
               conjunction with the Protocol for providing the
               Connectionless-mode Network Service (ISO 8473)".

    [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
               dual environments", RFC 1195, December 1990.

    [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.

    [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
               Marker", RFC 2697, September 1999.

    [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boudec,
               J., Courtney, W., Davari, S., Firoiu, V., and D.
               Stiliadis, "An Expedited Forwarding PHB (Per-Hop
               Behavior)", RFC 3246, March 2002.

    [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineering
               (TE) Extensions to OSPF Version 2", RFC 3630,
               September 2003.

    [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
               System (IS-IS) Extensions for Traffic Engineering (TE)",
               RFC 3784, June 2004.

    [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
               Shaffer, "Extensions to OSPF for Advertising Optional
               Router Capabilities", RFC 4970, July 2007.

    [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
               System to Intermediate System (IS-IS) Extensions for
               Advertising Router Information", RFC 4971, July 2007.

    [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
               Topology (MT) Routing in Intermediate System to
               Intermediate Systems (IS-ISs)", RFC 5120, February 2008.

    [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
               Engineering", RFC 5305, October 2008.

    [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
               "Traffic Engineering Extensions to OSPF Version 3",
               RFC 5329, September 2008.

    [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
               for IPv6", RFC 5340, July 2008.




Davari, et al.          Expires December 17, 2013              [Page 30]

Internet-Draft        Transporting Timing over MPLS            June 2013


    [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network
               Time Protocol Version 4: Protocol and Algorithms
               Specification", RFC 5905, June 2010.

    [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
               J., and S. Amante, "Flow-Aware Transport of Pseudowires
               over an MPLS Packet Switched Network", RFC 6391,
               November 2011.

    [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
               L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
               RFC 6790, November 2012.







































Davari, et al.          Expires December 17, 2013              [Page 31]

Internet-Draft        Transporting Timing over MPLS            June 2013


1.  Routing extensions for Timing-aware Routers

    MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] and
    IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (TE)
    link information used for constraint-based routing.

    Indeed, it is useful to advertise data plane TE router link
    capabilities, such as the capability for a router to be Timing-aware.
    This capability MUST then be taken into account during path
    computation to prefer or even require links that advertise themselves
    as Timing-aware.  In this way the path can ensure the entry and exit
    points into the LERs and, if desired, the links into the LSRs are
    able to perform port based time-stamping thus minimizing their impact
    on the performance of the slave clock.

    extensions are required to OSPF and IS-IS in order to advertise
    Timing-aware capabilities of a link.  Such extensions are outside the
    scope of this document; however such extension SHOULD be able to
    signal the following information per Router Link:

    o  Capable of processing PTP, NTP or other Timing flows

    o  Capable of performing Transparent Clock operation

    o  Capable of performing Boundary Clock operation


























Davari, et al.          Expires December 17, 2013              [Page 32]

Internet-Draft        Transporting Timing over MPLS            June 2013


2.  Signaling Extensions for Creating Timing LSPs

    RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-TE
    is used to setup Timing LSPs, some information that indicates that
    the LSP is carrying Timing flows MUST be included in the new
    Extensions to RSVP-TE:

    The following information MAY also be included in the new Extensions
    to RSVP-TE:

    o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
       field

    o  Number of VLANs in case of PW encapsulation

    o  Timestamp field Type

       *  Correction Field, Timestamp

    o  Timestamp Field format

       *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
          NTP, etc.

    Note that in case the above optional information is signaled with
    RSVP-TE for a Timing LSP, all the Timing packets carried in that LSP
    must have the same signaled characteristics.  For example if
    Timestamp format is signaled as 64-bit PTPv1, then all Timing packets
    must use 64-bit PTPv1 time-stamp.






















Davari, et al.          Expires December 17, 2013              [Page 33]

Internet-Draft        Transporting Timing over MPLS            June 2013


Authors' Addresses

    Shahram Davari
    Broadcom Corp.
    San Jose, CA  95134
    USA

    Email: davari@broadcom.com


    Amit Oren
    Broadcom Corp.
    San Jose, CA  95134
    USA

    Email: amito@broadcom.com


    Manav Bhatia
    Alcatel-Lucent
    Bangalore,
    India

    Email: manav.bhatia@alcatel-lucent.com


    Peter Roberts
    Alcatel-Lucent
    Kanata,
    Canada

    Email: peter.roberts@alcatel-lucent.com


    Laurent Montini
    Cisco Systems
    San Jose CA
    USA

    Email: lmontini@cisco.com











Davari, et al.          Expires December 17, 2013              [Page 34]

Internet-Draft        Transporting Timing over MPLS            June 2013


    Luca
    Cisco Systems
    San Jose CA
    USA

    Email: lmartini@cisco.com













































Davari, et al.          Expires December 17, 2013              [Page 35]


-- 
For corporate legal information go to:

http://www.cisco.com/web/about/doing_business/legal/cri/index.html


From stbryant@cisco.com  Fri Aug  2 03:02:15 2013
Return-Path: <stbryant@cisco.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 1C9D821E82DB; Fri,  2 Aug 2013 03:02:15 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -110.599
X-Spam-Level: 
X-Spam-Status: No, score=-110.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id W53Wdj8NZ1AT; Fri,  2 Aug 2013 03:02:09 -0700 (PDT)
Received: from ams-iport-1.cisco.com (ams-iport-1.cisco.com [144.254.224.140]) by ietfa.amsl.com (Postfix) with ESMTP id BA2BC21E82B8; Fri,  2 Aug 2013 03:02:07 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=58853; q=dns/txt; s=iport; t=1375437728; x=1376647328; h=message-id:date:from:reply-to:mime-version:to:cc:subject: content-transfer-encoding; bh=V6+bXfJLr6FMwuS+ayZ8inCnBSrHT8BTYWA4steJYCU=; b=TPf+yaEYvtj2Q8Xb+6gLRb0NrBp/ZyNjmPmJ/b9K2/cYWk4aVGxxDPe9 lIfaETsGeh3A7nEEx69W/6A9hDelp2NagGk1dlHsGl5NFpbGFZpBexelP hnQATbgDFcIYnRjxnYcPTdF2HLKz32KDpMmu4hONE5VecXGSTmya4Uv2D I=;
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AisFACuD+1GQ/khL/2dsb2JhbABQCoMGNb8igRwWdIJFAR0uAREBLQ8MChgDAgECATcUAQkDAQUCAQEXh3UMn0aZQ45FAgkGBYEbIhGEAwOLLoEShwdBg1eBKoVsijiDGIFoCBc
X-IronPort-AV: E=Sophos;i="4.89,800,1367971200"; d="scan'208";a="157898215"
Received: from ams-core-2.cisco.com ([144.254.72.75]) by ams-iport-1.cisco.com with ESMTP; 02 Aug 2013 10:02:02 +0000
Received: from cisco.com (mrwint.cisco.com [64.103.70.36]) by ams-core-2.cisco.com (8.14.5/8.14.5) with ESMTP id r72A1x1f013725 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 2 Aug 2013 10:01:59 GMT
Received: from [IPv6:::1] (localhost [127.0.0.1]) by cisco.com (8.14.4+Sun/8.8.8) with ESMTP id r72A1wA0005902; Fri, 2 Aug 2013 11:01:59 +0100 (BST)
Message-ID: <51FB8396.9020504@cisco.com>
Date: Fri, 02 Aug 2013 11:01:58 +0100
From: Stewart Bryant <stbryant@cisco.com>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:17.0) Gecko/20130620 Thunderbird/17.0.7
MIME-Version: 1.0
To: "mpls-chairs@tools.ietf.org" <mpls-chairs@tools.ietf.org>, "tictoc-chairs@tools.ietf.org" <tictoc-chairs@tools.ietf.org>, int-ads@tools.ietf.org, "rtg-ads@tools.ietf.org" <rtg-ads@tools.ietf.org>, "mpls-chairs@tools.ietf.org" <mpls-chairs@tools.ietf.org>, draft-ietf-tictoc-1588overmpls@tools.ietf.org, John E Drake <jdrake@juniper.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
Reply-To: stbryant@cisco.com
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 02 Aug 2013 10:02:15 -0000

Talking to John Drake about this, an alternative general
model is for the LSP is that it is an LSP that timestamps
the packet and then passes it to an application associated
with the LSP at that hop.

This has a lot of merit.

If however we think about it some more we have a
type of network service chaining going on here and
so we don't need an LSP per say, because the path
and instruction can be in the packet.

In other words a general solution  to the problem
is to define an LSP with the properties that the
packet is passed one hop, timestamped and delivered
to the associated application.

How this is MPLS construct is used to support time
tranfer is entirely within the scope of TICTOC, but
at the MPLS layer we have a clean reusable network
service.

- Stewart


============
SB> This draft does not seem to provide a precise definition
SB> the properties of the new LSP type that it wishes to
SB> define, in particular it does it define the PHB of those
SB> LSPs, nor the full interaction with the MPLS
SB> architecture.
SB>
SB> I have not tracked TICTOC for a while but I thought that
SB> the original plan was to define the concept of an offset
SB> into a packet to do the correction.
SB>
SB> It is disappointing that the opportunity was not taken
SB> to define a timing shim inside the timing LSP so that
SB> a time correction could be added to any packet such that
SB> the MPLS system was isolated from the details of the
SB> complexity of the particular time transfer type.

SB> I think that much more clarify is needed in terms of
SB> definition of the new LSP type, since it is unclear
SB> from this text how to implement one.
SB>
SB> There are a lot of other MPLS services such as
SB> LSP ping that need to be considered.
SB>
SB> Please see inline for more comments. However these
SB> comments are made in the context of the text as written
SB> whilst I have a fundamental concern that this approach
SB> lacks an MPLS architectural soundness that need
SB> greater thought with significant impact on the
SB> draft.

- Stewart


TICTOC Working Group                                           S. Davari
Internet-Draft                                                   A. Oren
Intended status: Standards Track                          Broadcom Corp.
Expires: December 17, 2013                                     M. Bhatia
                                                               P. Roberts
                                                           Alcatel-Lucent
                                                               L. Montini
                                                               L. Martini
                                                            Cisco Systems
                                                            June 15, 2013


             Transporting Timing messages over MPLS Networks
                    draft-ietf-tictoc-1588overmpls-05

Abstract

    This document defines the method for transporting Timing messages
    such as PTP and NTP over an MPLS network.  The method allows for the
    easy identification of these PDUs at the port level to allow for port

SB> What is a port

    level processing of these PDUs in both LERs and LSRs.

    The basic idea is to transport Timing messages inside dedicated MPLS
    LSPs.  These LSPs only carry Timing messages and possibly Control and
    Management packets, but they do not carry customer traffic.

SB> More specifically they only carry traffic associated with the
SB> timing service and its support.
SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
SB> also it gets carried in a structure that causes it to get
SB> timestamped.

    Two methods for transporting Timing messages over MPLS are defined.

SB> Perhaps the right approach is to define the new LSP type and then
SB> seperately to define  the mapping of the various timing services
SB> over that LSP type.

    The first method is to transport Timing messages directly over the
    dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
    MPLS networks.  The second method is to transport Timing messages
    inside a PW via Ethernet encapsulation.

SB> I think that we should note that there are some
SB> h/w reasons for this preference. A clean sheet approach
SB> would have been to use PTP over MPLS with no intermediate
SB> layers.

Status of this Memo

    This Internet-Draft is submitted in full conformance with the
    provisions of BCP 78 and BCP 79.

    Internet-Drafts are working documents of the Internet Engineering
    Task Force (IETF).  Note that other groups may also distribute
    working documents as Internet-Drafts.  The list of current Internet-
    Drafts is at http://datatracker.ietf.org/drafts/current/.

    Internet-Drafts are draft documents valid for a maximum of six months
    and may be updated, replaced, or obsoleted by other documents at any
    time.  It is inappropriate to use Internet-Drafts as reference
    material or to cite them other than as "work in progress."

    This Internet-Draft will expire on December 17, 2013.



Davari, et al.          Expires December 17, 2013               [Page 1]

Internet-Draft        Transporting Timing over MPLS            June 2013


Copyright Notice

    Copyright (c) 2013 IETF Trust and the persons identified as the
    document authors.  All rights reserved.

    This document is subject to BCP 78 and the IETF Trust's Legal
    Provisions Relating to IETF Documents
    (http://trustee.ietf.org/license-info) in effect on the date of
    publication of this document.  Please review these documents
    carefully, as they describe your rights and restrictions with respect
    to this document.  Code Components extracted from this document must
    include Simplified BSD License text as described in Section 4.e of
    the Trust Legal Provisions and are provided without warranty as
    described in the Simplified BSD License.


Table of Contents

    1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  5

    2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .  7

    3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .  8

    4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .  9

    5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . . 12

    6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . . 13
      6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . . 13
      6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . . 13
      6.3.  Other Timing Encapsulation methods . . . . . . . . . . . . 14

    7.  Timing message Processing  . . . . . . . . . . . . . . . . . . 15

    8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . . 16

    9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

    10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

    11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

    12. OAM, Control and Management  . . . . . . . . . . . . . . . . . 20

    13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . . 21

    14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . . 22



Davari, et al.          Expires December 17, 2013               [Page 2]

Internet-Draft        Transporting Timing over MPLS            June 2013


    15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . . 23
      15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . . 23
      15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . . 23
      15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . . 24

    16. Other considerations . . . . . . . . . . . . . . . . . . . . . 25

    17. Security Considerations  . . . . . . . . . . . . . . . . . . . 26

    18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 27

    19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 28

    20. References . . . . . . . . . . . . . . . . . . . . . . . . . . 29
      20.1. Normative References . . . . . . . . . . . . . . . . . . . 29
      20.2. Informative References . . . . . . . . . . . . . . . . . . 29

    Appendix 1.  Routing extensions for Timing-aware Routers . . . . . 32

    Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . . 33

    Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 34





























Davari, et al.          Expires December 17, 2013               [Page 3]

Internet-Draft        Transporting Timing over MPLS            June 2013


    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
    "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
    document are to be interpreted as described in RFC2119 [RFC2119].

    When used in lower case, these words convey their typical use in
    common language, and are not to be interpreted as described in
    RFC2119 [RFC2119].












































Davari, et al.          Expires December 17, 2013               [Page 4]

Internet-Draft        Transporting Timing over MPLS            June 2013


1.  Introduction

    The objective of Precision Time Protocol (PTP) and Network Timing
    Protocol (NTP) are to synchronize independent clocks running on
    separate nodes of a distributed system.

    [IEEE-1588] defines PTP messages for frequency, phase and time
    synchronization.  The PTP messages include PTP PDUs over UDP/IP
    (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F of
    [IEEE-1588]).

SB> Sure BUT it is acknowledged that IEEE is open to the definition
SB> of other PTP mappings if they provide better optimisation.

    This document defines mapping and transport of the PTP
    messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
    defines several clock types: ordinary clocks, boundary clocks, end-
    to-end transparent clocks, and peer-to-peer transparent clocks.
    Transparent clocks require intermediate nodes to update correction
    field inside PTP message that reflects the transit time in the node.

    [RFC5905] defines NTP messages for clock and time synchronization.
    The PTP messages (PDUs) are transported over UDP/IP.  This document
SB> Should that be NTP messages?
SB> It needs to be made clear as soon as you introduce NTP that
SB> they use different time representations.

    defines mapping and transport of the NTP messages defined in
    [RFC5905] over MPLS networks.

    One key attribute of all of these Timing messages is that the Time
    stamp processing should occur as close as possible to the actual
    transmission and reception at the physical port interface.  This
    targets optimal time and/or frequency recovery by avoiding variable
    delay introduced by queues internal to the clocks.

SB> As I recall NTP has no epoch point defined, and I am not sure
SB> where that point is in the case of PTP in this mapping
SB> Hopefully this will get defined in due course.

    To facilitate the fast and efficient recognition of Timing messages
    at the port level when the Timing messages are carried over MPLS
    LSPs,

SB> Over a new LSP type with time optimied characteristics

    this document defines the specific encapsulations that should
    be used.
SB> Hopefully it will also define the PHP

    In addition, it can be expected that there will exist LSR/
    LERs where only a subset of the physical ports will have the port-
    based Timing message processing capabilities.
SB> Do you need to clarify that this only works at base and not in
SB> a label heirarchy.


    In order to ensure
    that the LSPs carrying Timing packets always enter and exit ports
    with this capability, routing extensions are defined to advertise
    this capability on a port basis and to allow for the establishment of
    LSPs that only transit such ports.  While this path establishment
    restriction may be applied only at the LER Ingress and/or egress
    ports, it becomes more important when using transparent clock capable
    LSRs in the path.
SB> I do not understand the implications of the last
SB> sentences - starting ", it becomes"


    Port based Timing message processing involves Timing message
    recognition.  Once the Timing messages are recognized they can be
    modified based on the reception or transmission Time-stamp.

    This document provides two methods for transporting Timing messages
    over MPLS.  One is applicable to MPLS environment and the other one
    is applicable to MPLS/MPLS-TP environment

SB> I think the sentence is incomplete.



Davari, et al.          Expires December 17, 2013               [Page 5]

Internet-Draft        Transporting Timing over MPLS            June 2013


    The solution involves transporting Timing messages over dedicated
    LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
    carry Management and control messages, but not data plane client
    traffic.

SB> It is not clear why this restriction applies.

    Timing LSPs can be established statically or via signaling.
SB> s/statically/by provisioning/network management/

    Extensions to control plane (OSPF, ISIS, etc.) is required to enable
    routers to distribute their Timing processing capabilities over MPLS
    to other routers.  However such extensions are outside the scope of
    this document.

    When signaling is used to setup the PTP LSP, Extensions to signaling
SB> is it a PTP LSP or a Timing LSP?

    protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
    However such extensions are outside the scope of this document.

SB> for mpls-tp GMPLS is the signalling protocol

    While the techniques included herein allow for the establishment of
    paths optimized to include Time-stamping capable links, the
    performance of the Slave clocks is outside the scope of this
    document.

    At the time of publishing this specification, Transparent Clocking
    (TC) is only defined for PTP.  Therefore at this time any part of
    this specification that talks about Transparent Clocking applies only
    to PTP.





























Davari, et al.          Expires December 17, 2013               [Page 6]

Internet-Draft        Transporting Timing over MPLS            June 2013


2.  Terminology

    1588: The timing and synchronization as defined by IEEE 1588.

SB> I think that there is a more formal name for the 1588 group
SB> that needs to be used here.
SB> Also do we need to talk about 1588-200? as there is
SB> an update in progress

    NTP: The timing and synchronization protocol defined by IETF RFC-1305
    and RFC-5905.

    PTP: The timing and synchronization protocol used by 1588.
SB> need the proper name for 1588

    Master Clock: The source of 1588 timing to a set of slave clocks.

    Master Port: A port on a ordinary or boundary clock that is in Master
    state.  This is the source of timing toward slave ports.

SB> I am not sure the reader knows what a port is

    Slave Clock: A receiver of 1588 timing from a master clock.

    Slave Port: A port on a boundary clock or ordinary clock that is
    receiving timing from a master clock.

    Ordinary Clock: A device with a single PTP port.

    Transparent Clock.  A device that measures the time taken for a PTP
    event message to transit the device and then updates the
    correctionField of the message with this transit time.

    Boundary Clock: A device with more than one PTP port.  Generally
    boundary clocks will have one port in slave state to receive timing
    and then other ports in master state to re-distribute the timing.

    PTP LSP: An LSP dedicated to carry PTP messages

SB> PTP or timing?


    PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
    messages.

SB> Ah I don't think that PWE3 know what one of these is

    CW: Pseudowire Control Word

    LAG: Link Aggregation

    ECMP: Equal Cost Multipath

    CF: Correction Field, a field inside certain PTP messages (message
    type 0-3)that holds the accumulative transit time inside intermediate
    switches

    Timing messages: Timing Protocol messages that are exchanged between
    routers in order to establish a synchronized clock.

SB> A number of these definitions look like copies of IEEE1588
SB> definitions. We need to provide references and note the
SB> priority of the IEEE base reference.



Davari, et al.          Expires December 17, 2013               [Page 7]

Internet-Draft        Transporting Timing over MPLS            June 2013


3.  Problem Statement

    [IEEE-1588] has defined methods for transporting PTP messages over
    Ethernet and IP networks.  [RFC5905] has defined the method of
    transporting NTP messages over IP networks.  There is a need to
    transport Timing messages over MPLS networks while supporting the
    Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
    functionality in the LER and LSRs in the MPLS network.

    There are multiple ways of transporting Timing over MPLS.  However,
    there is a requirement to limit the possible encapsulation options to
    simplify the Timing message identification and processing required at
    the port level.

    When Timing-awareness is needed, Timing messages should not be
    transported over LSPs or PWs that are carrying customer traffic
    because LSRs perform Label switching based on the top label in the
    stack.

SB> Have you explained why?

    To detect Timing messages inside such LSPs require special
    hardware to do deep packet inspection at line rate.  Even if such
    hardware exists, the payload can't be deterministically identified by
    LSRs because the payload type is a context of the PW label, and the
    PW label and its context are only known to the Edge routers (PEs/
    LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CES,
    etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
    LSRs dont have the knowledge of whether PW Control Word (CW) is
    present or not and therefore can not deterministically identify the
    payload.

    A generic method is defined in this document that does not require
    deep packet inspection at line rate, and can deterministically
    identify Timing messages.  This method can be used to detect Timing
    Messages in both one-step and two-step clock implementations of
    ordinary, boundary and transparent clocks.

SB> Needs a ref and I am sure many MPLS specialists will not understand
SB> the msg types.


















Davari, et al.          Expires December 17, 2013               [Page 8]

Internet-Draft        Transporting Timing over MPLS            June 2013


4.  Timing over MPLS Architecture

    Timing messages are exchange between Timing ports on ordinary and

SB> Have you defined a timing port?

    boundary clocks.  Boundary clocks terminate the Timing messages and
    act as master for other boundary clocks or for slave clocks.  End-to-
    End Transparent clocks do not terminate the Timing messages but they
    do modify the contents of the Timing messages as they transit across
    the transparent clock.

    Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Clock

    (TC) could be implemented in either LERs or LSRs.

SB> LER and LSR need to be expanded

    An example is shown in Figure 1, where the LERs act as Ordinary Clock
    (OC) and are the initiating/terminating point for Timing messages.
    The ingress LER encapsulates the Timing messages in Timing LSP and
    the Egress LER terminates the Timing LSP.  The LSRs act as
    Transparent Clock (TC) and just update the Timing field in the Timing
    messages.


       +--------+     +-------+     +-------+     +-------+     +--------+
       |Switch, |     |       |     |       |     |       |     |Switch, |
       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
       |        |     |  OC   |     |  TC   |     |  OC   |     |        |
       +--------+     +-------+     +-------+     +-------+     +--------+
                      /                                 \
       +-------+     /                                   \     +-------+
       |  LER  |    /                                     \    |  LER  |
       | Master|---/                                       \---| Slave |
       | Clock |                                               | Clock |
       +-------+                                               +-------+

      Figure (1) - Deployment example 1 of timing over MPLS network

    Another example is shown in Figure2, where LERs terminate the Timing
    messages received from switch/routers that are outside of the MPLS
    network acting as OC or BC.  In this example LERs regenerate the
    clock and initiate timing messages encapsulated in Timing LSP toward
    the MPLS network, while the LSRs act as Transparent Clock (TC) and
    just update the Timing field in the Timing messages, which are
    already encapsulated in Timing LSPs.









Davari, et al.          Expires December 17, 2013               [Page 9]

Internet-Draft        Transporting Timing over MPLS            June 2013


      +--------+     +-------+     +-------+     +-------+     +--------+
      |Switch, |     |       |     |       |     |       |     |Switch, |
      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
      | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC  |
      +--------+     +-------+     +-------+     +-------+     +--------+

      Figure (2) - Deployment example 2 of timing over MPLS network


    Another example is shown in Figure 3, where LERs do not terminate the
    Timing messages received from switch/routers that are outside of the
    MPLS network acting as OC, TC or BC.  The LERs act as TC and update
    the Timing field in the Timing messages as they transit the LER,
    while encapsulating them in timing LSP.  The LSRs also act as
    Transparent Clock (TC) and just update the Timing field in the Timing
    messages which are already encapsulated in Timing LSPs.

       +--------+     +-------+     +-------+     +-------+     +--------+
       |Switch, |     |       |     |       |     |       |     |Switch, |
       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
       |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC/BC|
       +--------+     +-------+     +-------+     +-------+     +--------+

     Figure (3) - Deployment example 3 of timing over MPLS network

    Another example is shown in Figure 4, where LERs and LSRs support
    Boundary Clocks.  A single-hop LSP is created between two adjacent
    LSRs engaged in BC operation.  Other methods such as PTP transport
    over Ethernet MAY be used for transporting timing messages if the
    link between the two routers is Ethernet.

      +--------+     +-------+     +-------+     +-------+     +--------+
      |Switch, |     |       |     |       |     |       |     |Switch, |
      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
      | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC  |
      +--------+     +-------+     +-------+     +-------+     +--------+

    Figure (4) - Deployment example 3 of timing over MPLS network

    An MPLS domain MAY serve multiple customers.  In these cases the MPLS
    domain (maintained by a service provider) may provide timing services
    to multiple customers, each having their own Timing domain.

    The Timing over MPLS architecture assumes full mesh of Timing LSPs
    between all LERs supporting this specification.

SB> Note sure this is right - the salves surely do not need to
SB> exchange timing amongst themselves

    It supports
    Point-to- point (VPWS) and Multipoint (VPLS) services.

SB> What does that mean? You do not carry user data traffic?
SB> Maybe it's the ordering of the statemnets that is causing
SB> confusion.

    This means
    that a customer may purchase a Point-to-point Timing service between
    two customer sites or a Multipoint Timing service between more than



Davari, et al.          Expires December 17, 2013              [Page 10]

Internet-Draft        Transporting Timing over MPLS            June 2013


    two customer sites.

    The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
    This means that the Timing Multicast messages such as PTP Multicast
    event messages can be transported over P2MP Timing LSP or be
    replicated and transported over many P2P Timing LSPs.

SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
SB> nor a P2MP PW, although we are close.

    Timing messages, that do not require Time stamping or Correction
    Field update MAY be transported over Timing LSPs to simplify hardware
    and software.

    PTP Announce messages that determine the Timing LSP terminating point
    behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
    to simplify hardware and software.

SB> have you defined and referenced PTP announce msgs?





































Davari, et al.          Expires December 17, 2013              [Page 11]

Internet-Draft        Transporting Timing over MPLS            June 2013


5.  Dedicated LSPs for Timing messages

    Many methods have been considered for identifying the Timing messages
    when they are encapsulated in MPLS such as using GAL/G-ACH or a new
    reserved label.  These methods were not attractive since they either
    required deep packet inspection at line rate in the intermediate LSRs
    or they required use of a scarce new reserved label.  Also one of the
    goals was to reuse existing OAM mechanisms.

SB> RLs = SPLs are not so rare now. In any case needs a ref.

    The method defined in this document can be used by LER and LSRs to
    identify Timing messages in MPLS tunnels by just looking at the top
    label in the MPLS label stack, which only carry Timing messages as
    well as OAM, but not data plane client traffic.

    Compliant implementations MUST use dedicated LSPs to carry Timing
    messages over MPLS.

SB> I think that we need a definition of the properies of these LSPs

    These LSPs are herein referred to as "Timing
    LSPs" and the labels associated with these LSPs as "Timing LSP
    labels".  The Timing LSPs that runs between Ingress and Egress LERs
    MUST be co-routed.  Alternatively, a single bidirectional co-routed
    LSP can be used.

SB> I though that you said you could use M2MP LSPs - these are not
SB> bidirectional.

    Co-routing of the two directions is required to limit the difference
    in the delays in the Master clock to Slave clock direction compared
    to the Slave clock to Master clock direction.  The Timing LSP MAY be
    MPLS/MPLS-TP LSP.

    The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
    New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
    outside the scope of this document.

    The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such as
    BFD and LSP Ping but the LSP data plane client plane traffic MUST be
    Timing packets only.

SB> Why?


















Davari, et al.          Expires December 17, 2013              [Page 12]

Internet-Draft        Transporting Timing over MPLS            June 2013


6.  Timing over LSP Encapsulation

The encapsulations is not LSP is it?

    This document defines two methods for carrying Timing messages over
    MPLS.  The first method is carrying UDP/IP encapsulated Timing
    messages over Timing LSPs, and the second method, is carrying
    Ethernet encapsulated Timing messages over Ethernet PWs inside Timing
    LSPs.

6.1.  Timing over UDP/IP over MPLS Encapsulation

    The simplest method of transporting Timing messages over MPLS is to
    encapsulate Timing PDUs in UDP/IP and then encapsulate them in Timing
    LSP.  This format is shown in Figure 4.


                     +----------------------+
                     |   Timing LSP Label   |
                     +----------------------+
                     |        IPv4/6        |
                     +----------------------+
                     |         UDP          |
                     +----------------------+
                     |     Timing PDU       |
                     +----------------------+

       Figure (4) - Timing over UDP/IP over MPLS Encapsulation


    This encapsulation is very simple and is useful when the network
    between Timing Master Clock and Slave Clock is MPLS network.

SB> Simple is a judgement call

    In order for an LER/LSR to process Timing messages, the Timing LSP
    Label must be at the top label of the label stack.  The LER/LSR MUST
    know that the Timing LSP Label is used for carrying Timing messages.
    This can be accomplished via static configuration or via RSVP-TE
    signaling.

    The UDP/IP encapsulation of PTP MUST follow Annex D and E of
    [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
    [RFC5905].

6.2.  Timing over PW Encapsulation

    Another method of transporting Timing over MPLS networks is by
    encapsulating Timing PDUs in PW which in turn is transported over
    Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
    shown in Fig 5(A) MUST be used and the Ethernet encapsulation of PTP
    MUST follow Annex F of [IEEE-1588].



Davari, et al.          Expires December 17, 2013              [Page 13]

Internet-Draft        Transporting Timing over MPLS            June 2013


    The RAW mode or Tagged mode defined in [RFC4448] MAY be used and the
    Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
    Timing over PW encapsulation MUST use the Control Word (CW) as
    specified in [RFC4448] to ensure proper detection of PTP messages
    inside the MPLS packets for Timing over LSP and Timing over PW
    encapsulation.

SB> That needs explanation

    The use of Sequence Number in the CW is optional.

SB> Given that s/n are never in practice deployed, you could probably
SB> simplify things by sayig that they are not used.

    Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over PW
    (the IP PW discussed in [RFC4447]) shown in Fig 5(B).

                     +----------------+  +----------------+
                      |Timing LSP Label|  |Timing LSP Label|
                      +----------------+  +----------------+
                      |    PW Label    |  |    PW Label    |
                      +----------------+  +----------------+
                      |  Control Word  |  |      IP        |
                      +----------------+  +----------------+
                      |    Ethernet    |  |      UDP       |
                      |     Header     |  +----------------+
                      +----------------+  |   Timing PDU   |
                      |S-VLAN(Optional)|  |                |
                      +----------------+  +----------------+
                      |C-VLAN(Optional)|        (B)
                      +----------------+
                      |   Timing PDU   |
                      |                |
                      +----------------+
                             (A)

               Figure (5) - Timing over PW Encapsulations

    In order for an LSR to process PTP messages, the top label of the
    label stack (the Tunnel Label) MUST be a Timing label.

S> You said that before.

6.3.  Other Timing Encapsulation methods

    In future other timing encapsulation methods may be introduced, such
    as a new shim header after the Bottom of Stack to carry the Timing
    information.  Such new encapsulations are outside the scope of this
    document.


SB> Taking a pure MPLS pov, you can simplify a lot of the text
SB> out of the definition of the LSP








Davari, et al.          Expires December 17, 2013              [Page 14]

Internet-Draft        Transporting Timing over MPLS            June 2013


SB> I think we need a section on LSP processing

7.  Timing message Processing

    Each Timing protocol such as PTP and NTP, define their set of Timing
    messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
    FOLLOW_UP, etc messages.

    Some of the Timing messages require time stamping or correction field
    update at port level and some dont.  It is the job of the LER/LSR to
    parse the timing message and find out the type of the Timing message
    and decide whether and how to Time- stamp it (e.g., BC) or update
    correction field(e.g., TC).


SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
SB> function rather than the LER?

    For example the following PTP messages (called Event messages)
    require time-stamping or correction field update:

    o  SYNC

    o  DELAY_REQ (Delay Request)

    o  PDELAY_REQ (Peer Delay Request)

    o  PDELAY_RESP (Peer Delay Response)

    SYNC and DELAY_REQ are exchanged between Master Clock and Slave Clock
    and MUST be transported over PTP LSPs.  PDELAY_REQ and PDELAY_RESP
    are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
    Boundary, or Transparent) and SHOULD be transported over single hop
    PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
    and PDELAY_RESP_FOLLOW_UP messages MUST also be transported over the
    PTP LSPs.

    For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
    transported over two PTP LSPs that are in opposite directions.  These
    PTP LSPs, which are in opposite directions MUST be congruent and co-
    routed.  Alternatively, a single bidirectional co-routed LSP can be
    used.

    Except as indicated above for the two-step PTP clocks, Non-Event PTP
    message types do not need to be processed by intermediate routers.
    These message types MAY be carried in PTP Tunnel LSPs.

SB> Are you saying that a timing P router has to be msg type sensitive?









Davari, et al.          Expires December 17, 2013              [Page 15]

Internet-Draft        Transporting Timing over MPLS            June 2013


8.  Protection and Redundancy


SB> This is a bit of a jump - I don't know how the LSP itself works yet!

    In order to ensure continuous uninterrupted operation of slave
    clocks, usually as a general practice, slave clocks (or ports) track
    redundant master clocks.

    It is the responsibility of the network operator to ensure that
    physically disjoint Timing LSPs are established between a slave clock
    (or port) and redundant master clocks (or ports).

    When a slave clock (or port) listens to redundant master clocks or
    ports, any prolonged Timing LSP outage will trigger the slave clock
    or port to switch to a redundant master clock or port.

    LSP/PW protection such as Linear protection Switching (1:1, 1+1),
    Ring protection switching or MPLS Fast Reroute (FRR) generally switch
    alternative path that usually cause a change in delay, which if
    undetected by slave clock can reduce accuracy of the slave clock.

    Therefore protection switching MAY be used, as long as phase jumps
    upon switchover due to differences in path latency are detected and
    compensated for (such compensation not being required if BCs or peer-
    peer TCs are used throughout).

    Note that any protection or reroute mechanism that adds additional
    MPLS label to the label stack, such as Facility Backup Fast Reroute,
    MUST ensure that the pushed label is also a Timing Label to ensure
    recognition of the MPLS frame as containing Timing messages, as it
    transits the backup path.






















Davari, et al.          Expires December 17, 2013              [Page 16]

Internet-Draft        Transporting Timing over MPLS            June 2013


9.  ECMP

    To ensure the optimal operation of slave clocks and avoid error
    introduced by forward and reverse path delay asymmetry, the physical
    path for Timing messages from master clock to slave Clock and vice
    versa must be the same for all Event Timing messages listed in
    section 7.

    Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
    Multipath).









































Davari, et al.          Expires December 17, 2013              [Page 17]

Internet-Draft        Transporting Timing over MPLS            June 2013


10.  PHP

    To ensure that the label on the top of the label stack is the Timing
    LSP Label, PHP MUST not be used.















































Davari, et al.          Expires December 17, 2013              [Page 18]

Internet-Draft        Transporting Timing over MPLS            June 2013


11.  Entropy

    To ensure all Timing messages in a Timing LSP take the same path,
    Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
    Entropy Label MUST NOT be used for the PWs that are carried inside
    Timing LSP [RFC6391].

SB> This is incorrect - you mean that all msgs of the same timing
SB> flow need to have the same EL value.













































Davari, et al.          Expires December 17, 2013              [Page 19]

Internet-Draft        Transporting Timing over MPLS            June 2013


12.  OAM, Control and Management

    In order to monitor Timing LSPs and their encapsulated PWs, they MUST
    be able to carry OAM and management messages.  These management
    messages MUST be differentiated from Timing messages via already
    defined IETF methods.

    For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
    over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
    Management protocols can easily be identified by the UDP Destination
    Port number or by GAL/G-ACH respectively.

    Also BFD, LSP-Ping and other management messages MAY run over the PWs
    encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3 or
    4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is going
    to be deprecated by IETF).  In this case G-ACH, PW label (TTL=1) or
    GAL-ACH are used to identify such management messages.


































Davari, et al.          Expires December 17, 2013              [Page 20]

Internet-Draft        Transporting Timing over MPLS            June 2013


13.  QoS Considerations

    In network deployments where not every LSR/LER is Timing-aware, it is
    important to reduce the impact of the non-Timing-aware LSR/LERs on
    the timing recovery in the slave clock.  The Timing messages are time
    critical and must be treated with the highest priority.  Therefore
    Timing over MPLS messages must be treated with the highest priority
    in the routers.  This can be achieved by proper setup of Timing LSPs.

    It is recommended that the Timing LSPs are setup or configured
    properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC2697]
    for drop eligibility.







































Davari, et al.          Expires December 17, 2013              [Page 21]

Internet-Draft        Transporting Timing over MPLS            June 2013


14.  FCS and Checksum Recalculation

    When time-stamp generation and timing packet adjustment is performed
    near the physical port hardware, the process MUST include
    recalculation of the Ethernet FCS.

SB> The above is confusing - an LSR always recomputes the link layer
SB> CRC which may or may not be Ethernet.

    Also FCS retention for the
    payload Ethernet described in [RFC4720] MUST NOT be used.

    For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
    may be required as per UDP transport standards.

SB> You really need to be working on getting the IPv6 C?S computation
SB> removed from PTP msgs.

    When UDP checksum is used, each Timing-aware LER/LSR must either
    incrementally update the UDP checksum after Time stamping or
    Correction Field update or verify the UDP checksum on reception from
    upstream and recalculate the checksum completely on transmission to
    downstream node after Time stamping or Correction Field update.




































Davari, et al.          Expires December 17, 2013              [Page 22]

Internet-Draft        Transporting Timing over MPLS            June 2013


15.  Behavior of LER/LSR

    Timing-capable/aware LERs and LSRs are routers that have one or more

SB> You mean physical interfaces?

    interfaces that can perform Timing operations (OC/BC/TC) on Timing
    packets and are configured to do so.  Timing-capable/aware LERs and
    LSRs can advertise their Timing-capability per-interface via control
    plane such as OSPF or IS-IS.
SB> ISIS and OSPF are routing protocols.

   The Timing-capable/aware LERs can then
    signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timing
    capability of LER and LSRs may be configured in a centralized
    controller and the Timing LSP may be setup using manual configuration
    or other methods such as SDN.

SB> it can also be configured individually rather then through
SB> a cebtral controllwe

15.1.  Behavior of Timing-capable/aware LER

    When a Timing-capable/aware LER behaves as a Transparent clock and
    receives a Timing message from a Timing-capable/aware non-MPLS
    interface, the LER updates the Correction Field (CF) and encapsulates
    and forwards the timing message over previously established Timing
    LSP.

SB> You need to call out the details so that people properly
SB> understand the definition of the new LSP.

    Also when a Timing message is received from a Timing-capable/
    aware MPLS interface, LER updates the Correction Filed (CF) and
    decapsulates the MPLS encapsulation and forwards the timing message
    to a non-MPLS interface.

    When a Timing-capable/aware LER behaves as a Boundary clock and
    receives a Timing message from a Timing-capable/aware non MPLS
    interface, the LER Timestamps the Timing packet and sends it to the
    LERs Boundary clock processing module.  Also when a Timing message is
    received from a Timing- capable/aware MPLS interface, the LER
    Timestamps the Timing packet and sends it to the LERs Boundary clock
    processing module.

    When a Timing-capable/aware LER behaves as an Ordinary Clock toward
    the MPLS network, and receives a Timing message from a Timing-
    capable/aware MPLS interface, the LER Timestamps the Timing packet
    and sends it to the LERs Ordinary clock processing module.

15.2.  Behavior of Timing-capable/aware LSR

    When a Timing-capable/aware LSR behaves as a Transparent clock and
    receives a Timing message from a Timing-capable/aware MPLS interface,
    The LSR updates the Correction Filed (CF) and forwards the timing
    message over another MPLS interface.

    When a Timing-capable/aware LSR behaves as a Boundary clock and
    receives a Timing message from a Timing-capable/aware MPLS interface.
    The LSR performs the functions of a Boundary Clock in terminating the
    received Timing message and re-generating a new timing message over
    another (or the same) MPLS interface.



Davari, et al.          Expires December 17, 2013              [Page 23]

Internet-Draft        Transporting Timing over MPLS            June 2013


15.3.  Behavior of non-Timing-capable/aware LSR

    It is most beneficial when all LSRs in the path of a Timing LSP be
    timing-Capable/aware LSRs.  This would ensure the highest quality
    time and clock synchronization by Timing Slave Clocks.  However, this
    specification does not mandate that all LSRs in path of a Timing LSP
    be Timing- capable/aware.

    Non-Timing-capable/aware LSRs just switch the packets encapsulated in
    Timing LSPs and dont perform any Timing operation (TC or BC).
    However as explained in QoS section the Timing over MPLS packets MUST
    be still be treated with the highest priority based on their Traffic
    Class (TC) marking.






































Davari, et al.          Expires December 17, 2013              [Page 24]

Internet-Draft        Transporting Timing over MPLS            June 2013


16.  Other considerations

    [IEEE-1588] defines an optional peer-to-peer Transparent clocking
    that requires peer delay measurement between two adjacent Timing-
    capable/ aware routers/switches.  Peer delay measurement messages
    need to be time stamped and terminated by the Timing-capable/aware
    routers/ switches.  This means that two adjacent LSRs may be engaged
    in a peer delay measurement.

    For transporting such peer delay measurement messages a single-hop
    LSP SHOULD to be created between the two adjacent LSRs engaged in
    peer delay measurement to carry peer delay measurement messages.
    Other methods such as PTP transport over Ethernet MAY be used for
    transporting peer delay measurement messages if the link between the
    two routers is Ethernet.

    In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ ware
    routers/switches MUST maintain a list of all the neighbors it needs
    to send a PDelay_Req to, where each neighbor corresponds to a timing
    LSP.

    The use of Explicit Null Label (Label= 0 or 2) is acceptable as long
    as either the Explicit Null label is the bottom of stack label
    (applicable only to UDP/IP encapsulation) or the label below the
    Explicit Null label is a PTP label.


























Davari, et al.          Expires December 17, 2013              [Page 25]

Internet-Draft        Transporting Timing over MPLS            June 2013


17.  Security Considerations

    MPLS PW security considerations in general are discussed in [RFC3985]
    and [RFC4447],and those considerations also apply to this document.

    An experimental security protocol is defined in [IEEE-1588].The PTP
    security extension and protocol provides group source authentication,
    message integrity, and replay attack protection for PTP messages.

    When the MPLS network (provider network) serves multiple customers,
    it is important to maintain and process each customers clock and
    Timing messages separately from other customers to ensure there is no
    cross- customer effect.  For example if an LER BC is synchronized to
    a specific grandmaster, belonging to customer A, then the LER MUST
    use that BC clock only for customer A to ensure that customer A
    cannot attack other customers by manipulating its time.

    Timing messages MAY be encrypted or authenticated, provided that the
    LERs/LSRs that are Timing capable/aware can authenticate/ decrypt the
    timing messages.































Davari, et al.          Expires December 17, 2013              [Page 26]

Internet-Draft        Transporting Timing over MPLS            June 2013


18.  Acknowledgements

    The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizrahi,
    Stefano Ruffini, Peter Meyer, and other members of IETF for reviewing
    and providing feedback on this draft.














































Davari, et al.          Expires December 17, 2013              [Page 27]

Internet-Draft        Transporting Timing over MPLS            June 2013


19.  IANA Considerations

    There are no IANA requirements in this specification.



















































Davari, et al.          Expires December 17, 2013              [Page 28]

Internet-Draft        Transporting Timing over MPLS            June 2013


20.  References

20.1.  Normative References

    [IEEE-1588]
               IEEE 1588-2008, "IEEE Standard for a Precision Clock
               Synchronization Protocol for Networked Measurement and
               Control Systems".

    [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
               Requirement Levels", BCP 14, RFC 2119, March 1997.

    [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
               Edge (PWE3) Architecture", RFC 3985, March 2005.

    [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
               Proxies (ND Proxy)", RFC 4389, April 2006.

    [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
               Heron, "Pseudowire Setup and Maintenance Using the Label
               Distribution Protocol (LDP)", RFC 4447, April 2006.

    [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
               "Encapsulation Methods for Transport of Ethernet over MPLS
               Networks", RFC 4448, April 2006.

    [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
               Emulation Edge-to-Edge (PWE3) Frame Check Sequence
               Retention", RFC 4720, November 2006.

    [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
               Connectivity Verification (VCCV): A Control Channel for
               Pseudowires", RFC 5085, December 2007.

    [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detection
               (BFD)", RFC 5880, June 2010.

    [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
               "Bidirectional Forwarding Detection (BFD) for MPLS Label
               Switched Paths (LSPs)", RFC 5884, June 2010.

20.2.  Informative References

    [I-D.ietf-pwe3-fat-pw]
               Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
               J., and S. Amante, "Flow Aware Transport of Pseudowires
               over an MPLS Packet Switched Network",
               draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.



Davari, et al.          Expires December 17, 2013              [Page 29]

Internet-Draft        Transporting Timing over MPLS            June 2013


    [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
               system routeing information exchange protocol for use in
               conjunction with the Protocol for providing the
               Connectionless-mode Network Service (ISO 8473)".

    [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
               dual environments", RFC 1195, December 1990.

    [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.

    [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
               Marker", RFC 2697, September 1999.

    [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boudec,
               J., Courtney, W., Davari, S., Firoiu, V., and D.
               Stiliadis, "An Expedited Forwarding PHB (Per-Hop
               Behavior)", RFC 3246, March 2002.

    [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineering
               (TE) Extensions to OSPF Version 2", RFC 3630,
               September 2003.

    [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
               System (IS-IS) Extensions for Traffic Engineering (TE)",
               RFC 3784, June 2004.

    [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
               Shaffer, "Extensions to OSPF for Advertising Optional
               Router Capabilities", RFC 4970, July 2007.

    [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
               System to Intermediate System (IS-IS) Extensions for
               Advertising Router Information", RFC 4971, July 2007.

    [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
               Topology (MT) Routing in Intermediate System to
               Intermediate Systems (IS-ISs)", RFC 5120, February 2008.

    [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
               Engineering", RFC 5305, October 2008.

    [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
               "Traffic Engineering Extensions to OSPF Version 3",
               RFC 5329, September 2008.

    [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
               for IPv6", RFC 5340, July 2008.




Davari, et al.          Expires December 17, 2013              [Page 30]

Internet-Draft        Transporting Timing over MPLS            June 2013


    [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network
               Time Protocol Version 4: Protocol and Algorithms
               Specification", RFC 5905, June 2010.

    [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
               J., and S. Amante, "Flow-Aware Transport of Pseudowires
               over an MPLS Packet Switched Network", RFC 6391,
               November 2011.

    [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
               L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
               RFC 6790, November 2012.







































Davari, et al.          Expires December 17, 2013              [Page 31]

Internet-Draft        Transporting Timing over MPLS            June 2013


1.  Routing extensions for Timing-aware Routers

    MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] and
    IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (TE)
    link information used for constraint-based routing.

    Indeed, it is useful to advertise data plane TE router link
    capabilities, such as the capability for a router to be Timing-aware.
    This capability MUST then be taken into account during path
    computation to prefer or even require links that advertise themselves
    as Timing-aware.  In this way the path can ensure the entry and exit
    points into the LERs and, if desired, the links into the LSRs are
    able to perform port based time-stamping thus minimizing their impact
    on the performance of the slave clock.

    extensions are required to OSPF and IS-IS in order to advertise
    Timing-aware capabilities of a link.  Such extensions are outside the
    scope of this document; however such extension SHOULD be able to
    signal the following information per Router Link:

    o  Capable of processing PTP, NTP or other Timing flows

    o  Capable of performing Transparent Clock operation

    o  Capable of performing Boundary Clock operation


























Davari, et al.          Expires December 17, 2013              [Page 32]

Internet-Draft        Transporting Timing over MPLS            June 2013


2.  Signaling Extensions for Creating Timing LSPs

    RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-TE
    is used to setup Timing LSPs, some information that indicates that
    the LSP is carrying Timing flows MUST be included in the new
    Extensions to RSVP-TE:

    The following information MAY also be included in the new Extensions
    to RSVP-TE:

    o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
       field

    o  Number of VLANs in case of PW encapsulation

    o  Timestamp field Type

       *  Correction Field, Timestamp

    o  Timestamp Field format

       *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
          NTP, etc.

    Note that in case the above optional information is signaled with
    RSVP-TE for a Timing LSP, all the Timing packets carried in that LSP
    must have the same signaled characteristics.  For example if
    Timestamp format is signaled as 64-bit PTPv1, then all Timing packets
    must use 64-bit PTPv1 time-stamp.






















Davari, et al.          Expires December 17, 2013              [Page 33]

Internet-Draft        Transporting Timing over MPLS            June 2013


Authors' Addresses

    Shahram Davari
    Broadcom Corp.
    San Jose, CA  95134
    USA

    Email: davari@broadcom.com


    Amit Oren
    Broadcom Corp.
    San Jose, CA  95134
    USA

    Email: amito@broadcom.com


    Manav Bhatia
    Alcatel-Lucent
    Bangalore,
    India

    Email: manav.bhatia@alcatel-lucent.com


    Peter Roberts
    Alcatel-Lucent
    Kanata,
    Canada

    Email: peter.roberts@alcatel-lucent.com


    Laurent Montini
    Cisco Systems
    San Jose CA
    USA

    Email: lmontini@cisco.com











Davari, et al.          Expires December 17, 2013              [Page 34]

Internet-Draft        Transporting Timing over MPLS            June 2013


    Luca
    Cisco Systems
    San Jose CA
    USA

    Email: lmartini@cisco.com













































Davari, et al.          Expires December 17, 2013              [Page 35]


-- 
For corporate legal information go to:

http://www.cisco.com/web/about/doing_business/legal/cri/index.html


From davari@broadcom.com  Fri Aug  2 03:24:02 2013
Return-Path: <davari@broadcom.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9470D21E82D3; Fri,  2 Aug 2013 03:24:02 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.599
X-Spam-Level: 
X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id d7GBZieuU69G; Fri,  2 Aug 2013 03:23:58 -0700 (PDT)
Received: from mms3.broadcom.com (mms3.broadcom.com [216.31.210.19]) by ietfa.amsl.com (Postfix) with ESMTP id DF17121F8EB2; Fri,  2 Aug 2013 03:23:57 -0700 (PDT)
Received: from [10.9.208.53] by mms3.broadcom.com with ESMTP (Broadcom SMTP Relay (Email Firewall v6.5)); Fri, 02 Aug 2013 03:13:42 -0700
X-Server-Uuid: B86B6450-0931-4310-942E-F00ED04CA7AF
Received: from SJEXCHCAS04.corp.ad.broadcom.com (10.16.203.10) by IRVEXCHCAS06.corp.ad.broadcom.com (10.9.208.53) with Microsoft SMTP Server (TLS) id 14.1.438.0; Fri, 2 Aug 2013 03:23:33 -0700
Received: from SJEXCHMB12.corp.ad.broadcom.com ( [fe80::bc15:c1e1:c29a:36f7]) by SJEXCHCAS04.corp.ad.broadcom.com ( [::1]) with mapi id 14.01.0438.000; Fri, 2 Aug 2013 03:23:33 -0700
From: "Shahram Davari" <davari@broadcom.com>
To: "<stbryant@cisco.com>" <stbryant@cisco.com>
Thread-Topic: [mpls] Comments on draft-ietf-tictoc-1588overmpls
Thread-Index: AQHOj2dsoD26UTw5FECSgmrxo5m7wJmBtd9E
Date: Fri, 2 Aug 2013 10:23:33 +0000
Message-ID: <D2796BF4-4DDB-4816-B730-1289CAB11080@broadcom.com>
References: <51FB8396.9020504@cisco.com>
In-Reply-To: <51FB8396.9020504@cisco.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
MIME-Version: 1.0
X-WSS-ID: 7DE559DC2L869999094-01-01
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: quoted-printable
Cc: "tictoc-chairs@tools.ietf.org" <tictoc-chairs@tools.ietf.org>, "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>, "draft-ietf-tictoc-1588overmpls@tools.ietf.org" <draft-ietf-tictoc-1588overmpls@tools.ietf.org>, "rtg-ads@tools.ietf.org" <rtg-ads@tools.ietf.org>, "int-ads@tools.ietf.org" <int-ads@tools.ietf.org>, "mpls-chairs@tools.ietf.org" <mpls-chairs@tools.ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 02 Aug 2013 10:24:02 -0000

Hi Stewart

We have already looked at similar approach. The issue with one hop LSP is t=
hat not all LSRs are 1588 capable. One of the requirements is for non-1588 =
capable routers to just switch the packet normally.

Regards,
Shahram


On Aug 2, 2013, at 12:02 PM, "Stewart Bryant" <stbryant@cisco.com> wrote:

> Talking to John Drake about this, an alternative general
> model is for the LSP is that it is an LSP that timestamps
> the packet and then passes it to an application associated
> with the LSP at that hop.
>=20
> This has a lot of merit.
>=20
> If however we think about it some more we have a
> type of network service chaining going on here and
> so we don't need an LSP per say, because the path
> and instruction can be in the packet.
>=20
> In other words a general solution  to the problem
> is to define an LSP with the properties that the
> packet is passed one hop, timestamped and delivered
> to the associated application.
>=20
> How this is MPLS construct is used to support time
> tranfer is entirely within the scope of TICTOC, but
> at the MPLS layer we have a clean reusable network
> service.
>=20
> - Stewart
>=20
>=20
> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> SB> This draft does not seem to provide a precise definition
> SB> the properties of the new LSP type that it wishes to
> SB> define, in particular it does it define the PHB of those
> SB> LSPs, nor the full interaction with the MPLS
> SB> architecture.
> SB>
> SB> I have not tracked TICTOC for a while but I thought that
> SB> the original plan was to define the concept of an offset
> SB> into a packet to do the correction.
> SB>
> SB> It is disappointing that the opportunity was not taken
> SB> to define a timing shim inside the timing LSP so that
> SB> a time correction could be added to any packet such that
> SB> the MPLS system was isolated from the details of the
> SB> complexity of the particular time transfer type.
>=20
> SB> I think that much more clarify is needed in terms of
> SB> definition of the new LSP type, since it is unclear
> SB> from this text how to implement one.
> SB>
> SB> There are a lot of other MPLS services such as
> SB> LSP ping that need to be considered.
> SB>
> SB> Please see inline for more comments. However these
> SB> comments are made in the context of the text as written
> SB> whilst I have a fundamental concern that this approach
> SB> lacks an MPLS architectural soundness that need
> SB> greater thought with significant impact on the
> SB> draft.
>=20
> - Stewart
>=20
>=20
> TICTOC Working Group                                           S. Davari
> Internet-Draft                                                   A. Oren
> Intended status: Standards Track                          Broadcom Corp.
> Expires: December 17, 2013                                     M. Bhatia
>                                                              P. Roberts
>                                                          Alcatel-Lucent
>                                                              L. Montini
>                                                              L. Martini
>                                                           Cisco Systems
>                                                           June 15, 2013
>=20
>=20
>            Transporting Timing messages over MPLS Networks
>                   draft-ietf-tictoc-1588overmpls-05
>=20
> Abstract
>=20
>   This document defines the method for transporting Timing messages
>   such as PTP and NTP over an MPLS network.  The method allows for the
>   easy identification of these PDUs at the port level to allow for port
>=20
> SB> What is a port
>=20
>   level processing of these PDUs in both LERs and LSRs.
>=20
>   The basic idea is to transport Timing messages inside dedicated MPLS
>   LSPs.  These LSPs only carry Timing messages and possibly Control and
>   Management packets, but they do not carry customer traffic.
>=20
> SB> More specifically they only carry traffic associated with the
> SB> timing service and its support.
> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
> SB> also it gets carried in a structure that causes it to get
> SB> timestamped.
>=20
>   Two methods for transporting Timing messages over MPLS are defined.
>=20
> SB> Perhaps the right approach is to define the new LSP type and then
> SB> seperately to define  the mapping of the various timing services
> SB> over that LSP type.
>=20
>   The first method is to transport Timing messages directly over the
>   dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
>   MPLS networks.  The second method is to transport Timing messages
>   inside a PW via Ethernet encapsulation.
>=20
> SB> I think that we should note that there are some
> SB> h/w reasons for this preference. A clean sheet approach
> SB> would have been to use PTP over MPLS with no intermediate
> SB> layers.
>=20
> Status of this Memo
>=20
>   This Internet-Draft is submitted in full conformance with the
>   provisions of BCP 78 and BCP 79.
>=20
>   Internet-Drafts are working documents of the Internet Engineering
>   Task Force (IETF).  Note that other groups may also distribute
>   working documents as Internet-Drafts.  The list of current Internet-
>   Drafts is at http://datatracker.ietf.org/drafts/current/.
>=20
>   Internet-Drafts are draft documents valid for a maximum of six months
>   and may be updated, replaced, or obsoleted by other documents at any
>   time.  It is inappropriate to use Internet-Drafts as reference
>   material or to cite them other than as "work in progress."
>=20
>   This Internet-Draft will expire on December 17, 2013.
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013               [Page 1]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> Copyright Notice
>=20
>   Copyright (c) 2013 IETF Trust and the persons identified as the
>   document authors.  All rights reserved.
>=20
>   This document is subject to BCP 78 and the IETF Trust's Legal
>   Provisions Relating to IETF Documents
>   (http://trustee.ietf.org/license-info) in effect on the date of
>   publication of this document.  Please review these documents
>   carefully, as they describe your rights and restrictions with respect
>   to this document.  Code Components extracted from this document must
>   include Simplified BSD License text as described in Section 4.e of
>   the Trust Legal Provisions and are provided without warranty as
>   described in the Simplified BSD License.
>=20
>=20
> Table of Contents
>=20
>   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  5
>=20
>   2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .  7
>=20
>   3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .  8
>=20
>   4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .  9
>=20
>   5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . . 12
>=20
>   6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . . 13
>     6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . . 13
>     6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . . 13
>     6.3.  Other Timing Encapsulation methods . . . . . . . . . . . . 14
>=20
>   7.  Timing message Processing  . . . . . . . . . . . . . . . . . . 15
>=20
>   8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . . 16
>=20
>   9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
>=20
>   10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
>=20
>   11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
>=20
>   12. OAM, Control and Management  . . . . . . . . . . . . . . . . . 20
>=20
>   13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . . 21
>=20
>   14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . . 22
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013               [Page 2]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
>   15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . . 23
>     15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . . 23
>     15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . . 23
>     15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . . 24
>=20
>   16. Other considerations . . . . . . . . . . . . . . . . . . . . . 25
>=20
>   17. Security Considerations  . . . . . . . . . . . . . . . . . . . 26
>=20
>   18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 27
>=20
>   19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 28
>=20
>   20. References . . . . . . . . . . . . . . . . . . . . . . . . . . 29
>     20.1. Normative References . . . . . . . . . . . . . . . . . . . 29
>     20.2. Informative References . . . . . . . . . . . . . . . . . . 29
>=20
>   Appendix 1.  Routing extensions for Timing-aware Routers . . . . . 32
>=20
>   Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . . 33
>=20
>   Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 34
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013               [Page 3]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
>   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
>   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
>   document are to be interpreted as described in RFC2119 [RFC2119].
>=20
>   When used in lower case, these words convey their typical use in
>   common language, and are not to be interpreted as described in
>   RFC2119 [RFC2119].
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013               [Page 4]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 1.  Introduction
>=20
>   The objective of Precision Time Protocol (PTP) and Network Timing
>   Protocol (NTP) are to synchronize independent clocks running on
>   separate nodes of a distributed system.
>=20
>   [IEEE-1588] defines PTP messages for frequency, phase and time
>   synchronization.  The PTP messages include PTP PDUs over UDP/IP
>   (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F of
>   [IEEE-1588]).
>=20
> SB> Sure BUT it is acknowledged that IEEE is open to the definition
> SB> of other PTP mappings if they provide better optimisation.
>=20
>   This document defines mapping and transport of the PTP
>   messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>   defines several clock types: ordinary clocks, boundary clocks, end-
>   to-end transparent clocks, and peer-to-peer transparent clocks.
>   Transparent clocks require intermediate nodes to update correction
>   field inside PTP message that reflects the transit time in the node.
>=20
>   [RFC5905] defines NTP messages for clock and time synchronization.
>   The PTP messages (PDUs) are transported over UDP/IP.  This document
> SB> Should that be NTP messages?
> SB> It needs to be made clear as soon as you introduce NTP that
> SB> they use different time representations.
>=20
>   defines mapping and transport of the NTP messages defined in
>   [RFC5905] over MPLS networks.
>=20
>   One key attribute of all of these Timing messages is that the Time
>   stamp processing should occur as close as possible to the actual
>   transmission and reception at the physical port interface.  This
>   targets optimal time and/or frequency recovery by avoiding variable
>   delay introduced by queues internal to the clocks.
>=20
> SB> As I recall NTP has no epoch point defined, and I am not sure
> SB> where that point is in the case of PTP in this mapping
> SB> Hopefully this will get defined in due course.
>=20
>   To facilitate the fast and efficient recognition of Timing messages
>   at the port level when the Timing messages are carried over MPLS
>   LSPs,
>=20
> SB> Over a new LSP type with time optimied characteristics
>=20
>   this document defines the specific encapsulations that should
>   be used.
> SB> Hopefully it will also define the PHP
>=20
>   In addition, it can be expected that there will exist LSR/
>   LERs where only a subset of the physical ports will have the port-
>   based Timing message processing capabilities.
> SB> Do you need to clarify that this only works at base and not in
> SB> a label heirarchy.
>=20
>=20
>   In order to ensure
>   that the LSPs carrying Timing packets always enter and exit ports
>   with this capability, routing extensions are defined to advertise
>   this capability on a port basis and to allow for the establishment of
>   LSPs that only transit such ports.  While this path establishment
>   restriction may be applied only at the LER Ingress and/or egress
>   ports, it becomes more important when using transparent clock capable
>   LSRs in the path.
> SB> I do not understand the implications of the last
> SB> sentences - starting ", it becomes"
>=20
>=20
>   Port based Timing message processing involves Timing message
>   recognition.  Once the Timing messages are recognized they can be
>   modified based on the reception or transmission Time-stamp.
>=20
>   This document provides two methods for transporting Timing messages
>   over MPLS.  One is applicable to MPLS environment and the other one
>   is applicable to MPLS/MPLS-TP environment
>=20
> SB> I think the sentence is incomplete.
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013               [Page 5]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
>   The solution involves transporting Timing messages over dedicated
>   LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
>   carry Management and control messages, but not data plane client
>   traffic.
>=20
> SB> It is not clear why this restriction applies.
>=20
>   Timing LSPs can be established statically or via signaling.
> SB> s/statically/by provisioning/network management/
>=20
>   Extensions to control plane (OSPF, ISIS, etc.) is required to enable
>   routers to distribute their Timing processing capabilities over MPLS
>   to other routers.  However such extensions are outside the scope of
>   this document.
>=20
>   When signaling is used to setup the PTP LSP, Extensions to signaling
> SB> is it a PTP LSP or a Timing LSP?
>=20
>   protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
>   However such extensions are outside the scope of this document.
>=20
> SB> for mpls-tp GMPLS is the signalling protocol
>=20
>   While the techniques included herein allow for the establishment of
>   paths optimized to include Time-stamping capable links, the
>   performance of the Slave clocks is outside the scope of this
>   document.
>=20
>   At the time of publishing this specification, Transparent Clocking
>   (TC) is only defined for PTP.  Therefore at this time any part of
>   this specification that talks about Transparent Clocking applies only
>   to PTP.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013               [Page 6]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 2.  Terminology
>=20
>   1588: The timing and synchronization as defined by IEEE 1588.
>=20
> SB> I think that there is a more formal name for the 1588 group
> SB> that needs to be used here.
> SB> Also do we need to talk about 1588-200? as there is
> SB> an update in progress
>=20
>   NTP: The timing and synchronization protocol defined by IETF RFC-1305
>   and RFC-5905.
>=20
>   PTP: The timing and synchronization protocol used by 1588.
> SB> need the proper name for 1588
>=20
>   Master Clock: The source of 1588 timing to a set of slave clocks.
>=20
>   Master Port: A port on a ordinary or boundary clock that is in Master
>   state.  This is the source of timing toward slave ports.
>=20
> SB> I am not sure the reader knows what a port is
>=20
>   Slave Clock: A receiver of 1588 timing from a master clock.
>=20
>   Slave Port: A port on a boundary clock or ordinary clock that is
>   receiving timing from a master clock.
>=20
>   Ordinary Clock: A device with a single PTP port.
>=20
>   Transparent Clock.  A device that measures the time taken for a PTP
>   event message to transit the device and then updates the
>   correctionField of the message with this transit time.
>=20
>   Boundary Clock: A device with more than one PTP port.  Generally
>   boundary clocks will have one port in slave state to receive timing
>   and then other ports in master state to re-distribute the timing.
>=20
>   PTP LSP: An LSP dedicated to carry PTP messages
>=20
> SB> PTP or timing?
>=20
>=20
>   PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>   messages.
>=20
> SB> Ah I don't think that PWE3 know what one of these is
>=20
>   CW: Pseudowire Control Word
>=20
>   LAG: Link Aggregation
>=20
>   ECMP: Equal Cost Multipath
>=20
>   CF: Correction Field, a field inside certain PTP messages (message
>   type 0-3)that holds the accumulative transit time inside intermediate
>   switches
>=20
>   Timing messages: Timing Protocol messages that are exchanged between
>   routers in order to establish a synchronized clock.
>=20
> SB> A number of these definitions look like copies of IEEE1588
> SB> definitions. We need to provide references and note the
> SB> priority of the IEEE base reference.
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013               [Page 7]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 3.  Problem Statement
>=20
>   [IEEE-1588] has defined methods for transporting PTP messages over
>   Ethernet and IP networks.  [RFC5905] has defined the method of
>   transporting NTP messages over IP networks.  There is a need to
>   transport Timing messages over MPLS networks while supporting the
>   Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
>   functionality in the LER and LSRs in the MPLS network.
>=20
>   There are multiple ways of transporting Timing over MPLS.  However,
>   there is a requirement to limit the possible encapsulation options to
>   simplify the Timing message identification and processing required at
>   the port level.
>=20
>   When Timing-awareness is needed, Timing messages should not be
>   transported over LSPs or PWs that are carrying customer traffic
>   because LSRs perform Label switching based on the top label in the
>   stack.
>=20
> SB> Have you explained why?
>=20
>   To detect Timing messages inside such LSPs require special
>   hardware to do deep packet inspection at line rate.  Even if such
>   hardware exists, the payload can't be deterministically identified by
>   LSRs because the payload type is a context of the PW label, and the
>   PW label and its context are only known to the Edge routers (PEs/
>   LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CES,
>   etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
>   LSRs dont have the knowledge of whether PW Control Word (CW) is
>   present or not and therefore can not deterministically identify the
>   payload.
>=20
>   A generic method is defined in this document that does not require
>   deep packet inspection at line rate, and can deterministically
>   identify Timing messages.  This method can be used to detect Timing
>   Messages in both one-step and two-step clock implementations of
>   ordinary, boundary and transparent clocks.
>=20
> SB> Needs a ref and I am sure many MPLS specialists will not understand
> SB> the msg types.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013               [Page 8]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 4.  Timing over MPLS Architecture
>=20
>   Timing messages are exchange between Timing ports on ordinary and
>=20
> SB> Have you defined a timing port?
>=20
>   boundary clocks.  Boundary clocks terminate the Timing messages and
>   act as master for other boundary clocks or for slave clocks.  End-to-
>   End Transparent clocks do not terminate the Timing messages but they
>   do modify the contents of the Timing messages as they transit across
>   the transparent clock.
>=20
>   Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Clock
>=20
>   (TC) could be implemented in either LERs or LSRs.
>=20
> SB> LER and LSR need to be expanded
>=20
>   An example is shown in Figure 1, where the LERs act as Ordinary Clock
>   (OC) and are the initiating/terminating point for Timing messages.
>   The ingress LER encapsulates the Timing messages in Timing LSP and
>   the Egress LER terminates the Timing LSP.  The LSRs act as
>   Transparent Clock (TC) and just update the Timing field in the Timing
>   messages.
>=20
>=20
>      +--------+     +-------+     +-------+     +-------+     +--------+
>      |Switch, |     |       |     |       |     |       |     |Switch, |
>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>      |        |     |  OC   |     |  TC   |     |  OC   |     |        |
>      +--------+     +-------+     +-------+     +-------+     +--------+
>                     /                                 \
>      +-------+     /                                   \     +-------+
>      |  LER  |    /                                     \    |  LER  |
>      | Master|---/                                       \---| Slave |
>      | Clock |                                               | Clock |
>      +-------+                                               +-------+
>=20
>     Figure (1) - Deployment example 1 of timing over MPLS network
>=20
>   Another example is shown in Figure2, where LERs terminate the Timing
>   messages received from switch/routers that are outside of the MPLS
>   network acting as OC or BC.  In this example LERs regenerate the
>   clock and initiate timing messages encapsulated in Timing LSP toward
>   the MPLS network, while the LSRs act as Transparent Clock (TC) and
>   just update the Timing field in the Timing messages, which are
>   already encapsulated in Timing LSPs.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013               [Page 9]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
>     +--------+     +-------+     +-------+     +-------+     +--------+
>     |Switch, |     |       |     |       |     |       |     |Switch, |
>     | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>     | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC  |
>     +--------+     +-------+     +-------+     +-------+     +--------+
>=20
>     Figure (2) - Deployment example 2 of timing over MPLS network
>=20
>=20
>   Another example is shown in Figure 3, where LERs do not terminate the
>   Timing messages received from switch/routers that are outside of the
>   MPLS network acting as OC, TC or BC.  The LERs act as TC and update
>   the Timing field in the Timing messages as they transit the LER,
>   while encapsulating them in timing LSP.  The LSRs also act as
>   Transparent Clock (TC) and just update the Timing field in the Timing
>   messages which are already encapsulated in Timing LSPs.
>=20
>      +--------+     +-------+     +-------+     +-------+     +--------+
>      |Switch, |     |       |     |       |     |       |     |Switch, |
>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>      |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC/BC|
>      +--------+     +-------+     +-------+     +-------+     +--------+
>=20
>    Figure (3) - Deployment example 3 of timing over MPLS network
>=20
>   Another example is shown in Figure 4, where LERs and LSRs support
>   Boundary Clocks.  A single-hop LSP is created between two adjacent
>   LSRs engaged in BC operation.  Other methods such as PTP transport
>   over Ethernet MAY be used for transporting timing messages if the
>   link between the two routers is Ethernet.
>=20
>     +--------+     +-------+     +-------+     +-------+     +--------+
>     |Switch, |     |       |     |       |     |       |     |Switch, |
>     | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>     | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC  |
>     +--------+     +-------+     +-------+     +-------+     +--------+
>=20
>   Figure (4) - Deployment example 3 of timing over MPLS network
>=20
>   An MPLS domain MAY serve multiple customers.  In these cases the MPLS
>   domain (maintained by a service provider) may provide timing services
>   to multiple customers, each having their own Timing domain.
>=20
>   The Timing over MPLS architecture assumes full mesh of Timing LSPs
>   between all LERs supporting this specification.
>=20
> SB> Note sure this is right - the salves surely do not need to
> SB> exchange timing amongst themselves
>=20
>   It supports
>   Point-to- point (VPWS) and Multipoint (VPLS) services.
>=20
> SB> What does that mean? You do not carry user data traffic?
> SB> Maybe it's the ordering of the statemnets that is causing
> SB> confusion.
>=20
>   This means
>   that a customer may purchase a Point-to-point Timing service between
>   two customer sites or a Multipoint Timing service between more than
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 10]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
>   two customer sites.
>=20
>   The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
>   This means that the Timing Multicast messages such as PTP Multicast
>   event messages can be transported over P2MP Timing LSP or be
>   replicated and transported over many P2P Timing LSPs.
>=20
> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
> SB> nor a P2MP PW, although we are close.
>=20
>   Timing messages, that do not require Time stamping or Correction
>   Field update MAY be transported over Timing LSPs to simplify hardware
>   and software.
>=20
>   PTP Announce messages that determine the Timing LSP terminating point
>   behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
>   to simplify hardware and software.
>=20
> SB> have you defined and referenced PTP announce msgs?
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 11]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 5.  Dedicated LSPs for Timing messages
>=20
>   Many methods have been considered for identifying the Timing messages
>   when they are encapsulated in MPLS such as using GAL/G-ACH or a new
>   reserved label.  These methods were not attractive since they either
>   required deep packet inspection at line rate in the intermediate LSRs
>   or they required use of a scarce new reserved label.  Also one of the
>   goals was to reuse existing OAM mechanisms.
>=20
> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
>=20
>   The method defined in this document can be used by LER and LSRs to
>   identify Timing messages in MPLS tunnels by just looking at the top
>   label in the MPLS label stack, which only carry Timing messages as
>   well as OAM, but not data plane client traffic.
>=20
>   Compliant implementations MUST use dedicated LSPs to carry Timing
>   messages over MPLS.
>=20
> SB> I think that we need a definition of the properies of these LSPs
>=20
>   These LSPs are herein referred to as "Timing
>   LSPs" and the labels associated with these LSPs as "Timing LSP
>   labels".  The Timing LSPs that runs between Ingress and Egress LERs
>   MUST be co-routed.  Alternatively, a single bidirectional co-routed
>   LSP can be used.
>=20
> SB> I though that you said you could use M2MP LSPs - these are not
> SB> bidirectional.
>=20
>   Co-routing of the two directions is required to limit the difference
>   in the delays in the Master clock to Slave clock direction compared
>   to the Slave clock to Master clock direction.  The Timing LSP MAY be
>   MPLS/MPLS-TP LSP.
>=20
>   The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
>   New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
>   outside the scope of this document.
>=20
>   The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such as
>   BFD and LSP Ping but the LSP data plane client plane traffic MUST be
>   Timing packets only.
>=20
> SB> Why?
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 12]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 6.  Timing over LSP Encapsulation
>=20
> The encapsulations is not LSP is it?
>=20
>   This document defines two methods for carrying Timing messages over
>   MPLS.  The first method is carrying UDP/IP encapsulated Timing
>   messages over Timing LSPs, and the second method, is carrying
>   Ethernet encapsulated Timing messages over Ethernet PWs inside Timing
>   LSPs.
>=20
> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>=20
>   The simplest method of transporting Timing messages over MPLS is to
>   encapsulate Timing PDUs in UDP/IP and then encapsulate them in Timing
>   LSP.  This format is shown in Figure 4.
>=20
>=20
>                    +----------------------+
>                    |   Timing LSP Label   |
>                    +----------------------+
>                    |        IPv4/6        |
>                    +----------------------+
>                    |         UDP          |
>                    +----------------------+
>                    |     Timing PDU       |
>                    +----------------------+
>=20
>      Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>=20
>=20
>   This encapsulation is very simple and is useful when the network
>   between Timing Master Clock and Slave Clock is MPLS network.
>=20
> SB> Simple is a judgement call
>=20
>   In order for an LER/LSR to process Timing messages, the Timing LSP
>   Label must be at the top label of the label stack.  The LER/LSR MUST
>   know that the Timing LSP Label is used for carrying Timing messages.
>   This can be accomplished via static configuration or via RSVP-TE
>   signaling.
>=20
>   The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>   [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>   [RFC5905].
>=20
> 6.2.  Timing over PW Encapsulation
>=20
>   Another method of transporting Timing over MPLS networks is by
>   encapsulating Timing PDUs in PW which in turn is transported over
>   Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
>   shown in Fig 5(A) MUST be used and the Ethernet encapsulation of PTP
>   MUST follow Annex F of [IEEE-1588].
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 13]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
>   The RAW mode or Tagged mode defined in [RFC4448] MAY be used and the
>   Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>   Timing over PW encapsulation MUST use the Control Word (CW) as
>   specified in [RFC4448] to ensure proper detection of PTP messages
>   inside the MPLS packets for Timing over LSP and Timing over PW
>   encapsulation.
>=20
> SB> That needs explanation
>=20
>   The use of Sequence Number in the CW is optional.
>=20
> SB> Given that s/n are never in practice deployed, you could probably
> SB> simplify things by sayig that they are not used.
>=20
>   Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over PW
>   (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>=20
>                    +----------------+  +----------------+
>                     |Timing LSP Label|  |Timing LSP Label|
>                     +----------------+  +----------------+
>                     |    PW Label    |  |    PW Label    |
>                     +----------------+  +----------------+
>                     |  Control Word  |  |      IP        |
>                     +----------------+  +----------------+
>                     |    Ethernet    |  |      UDP       |
>                     |     Header     |  +----------------+
>                     +----------------+  |   Timing PDU   |
>                     |S-VLAN(Optional)|  |                |
>                     +----------------+  +----------------+
>                     |C-VLAN(Optional)|        (B)
>                     +----------------+
>                     |   Timing PDU   |
>                     |                |
>                     +----------------+
>                            (A)
>=20
>              Figure (5) - Timing over PW Encapsulations
>=20
>   In order for an LSR to process PTP messages, the top label of the
>   label stack (the Tunnel Label) MUST be a Timing label.
>=20
> S> You said that before.
>=20
> 6.3.  Other Timing Encapsulation methods
>=20
>   In future other timing encapsulation methods may be introduced, such
>   as a new shim header after the Bottom of Stack to carry the Timing
>   information.  Such new encapsulations are outside the scope of this
>   document.
>=20
>=20
> SB> Taking a pure MPLS pov, you can simplify a lot of the text
> SB> out of the definition of the LSP
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 14]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> SB> I think we need a section on LSP processing
>=20
> 7.  Timing message Processing
>=20
>   Each Timing protocol such as PTP and NTP, define their set of Timing
>   messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>   FOLLOW_UP, etc messages.
>=20
>   Some of the Timing messages require time stamping or correction field
>   update at port level and some dont.  It is the job of the LER/LSR to
>   parse the timing message and find out the type of the Timing message
>   and decide whether and how to Time- stamp it (e.g., BC) or update
>   correction field(e.g., TC).
>=20
>=20
> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
> SB> function rather than the LER?
>=20
>   For example the following PTP messages (called Event messages)
>   require time-stamping or correction field update:
>=20
>   o  SYNC
>=20
>   o  DELAY_REQ (Delay Request)
>=20
>   o  PDELAY_REQ (Peer Delay Request)
>=20
>   o  PDELAY_RESP (Peer Delay Response)
>=20
>   SYNC and DELAY_REQ are exchanged between Master Clock and Slave Clock
>   and MUST be transported over PTP LSPs.  PDELAY_REQ and PDELAY_RESP
>   are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>   Boundary, or Transparent) and SHOULD be transported over single hop
>   PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
>   and PDELAY_RESP_FOLLOW_UP messages MUST also be transported over the
>   PTP LSPs.
>=20
>   For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
>   transported over two PTP LSPs that are in opposite directions.  These
>   PTP LSPs, which are in opposite directions MUST be congruent and co-
>   routed.  Alternatively, a single bidirectional co-routed LSP can be
>   used.
>=20
>   Except as indicated above for the two-step PTP clocks, Non-Event PTP
>   message types do not need to be processed by intermediate routers.
>   These message types MAY be carried in PTP Tunnel LSPs.
>=20
> SB> Are you saying that a timing P router has to be msg type sensitive?
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 15]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 8.  Protection and Redundancy
>=20
>=20
> SB> This is a bit of a jump - I don't know how the LSP itself works yet!
>=20
>   In order to ensure continuous uninterrupted operation of slave
>   clocks, usually as a general practice, slave clocks (or ports) track
>   redundant master clocks.
>=20
>   It is the responsibility of the network operator to ensure that
>   physically disjoint Timing LSPs are established between a slave clock
>   (or port) and redundant master clocks (or ports).
>=20
>   When a slave clock (or port) listens to redundant master clocks or
>   ports, any prolonged Timing LSP outage will trigger the slave clock
>   or port to switch to a redundant master clock or port.
>=20
>   LSP/PW protection such as Linear protection Switching (1:1, 1+1),
>   Ring protection switching or MPLS Fast Reroute (FRR) generally switch
>   alternative path that usually cause a change in delay, which if
>   undetected by slave clock can reduce accuracy of the slave clock.
>=20
>   Therefore protection switching MAY be used, as long as phase jumps
>   upon switchover due to differences in path latency are detected and
>   compensated for (such compensation not being required if BCs or peer-
>   peer TCs are used throughout).
>=20
>   Note that any protection or reroute mechanism that adds additional
>   MPLS label to the label stack, such as Facility Backup Fast Reroute,
>   MUST ensure that the pushed label is also a Timing Label to ensure
>   recognition of the MPLS frame as containing Timing messages, as it
>   transits the backup path.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 16]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 9.  ECMP
>=20
>   To ensure the optimal operation of slave clocks and avoid error
>   introduced by forward and reverse path delay asymmetry, the physical
>   path for Timing messages from master clock to slave Clock and vice
>   versa must be the same for all Event Timing messages listed in
>   section 7.
>=20
>   Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
>   Multipath).
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 17]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 10.  PHP
>=20
>   To ensure that the label on the top of the label stack is the Timing
>   LSP Label, PHP MUST not be used.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 18]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 11.  Entropy
>=20
>   To ensure all Timing messages in a Timing LSP take the same path,
>   Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>   Entropy Label MUST NOT be used for the PWs that are carried inside
>   Timing LSP [RFC6391].
>=20
> SB> This is incorrect - you mean that all msgs of the same timing
> SB> flow need to have the same EL value.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 19]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 12.  OAM, Control and Management
>=20
>   In order to monitor Timing LSPs and their encapsulated PWs, they MUST
>   be able to carry OAM and management messages.  These management
>   messages MUST be differentiated from Timing messages via already
>   defined IETF methods.
>=20
>   For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
>   over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>   Management protocols can easily be identified by the UDP Destination
>   Port number or by GAL/G-ACH respectively.
>=20
>   Also BFD, LSP-Ping and other management messages MAY run over the PWs
>   encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3 or
>   4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is going
>   to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1) or
>   GAL-ACH are used to identify such management messages.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 20]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 13.  QoS Considerations
>=20
>   In network deployments where not every LSR/LER is Timing-aware, it is
>   important to reduce the impact of the non-Timing-aware LSR/LERs on
>   the timing recovery in the slave clock.  The Timing messages are time
>   critical and must be treated with the highest priority.  Therefore
>   Timing over MPLS messages must be treated with the highest priority
>   in the routers.  This can be achieved by proper setup of Timing LSPs.
>=20
>   It is recommended that the Timing LSPs are setup or configured
>   properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC2697]
>   for drop eligibility.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 21]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 14.  FCS and Checksum Recalculation
>=20
>   When time-stamp generation and timing packet adjustment is performed
>   near the physical port hardware, the process MUST include
>   recalculation of the Ethernet FCS.
>=20
> SB> The above is confusing - an LSR always recomputes the link layer
> SB> CRC which may or may not be Ethernet.
>=20
>   Also FCS retention for the
>   payload Ethernet described in [RFC4720] MUST NOT be used.
>=20
>   For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
>   may be required as per UDP transport standards.
>=20
> SB> You really need to be working on getting the IPv6 C?S computation
> SB> removed from PTP msgs.
>=20
>   When UDP checksum is used, each Timing-aware LER/LSR must either
>   incrementally update the UDP checksum after Time stamping or
>   Correction Field update or verify the UDP checksum on reception from
>   upstream and recalculate the checksum completely on transmission to
>   downstream node after Time stamping or Correction Field update.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 22]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 15.  Behavior of LER/LSR
>=20
>   Timing-capable/aware LERs and LSRs are routers that have one or more
>=20
> SB> You mean physical interfaces?
>=20
>   interfaces that can perform Timing operations (OC/BC/TC) on Timing
>   packets and are configured to do so.  Timing-capable/aware LERs and
>   LSRs can advertise their Timing-capability per-interface via control
>   plane such as OSPF or IS-IS.
> SB> ISIS and OSPF are routing protocols.
>=20
>  The Timing-capable/aware LERs can then
>   signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timing
>   capability of LER and LSRs may be configured in a centralized
>   controller and the Timing LSP may be setup using manual configuration
>   or other methods such as SDN.
>=20
> SB> it can also be configured individually rather then through
> SB> a cebtral controllwe
>=20
> 15.1.  Behavior of Timing-capable/aware LER
>=20
>   When a Timing-capable/aware LER behaves as a Transparent clock and
>   receives a Timing message from a Timing-capable/aware non-MPLS
>   interface, the LER updates the Correction Field (CF) and encapsulates
>   and forwards the timing message over previously established Timing
>   LSP.
>=20
> SB> You need to call out the details so that people properly
> SB> understand the definition of the new LSP.
>=20
>   Also when a Timing message is received from a Timing-capable/
>   aware MPLS interface, LER updates the Correction Filed (CF) and
>   decapsulates the MPLS encapsulation and forwards the timing message
>   to a non-MPLS interface.
>=20
>   When a Timing-capable/aware LER behaves as a Boundary clock and
>   receives a Timing message from a Timing-capable/aware non MPLS
>   interface, the LER Timestamps the Timing packet and sends it to the
>   LERs Boundary clock processing module.  Also when a Timing message is
>   received from a Timing- capable/aware MPLS interface, the LER
>   Timestamps the Timing packet and sends it to the LERs Boundary clock
>   processing module.
>=20
>   When a Timing-capable/aware LER behaves as an Ordinary Clock toward
>   the MPLS network, and receives a Timing message from a Timing-
>   capable/aware MPLS interface, the LER Timestamps the Timing packet
>   and sends it to the LERs Ordinary clock processing module.
>=20
> 15.2.  Behavior of Timing-capable/aware LSR
>=20
>   When a Timing-capable/aware LSR behaves as a Transparent clock and
>   receives a Timing message from a Timing-capable/aware MPLS interface,
>   The LSR updates the Correction Filed (CF) and forwards the timing
>   message over another MPLS interface.
>=20
>   When a Timing-capable/aware LSR behaves as a Boundary clock and
>   receives a Timing message from a Timing-capable/aware MPLS interface.
>   The LSR performs the functions of a Boundary Clock in terminating the
>   received Timing message and re-generating a new timing message over
>   another (or the same) MPLS interface.
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 23]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 15.3.  Behavior of non-Timing-capable/aware LSR
>=20
>   It is most beneficial when all LSRs in the path of a Timing LSP be
>   timing-Capable/aware LSRs.  This would ensure the highest quality
>   time and clock synchronization by Timing Slave Clocks.  However, this
>   specification does not mandate that all LSRs in path of a Timing LSP
>   be Timing- capable/aware.
>=20
>   Non-Timing-capable/aware LSRs just switch the packets encapsulated in
>   Timing LSPs and dont perform any Timing operation (TC or BC).
>   However as explained in QoS section the Timing over MPLS packets MUST
>   be still be treated with the highest priority based on their Traffic
>   Class (TC) marking.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 24]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 16.  Other considerations
>=20
>   [IEEE-1588] defines an optional peer-to-peer Transparent clocking
>   that requires peer delay measurement between two adjacent Timing-
>   capable/ aware routers/switches.  Peer delay measurement messages
>   need to be time stamped and terminated by the Timing-capable/aware
>   routers/ switches.  This means that two adjacent LSRs may be engaged
>   in a peer delay measurement.
>=20
>   For transporting such peer delay measurement messages a single-hop
>   LSP SHOULD to be created between the two adjacent LSRs engaged in
>   peer delay measurement to carry peer delay measurement messages.
>   Other methods such as PTP transport over Ethernet MAY be used for
>   transporting peer delay measurement messages if the link between the
>   two routers is Ethernet.
>=20
>   In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ ware
>   routers/switches MUST maintain a list of all the neighbors it needs
>   to send a PDelay_Req to, where each neighbor corresponds to a timing
>   LSP.
>=20
>   The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as long
>   as either the Explicit Null label is the bottom of stack label
>   (applicable only to UDP/IP encapsulation) or the label below the
>   Explicit Null label is a PTP label.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 25]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 17.  Security Considerations
>=20
>   MPLS PW security considerations in general are discussed in [RFC3985]
>   and [RFC4447],and those considerations also apply to this document.
>=20
>   An experimental security protocol is defined in [IEEE-1588].The PTP
>   security extension and protocol provides group source authentication,
>   message integrity, and replay attack protection for PTP messages.
>=20
>   When the MPLS network (provider network) serves multiple customers,
>   it is important to maintain and process each customers clock and
>   Timing messages separately from other customers to ensure there is no
>   cross- customer effect.  For example if an LER BC is synchronized to
>   a specific grandmaster, belonging to customer A, then the LER MUST
>   use that BC clock only for customer A to ensure that customer A
>   cannot attack other customers by manipulating its time.
>=20
>   Timing messages MAY be encrypted or authenticated, provided that the
>   LERs/LSRs that are Timing capable/aware can authenticate/ decrypt the
>   timing messages.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 26]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 18.  Acknowledgements
>=20
>   The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizrahi,
>   Stefano Ruffini, Peter Meyer, and other members of IETF for reviewing
>   and providing feedback on this draft.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 27]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 19.  IANA Considerations
>=20
>   There are no IANA requirements in this specification.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 28]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 20.  References
>=20
> 20.1.  Normative References
>=20
>   [IEEE-1588]
>              IEEE 1588-2008, "IEEE Standard for a Precision Clock
>              Synchronization Protocol for Networked Measurement and
>              Control Systems".
>=20
>   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>              Requirement Levels", BCP 14, RFC 2119, March 1997.
>=20
>   [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
>              Edge (PWE3) Architecture", RFC 3985, March 2005.
>=20
>   [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
>              Proxies (ND Proxy)", RFC 4389, April 2006.
>=20
>   [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
>              Heron, "Pseudowire Setup and Maintenance Using the Label
>              Distribution Protocol (LDP)", RFC 4447, April 2006.
>=20
>   [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>              "Encapsulation Methods for Transport of Ethernet over MPLS
>              Networks", RFC 4448, April 2006.
>=20
>   [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>              Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>              Retention", RFC 4720, November 2006.
>=20
>   [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
>              Connectivity Verification (VCCV): A Control Channel for
>              Pseudowires", RFC 5085, December 2007.
>=20
>   [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detection
>              (BFD)", RFC 5880, June 2010.
>=20
>   [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
>              "Bidirectional Forwarding Detection (BFD) for MPLS Label
>              Switched Paths (LSPs)", RFC 5884, June 2010.
>=20
> 20.2.  Informative References
>=20
>   [I-D.ietf-pwe3-fat-pw]
>              Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>              J., and S. Amante, "Flow Aware Transport of Pseudowires
>              over an MPLS Packet Switched Network",
>              draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 29]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
>   [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
>              system routeing information exchange protocol for use in
>              conjunction with the Protocol for providing the
>              Connectionless-mode Network Service (ISO 8473)".
>=20
>   [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
>              dual environments", RFC 1195, December 1990.
>=20
>   [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
>=20
>   [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
>              Marker", RFC 2697, September 1999.
>=20
>   [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boudec,
>              J., Courtney, W., Davari, S., Firoiu, V., and D.
>              Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>              Behavior)", RFC 3246, March 2002.
>=20
>   [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineering
>              (TE) Extensions to OSPF Version 2", RFC 3630,
>              September 2003.
>=20
>   [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
>              System (IS-IS) Extensions for Traffic Engineering (TE)",
>              RFC 3784, June 2004.
>=20
>   [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
>              Shaffer, "Extensions to OSPF for Advertising Optional
>              Router Capabilities", RFC 4970, July 2007.
>=20
>   [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
>              System to Intermediate System (IS-IS) Extensions for
>              Advertising Router Information", RFC 4971, July 2007.
>=20
>   [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
>              Topology (MT) Routing in Intermediate System to
>              Intermediate Systems (IS-ISs)", RFC 5120, February 2008.
>=20
>   [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>              Engineering", RFC 5305, October 2008.
>=20
>   [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>              "Traffic Engineering Extensions to OSPF Version 3",
>              RFC 5329, September 2008.
>=20
>   [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
>              for IPv6", RFC 5340, July 2008.
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 30]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
>   [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network
>              Time Protocol Version 4: Protocol and Algorithms
>              Specification", RFC 5905, June 2010.
>=20
>   [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>              J., and S. Amante, "Flow-Aware Transport of Pseudowires
>              over an MPLS Packet Switched Network", RFC 6391,
>              November 2011.
>=20
>   [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
>              L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
>              RFC 6790, November 2012.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 31]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 1.  Routing extensions for Timing-aware Routers
>=20
>   MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] and
>   IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (TE)
>   link information used for constraint-based routing.
>=20
>   Indeed, it is useful to advertise data plane TE router link
>   capabilities, such as the capability for a router to be Timing-aware.
>   This capability MUST then be taken into account during path
>   computation to prefer or even require links that advertise themselves
>   as Timing-aware.  In this way the path can ensure the entry and exit
>   points into the LERs and, if desired, the links into the LSRs are
>   able to perform port based time-stamping thus minimizing their impact
>   on the performance of the slave clock.
>=20
>   extensions are required to OSPF and IS-IS in order to advertise
>   Timing-aware capabilities of a link.  Such extensions are outside the
>   scope of this document; however such extension SHOULD be able to
>   signal the following information per Router Link:
>=20
>   o  Capable of processing PTP, NTP or other Timing flows
>=20
>   o  Capable of performing Transparent Clock operation
>=20
>   o  Capable of performing Boundary Clock operation
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 32]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> 2.  Signaling Extensions for Creating Timing LSPs
>=20
>   RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-TE
>   is used to setup Timing LSPs, some information that indicates that
>   the LSP is carrying Timing flows MUST be included in the new
>   Extensions to RSVP-TE:
>=20
>   The following information MAY also be included in the new Extensions
>   to RSVP-TE:
>=20
>   o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
>      field
>=20
>   o  Number of VLANs in case of PW encapsulation
>=20
>   o  Timestamp field Type
>=20
>      *  Correction Field, Timestamp
>=20
>   o  Timestamp Field format
>=20
>      *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
>         NTP, etc.
>=20
>   Note that in case the above optional information is signaled with
>   RSVP-TE for a Timing LSP, all the Timing packets carried in that LSP
>   must have the same signaled characteristics.  For example if
>   Timestamp format is signaled as 64-bit PTPv1, then all Timing packets
>   must use 64-bit PTPv1 time-stamp.
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 33]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
> Authors' Addresses
>=20
>   Shahram Davari
>   Broadcom Corp.
>   San Jose, CA  95134
>   USA
>=20
>   Email: davari@broadcom.com
>=20
>=20
>   Amit Oren
>   Broadcom Corp.
>   San Jose, CA  95134
>   USA
>=20
>   Email: amito@broadcom.com
>=20
>=20
>   Manav Bhatia
>   Alcatel-Lucent
>   Bangalore,
>   India
>=20
>   Email: manav.bhatia@alcatel-lucent.com
>=20
>=20
>   Peter Roberts
>   Alcatel-Lucent
>   Kanata,
>   Canada
>=20
>   Email: peter.roberts@alcatel-lucent.com
>=20
>=20
>   Laurent Montini
>   Cisco Systems
>   San Jose CA
>   USA
>=20
>   Email: lmontini@cisco.com
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 34]
> =0C
> Internet-Draft        Transporting Timing over MPLS            June 2013
>=20
>=20
>   Luca
>   Cisco Systems
>   San Jose CA
>   USA
>=20
>   Email: lmartini@cisco.com
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
>=20
> Davari, et al.          Expires December 17, 2013              [Page 35]
> =0C
>=20
> --=20
> For corporate legal information go to:
>=20
> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>=20
> _______________________________________________
> mpls mailing list
> mpls@ietf.org
> https://www.ietf.org/mailman/listinfo/mpls
>=20


From jdrake@juniper.net  Fri Aug  2 03:52:23 2013
Return-Path: <jdrake@juniper.net>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id A99F311E8241; Fri,  2 Aug 2013 03:52:23 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 1.533
X-Spam-Level: *
X-Spam-Status: No, score=1.533 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_LOW=-1, SARE_RAND_6=2, UNRESOLVED_TEMPLATE=3.132]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id hVBUASLshY42; Fri,  2 Aug 2013 03:52:13 -0700 (PDT)
Received: from am1outboundpool.messaging.microsoft.com (am1ehsobe005.messaging.microsoft.com [213.199.154.208]) by ietfa.amsl.com (Postfix) with ESMTP id 2585711E8210; Fri,  2 Aug 2013 03:52:12 -0700 (PDT)
Received: from mail13-am1-R.bigfish.com (10.3.201.244) by AM1EHSOBE019.bigfish.com (10.3.207.141) with Microsoft SMTP Server id 14.1.225.22; Fri, 2 Aug 2013 10:52:11 +0000
Received: from mail13-am1 (localhost [127.0.0.1])	by mail13-am1-R.bigfish.com (Postfix) with ESMTP id 094911A00F5; Fri,  2 Aug 2013 10:52:11 +0000 (UTC)
X-Forefront-Antispam-Report: CIP:66.129.224.54; KIP:(null); UIP:(null); IPV:NLI; H:P-EMF01-SAC.jnpr.net; RD:none; EFVD:NLI
X-SpamScore: -30
X-BigFish: VPS-30(zz62a3I98dI9371I601I542Iec9I1432I1418I168aJzz1f42h208ch1ee6h1de0h1fdah2073h1202h1e76h1d1ah1d2ah1fc6hzz1de098h1033IL17326ah1de096h8275bh8275dh1de097hz2fh2a8h683h839h944hd25hf0ah1220h1288h12a5h12a9h12bdh137ah13b6h1441h1504h1537h153bh15d0h162dh1631h1758h18e1h1946h19b5h19ceh1ad9h1b0ah1b2fh1fb3h1d07h1d0ch1d2eh1d3fh1de9h1dfeh1dffh1e1dh1155h)
Received-SPF: pass (mail13-am1: domain of juniper.net designates 66.129.224.54 as permitted sender) client-ip=66.129.224.54; envelope-from=jdrake@juniper.net; helo=P-EMF01-SAC.jnpr.net ; SAC.jnpr.net ; 
X-Forefront-Antispam-Report-Untrusted: CIP:157.56.240.101; KIP:(null); UIP:(null); (null); H:BL2PRD0510HT001.namprd05.prod.outlook.com; R:internal; EFV:INT
Received: from mail13-am1 (localhost.localdomain [127.0.0.1]) by mail13-am1 (MessageSwitch) id 137544072793821_1236; Fri,  2 Aug 2013 10:52:07 +0000 (UTC)
Received: from AM1EHSMHS016.bigfish.com (unknown [10.3.201.228])	by mail13-am1.bigfish.com (Postfix) with ESMTP id 12E283C0046; Fri,  2 Aug 2013 10:52:07 +0000 (UTC)
Received: from P-EMF01-SAC.jnpr.net (66.129.224.54) by AM1EHSMHS016.bigfish.com (10.3.207.154) with Microsoft SMTP Server (TLS) id 14.16.227.3; Fri, 2 Aug 2013 10:52:06 +0000
Received: from P-CLDFE02-HQ.jnpr.net (172.24.192.60) by P-EMF01-SAC.jnpr.net (172.24.192.17) with Microsoft SMTP Server (TLS) id 14.3.146.0; Fri, 2 Aug 2013 03:52:04 -0700
Received: from o365mail.juniper.net (207.17.137.224) by o365mail.juniper.net (172.24.192.60) with Microsoft SMTP Server id 14.3.146.0; Fri, 2 Aug 2013 03:52:04 -0700
Received: from DB8EHSOBE022.bigfish.com (213.199.154.187) by o365mail.juniper.net (207.17.137.224) with Microsoft SMTP Server (TLS) id 14.3.146.0; Fri, 2 Aug 2013 04:04:59 -0700
Received: from mail83-db8-R.bigfish.com (10.174.8.239) by DB8EHSOBE022.bigfish.com (10.174.4.85) with Microsoft SMTP Server id 14.1.225.22; Fri, 2 Aug 2013 10:52:01 +0000
Received: from mail83-db8 (localhost [127.0.0.1])	by mail83-db8-R.bigfish.com (Postfix) with ESMTP id 347825402B8; Fri,  2 Aug 2013 10:52:01 +0000 (UTC)
Received: from mail83-db8 (localhost.localdomain [127.0.0.1]) by mail83-db8 (MessageSwitch) id 1375440705320831_27751; Fri,  2 Aug 2013 10:51:45 +0000 (UTC)
Received: from DB8EHSMHS024.bigfish.com (unknown [10.174.8.232])	by mail83-db8.bigfish.com (Postfix) with ESMTP id 3E7CFCC0047; Fri,  2 Aug 2013 10:51:45 +0000 (UTC)
Received: from BL2PRD0510HT001.namprd05.prod.outlook.com (157.56.240.101) by DB8EHSMHS024.bigfish.com (10.174.4.34) with Microsoft SMTP Server (TLS) id 14.16.227.3; Fri, 2 Aug 2013 10:51:43 +0000
Received: from BL2PRD0510MB349.namprd05.prod.outlook.com ([169.254.1.45]) by BL2PRD0510HT001.namprd05.prod.outlook.com ([10.255.100.36]) with mapi id 14.16.0341.000; Fri, 2 Aug 2013 10:51:42 +0000
From: John E Drake <jdrake@juniper.net>
To: Shahram Davari <davari@broadcom.com>, "<stbryant@cisco.com>" <stbryant@cisco.com>
Thread-Topic: [mpls] Comments on draft-ietf-tictoc-1588overmpls
Thread-Index: AQHOj2d0A3VaSYY+akqLLdxXI6C3DpmBtd+AgAAHkOA=
Date: Fri, 2 Aug 2013 10:51:41 +0000
Message-ID: <0182DEA5604B3A44A2EE61F3EE3ED69E2076CBA4@BL2PRD0510MB349.namprd05.prod.outlook.com>
References: <51FB8396.9020504@cisco.com> <D2796BF4-4DDB-4816-B730-1289CAB11080@broadcom.com>
In-Reply-To: <D2796BF4-4DDB-4816-B730-1289CAB11080@broadcom.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-originating-ip: [66.129.224.52]
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-FOPE-CONNECTOR: Id%0$Dn%*$RO%0$TLS%0$FQDN%$TlsDn%
X-FOPE-CONNECTOR: Id%12219$Dn%BROADCOM.COM$RO%2$TLS%5$FQDN%onpremiseedge-1018244.customer.frontbridge.com$TlsDn%o365mail.juniper.net
X-FOPE-CONNECTOR: Id%12219$Dn%CISCO.COM$RO%2$TLS%5$FQDN%onpremiseedge-1018244.customer.frontbridge.com$TlsDn%o365mail.juniper.net
X-FOPE-CONNECTOR: Id%12219$Dn%TOOLS.IETF.ORG$RO%2$TLS%5$FQDN%onpremiseedge-1018244.customer.frontbridge.com$TlsDn%o365mail.juniper.net
X-FOPE-CONNECTOR: Id%12219$Dn%IETF.ORG$RO%2$TLS%5$FQDN%onpremiseedge-1018244.customer.frontbridge.com$TlsDn%o365mail.juniper.net
X-OriginatorOrg: juniper.net
X-FOPE-CONNECTOR: Id%0$Dn%*$RO%0$TLS%0$FQDN%$TlsDn%
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc-chairs@tools.ietf.org" <tictoc-chairs@tools.ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>, "draft-ietf-tictoc-1588overmpls@tools.ietf.org" <draft-ietf-tictoc-1588overmpls@tools.ietf.org>, "rtg-ads@tools.ietf.org" <rtg-ads@tools.ietf.org>, "int-ads@tools.ietf.org" <int-ads@tools.ietf.org>, "mpls-chairs@tools.ietf.org" <mpls-chairs@tools.ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 02 Aug 2013 10:52:24 -0000

Shahram,

The LSPs would be between LSRs that are 1588 capable, so they could be one =
or more hops.

Yours Irrespectively,

John


> -----Original Message-----
> From: Shahram Davari [mailto:davari@broadcom.com]
> Sent: Friday, August 02, 2013 3:24 AM
> To: <stbryant@cisco.com>
> Cc: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
> ads@tools.ietf.org; rtg-ads@tools.ietf.org; draft-ietf-tictoc-
> 1588overmpls@tools.ietf.org; John E Drake; mpls@ietf.org;
> tictoc@ietf.org
> Subject: Re: [mpls] Comments on draft-ietf-tictoc-1588overmpls
>=20
> Hi Stewart
>=20
> We have already looked at similar approach. The issue with one hop LSP
> is that not all LSRs are 1588 capable. One of the requirements is for
> non-1588 capable routers to just switch the packet normally.
>=20
> Regards,
> Shahram
>=20
>=20
> On Aug 2, 2013, at 12:02 PM, "Stewart Bryant" <stbryant@cisco.com>
> wrote:
>=20
> > Talking to John Drake about this, an alternative general
> > model is for the LSP is that it is an LSP that timestamps
> > the packet and then passes it to an application associated
> > with the LSP at that hop.
> >
> > This has a lot of merit.
> >
> > If however we think about it some more we have a
> > type of network service chaining going on here and
> > so we don't need an LSP per say, because the path
> > and instruction can be in the packet.
> >
> > In other words a general solution  to the problem
> > is to define an LSP with the properties that the
> > packet is passed one hop, timestamped and delivered
> > to the associated application.
> >
> > How this is MPLS construct is used to support time
> > tranfer is entirely within the scope of TICTOC, but
> > at the MPLS layer we have a clean reusable network
> > service.
> >
> > - Stewart
> >
> >
> > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> > SB> This draft does not seem to provide a precise definition
> > SB> the properties of the new LSP type that it wishes to
> > SB> define, in particular it does it define the PHB of those
> > SB> LSPs, nor the full interaction with the MPLS
> > SB> architecture.
> > SB>
> > SB> I have not tracked TICTOC for a while but I thought that
> > SB> the original plan was to define the concept of an offset
> > SB> into a packet to do the correction.
> > SB>
> > SB> It is disappointing that the opportunity was not taken
> > SB> to define a timing shim inside the timing LSP so that
> > SB> a time correction could be added to any packet such that
> > SB> the MPLS system was isolated from the details of the
> > SB> complexity of the particular time transfer type.
> >
> > SB> I think that much more clarify is needed in terms of
> > SB> definition of the new LSP type, since it is unclear
> > SB> from this text how to implement one.
> > SB>
> > SB> There are a lot of other MPLS services such as
> > SB> LSP ping that need to be considered.
> > SB>
> > SB> Please see inline for more comments. However these
> > SB> comments are made in the context of the text as written
> > SB> whilst I have a fundamental concern that this approach
> > SB> lacks an MPLS architectural soundness that need
> > SB> greater thought with significant impact on the
> > SB> draft.
> >
> > - Stewart
> >
> >
> > TICTOC Working Group                                           S.
> Davari
> > Internet-Draft                                                   A.
> Oren
> > Intended status: Standards Track                          Broadcom
> Corp.
> > Expires: December 17, 2013                                     M.
> Bhatia
> >                                                              P.
> Roberts
> >                                                          Alcatel-
> Lucent
> >                                                              L.
> Montini
> >                                                              L.
> Martini
> >                                                           Cisco
> Systems
> >                                                           June 15,
> 2013
> >
> >
> >            Transporting Timing messages over MPLS Networks
> >                   draft-ietf-tictoc-1588overmpls-05
> >
> > Abstract
> >
> >   This document defines the method for transporting Timing messages
> >   such as PTP and NTP over an MPLS network.  The method allows for
> the
> >   easy identification of these PDUs at the port level to allow for
> port
> >
> > SB> What is a port
> >
> >   level processing of these PDUs in both LERs and LSRs.
> >
> >   The basic idea is to transport Timing messages inside dedicated
> MPLS
> >   LSPs.  These LSPs only carry Timing messages and possibly Control
> and
> >   Management packets, but they do not carry customer traffic.
> >
> > SB> More specifically they only carry traffic associated with the
> > SB> timing service and its support.
> > SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
> > SB> also it gets carried in a structure that causes it to get
> > SB> timestamped.
> >
> >   Two methods for transporting Timing messages over MPLS are defined.
> >
> > SB> Perhaps the right approach is to define the new LSP type and then
> > SB> seperately to define  the mapping of the various timing services
> > SB> over that LSP type.
> >
> >   The first method is to transport Timing messages directly over the
> >   dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
> >   MPLS networks.  The second method is to transport Timing messages
> >   inside a PW via Ethernet encapsulation.
> >
> > SB> I think that we should note that there are some
> > SB> h/w reasons for this preference. A clean sheet approach
> > SB> would have been to use PTP over MPLS with no intermediate
> > SB> layers.
> >
> > Status of this Memo
> >
> >   This Internet-Draft is submitted in full conformance with the
> >   provisions of BCP 78 and BCP 79.
> >
> >   Internet-Drafts are working documents of the Internet Engineering
> >   Task Force (IETF).  Note that other groups may also distribute
> >   working documents as Internet-Drafts.  The list of current
> Internet-
> >   Drafts is at http://datatracker.ietf.org/drafts/current/.
> >
> >   Internet-Drafts are draft documents valid for a maximum of six
> months
> >   and may be updated, replaced, or obsoleted by other documents at
> any
> >   time.  It is inappropriate to use Internet-Drafts as reference
> >   material or to cite them other than as "work in progress."
> >
> >   This Internet-Draft will expire on December 17, 2013.
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013               [Page
> 1]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > Copyright Notice
> >
> >   Copyright (c) 2013 IETF Trust and the persons identified as the
> >   document authors.  All rights reserved.
> >
> >   This document is subject to BCP 78 and the IETF Trust's Legal
> >   Provisions Relating to IETF Documents
> >   (http://trustee.ietf.org/license-info) in effect on the date of
> >   publication of this document.  Please review these documents
> >   carefully, as they describe your rights and restrictions with
> respect
> >   to this document.  Code Components extracted from this document
> must
> >   include Simplified BSD License text as described in Section 4.e of
> >   the Trust Legal Provisions and are provided without warranty as
> >   described in the Simplified BSD License.
> >
> >
> > Table of Contents
> >
> >   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .
> 5
> >
> >   2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .
> 7
> >
> >   3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .
> 8
> >
> >   4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .
> 9
> >
> >   5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . .
> 12
> >
> >   6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . .
> 13
> >     6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . .
> 13
> >     6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . .
> 13
> >     6.3.  Other Timing Encapsulation methods . . . . . . . . . . . .
> 14
> >
> >   7.  Timing message Processing  . . . . . . . . . . . . . . . . . .
> 15
> >
> >   8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . .
> 16
> >
> >   9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
> 17
> >
> >   10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
> 18
> >
> >   11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . .
> 19
> >
> >   12. OAM, Control and Management  . . . . . . . . . . . . . . . . .
> 20
> >
> >   13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . .
> 21
> >
> >   14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . .
> 22
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013               [Page
> 2]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> >   15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . .
> 23
> >     15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . .
> 23
> >     15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . .
> 23
> >     15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . .
> 24
> >
> >   16. Other considerations . . . . . . . . . . . . . . . . . . . . .
> 25
> >
> >   17. Security Considerations  . . . . . . . . . . . . . . . . . . .
> 26
> >
> >   18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . .
> 27
> >
> >   19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . .
> 28
> >
> >   20. References . . . . . . . . . . . . . . . . . . . . . . . . . .
> 29
> >     20.1. Normative References . . . . . . . . . . . . . . . . . . .
> 29
> >     20.2. Informative References . . . . . . . . . . . . . . . . . .
> 29
> >
> >   Appendix 1.  Routing extensions for Timing-aware Routers . . . . .
> 32
> >
> >   Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . .
> 33
> >
> >   Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . .
> 34
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013               [Page
> 3]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> >   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
> >   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in
> this
> >   document are to be interpreted as described in RFC2119 [RFC2119].
> >
> >   When used in lower case, these words convey their typical use in
> >   common language, and are not to be interpreted as described in
> >   RFC2119 [RFC2119].
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013               [Page
> 4]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 1.  Introduction
> >
> >   The objective of Precision Time Protocol (PTP) and Network Timing
> >   Protocol (NTP) are to synchronize independent clocks running on
> >   separate nodes of a distributed system.
> >
> >   [IEEE-1588] defines PTP messages for frequency, phase and time
> >   synchronization.  The PTP messages include PTP PDUs over UDP/IP
> >   (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F
> of
> >   [IEEE-1588]).
> >
> > SB> Sure BUT it is acknowledged that IEEE is open to the definition
> > SB> of other PTP mappings if they provide better optimisation.
> >
> >   This document defines mapping and transport of the PTP
> >   messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
> >   defines several clock types: ordinary clocks, boundary clocks, end-
> >   to-end transparent clocks, and peer-to-peer transparent clocks.
> >   Transparent clocks require intermediate nodes to update correction
> >   field inside PTP message that reflects the transit time in the
> node.
> >
> >   [RFC5905] defines NTP messages for clock and time synchronization.
> >   The PTP messages (PDUs) are transported over UDP/IP.  This document
> > SB> Should that be NTP messages?
> > SB> It needs to be made clear as soon as you introduce NTP that
> > SB> they use different time representations.
> >
> >   defines mapping and transport of the NTP messages defined in
> >   [RFC5905] over MPLS networks.
> >
> >   One key attribute of all of these Timing messages is that the Time
> >   stamp processing should occur as close as possible to the actual
> >   transmission and reception at the physical port interface.  This
> >   targets optimal time and/or frequency recovery by avoiding variable
> >   delay introduced by queues internal to the clocks.
> >
> > SB> As I recall NTP has no epoch point defined, and I am not sure
> > SB> where that point is in the case of PTP in this mapping
> > SB> Hopefully this will get defined in due course.
> >
> >   To facilitate the fast and efficient recognition of Timing messages
> >   at the port level when the Timing messages are carried over MPLS
> >   LSPs,
> >
> > SB> Over a new LSP type with time optimied characteristics
> >
> >   this document defines the specific encapsulations that should
> >   be used.
> > SB> Hopefully it will also define the PHP
> >
> >   In addition, it can be expected that there will exist LSR/
> >   LERs where only a subset of the physical ports will have the port-
> >   based Timing message processing capabilities.
> > SB> Do you need to clarify that this only works at base and not in
> > SB> a label heirarchy.
> >
> >
> >   In order to ensure
> >   that the LSPs carrying Timing packets always enter and exit ports
> >   with this capability, routing extensions are defined to advertise
> >   this capability on a port basis and to allow for the establishment
> of
> >   LSPs that only transit such ports.  While this path establishment
> >   restriction may be applied only at the LER Ingress and/or egress
> >   ports, it becomes more important when using transparent clock
> capable
> >   LSRs in the path.
> > SB> I do not understand the implications of the last
> > SB> sentences - starting ", it becomes"
> >
> >
> >   Port based Timing message processing involves Timing message
> >   recognition.  Once the Timing messages are recognized they can be
> >   modified based on the reception or transmission Time-stamp.
> >
> >   This document provides two methods for transporting Timing messages
> >   over MPLS.  One is applicable to MPLS environment and the other one
> >   is applicable to MPLS/MPLS-TP environment
> >
> > SB> I think the sentence is incomplete.
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013               [Page
> 5]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> >   The solution involves transporting Timing messages over dedicated
> >   LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
> >   carry Management and control messages, but not data plane client
> >   traffic.
> >
> > SB> It is not clear why this restriction applies.
> >
> >   Timing LSPs can be established statically or via signaling.
> > SB> s/statically/by provisioning/network management/
> >
> >   Extensions to control plane (OSPF, ISIS, etc.) is required to
> enable
> >   routers to distribute their Timing processing capabilities over
> MPLS
> >   to other routers.  However such extensions are outside the scope of
> >   this document.
> >
> >   When signaling is used to setup the PTP LSP, Extensions to
> signaling
> > SB> is it a PTP LSP or a Timing LSP?
> >
> >   protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
> >   However such extensions are outside the scope of this document.
> >
> > SB> for mpls-tp GMPLS is the signalling protocol
> >
> >   While the techniques included herein allow for the establishment of
> >   paths optimized to include Time-stamping capable links, the
> >   performance of the Slave clocks is outside the scope of this
> >   document.
> >
> >   At the time of publishing this specification, Transparent Clocking
> >   (TC) is only defined for PTP.  Therefore at this time any part of
> >   this specification that talks about Transparent Clocking applies
> only
> >   to PTP.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013               [Page
> 6]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 2.  Terminology
> >
> >   1588: The timing and synchronization as defined by IEEE 1588.
> >
> > SB> I think that there is a more formal name for the 1588 group
> > SB> that needs to be used here.
> > SB> Also do we need to talk about 1588-200? as there is
> > SB> an update in progress
> >
> >   NTP: The timing and synchronization protocol defined by IETF RFC-
> 1305
> >   and RFC-5905.
> >
> >   PTP: The timing and synchronization protocol used by 1588.
> > SB> need the proper name for 1588
> >
> >   Master Clock: The source of 1588 timing to a set of slave clocks.
> >
> >   Master Port: A port on a ordinary or boundary clock that is in
> Master
> >   state.  This is the source of timing toward slave ports.
> >
> > SB> I am not sure the reader knows what a port is
> >
> >   Slave Clock: A receiver of 1588 timing from a master clock.
> >
> >   Slave Port: A port on a boundary clock or ordinary clock that is
> >   receiving timing from a master clock.
> >
> >   Ordinary Clock: A device with a single PTP port.
> >
> >   Transparent Clock.  A device that measures the time taken for a PTP
> >   event message to transit the device and then updates the
> >   correctionField of the message with this transit time.
> >
> >   Boundary Clock: A device with more than one PTP port.  Generally
> >   boundary clocks will have one port in slave state to receive timing
> >   and then other ports in master state to re-distribute the timing.
> >
> >   PTP LSP: An LSP dedicated to carry PTP messages
> >
> > SB> PTP or timing?
> >
> >
> >   PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
> >   messages.
> >
> > SB> Ah I don't think that PWE3 know what one of these is
> >
> >   CW: Pseudowire Control Word
> >
> >   LAG: Link Aggregation
> >
> >   ECMP: Equal Cost Multipath
> >
> >   CF: Correction Field, a field inside certain PTP messages (message
> >   type 0-3)that holds the accumulative transit time inside
> intermediate
> >   switches
> >
> >   Timing messages: Timing Protocol messages that are exchanged
> between
> >   routers in order to establish a synchronized clock.
> >
> > SB> A number of these definitions look like copies of IEEE1588
> > SB> definitions. We need to provide references and note the
> > SB> priority of the IEEE base reference.
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013               [Page
> 7]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 3.  Problem Statement
> >
> >   [IEEE-1588] has defined methods for transporting PTP messages over
> >   Ethernet and IP networks.  [RFC5905] has defined the method of
> >   transporting NTP messages over IP networks.  There is a need to
> >   transport Timing messages over MPLS networks while supporting the
> >   Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
> >   functionality in the LER and LSRs in the MPLS network.
> >
> >   There are multiple ways of transporting Timing over MPLS.  However,
> >   there is a requirement to limit the possible encapsulation options
> to
> >   simplify the Timing message identification and processing required
> at
> >   the port level.
> >
> >   When Timing-awareness is needed, Timing messages should not be
> >   transported over LSPs or PWs that are carrying customer traffic
> >   because LSRs perform Label switching based on the top label in the
> >   stack.
> >
> > SB> Have you explained why?
> >
> >   To detect Timing messages inside such LSPs require special
> >   hardware to do deep packet inspection at line rate.  Even if such
> >   hardware exists, the payload can't be deterministically identified
> by
> >   LSRs because the payload type is a context of the PW label, and the
> >   PW label and its context are only known to the Edge routers (PEs/
> >   LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR,
> CES,
> >   etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
> >   LSRs dont have the knowledge of whether PW Control Word (CW) is
> >   present or not and therefore can not deterministically identify the
> >   payload.
> >
> >   A generic method is defined in this document that does not require
> >   deep packet inspection at line rate, and can deterministically
> >   identify Timing messages.  This method can be used to detect Timing
> >   Messages in both one-step and two-step clock implementations of
> >   ordinary, boundary and transparent clocks.
> >
> > SB> Needs a ref and I am sure many MPLS specialists will not
> understand
> > SB> the msg types.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013               [Page
> 8]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 4.  Timing over MPLS Architecture
> >
> >   Timing messages are exchange between Timing ports on ordinary and
> >
> > SB> Have you defined a timing port?
> >
> >   boundary clocks.  Boundary clocks terminate the Timing messages and
> >   act as master for other boundary clocks or for slave clocks.  End-
> to-
> >   End Transparent clocks do not terminate the Timing messages but
> they
> >   do modify the contents of the Timing messages as they transit
> across
> >   the transparent clock.
> >
> >   Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent
> Clock
> >
> >   (TC) could be implemented in either LERs or LSRs.
> >
> > SB> LER and LSR need to be expanded
> >
> >   An example is shown in Figure 1, where the LERs act as Ordinary
> Clock
> >   (OC) and are the initiating/terminating point for Timing messages.
> >   The ingress LER encapsulates the Timing messages in Timing LSP and
> >   the Egress LER terminates the Timing LSP.  The LSRs act as
> >   Transparent Clock (TC) and just update the Timing field in the
> Timing
> >   messages.
> >
> >
> >      +--------+     +-------+     +-------+     +-------+     +------
> --+
> >      |Switch, |     |       |     |       |     |       |
> |Switch, |
> >      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----|
> Router |
> >      |        |     |  OC   |     |  TC   |     |  OC   |     |
> |
> >      +--------+     +-------+     +-------+     +-------+     +------
> --+
> >                     /                                 \
> >      +-------+     /                                   \     +-------
> +
> >      |  LER  |    /                                     \    |  LER
> |
> >      | Master|---/                                       \---| Slave
> |
> >      | Clock |                                               | Clock
> |
> >      +-------+                                               +-------
> +
> >
> >     Figure (1) - Deployment example 1 of timing over MPLS network
> >
> >   Another example is shown in Figure2, where LERs terminate the
> Timing
> >   messages received from switch/routers that are outside of the MPLS
> >   network acting as OC or BC.  In this example LERs regenerate the
> >   clock and initiate timing messages encapsulated in Timing LSP
> toward
> >   the MPLS network, while the LSRs act as Transparent Clock (TC) and
> >   just update the Timing field in the Timing messages, which are
> >   already encapsulated in Timing LSPs.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013               [Page
> 9]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> >     +--------+     +-------+     +-------+     +-------+     +-------
> -+
> >     |Switch, |     |       |     |       |     |       |     |Switch,
> |
> >     | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router
> |
> >     | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC
> |
> >     +--------+     +-------+     +-------+     +-------+     +-------
> -+
> >
> >     Figure (2) - Deployment example 2 of timing over MPLS network
> >
> >
> >   Another example is shown in Figure 3, where LERs do not terminate
> the
> >   Timing messages received from switch/routers that are outside of
> the
> >   MPLS network acting as OC, TC or BC.  The LERs act as TC and update
> >   the Timing field in the Timing messages as they transit the LER,
> >   while encapsulating them in timing LSP.  The LSRs also act as
> >   Transparent Clock (TC) and just update the Timing field in the
> Timing
> >   messages which are already encapsulated in Timing LSPs.
> >
> >      +--------+     +-------+     +-------+     +-------+     +------
> --+
> >      |Switch, |     |       |     |       |     |       |
> |Switch, |
> >      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----|
> Router |
> >      |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |
> |OC/TC/BC|
> >      +--------+     +-------+     +-------+     +-------+     +------
> --+
> >
> >    Figure (3) - Deployment example 3 of timing over MPLS network
> >
> >   Another example is shown in Figure 4, where LERs and LSRs support
> >   Boundary Clocks.  A single-hop LSP is created between two adjacent
> >   LSRs engaged in BC operation.  Other methods such as PTP transport
> >   over Ethernet MAY be used for transporting timing messages if the
> >   link between the two routers is Ethernet.
> >
> >     +--------+     +-------+     +-------+     +-------+     +-------
> -+
> >     |Switch, |     |       |     |       |     |       |     |Switch,
> |
> >     | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router
> |
> >     | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC
> |
> >     +--------+     +-------+     +-------+     +-------+     +-------
> -+
> >
> >   Figure (4) - Deployment example 3 of timing over MPLS network
> >
> >   An MPLS domain MAY serve multiple customers.  In these cases the
> MPLS
> >   domain (maintained by a service provider) may provide timing
> services
> >   to multiple customers, each having their own Timing domain.
> >
> >   The Timing over MPLS architecture assumes full mesh of Timing LSPs
> >   between all LERs supporting this specification.
> >
> > SB> Note sure this is right - the salves surely do not need to
> > SB> exchange timing amongst themselves
> >
> >   It supports
> >   Point-to- point (VPWS) and Multipoint (VPLS) services.
> >
> > SB> What does that mean? You do not carry user data traffic?
> > SB> Maybe it's the ordering of the statemnets that is causing
> > SB> confusion.
> >
> >   This means
> >   that a customer may purchase a Point-to-point Timing service
> between
> >   two customer sites or a Multipoint Timing service between more than
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 10]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> >   two customer sites.
> >
> >   The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
> >   This means that the Timing Multicast messages such as PTP Multicast
> >   event messages can be transported over P2MP Timing LSP or be
> >   replicated and transported over many P2P Timing LSPs.
> >
> > SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
> > SB> nor a P2MP PW, although we are close.
> >
> >   Timing messages, that do not require Time stamping or Correction
> >   Field update MAY be transported over Timing LSPs to simplify
> hardware
> >   and software.
> >
> >   PTP Announce messages that determine the Timing LSP terminating
> point
> >   behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
> >   to simplify hardware and software.
> >
> > SB> have you defined and referenced PTP announce msgs?
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 11]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 5.  Dedicated LSPs for Timing messages
> >
> >   Many methods have been considered for identifying the Timing
> messages
> >   when they are encapsulated in MPLS such as using GAL/G-ACH or a new
> >   reserved label.  These methods were not attractive since they
> either
> >   required deep packet inspection at line rate in the intermediate
> LSRs
> >   or they required use of a scarce new reserved label.  Also one of
> the
> >   goals was to reuse existing OAM mechanisms.
> >
> > SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
> >
> >   The method defined in this document can be used by LER and LSRs to
> >   identify Timing messages in MPLS tunnels by just looking at the top
> >   label in the MPLS label stack, which only carry Timing messages as
> >   well as OAM, but not data plane client traffic.
> >
> >   Compliant implementations MUST use dedicated LSPs to carry Timing
> >   messages over MPLS.
> >
> > SB> I think that we need a definition of the properies of these LSPs
> >
> >   These LSPs are herein referred to as "Timing
> >   LSPs" and the labels associated with these LSPs as "Timing LSP
> >   labels".  The Timing LSPs that runs between Ingress and Egress LERs
> >   MUST be co-routed.  Alternatively, a single bidirectional co-routed
> >   LSP can be used.
> >
> > SB> I though that you said you could use M2MP LSPs - these are not
> > SB> bidirectional.
> >
> >   Co-routing of the two directions is required to limit the
> difference
> >   in the delays in the Master clock to Slave clock direction compared
> >   to the Slave clock to Master clock direction.  The Timing LSP MAY
> be
> >   MPLS/MPLS-TP LSP.
> >
> >   The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
> >   New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
> >   outside the scope of this document.
> >
> >   The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such
> as
> >   BFD and LSP Ping but the LSP data plane client plane traffic MUST
> be
> >   Timing packets only.
> >
> > SB> Why?
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 12]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 6.  Timing over LSP Encapsulation
> >
> > The encapsulations is not LSP is it?
> >
> >   This document defines two methods for carrying Timing messages over
> >   MPLS.  The first method is carrying UDP/IP encapsulated Timing
> >   messages over Timing LSPs, and the second method, is carrying
> >   Ethernet encapsulated Timing messages over Ethernet PWs inside
> Timing
> >   LSPs.
> >
> > 6.1.  Timing over UDP/IP over MPLS Encapsulation
> >
> >   The simplest method of transporting Timing messages over MPLS is to
> >   encapsulate Timing PDUs in UDP/IP and then encapsulate them in
> Timing
> >   LSP.  This format is shown in Figure 4.
> >
> >
> >                    +----------------------+
> >                    |   Timing LSP Label   |
> >                    +----------------------+
> >                    |        IPv4/6        |
> >                    +----------------------+
> >                    |         UDP          |
> >                    +----------------------+
> >                    |     Timing PDU       |
> >                    +----------------------+
> >
> >      Figure (4) - Timing over UDP/IP over MPLS Encapsulation
> >
> >
> >   This encapsulation is very simple and is useful when the network
> >   between Timing Master Clock and Slave Clock is MPLS network.
> >
> > SB> Simple is a judgement call
> >
> >   In order for an LER/LSR to process Timing messages, the Timing LSP
> >   Label must be at the top label of the label stack.  The LER/LSR
> MUST
> >   know that the Timing LSP Label is used for carrying Timing
> messages.
> >   This can be accomplished via static configuration or via RSVP-TE
> >   signaling.
> >
> >   The UDP/IP encapsulation of PTP MUST follow Annex D and E of
> >   [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
> >   [RFC5905].
> >
> > 6.2.  Timing over PW Encapsulation
> >
> >   Another method of transporting Timing over MPLS networks is by
> >   encapsulating Timing PDUs in PW which in turn is transported over
> >   Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
> >   shown in Fig 5(A) MUST be used and the Ethernet encapsulation of
> PTP
> >   MUST follow Annex F of [IEEE-1588].
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 13]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> >   The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
> the
> >   Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
> >   Timing over PW encapsulation MUST use the Control Word (CW) as
> >   specified in [RFC4448] to ensure proper detection of PTP messages
> >   inside the MPLS packets for Timing over LSP and Timing over PW
> >   encapsulation.
> >
> > SB> That needs explanation
> >
> >   The use of Sequence Number in the CW is optional.
> >
> > SB> Given that s/n are never in practice deployed, you could probably
> > SB> simplify things by sayig that they are not used.
> >
> >   Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over
> PW
> >   (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
> >
> >                    +----------------+  +----------------+
> >                     |Timing LSP Label|  |Timing LSP Label|
> >                     +----------------+  +----------------+
> >                     |    PW Label    |  |    PW Label    |
> >                     +----------------+  +----------------+
> >                     |  Control Word  |  |      IP        |
> >                     +----------------+  +----------------+
> >                     |    Ethernet    |  |      UDP       |
> >                     |     Header     |  +----------------+
> >                     +----------------+  |   Timing PDU   |
> >                     |S-VLAN(Optional)|  |                |
> >                     +----------------+  +----------------+
> >                     |C-VLAN(Optional)|        (B)
> >                     +----------------+
> >                     |   Timing PDU   |
> >                     |                |
> >                     +----------------+
> >                            (A)
> >
> >              Figure (5) - Timing over PW Encapsulations
> >
> >   In order for an LSR to process PTP messages, the top label of the
> >   label stack (the Tunnel Label) MUST be a Timing label.
> >
> > S> You said that before.
> >
> > 6.3.  Other Timing Encapsulation methods
> >
> >   In future other timing encapsulation methods may be introduced,
> such
> >   as a new shim header after the Bottom of Stack to carry the Timing
> >   information.  Such new encapsulations are outside the scope of this
> >   document.
> >
> >
> > SB> Taking a pure MPLS pov, you can simplify a lot of the text
> > SB> out of the definition of the LSP
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 14]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > SB> I think we need a section on LSP processing
> >
> > 7.  Timing message Processing
> >
> >   Each Timing protocol such as PTP and NTP, define their set of
> Timing
> >   messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
> >   FOLLOW_UP, etc messages.
> >
> >   Some of the Timing messages require time stamping or correction
> field
> >   update at port level and some dont.  It is the job of the LER/LSR
> to
> >   parse the timing message and find out the type of the Timing
> message
> >   and decide whether and how to Time- stamp it (e.g., BC) or update
> >   correction field(e.g., TC).
> >
> >
> > SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
> > SB> function rather than the LER?
> >
> >   For example the following PTP messages (called Event messages)
> >   require time-stamping or correction field update:
> >
> >   o  SYNC
> >
> >   o  DELAY_REQ (Delay Request)
> >
> >   o  PDELAY_REQ (Peer Delay Request)
> >
> >   o  PDELAY_RESP (Peer Delay Response)
> >
> >   SYNC and DELAY_REQ are exchanged between Master Clock and Slave
> Clock
> >   and MUST be transported over PTP LSPs.  PDELAY_REQ and PDELAY_RESP
> >   are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
> >   Boundary, or Transparent) and SHOULD be transported over single hop
> >   PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
> >   and PDELAY_RESP_FOLLOW_UP messages MUST also be transported over
> the
> >   PTP LSPs.
> >
> >   For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
> >   transported over two PTP LSPs that are in opposite directions.
> These
> >   PTP LSPs, which are in opposite directions MUST be congruent and
> co-
> >   routed.  Alternatively, a single bidirectional co-routed LSP can be
> >   used.
> >
> >   Except as indicated above for the two-step PTP clocks, Non-Event
> PTP
> >   message types do not need to be processed by intermediate routers.
> >   These message types MAY be carried in PTP Tunnel LSPs.
> >
> > SB> Are you saying that a timing P router has to be msg type
> sensitive?
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 15]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 8.  Protection and Redundancy
> >
> >
> > SB> This is a bit of a jump - I don't know how the LSP itself works
> yet!
> >
> >   In order to ensure continuous uninterrupted operation of slave
> >   clocks, usually as a general practice, slave clocks (or ports)
> track
> >   redundant master clocks.
> >
> >   It is the responsibility of the network operator to ensure that
> >   physically disjoint Timing LSPs are established between a slave
> clock
> >   (or port) and redundant master clocks (or ports).
> >
> >   When a slave clock (or port) listens to redundant master clocks or
> >   ports, any prolonged Timing LSP outage will trigger the slave clock
> >   or port to switch to a redundant master clock or port.
> >
> >   LSP/PW protection such as Linear protection Switching (1:1, 1+1),
> >   Ring protection switching or MPLS Fast Reroute (FRR) generally
> switch
> >   alternative path that usually cause a change in delay, which if
> >   undetected by slave clock can reduce accuracy of the slave clock.
> >
> >   Therefore protection switching MAY be used, as long as phase jumps
> >   upon switchover due to differences in path latency are detected and
> >   compensated for (such compensation not being required if BCs or
> peer-
> >   peer TCs are used throughout).
> >
> >   Note that any protection or reroute mechanism that adds additional
> >   MPLS label to the label stack, such as Facility Backup Fast
> Reroute,
> >   MUST ensure that the pushed label is also a Timing Label to ensure
> >   recognition of the MPLS frame as containing Timing messages, as it
> >   transits the backup path.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 16]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 9.  ECMP
> >
> >   To ensure the optimal operation of slave clocks and avoid error
> >   introduced by forward and reverse path delay asymmetry, the
> physical
> >   path for Timing messages from master clock to slave Clock and vice
> >   versa must be the same for all Event Timing messages listed in
> >   section 7.
> >
> >   Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
> >   Multipath).
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 17]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 10.  PHP
> >
> >   To ensure that the label on the top of the label stack is the
> Timing
> >   LSP Label, PHP MUST not be used.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 18]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 11.  Entropy
> >
> >   To ensure all Timing messages in a Timing LSP take the same path,
> >   Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
> >   Entropy Label MUST NOT be used for the PWs that are carried inside
> >   Timing LSP [RFC6391].
> >
> > SB> This is incorrect - you mean that all msgs of the same timing
> > SB> flow need to have the same EL value.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 19]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 12.  OAM, Control and Management
> >
> >   In order to monitor Timing LSPs and their encapsulated PWs, they
> MUST
> >   be able to carry OAM and management messages.  These management
> >   messages MUST be differentiated from Timing messages via already
> >   defined IETF methods.
> >
> >   For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
> >   over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
> >   Management protocols can easily be identified by the UDP
> Destination
> >   Port number or by GAL/G-ACH respectively.
> >
> >   Also BFD, LSP-Ping and other management messages MAY run over the
> PWs
> >   encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3
> or
> >   4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is
> going
> >   to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1) or
> >   GAL-ACH are used to identify such management messages.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 20]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 13.  QoS Considerations
> >
> >   In network deployments where not every LSR/LER is Timing-aware, it
> is
> >   important to reduce the impact of the non-Timing-aware LSR/LERs on
> >   the timing recovery in the slave clock.  The Timing messages are
> time
> >   critical and must be treated with the highest priority.  Therefore
> >   Timing over MPLS messages must be treated with the highest priority
> >   in the routers.  This can be achieved by proper setup of Timing
> LSPs.
> >
> >   It is recommended that the Timing LSPs are setup or configured
> >   properly to indicate EF-PHB [RFC3246]for the CoS and Green
> [RFC2697]
> >   for drop eligibility.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 21]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 14.  FCS and Checksum Recalculation
> >
> >   When time-stamp generation and timing packet adjustment is
> performed
> >   near the physical port hardware, the process MUST include
> >   recalculation of the Ethernet FCS.
> >
> > SB> The above is confusing - an LSR always recomputes the link layer
> > SB> CRC which may or may not be Ethernet.
> >
> >   Also FCS retention for the
> >   payload Ethernet described in [RFC4720] MUST NOT be used.
> >
> >   For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
> >   may be required as per UDP transport standards.
> >
> > SB> You really need to be working on getting the IPv6 C?S computation
> > SB> removed from PTP msgs.
> >
> >   When UDP checksum is used, each Timing-aware LER/LSR must either
> >   incrementally update the UDP checksum after Time stamping or
> >   Correction Field update or verify the UDP checksum on reception
> from
> >   upstream and recalculate the checksum completely on transmission to
> >   downstream node after Time stamping or Correction Field update.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 22]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 15.  Behavior of LER/LSR
> >
> >   Timing-capable/aware LERs and LSRs are routers that have one or
> more
> >
> > SB> You mean physical interfaces?
> >
> >   interfaces that can perform Timing operations (OC/BC/TC) on Timing
> >   packets and are configured to do so.  Timing-capable/aware LERs and
> >   LSRs can advertise their Timing-capability per-interface via
> control
> >   plane such as OSPF or IS-IS.
> > SB> ISIS and OSPF are routing protocols.
> >
> >  The Timing-capable/aware LERs can then
> >   signals Timing LSPs via RSVP-TE signaling.  Alternatively the
> Timing
> >   capability of LER and LSRs may be configured in a centralized
> >   controller and the Timing LSP may be setup using manual
> configuration
> >   or other methods such as SDN.
> >
> > SB> it can also be configured individually rather then through
> > SB> a cebtral controllwe
> >
> > 15.1.  Behavior of Timing-capable/aware LER
> >
> >   When a Timing-capable/aware LER behaves as a Transparent clock and
> >   receives a Timing message from a Timing-capable/aware non-MPLS
> >   interface, the LER updates the Correction Field (CF) and
> encapsulates
> >   and forwards the timing message over previously established Timing
> >   LSP.
> >
> > SB> You need to call out the details so that people properly
> > SB> understand the definition of the new LSP.
> >
> >   Also when a Timing message is received from a Timing-capable/
> >   aware MPLS interface, LER updates the Correction Filed (CF) and
> >   decapsulates the MPLS encapsulation and forwards the timing message
> >   to a non-MPLS interface.
> >
> >   When a Timing-capable/aware LER behaves as a Boundary clock and
> >   receives a Timing message from a Timing-capable/aware non MPLS
> >   interface, the LER Timestamps the Timing packet and sends it to the
> >   LERs Boundary clock processing module.  Also when a Timing message
> is
> >   received from a Timing- capable/aware MPLS interface, the LER
> >   Timestamps the Timing packet and sends it to the LERs Boundary
> clock
> >   processing module.
> >
> >   When a Timing-capable/aware LER behaves as an Ordinary Clock toward
> >   the MPLS network, and receives a Timing message from a Timing-
> >   capable/aware MPLS interface, the LER Timestamps the Timing packet
> >   and sends it to the LERs Ordinary clock processing module.
> >
> > 15.2.  Behavior of Timing-capable/aware LSR
> >
> >   When a Timing-capable/aware LSR behaves as a Transparent clock and
> >   receives a Timing message from a Timing-capable/aware MPLS
> interface,
> >   The LSR updates the Correction Filed (CF) and forwards the timing
> >   message over another MPLS interface.
> >
> >   When a Timing-capable/aware LSR behaves as a Boundary clock and
> >   receives a Timing message from a Timing-capable/aware MPLS
> interface.
> >   The LSR performs the functions of a Boundary Clock in terminating
> the
> >   received Timing message and re-generating a new timing message over
> >   another (or the same) MPLS interface.
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 23]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 15.3.  Behavior of non-Timing-capable/aware LSR
> >
> >   It is most beneficial when all LSRs in the path of a Timing LSP be
> >   timing-Capable/aware LSRs.  This would ensure the highest quality
> >   time and clock synchronization by Timing Slave Clocks.  However,
> this
> >   specification does not mandate that all LSRs in path of a Timing
> LSP
> >   be Timing- capable/aware.
> >
> >   Non-Timing-capable/aware LSRs just switch the packets encapsulated
> in
> >   Timing LSPs and dont perform any Timing operation (TC or BC).
> >   However as explained in QoS section the Timing over MPLS packets
> MUST
> >   be still be treated with the highest priority based on their
> Traffic
> >   Class (TC) marking.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 24]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 16.  Other considerations
> >
> >   [IEEE-1588] defines an optional peer-to-peer Transparent clocking
> >   that requires peer delay measurement between two adjacent Timing-
> >   capable/ aware routers/switches.  Peer delay measurement messages
> >   need to be time stamped and terminated by the Timing-capable/aware
> >   routers/ switches.  This means that two adjacent LSRs may be
> engaged
> >   in a peer delay measurement.
> >
> >   For transporting such peer delay measurement messages a single-hop
> >   LSP SHOULD to be created between the two adjacent LSRs engaged in
> >   peer delay measurement to carry peer delay measurement messages.
> >   Other methods such as PTP transport over Ethernet MAY be used for
> >   transporting peer delay measurement messages if the link between
> the
> >   two routers is Ethernet.
> >
> >   In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/
> ware
> >   routers/switches MUST maintain a list of all the neighbors it needs
> >   to send a PDelay_Req to, where each neighbor corresponds to a
> timing
> >   LSP.
> >
> >   The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as
> long
> >   as either the Explicit Null label is the bottom of stack label
> >   (applicable only to UDP/IP encapsulation) or the label below the
> >   Explicit Null label is a PTP label.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 25]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 17.  Security Considerations
> >
> >   MPLS PW security considerations in general are discussed in
> [RFC3985]
> >   and [RFC4447],and those considerations also apply to this document.
> >
> >   An experimental security protocol is defined in [IEEE-1588].The PTP
> >   security extension and protocol provides group source
> authentication,
> >   message integrity, and replay attack protection for PTP messages.
> >
> >   When the MPLS network (provider network) serves multiple customers,
> >   it is important to maintain and process each customers clock and
> >   Timing messages separately from other customers to ensure there is
> no
> >   cross- customer effect.  For example if an LER BC is synchronized
> to
> >   a specific grandmaster, belonging to customer A, then the LER MUST
> >   use that BC clock only for customer A to ensure that customer A
> >   cannot attack other customers by manipulating its time.
> >
> >   Timing messages MAY be encrypted or authenticated, provided that
> the
> >   LERs/LSRs that are Timing capable/aware can authenticate/ decrypt
> the
> >   timing messages.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 26]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 18.  Acknowledgements
> >
> >   The authors would like to thank Ron Cohen, Yaakov Stein, Tal
> Mizrahi,
> >   Stefano Ruffini, Peter Meyer, and other members of IETF for
> reviewing
> >   and providing feedback on this draft.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 27]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 19.  IANA Considerations
> >
> >   There are no IANA requirements in this specification.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 28]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 20.  References
> >
> > 20.1.  Normative References
> >
> >   [IEEE-1588]
> >              IEEE 1588-2008, "IEEE Standard for a Precision Clock
> >              Synchronization Protocol for Networked Measurement and
> >              Control Systems".
> >
> >   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
> >              Requirement Levels", BCP 14, RFC 2119, March 1997.
> >
> >   [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
> >              Edge (PWE3) Architecture", RFC 3985, March 2005.
> >
> >   [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor
> Discovery
> >              Proxies (ND Proxy)", RFC 4389, April 2006.
> >
> >   [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
> >              Heron, "Pseudowire Setup and Maintenance Using the Label
> >              Distribution Protocol (LDP)", RFC 4447, April 2006.
> >
> >   [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
> >              "Encapsulation Methods for Transport of Ethernet over
> MPLS
> >              Networks", RFC 4448, April 2006.
> >
> >   [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
> >              Emulation Edge-to-Edge (PWE3) Frame Check Sequence
> >              Retention", RFC 4720, November 2006.
> >
> >   [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
> >              Connectivity Verification (VCCV): A Control Channel for
> >              Pseudowires", RFC 5085, December 2007.
> >
> >   [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding
> Detection
> >              (BFD)", RFC 5880, June 2010.
> >
> >   [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
> >              "Bidirectional Forwarding Detection (BFD) for MPLS Label
> >              Switched Paths (LSPs)", RFC 5884, June 2010.
> >
> > 20.2.  Informative References
> >
> >   [I-D.ietf-pwe3-fat-pw]
> >              Bryant, S., Filsfils, C., Drafz, U., Kompella, V.,
> Regan,
> >              J., and S. Amante, "Flow Aware Transport of Pseudowires
> >              over an MPLS Packet Switched Network",
> >              draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 29]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> >   [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
> >              system routeing information exchange protocol for use in
> >              conjunction with the Protocol for providing the
> >              Connectionless-mode Network Service (ISO 8473)".
> >
> >   [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
> >              dual environments", RFC 1195, December 1990.
> >
> >   [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
> >
> >   [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
> >              Marker", RFC 2697, September 1999.
> >
> >   [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le
> Boudec,
> >              J., Courtney, W., Davari, S., Firoiu, V., and D.
> >              Stiliadis, "An Expedited Forwarding PHB (Per-Hop
> >              Behavior)", RFC 3246, March 2002.
> >
> >   [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic
> Engineering
> >              (TE) Extensions to OSPF Version 2", RFC 3630,
> >              September 2003.
> >
> >   [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
> >              System (IS-IS) Extensions for Traffic Engineering (TE)",
> >              RFC 3784, June 2004.
> >
> >   [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
> >              Shaffer, "Extensions to OSPF for Advertising Optional
> >              Router Capabilities", RFC 4970, July 2007.
> >
> >   [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
> >              System to Intermediate System (IS-IS) Extensions for
> >              Advertising Router Information", RFC 4971, July 2007.
> >
> >   [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
> >              Topology (MT) Routing in Intermediate System to
> >              Intermediate Systems (IS-ISs)", RFC 5120, February 2008.
> >
> >   [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
> >              Engineering", RFC 5305, October 2008.
> >
> >   [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
> >              "Traffic Engineering Extensions to OSPF Version 3",
> >              RFC 5329, September 2008.
> >
> >   [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
> >              for IPv6", RFC 5340, July 2008.
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 30]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> >   [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch,
> "Network
> >              Time Protocol Version 4: Protocol and Algorithms
> >              Specification", RFC 5905, June 2010.
> >
> >   [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V.,
> Regan,
> >              J., and S. Amante, "Flow-Aware Transport of Pseudowires
> >              over an MPLS Packet Switched Network", RFC 6391,
> >              November 2011.
> >
> >   [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
> >              L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
> >              RFC 6790, November 2012.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 31]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 1.  Routing extensions for Timing-aware Routers
> >
> >   MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340]
> and
> >   IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering
> (TE)
> >   link information used for constraint-based routing.
> >
> >   Indeed, it is useful to advertise data plane TE router link
> >   capabilities, such as the capability for a router to be Timing-
> aware.
> >   This capability MUST then be taken into account during path
> >   computation to prefer or even require links that advertise
> themselves
> >   as Timing-aware.  In this way the path can ensure the entry and
> exit
> >   points into the LERs and, if desired, the links into the LSRs are
> >   able to perform port based time-stamping thus minimizing their
> impact
> >   on the performance of the slave clock.
> >
> >   extensions are required to OSPF and IS-IS in order to advertise
> >   Timing-aware capabilities of a link.  Such extensions are outside
> the
> >   scope of this document; however such extension SHOULD be able to
> >   signal the following information per Router Link:
> >
> >   o  Capable of processing PTP, NTP or other Timing flows
> >
> >   o  Capable of performing Transparent Clock operation
> >
> >   o  Capable of performing Boundary Clock operation
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 32]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > 2.  Signaling Extensions for Creating Timing LSPs
> >
> >   RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-
> TE
> >   is used to setup Timing LSPs, some information that indicates that
> >   the LSP is carrying Timing flows MUST be included in the new
> >   Extensions to RSVP-TE:
> >
> >   The following information MAY also be included in the new
> Extensions
> >   to RSVP-TE:
> >
> >   o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
> >      field
> >
> >   o  Number of VLANs in case of PW encapsulation
> >
> >   o  Timestamp field Type
> >
> >      *  Correction Field, Timestamp
> >
> >   o  Timestamp Field format
> >
> >      *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
> >         NTP, etc.
> >
> >   Note that in case the above optional information is signaled with
> >   RSVP-TE for a Timing LSP, all the Timing packets carried in that
> LSP
> >   must have the same signaled characteristics.  For example if
> >   Timestamp format is signaled as 64-bit PTPv1, then all Timing
> packets
> >   must use 64-bit PTPv1 time-stamp.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 33]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> > Authors' Addresses
> >
> >   Shahram Davari
> >   Broadcom Corp.
> >   San Jose, CA  95134
> >   USA
> >
> >   Email: davari@broadcom.com
> >
> >
> >   Amit Oren
> >   Broadcom Corp.
> >   San Jose, CA  95134
> >   USA
> >
> >   Email: amito@broadcom.com
> >
> >
> >   Manav Bhatia
> >   Alcatel-Lucent
> >   Bangalore,
> >   India
> >
> >   Email: manav.bhatia@alcatel-lucent.com
> >
> >
> >   Peter Roberts
> >   Alcatel-Lucent
> >   Kanata,
> >   Canada
> >
> >   Email: peter.roberts@alcatel-lucent.com
> >
> >
> >   Laurent Montini
> >   Cisco Systems
> >   San Jose CA
> >   USA
> >
> >   Email: lmontini@cisco.com
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 34]
> >=20

> > Internet-Draft        Transporting Timing over MPLS            June
> 2013
> >
> >
> >   Luca
> >   Cisco Systems
> >   San Jose CA
> >   USA
> >
> >   Email: lmartini@cisco.com
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Davari, et al.          Expires December 17, 2013              [Page
> 35]
> >=20

> >
> > --
> > For corporate legal information go to:
> >
> > http://www.cisco.com/web/about/doing_business/legal/cri/index.html
> >
> > _______________________________________________
> > mpls mailing list
> > mpls@ietf.org
> > https://www.ietf.org/mailman/listinfo/mpls
> >
>=20
>=20




From davari@broadcom.com  Fri Aug  2 06:42:47 2013
Return-Path: <davari@broadcom.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CD2AE21E80AA; Fri,  2 Aug 2013 06:42:46 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.599
X-Spam-Level: 
X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Mrat-QwE5wud; Fri,  2 Aug 2013 06:42:40 -0700 (PDT)
Received: from mms3.broadcom.com (mms3.broadcom.com [216.31.210.19]) by ietfa.amsl.com (Postfix) with ESMTP id 9747D11E8321; Fri,  2 Aug 2013 06:42:35 -0700 (PDT)
Received: from [10.9.208.57] by mms3.broadcom.com with ESMTP (Broadcom SMTP Relay (Email Firewall v6.5)); Fri, 02 Aug 2013 06:32:28 -0700
X-Server-Uuid: B86B6450-0931-4310-942E-F00ED04CA7AF
Received: from SJEXCHCAS06.corp.ad.broadcom.com (10.16.203.14) by IRVEXCHCAS08.corp.ad.broadcom.com (10.9.208.57) with Microsoft SMTP Server (TLS) id 14.1.438.0; Fri, 2 Aug 2013 06:42:19 -0700
Received: from SJEXCHMB12.corp.ad.broadcom.com ( [fe80::bc15:c1e1:c29a:36f7]) by SJEXCHCAS06.corp.ad.broadcom.com ( [::1]) with mapi id 14.01.0438.000; Fri, 2 Aug 2013 06:42:19 -0700
From: "Shahram Davari" <davari@broadcom.com>
To: "John E Drake" <jdrake@juniper.net>
Thread-Topic: [mpls] Comments on draft-ietf-tictoc-1588overmpls
Thread-Index: AQHOj2dsoD26UTw5FECSgmrxo5m7wJmBtd9EgAB9NYD//7pTBA==
Date: Fri, 2 Aug 2013 13:42:18 +0000
Message-ID: <BD041262-27FA-4C2B-888C-9F7DC2D99794@broadcom.com>
References: <51FB8396.9020504@cisco.com> <D2796BF4-4DDB-4816-B730-1289CAB11080@broadcom.com>, <0182DEA5604B3A44A2EE61F3EE3ED69E2076CBA4@BL2PRD0510MB349.namprd05.prod.outlook.com>
In-Reply-To: <0182DEA5604B3A44A2EE61F3EE3ED69E2076CBA4@BL2PRD0510MB349.namprd05.prod.outlook.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
MIME-Version: 1.0
X-WSS-ID: 7DE56B662L870061941-01-01
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: quoted-printable
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc-chairs@tools.ietf.org" <tictoc-chairs@tools.ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>, "draft-ietf-tictoc-1588overmpls@tools.ietf.org" <draft-ietf-tictoc-1588overmpls@tools.ietf.org>, "rtg-ads@tools.ietf.org" <rtg-ads@tools.ietf.org>, "int-ads@tools.ietf.org" <int-ads@tools.ietf.org>, "mpls-chairs@tools.ietf.org" <mpls-chairs@tools.ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 02 Aug 2013 13:42:47 -0000

Hi John

In that case I fail to understand the difference between your proposal and =
the current draft.

In both cases you need a special LSP to tell the HW to Timestamp the packet=
 and I assume you need to use RSVPTE to sign it. That is what the draft doe=
s.

Regards,
Shahram


On Aug 2, 2013, at 12:52 PM, "John E Drake" <jdrake@juniper.net> wrote:

> Shahram,
>=20
> The LSPs would be between LSRs that are 1588 capable, so they could be on=
e or more hops.
>=20
> Yours Irrespectively,
>=20
> John
>=20
>=20
>> -----Original Message-----
>> From: Shahram Davari [mailto:davari@broadcom.com]
>> Sent: Friday, August 02, 2013 3:24 AM
>> To: <stbryant@cisco.com>
>> Cc: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; draft-ietf-tictoc-
>> 1588overmpls@tools.ietf.org; John E Drake; mpls@ietf.org;
>> tictoc@ietf.org
>> Subject: Re: [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>=20
>> Hi Stewart
>>=20
>> We have already looked at similar approach. The issue with one hop LSP
>> is that not all LSRs are 1588 capable. One of the requirements is for
>> non-1588 capable routers to just switch the packet normally.
>>=20
>> Regards,
>> Shahram
>>=20
>>=20
>> On Aug 2, 2013, at 12:02 PM, "Stewart Bryant" <stbryant@cisco.com>
>> wrote:
>>=20
>>> Talking to John Drake about this, an alternative general
>>> model is for the LSP is that it is an LSP that timestamps
>>> the packet and then passes it to an application associated
>>> with the LSP at that hop.
>>>=20
>>> This has a lot of merit.
>>>=20
>>> If however we think about it some more we have a
>>> type of network service chaining going on here and
>>> so we don't need an LSP per say, because the path
>>> and instruction can be in the packet.
>>>=20
>>> In other words a general solution  to the problem
>>> is to define an LSP with the properties that the
>>> packet is passed one hop, timestamped and delivered
>>> to the associated application.
>>>=20
>>> How this is MPLS construct is used to support time
>>> tranfer is entirely within the scope of TICTOC, but
>>> at the MPLS layer we have a clean reusable network
>>> service.
>>>=20
>>> - Stewart
>>>=20
>>>=20
>>> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
>>> SB> This draft does not seem to provide a precise definition
>>> SB> the properties of the new LSP type that it wishes to
>>> SB> define, in particular it does it define the PHB of those
>>> SB> LSPs, nor the full interaction with the MPLS
>>> SB> architecture.
>>> SB>
>>> SB> I have not tracked TICTOC for a while but I thought that
>>> SB> the original plan was to define the concept of an offset
>>> SB> into a packet to do the correction.
>>> SB>
>>> SB> It is disappointing that the opportunity was not taken
>>> SB> to define a timing shim inside the timing LSP so that
>>> SB> a time correction could be added to any packet such that
>>> SB> the MPLS system was isolated from the details of the
>>> SB> complexity of the particular time transfer type.
>>>=20
>>> SB> I think that much more clarify is needed in terms of
>>> SB> definition of the new LSP type, since it is unclear
>>> SB> from this text how to implement one.
>>> SB>
>>> SB> There are a lot of other MPLS services such as
>>> SB> LSP ping that need to be considered.
>>> SB>
>>> SB> Please see inline for more comments. However these
>>> SB> comments are made in the context of the text as written
>>> SB> whilst I have a fundamental concern that this approach
>>> SB> lacks an MPLS architectural soundness that need
>>> SB> greater thought with significant impact on the
>>> SB> draft.
>>>=20
>>> - Stewart
>>>=20
>>>=20
>>> TICTOC Working Group                                           S.
>> Davari
>>> Internet-Draft                                                   A.
>> Oren
>>> Intended status: Standards Track                          Broadcom
>> Corp.
>>> Expires: December 17, 2013                                     M.
>> Bhatia
>>>                                                             P.
>> Roberts
>>>                                                         Alcatel-
>> Lucent
>>>                                                             L.
>> Montini
>>>                                                             L.
>> Martini
>>>                                                          Cisco
>> Systems
>>>                                                          June 15,
>> 2013
>>>=20
>>>=20
>>>           Transporting Timing messages over MPLS Networks
>>>                  draft-ietf-tictoc-1588overmpls-05
>>>=20
>>> Abstract
>>>=20
>>>  This document defines the method for transporting Timing messages
>>>  such as PTP and NTP over an MPLS network.  The method allows for
>> the
>>>  easy identification of these PDUs at the port level to allow for
>> port
>>>=20
>>> SB> What is a port
>>>=20
>>>  level processing of these PDUs in both LERs and LSRs.
>>>=20
>>>  The basic idea is to transport Timing messages inside dedicated
>> MPLS
>>>  LSPs.  These LSPs only carry Timing messages and possibly Control
>> and
>>>  Management packets, but they do not carry customer traffic.
>>>=20
>>> SB> More specifically they only carry traffic associated with the
>>> SB> timing service and its support.
>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
>>> SB> also it gets carried in a structure that causes it to get
>>> SB> timestamped.
>>>=20
>>>  Two methods for transporting Timing messages over MPLS are defined.
>>>=20
>>> SB> Perhaps the right approach is to define the new LSP type and then
>>> SB> seperately to define  the mapping of the various timing services
>>> SB> over that LSP type.
>>>=20
>>>  The first method is to transport Timing messages directly over the
>>>  dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
>>>  MPLS networks.  The second method is to transport Timing messages
>>>  inside a PW via Ethernet encapsulation.
>>>=20
>>> SB> I think that we should note that there are some
>>> SB> h/w reasons for this preference. A clean sheet approach
>>> SB> would have been to use PTP over MPLS with no intermediate
>>> SB> layers.
>>>=20
>>> Status of this Memo
>>>=20
>>>  This Internet-Draft is submitted in full conformance with the
>>>  provisions of BCP 78 and BCP 79.
>>>=20
>>>  Internet-Drafts are working documents of the Internet Engineering
>>>  Task Force (IETF).  Note that other groups may also distribute
>>>  working documents as Internet-Drafts.  The list of current
>> Internet-
>>>  Drafts is at http://datatracker.ietf.org/drafts/current/.
>>>=20
>>>  Internet-Drafts are draft documents valid for a maximum of six
>> months
>>>  and may be updated, replaced, or obsoleted by other documents at
>> any
>>>  time.  It is inappropriate to use Internet-Drafts as reference
>>>  material or to cite them other than as "work in progress."
>>>=20
>>>  This Internet-Draft will expire on December 17, 2013.
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013               [Page
>> 1]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> Copyright Notice
>>>=20
>>>  Copyright (c) 2013 IETF Trust and the persons identified as the
>>>  document authors.  All rights reserved.
>>>=20
>>>  This document is subject to BCP 78 and the IETF Trust's Legal
>>>  Provisions Relating to IETF Documents
>>>  (http://trustee.ietf.org/license-info) in effect on the date of
>>>  publication of this document.  Please review these documents
>>>  carefully, as they describe your rights and restrictions with
>> respect
>>>  to this document.  Code Components extracted from this document
>> must
>>>  include Simplified BSD License text as described in Section 4.e of
>>>  the Trust Legal Provisions and are provided without warranty as
>>>  described in the Simplified BSD License.
>>>=20
>>>=20
>>> Table of Contents
>>>=20
>>>  1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .
>> 5
>>>=20
>>>  2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .
>> 7
>>>=20
>>>  3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .
>> 8
>>>=20
>>>  4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .
>> 9
>>>=20
>>>  5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . .
>> 12
>>>=20
>>>  6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . .
>> 13
>>>    6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . .
>> 13
>>>    6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . .
>> 13
>>>    6.3.  Other Timing Encapsulation methods . . . . . . . . . . . .
>> 14
>>>=20
>>>  7.  Timing message Processing  . . . . . . . . . . . . . . . . . .
>> 15
>>>=20
>>>  8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . .
>> 16
>>>=20
>>>  9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
>> 17
>>>=20
>>>  10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
>> 18
>>>=20
>>>  11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . .
>> 19
>>>=20
>>>  12. OAM, Control and Management  . . . . . . . . . . . . . . . . .
>> 20
>>>=20
>>>  13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . .
>> 21
>>>=20
>>>  14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . .
>> 22
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013               [Page
>> 2]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>>  15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . .
>> 23
>>>    15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . .
>> 23
>>>    15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . .
>> 23
>>>    15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . .
>> 24
>>>=20
>>>  16. Other considerations . . . . . . . . . . . . . . . . . . . . .
>> 25
>>>=20
>>>  17. Security Considerations  . . . . . . . . . . . . . . . . . . .
>> 26
>>>=20
>>>  18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . .
>> 27
>>>=20
>>>  19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . .
>> 28
>>>=20
>>>  20. References . . . . . . . . . . . . . . . . . . . . . . . . . .
>> 29
>>>    20.1. Normative References . . . . . . . . . . . . . . . . . . .
>> 29
>>>    20.2. Informative References . . . . . . . . . . . . . . . . . .
>> 29
>>>=20
>>>  Appendix 1.  Routing extensions for Timing-aware Routers . . . . .
>> 32
>>>=20
>>>  Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . .
>> 33
>>>=20
>>>  Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . .
>> 34
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013               [Page
>> 3]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>>  The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
>>>  "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in
>> this
>>>  document are to be interpreted as described in RFC2119 [RFC2119].
>>>=20
>>>  When used in lower case, these words convey their typical use in
>>>  common language, and are not to be interpreted as described in
>>>  RFC2119 [RFC2119].
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013               [Page
>> 4]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 1.  Introduction
>>>=20
>>>  The objective of Precision Time Protocol (PTP) and Network Timing
>>>  Protocol (NTP) are to synchronize independent clocks running on
>>>  separate nodes of a distributed system.
>>>=20
>>>  [IEEE-1588] defines PTP messages for frequency, phase and time
>>>  synchronization.  The PTP messages include PTP PDUs over UDP/IP
>>>  (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F
>> of
>>>  [IEEE-1588]).
>>>=20
>>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
>>> SB> of other PTP mappings if they provide better optimisation.
>>>=20
>>>  This document defines mapping and transport of the PTP
>>>  messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>>>  defines several clock types: ordinary clocks, boundary clocks, end-
>>>  to-end transparent clocks, and peer-to-peer transparent clocks.
>>>  Transparent clocks require intermediate nodes to update correction
>>>  field inside PTP message that reflects the transit time in the
>> node.
>>>=20
>>>  [RFC5905] defines NTP messages for clock and time synchronization.
>>>  The PTP messages (PDUs) are transported over UDP/IP.  This document
>>> SB> Should that be NTP messages?
>>> SB> It needs to be made clear as soon as you introduce NTP that
>>> SB> they use different time representations.
>>>=20
>>>  defines mapping and transport of the NTP messages defined in
>>>  [RFC5905] over MPLS networks.
>>>=20
>>>  One key attribute of all of these Timing messages is that the Time
>>>  stamp processing should occur as close as possible to the actual
>>>  transmission and reception at the physical port interface.  This
>>>  targets optimal time and/or frequency recovery by avoiding variable
>>>  delay introduced by queues internal to the clocks.
>>>=20
>>> SB> As I recall NTP has no epoch point defined, and I am not sure
>>> SB> where that point is in the case of PTP in this mapping
>>> SB> Hopefully this will get defined in due course.
>>>=20
>>>  To facilitate the fast and efficient recognition of Timing messages
>>>  at the port level when the Timing messages are carried over MPLS
>>>  LSPs,
>>>=20
>>> SB> Over a new LSP type with time optimied characteristics
>>>=20
>>>  this document defines the specific encapsulations that should
>>>  be used.
>>> SB> Hopefully it will also define the PHP
>>>=20
>>>  In addition, it can be expected that there will exist LSR/
>>>  LERs where only a subset of the physical ports will have the port-
>>>  based Timing message processing capabilities.
>>> SB> Do you need to clarify that this only works at base and not in
>>> SB> a label heirarchy.
>>>=20
>>>=20
>>>  In order to ensure
>>>  that the LSPs carrying Timing packets always enter and exit ports
>>>  with this capability, routing extensions are defined to advertise
>>>  this capability on a port basis and to allow for the establishment
>> of
>>>  LSPs that only transit such ports.  While this path establishment
>>>  restriction may be applied only at the LER Ingress and/or egress
>>>  ports, it becomes more important when using transparent clock
>> capable
>>>  LSRs in the path.
>>> SB> I do not understand the implications of the last
>>> SB> sentences - starting ", it becomes"
>>>=20
>>>=20
>>>  Port based Timing message processing involves Timing message
>>>  recognition.  Once the Timing messages are recognized they can be
>>>  modified based on the reception or transmission Time-stamp.
>>>=20
>>>  This document provides two methods for transporting Timing messages
>>>  over MPLS.  One is applicable to MPLS environment and the other one
>>>  is applicable to MPLS/MPLS-TP environment
>>>=20
>>> SB> I think the sentence is incomplete.
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013               [Page
>> 5]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>>  The solution involves transporting Timing messages over dedicated
>>>  LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
>>>  carry Management and control messages, but not data plane client
>>>  traffic.
>>>=20
>>> SB> It is not clear why this restriction applies.
>>>=20
>>>  Timing LSPs can be established statically or via signaling.
>>> SB> s/statically/by provisioning/network management/
>>>=20
>>>  Extensions to control plane (OSPF, ISIS, etc.) is required to
>> enable
>>>  routers to distribute their Timing processing capabilities over
>> MPLS
>>>  to other routers.  However such extensions are outside the scope of
>>>  this document.
>>>=20
>>>  When signaling is used to setup the PTP LSP, Extensions to
>> signaling
>>> SB> is it a PTP LSP or a Timing LSP?
>>>=20
>>>  protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
>>>  However such extensions are outside the scope of this document.
>>>=20
>>> SB> for mpls-tp GMPLS is the signalling protocol
>>>=20
>>>  While the techniques included herein allow for the establishment of
>>>  paths optimized to include Time-stamping capable links, the
>>>  performance of the Slave clocks is outside the scope of this
>>>  document.
>>>=20
>>>  At the time of publishing this specification, Transparent Clocking
>>>  (TC) is only defined for PTP.  Therefore at this time any part of
>>>  this specification that talks about Transparent Clocking applies
>> only
>>>  to PTP.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013               [Page
>> 6]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 2.  Terminology
>>>=20
>>>  1588: The timing and synchronization as defined by IEEE 1588.
>>>=20
>>> SB> I think that there is a more formal name for the 1588 group
>>> SB> that needs to be used here.
>>> SB> Also do we need to talk about 1588-200? as there is
>>> SB> an update in progress
>>>=20
>>>  NTP: The timing and synchronization protocol defined by IETF RFC-
>> 1305
>>>  and RFC-5905.
>>>=20
>>>  PTP: The timing and synchronization protocol used by 1588.
>>> SB> need the proper name for 1588
>>>=20
>>>  Master Clock: The source of 1588 timing to a set of slave clocks.
>>>=20
>>>  Master Port: A port on a ordinary or boundary clock that is in
>> Master
>>>  state.  This is the source of timing toward slave ports.
>>>=20
>>> SB> I am not sure the reader knows what a port is
>>>=20
>>>  Slave Clock: A receiver of 1588 timing from a master clock.
>>>=20
>>>  Slave Port: A port on a boundary clock or ordinary clock that is
>>>  receiving timing from a master clock.
>>>=20
>>>  Ordinary Clock: A device with a single PTP port.
>>>=20
>>>  Transparent Clock.  A device that measures the time taken for a PTP
>>>  event message to transit the device and then updates the
>>>  correctionField of the message with this transit time.
>>>=20
>>>  Boundary Clock: A device with more than one PTP port.  Generally
>>>  boundary clocks will have one port in slave state to receive timing
>>>  and then other ports in master state to re-distribute the timing.
>>>=20
>>>  PTP LSP: An LSP dedicated to carry PTP messages
>>>=20
>>> SB> PTP or timing?
>>>=20
>>>=20
>>>  PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>>  messages.
>>>=20
>>> SB> Ah I don't think that PWE3 know what one of these is
>>>=20
>>>  CW: Pseudowire Control Word
>>>=20
>>>  LAG: Link Aggregation
>>>=20
>>>  ECMP: Equal Cost Multipath
>>>=20
>>>  CF: Correction Field, a field inside certain PTP messages (message
>>>  type 0-3)that holds the accumulative transit time inside
>> intermediate
>>>  switches
>>>=20
>>>  Timing messages: Timing Protocol messages that are exchanged
>> between
>>>  routers in order to establish a synchronized clock.
>>>=20
>>> SB> A number of these definitions look like copies of IEEE1588
>>> SB> definitions. We need to provide references and note the
>>> SB> priority of the IEEE base reference.
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013               [Page
>> 7]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 3.  Problem Statement
>>>=20
>>>  [IEEE-1588] has defined methods for transporting PTP messages over
>>>  Ethernet and IP networks.  [RFC5905] has defined the method of
>>>  transporting NTP messages over IP networks.  There is a need to
>>>  transport Timing messages over MPLS networks while supporting the
>>>  Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
>>>  functionality in the LER and LSRs in the MPLS network.
>>>=20
>>>  There are multiple ways of transporting Timing over MPLS.  However,
>>>  there is a requirement to limit the possible encapsulation options
>> to
>>>  simplify the Timing message identification and processing required
>> at
>>>  the port level.
>>>=20
>>>  When Timing-awareness is needed, Timing messages should not be
>>>  transported over LSPs or PWs that are carrying customer traffic
>>>  because LSRs perform Label switching based on the top label in the
>>>  stack.
>>>=20
>>> SB> Have you explained why?
>>>=20
>>>  To detect Timing messages inside such LSPs require special
>>>  hardware to do deep packet inspection at line rate.  Even if such
>>>  hardware exists, the payload can't be deterministically identified
>> by
>>>  LSRs because the payload type is a context of the PW label, and the
>>>  PW label and its context are only known to the Edge routers (PEs/
>>>  LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR,
>> CES,
>>>  etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
>>>  LSRs dont have the knowledge of whether PW Control Word (CW) is
>>>  present or not and therefore can not deterministically identify the
>>>  payload.
>>>=20
>>>  A generic method is defined in this document that does not require
>>>  deep packet inspection at line rate, and can deterministically
>>>  identify Timing messages.  This method can be used to detect Timing
>>>  Messages in both one-step and two-step clock implementations of
>>>  ordinary, boundary and transparent clocks.
>>>=20
>>> SB> Needs a ref and I am sure many MPLS specialists will not
>> understand
>>> SB> the msg types.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013               [Page
>> 8]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 4.  Timing over MPLS Architecture
>>>=20
>>>  Timing messages are exchange between Timing ports on ordinary and
>>>=20
>>> SB> Have you defined a timing port?
>>>=20
>>>  boundary clocks.  Boundary clocks terminate the Timing messages and
>>>  act as master for other boundary clocks or for slave clocks.  End-
>> to-
>>>  End Transparent clocks do not terminate the Timing messages but
>> they
>>>  do modify the contents of the Timing messages as they transit
>> across
>>>  the transparent clock.
>>>=20
>>>  Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent
>> Clock
>>>=20
>>>  (TC) could be implemented in either LERs or LSRs.
>>>=20
>>> SB> LER and LSR need to be expanded
>>>=20
>>>  An example is shown in Figure 1, where the LERs act as Ordinary
>> Clock
>>>  (OC) and are the initiating/terminating point for Timing messages.
>>>  The ingress LER encapsulates the Timing messages in Timing LSP and
>>>  the Egress LER terminates the Timing LSP.  The LSRs act as
>>>  Transparent Clock (TC) and just update the Timing field in the
>> Timing
>>>  messages.
>>>=20
>>>=20
>>>     +--------+     +-------+     +-------+     +-------+     +------
>> --+
>>>     |Switch, |     |       |     |       |     |       |
>> |Switch, |
>>>     | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----|
>> Router |
>>>     |        |     |  OC   |     |  TC   |     |  OC   |     |
>> |
>>>     +--------+     +-------+     +-------+     +-------+     +------
>> --+
>>>                    /                                 \
>>>     +-------+     /                                   \     +-------
>> +
>>>     |  LER  |    /                                     \    |  LER
>> |
>>>     | Master|---/                                       \---| Slave
>> |
>>>     | Clock |                                               | Clock
>> |
>>>     +-------+                                               +-------
>> +
>>>=20
>>>    Figure (1) - Deployment example 1 of timing over MPLS network
>>>=20
>>>  Another example is shown in Figure2, where LERs terminate the
>> Timing
>>>  messages received from switch/routers that are outside of the MPLS
>>>  network acting as OC or BC.  In this example LERs regenerate the
>>>  clock and initiate timing messages encapsulated in Timing LSP
>> toward
>>>  the MPLS network, while the LSRs act as Transparent Clock (TC) and
>>>  just update the Timing field in the Timing messages, which are
>>>  already encapsulated in Timing LSPs.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013               [Page
>> 9]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>>    +--------+     +-------+     +-------+     +-------+     +-------
>> -+
>>>    |Switch, |     |       |     |       |     |       |     |Switch,
>> |
>>>    | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router
>> |
>>>    | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC
>> |
>>>    +--------+     +-------+     +-------+     +-------+     +-------
>> -+
>>>=20
>>>    Figure (2) - Deployment example 2 of timing over MPLS network
>>>=20
>>>=20
>>>  Another example is shown in Figure 3, where LERs do not terminate
>> the
>>>  Timing messages received from switch/routers that are outside of
>> the
>>>  MPLS network acting as OC, TC or BC.  The LERs act as TC and update
>>>  the Timing field in the Timing messages as they transit the LER,
>>>  while encapsulating them in timing LSP.  The LSRs also act as
>>>  Transparent Clock (TC) and just update the Timing field in the
>> Timing
>>>  messages which are already encapsulated in Timing LSPs.
>>>=20
>>>     +--------+     +-------+     +-------+     +-------+     +------
>> --+
>>>     |Switch, |     |       |     |       |     |       |
>> |Switch, |
>>>     | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----|
>> Router |
>>>     |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |
>> |OC/TC/BC|
>>>     +--------+     +-------+     +-------+     +-------+     +------
>> --+
>>>=20
>>>   Figure (3) - Deployment example 3 of timing over MPLS network
>>>=20
>>>  Another example is shown in Figure 4, where LERs and LSRs support
>>>  Boundary Clocks.  A single-hop LSP is created between two adjacent
>>>  LSRs engaged in BC operation.  Other methods such as PTP transport
>>>  over Ethernet MAY be used for transporting timing messages if the
>>>  link between the two routers is Ethernet.
>>>=20
>>>    +--------+     +-------+     +-------+     +-------+     +-------
>> -+
>>>    |Switch, |     |       |     |       |     |       |     |Switch,
>> |
>>>    | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router
>> |
>>>    | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC
>> |
>>>    +--------+     +-------+     +-------+     +-------+     +-------
>> -+
>>>=20
>>>  Figure (4) - Deployment example 3 of timing over MPLS network
>>>=20
>>>  An MPLS domain MAY serve multiple customers.  In these cases the
>> MPLS
>>>  domain (maintained by a service provider) may provide timing
>> services
>>>  to multiple customers, each having their own Timing domain.
>>>=20
>>>  The Timing over MPLS architecture assumes full mesh of Timing LSPs
>>>  between all LERs supporting this specification.
>>>=20
>>> SB> Note sure this is right - the salves surely do not need to
>>> SB> exchange timing amongst themselves
>>>=20
>>>  It supports
>>>  Point-to- point (VPWS) and Multipoint (VPLS) services.
>>>=20
>>> SB> What does that mean? You do not carry user data traffic?
>>> SB> Maybe it's the ordering of the statemnets that is causing
>>> SB> confusion.
>>>=20
>>>  This means
>>>  that a customer may purchase a Point-to-point Timing service
>> between
>>>  two customer sites or a Multipoint Timing service between more than
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 10]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>>  two customer sites.
>>>=20
>>>  The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
>>>  This means that the Timing Multicast messages such as PTP Multicast
>>>  event messages can be transported over P2MP Timing LSP or be
>>>  replicated and transported over many P2P Timing LSPs.
>>>=20
>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>>> SB> nor a P2MP PW, although we are close.
>>>=20
>>>  Timing messages, that do not require Time stamping or Correction
>>>  Field update MAY be transported over Timing LSPs to simplify
>> hardware
>>>  and software.
>>>=20
>>>  PTP Announce messages that determine the Timing LSP terminating
>> point
>>>  behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
>>>  to simplify hardware and software.
>>>=20
>>> SB> have you defined and referenced PTP announce msgs?
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 11]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 5.  Dedicated LSPs for Timing messages
>>>=20
>>>  Many methods have been considered for identifying the Timing
>> messages
>>>  when they are encapsulated in MPLS such as using GAL/G-ACH or a new
>>>  reserved label.  These methods were not attractive since they
>> either
>>>  required deep packet inspection at line rate in the intermediate
>> LSRs
>>>  or they required use of a scarce new reserved label.  Also one of
>> the
>>>  goals was to reuse existing OAM mechanisms.
>>>=20
>>> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
>>>=20
>>>  The method defined in this document can be used by LER and LSRs to
>>>  identify Timing messages in MPLS tunnels by just looking at the top
>>>  label in the MPLS label stack, which only carry Timing messages as
>>>  well as OAM, but not data plane client traffic.
>>>=20
>>>  Compliant implementations MUST use dedicated LSPs to carry Timing
>>>  messages over MPLS.
>>>=20
>>> SB> I think that we need a definition of the properies of these LSPs
>>>=20
>>>  These LSPs are herein referred to as "Timing
>>>  LSPs" and the labels associated with these LSPs as "Timing LSP
>>>  labels".  The Timing LSPs that runs between Ingress and Egress LERs
>>>  MUST be co-routed.  Alternatively, a single bidirectional co-routed
>>>  LSP can be used.
>>>=20
>>> SB> I though that you said you could use M2MP LSPs - these are not
>>> SB> bidirectional.
>>>=20
>>>  Co-routing of the two directions is required to limit the
>> difference
>>>  in the delays in the Master clock to Slave clock direction compared
>>>  to the Slave clock to Master clock direction.  The Timing LSP MAY
>> be
>>>  MPLS/MPLS-TP LSP.
>>>=20
>>>  The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
>>>  New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
>>>  outside the scope of this document.
>>>=20
>>>  The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such
>> as
>>>  BFD and LSP Ping but the LSP data plane client plane traffic MUST
>> be
>>>  Timing packets only.
>>>=20
>>> SB> Why?
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 12]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 6.  Timing over LSP Encapsulation
>>>=20
>>> The encapsulations is not LSP is it?
>>>=20
>>>  This document defines two methods for carrying Timing messages over
>>>  MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>>  messages over Timing LSPs, and the second method, is carrying
>>>  Ethernet encapsulated Timing messages over Ethernet PWs inside
>> Timing
>>>  LSPs.
>>>=20
>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>>=20
>>>  The simplest method of transporting Timing messages over MPLS is to
>>>  encapsulate Timing PDUs in UDP/IP and then encapsulate them in
>> Timing
>>>  LSP.  This format is shown in Figure 4.
>>>=20
>>>=20
>>>                   +----------------------+
>>>                   |   Timing LSP Label   |
>>>                   +----------------------+
>>>                   |        IPv4/6        |
>>>                   +----------------------+
>>>                   |         UDP          |
>>>                   +----------------------+
>>>                   |     Timing PDU       |
>>>                   +----------------------+
>>>=20
>>>     Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>>=20
>>>=20
>>>  This encapsulation is very simple and is useful when the network
>>>  between Timing Master Clock and Slave Clock is MPLS network.
>>>=20
>>> SB> Simple is a judgement call
>>>=20
>>>  In order for an LER/LSR to process Timing messages, the Timing LSP
>>>  Label must be at the top label of the label stack.  The LER/LSR
>> MUST
>>>  know that the Timing LSP Label is used for carrying Timing
>> messages.
>>>  This can be accomplished via static configuration or via RSVP-TE
>>>  signaling.
>>>=20
>>>  The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>>  [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>>>  [RFC5905].
>>>=20
>>> 6.2.  Timing over PW Encapsulation
>>>=20
>>>  Another method of transporting Timing over MPLS networks is by
>>>  encapsulating Timing PDUs in PW which in turn is transported over
>>>  Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
>>>  shown in Fig 5(A) MUST be used and the Ethernet encapsulation of
>> PTP
>>>  MUST follow Annex F of [IEEE-1588].
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 13]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>>  The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
>> the
>>>  Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>>>  Timing over PW encapsulation MUST use the Control Word (CW) as
>>>  specified in [RFC4448] to ensure proper detection of PTP messages
>>>  inside the MPLS packets for Timing over LSP and Timing over PW
>>>  encapsulation.
>>>=20
>>> SB> That needs explanation
>>>=20
>>>  The use of Sequence Number in the CW is optional.
>>>=20
>>> SB> Given that s/n are never in practice deployed, you could probably
>>> SB> simplify things by sayig that they are not used.
>>>=20
>>>  Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over
>> PW
>>>  (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>>=20
>>>                   +----------------+  +----------------+
>>>                    |Timing LSP Label|  |Timing LSP Label|
>>>                    +----------------+  +----------------+
>>>                    |    PW Label    |  |    PW Label    |
>>>                    +----------------+  +----------------+
>>>                    |  Control Word  |  |      IP        |
>>>                    +----------------+  +----------------+
>>>                    |    Ethernet    |  |      UDP       |
>>>                    |     Header     |  +----------------+
>>>                    +----------------+  |   Timing PDU   |
>>>                    |S-VLAN(Optional)|  |                |
>>>                    +----------------+  +----------------+
>>>                    |C-VLAN(Optional)|        (B)
>>>                    +----------------+
>>>                    |   Timing PDU   |
>>>                    |                |
>>>                    +----------------+
>>>                           (A)
>>>=20
>>>             Figure (5) - Timing over PW Encapsulations
>>>=20
>>>  In order for an LSR to process PTP messages, the top label of the
>>>  label stack (the Tunnel Label) MUST be a Timing label.
>>>=20
>>> S> You said that before.
>>>=20
>>> 6.3.  Other Timing Encapsulation methods
>>>=20
>>>  In future other timing encapsulation methods may be introduced,
>> such
>>>  as a new shim header after the Bottom of Stack to carry the Timing
>>>  information.  Such new encapsulations are outside the scope of this
>>>  document.
>>>=20
>>>=20
>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>>> SB> out of the definition of the LSP
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 14]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> SB> I think we need a section on LSP processing
>>>=20
>>> 7.  Timing message Processing
>>>=20
>>>  Each Timing protocol such as PTP and NTP, define their set of
>> Timing
>>>  messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>>>  FOLLOW_UP, etc messages.
>>>=20
>>>  Some of the Timing messages require time stamping or correction
>> field
>>>  update at port level and some dont.  It is the job of the LER/LSR
>> to
>>>  parse the timing message and find out the type of the Timing
>> message
>>>  and decide whether and how to Time- stamp it (e.g., BC) or update
>>>  correction field(e.g., TC).
>>>=20
>>>=20
>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
>>> SB> function rather than the LER?
>>>=20
>>>  For example the following PTP messages (called Event messages)
>>>  require time-stamping or correction field update:
>>>=20
>>>  o  SYNC
>>>=20
>>>  o  DELAY_REQ (Delay Request)
>>>=20
>>>  o  PDELAY_REQ (Peer Delay Request)
>>>=20
>>>  o  PDELAY_RESP (Peer Delay Response)
>>>=20
>>>  SYNC and DELAY_REQ are exchanged between Master Clock and Slave
>> Clock
>>>  and MUST be transported over PTP LSPs.  PDELAY_REQ and PDELAY_RESP
>>>  are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>>>  Boundary, or Transparent) and SHOULD be transported over single hop
>>>  PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
>>>  and PDELAY_RESP_FOLLOW_UP messages MUST also be transported over
>> the
>>>  PTP LSPs.
>>>=20
>>>  For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
>>>  transported over two PTP LSPs that are in opposite directions.
>> These
>>>  PTP LSPs, which are in opposite directions MUST be congruent and
>> co-
>>>  routed.  Alternatively, a single bidirectional co-routed LSP can be
>>>  used.
>>>=20
>>>  Except as indicated above for the two-step PTP clocks, Non-Event
>> PTP
>>>  message types do not need to be processed by intermediate routers.
>>>  These message types MAY be carried in PTP Tunnel LSPs.
>>>=20
>>> SB> Are you saying that a timing P router has to be msg type
>> sensitive?
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 15]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 8.  Protection and Redundancy
>>>=20
>>>=20
>>> SB> This is a bit of a jump - I don't know how the LSP itself works
>> yet!
>>>=20
>>>  In order to ensure continuous uninterrupted operation of slave
>>>  clocks, usually as a general practice, slave clocks (or ports)
>> track
>>>  redundant master clocks.
>>>=20
>>>  It is the responsibility of the network operator to ensure that
>>>  physically disjoint Timing LSPs are established between a slave
>> clock
>>>  (or port) and redundant master clocks (or ports).
>>>=20
>>>  When a slave clock (or port) listens to redundant master clocks or
>>>  ports, any prolonged Timing LSP outage will trigger the slave clock
>>>  or port to switch to a redundant master clock or port.
>>>=20
>>>  LSP/PW protection such as Linear protection Switching (1:1, 1+1),
>>>  Ring protection switching or MPLS Fast Reroute (FRR) generally
>> switch
>>>  alternative path that usually cause a change in delay, which if
>>>  undetected by slave clock can reduce accuracy of the slave clock.
>>>=20
>>>  Therefore protection switching MAY be used, as long as phase jumps
>>>  upon switchover due to differences in path latency are detected and
>>>  compensated for (such compensation not being required if BCs or
>> peer-
>>>  peer TCs are used throughout).
>>>=20
>>>  Note that any protection or reroute mechanism that adds additional
>>>  MPLS label to the label stack, such as Facility Backup Fast
>> Reroute,
>>>  MUST ensure that the pushed label is also a Timing Label to ensure
>>>  recognition of the MPLS frame as containing Timing messages, as it
>>>  transits the backup path.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 16]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 9.  ECMP
>>>=20
>>>  To ensure the optimal operation of slave clocks and avoid error
>>>  introduced by forward and reverse path delay asymmetry, the
>> physical
>>>  path for Timing messages from master clock to slave Clock and vice
>>>  versa must be the same for all Event Timing messages listed in
>>>  section 7.
>>>=20
>>>  Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
>>>  Multipath).
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 17]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 10.  PHP
>>>=20
>>>  To ensure that the label on the top of the label stack is the
>> Timing
>>>  LSP Label, PHP MUST not be used.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 18]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 11.  Entropy
>>>=20
>>>  To ensure all Timing messages in a Timing LSP take the same path,
>>>  Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>>  Entropy Label MUST NOT be used for the PWs that are carried inside
>>>  Timing LSP [RFC6391].
>>>=20
>>> SB> This is incorrect - you mean that all msgs of the same timing
>>> SB> flow need to have the same EL value.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 19]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 12.  OAM, Control and Management
>>>=20
>>>  In order to monitor Timing LSPs and their encapsulated PWs, they
>> MUST
>>>  be able to carry OAM and management messages.  These management
>>>  messages MUST be differentiated from Timing messages via already
>>>  defined IETF methods.
>>>=20
>>>  For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
>>>  over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>>>  Management protocols can easily be identified by the UDP
>> Destination
>>>  Port number or by GAL/G-ACH respectively.
>>>=20
>>>  Also BFD, LSP-Ping and other management messages MAY run over the
>> PWs
>>>  encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3
>> or
>>>  4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is
>> going
>>>  to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1) or
>>>  GAL-ACH are used to identify such management messages.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 20]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 13.  QoS Considerations
>>>=20
>>>  In network deployments where not every LSR/LER is Timing-aware, it
>> is
>>>  important to reduce the impact of the non-Timing-aware LSR/LERs on
>>>  the timing recovery in the slave clock.  The Timing messages are
>> time
>>>  critical and must be treated with the highest priority.  Therefore
>>>  Timing over MPLS messages must be treated with the highest priority
>>>  in the routers.  This can be achieved by proper setup of Timing
>> LSPs.
>>>=20
>>>  It is recommended that the Timing LSPs are setup or configured
>>>  properly to indicate EF-PHB [RFC3246]for the CoS and Green
>> [RFC2697]
>>>  for drop eligibility.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 21]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 14.  FCS and Checksum Recalculation
>>>=20
>>>  When time-stamp generation and timing packet adjustment is
>> performed
>>>  near the physical port hardware, the process MUST include
>>>  recalculation of the Ethernet FCS.
>>>=20
>>> SB> The above is confusing - an LSR always recomputes the link layer
>>> SB> CRC which may or may not be Ethernet.
>>>=20
>>>  Also FCS retention for the
>>>  payload Ethernet described in [RFC4720] MUST NOT be used.
>>>=20
>>>  For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
>>>  may be required as per UDP transport standards.
>>>=20
>>> SB> You really need to be working on getting the IPv6 C?S computation
>>> SB> removed from PTP msgs.
>>>=20
>>>  When UDP checksum is used, each Timing-aware LER/LSR must either
>>>  incrementally update the UDP checksum after Time stamping or
>>>  Correction Field update or verify the UDP checksum on reception
>> from
>>>  upstream and recalculate the checksum completely on transmission to
>>>  downstream node after Time stamping or Correction Field update.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 22]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 15.  Behavior of LER/LSR
>>>=20
>>>  Timing-capable/aware LERs and LSRs are routers that have one or
>> more
>>>=20
>>> SB> You mean physical interfaces?
>>>=20
>>>  interfaces that can perform Timing operations (OC/BC/TC) on Timing
>>>  packets and are configured to do so.  Timing-capable/aware LERs and
>>>  LSRs can advertise their Timing-capability per-interface via
>> control
>>>  plane such as OSPF or IS-IS.
>>> SB> ISIS and OSPF are routing protocols.
>>>=20
>>> The Timing-capable/aware LERs can then
>>>  signals Timing LSPs via RSVP-TE signaling.  Alternatively the
>> Timing
>>>  capability of LER and LSRs may be configured in a centralized
>>>  controller and the Timing LSP may be setup using manual
>> configuration
>>>  or other methods such as SDN.
>>>=20
>>> SB> it can also be configured individually rather then through
>>> SB> a cebtral controllwe
>>>=20
>>> 15.1.  Behavior of Timing-capable/aware LER
>>>=20
>>>  When a Timing-capable/aware LER behaves as a Transparent clock and
>>>  receives a Timing message from a Timing-capable/aware non-MPLS
>>>  interface, the LER updates the Correction Field (CF) and
>> encapsulates
>>>  and forwards the timing message over previously established Timing
>>>  LSP.
>>>=20
>>> SB> You need to call out the details so that people properly
>>> SB> understand the definition of the new LSP.
>>>=20
>>>  Also when a Timing message is received from a Timing-capable/
>>>  aware MPLS interface, LER updates the Correction Filed (CF) and
>>>  decapsulates the MPLS encapsulation and forwards the timing message
>>>  to a non-MPLS interface.
>>>=20
>>>  When a Timing-capable/aware LER behaves as a Boundary clock and
>>>  receives a Timing message from a Timing-capable/aware non MPLS
>>>  interface, the LER Timestamps the Timing packet and sends it to the
>>>  LERs Boundary clock processing module.  Also when a Timing message
>> is
>>>  received from a Timing- capable/aware MPLS interface, the LER
>>>  Timestamps the Timing packet and sends it to the LERs Boundary
>> clock
>>>  processing module.
>>>=20
>>>  When a Timing-capable/aware LER behaves as an Ordinary Clock toward
>>>  the MPLS network, and receives a Timing message from a Timing-
>>>  capable/aware MPLS interface, the LER Timestamps the Timing packet
>>>  and sends it to the LERs Ordinary clock processing module.
>>>=20
>>> 15.2.  Behavior of Timing-capable/aware LSR
>>>=20
>>>  When a Timing-capable/aware LSR behaves as a Transparent clock and
>>>  receives a Timing message from a Timing-capable/aware MPLS
>> interface,
>>>  The LSR updates the Correction Filed (CF) and forwards the timing
>>>  message over another MPLS interface.
>>>=20
>>>  When a Timing-capable/aware LSR behaves as a Boundary clock and
>>>  receives a Timing message from a Timing-capable/aware MPLS
>> interface.
>>>  The LSR performs the functions of a Boundary Clock in terminating
>> the
>>>  received Timing message and re-generating a new timing message over
>>>  another (or the same) MPLS interface.
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 23]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>>=20
>>>  It is most beneficial when all LSRs in the path of a Timing LSP be
>>>  timing-Capable/aware LSRs.  This would ensure the highest quality
>>>  time and clock synchronization by Timing Slave Clocks.  However,
>> this
>>>  specification does not mandate that all LSRs in path of a Timing
>> LSP
>>>  be Timing- capable/aware.
>>>=20
>>>  Non-Timing-capable/aware LSRs just switch the packets encapsulated
>> in
>>>  Timing LSPs and dont perform any Timing operation (TC or BC).
>>>  However as explained in QoS section the Timing over MPLS packets
>> MUST
>>>  be still be treated with the highest priority based on their
>> Traffic
>>>  Class (TC) marking.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 24]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 16.  Other considerations
>>>=20
>>>  [IEEE-1588] defines an optional peer-to-peer Transparent clocking
>>>  that requires peer delay measurement between two adjacent Timing-
>>>  capable/ aware routers/switches.  Peer delay measurement messages
>>>  need to be time stamped and terminated by the Timing-capable/aware
>>>  routers/ switches.  This means that two adjacent LSRs may be
>> engaged
>>>  in a peer delay measurement.
>>>=20
>>>  For transporting such peer delay measurement messages a single-hop
>>>  LSP SHOULD to be created between the two adjacent LSRs engaged in
>>>  peer delay measurement to carry peer delay measurement messages.
>>>  Other methods such as PTP transport over Ethernet MAY be used for
>>>  transporting peer delay measurement messages if the link between
>> the
>>>  two routers is Ethernet.
>>>=20
>>>  In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/
>> ware
>>>  routers/switches MUST maintain a list of all the neighbors it needs
>>>  to send a PDelay_Req to, where each neighbor corresponds to a
>> timing
>>>  LSP.
>>>=20
>>>  The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as
>> long
>>>  as either the Explicit Null label is the bottom of stack label
>>>  (applicable only to UDP/IP encapsulation) or the label below the
>>>  Explicit Null label is a PTP label.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 25]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 17.  Security Considerations
>>>=20
>>>  MPLS PW security considerations in general are discussed in
>> [RFC3985]
>>>  and [RFC4447],and those considerations also apply to this document.
>>>=20
>>>  An experimental security protocol is defined in [IEEE-1588].The PTP
>>>  security extension and protocol provides group source
>> authentication,
>>>  message integrity, and replay attack protection for PTP messages.
>>>=20
>>>  When the MPLS network (provider network) serves multiple customers,
>>>  it is important to maintain and process each customers clock and
>>>  Timing messages separately from other customers to ensure there is
>> no
>>>  cross- customer effect.  For example if an LER BC is synchronized
>> to
>>>  a specific grandmaster, belonging to customer A, then the LER MUST
>>>  use that BC clock only for customer A to ensure that customer A
>>>  cannot attack other customers by manipulating its time.
>>>=20
>>>  Timing messages MAY be encrypted or authenticated, provided that
>> the
>>>  LERs/LSRs that are Timing capable/aware can authenticate/ decrypt
>> the
>>>  timing messages.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 26]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 18.  Acknowledgements
>>>=20
>>>  The authors would like to thank Ron Cohen, Yaakov Stein, Tal
>> Mizrahi,
>>>  Stefano Ruffini, Peter Meyer, and other members of IETF for
>> reviewing
>>>  and providing feedback on this draft.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 27]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 19.  IANA Considerations
>>>=20
>>>  There are no IANA requirements in this specification.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 28]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 20.  References
>>>=20
>>> 20.1.  Normative References
>>>=20
>>>  [IEEE-1588]
>>>             IEEE 1588-2008, "IEEE Standard for a Precision Clock
>>>             Synchronization Protocol for Networked Measurement and
>>>             Control Systems".
>>>=20
>>>  [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>>             Requirement Levels", BCP 14, RFC 2119, March 1997.
>>>=20
>>>  [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
>>>             Edge (PWE3) Architecture", RFC 3985, March 2005.
>>>=20
>>>  [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor
>> Discovery
>>>             Proxies (ND Proxy)", RFC 4389, April 2006.
>>>=20
>>>  [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
>>>             Heron, "Pseudowire Setup and Maintenance Using the Label
>>>             Distribution Protocol (LDP)", RFC 4447, April 2006.
>>>=20
>>>  [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>>             "Encapsulation Methods for Transport of Ethernet over
>> MPLS
>>>             Networks", RFC 4448, April 2006.
>>>=20
>>>  [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>>             Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>>             Retention", RFC 4720, November 2006.
>>>=20
>>>  [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
>>>             Connectivity Verification (VCCV): A Control Channel for
>>>             Pseudowires", RFC 5085, December 2007.
>>>=20
>>>  [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding
>> Detection
>>>             (BFD)", RFC 5880, June 2010.
>>>=20
>>>  [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
>>>             "Bidirectional Forwarding Detection (BFD) for MPLS Label
>>>             Switched Paths (LSPs)", RFC 5884, June 2010.
>>>=20
>>> 20.2.  Informative References
>>>=20
>>>  [I-D.ietf-pwe3-fat-pw]
>>>             Bryant, S., Filsfils, C., Drafz, U., Kompella, V.,
>> Regan,
>>>             J., and S. Amante, "Flow Aware Transport of Pseudowires
>>>             over an MPLS Packet Switched Network",
>>>             draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 29]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>>  [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
>>>             system routeing information exchange protocol for use in
>>>             conjunction with the Protocol for providing the
>>>             Connectionless-mode Network Service (ISO 8473)".
>>>=20
>>>  [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
>>>             dual environments", RFC 1195, December 1990.
>>>=20
>>>  [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
>>>=20
>>>  [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
>>>             Marker", RFC 2697, September 1999.
>>>=20
>>>  [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le
>> Boudec,
>>>             J., Courtney, W., Davari, S., Firoiu, V., and D.
>>>             Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>>             Behavior)", RFC 3246, March 2002.
>>>=20
>>>  [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic
>> Engineering
>>>             (TE) Extensions to OSPF Version 2", RFC 3630,
>>>             September 2003.
>>>=20
>>>  [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
>>>             System (IS-IS) Extensions for Traffic Engineering (TE)",
>>>             RFC 3784, June 2004.
>>>=20
>>>  [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
>>>             Shaffer, "Extensions to OSPF for Advertising Optional
>>>             Router Capabilities", RFC 4970, July 2007.
>>>=20
>>>  [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
>>>             System to Intermediate System (IS-IS) Extensions for
>>>             Advertising Router Information", RFC 4971, July 2007.
>>>=20
>>>  [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
>>>             Topology (MT) Routing in Intermediate System to
>>>             Intermediate Systems (IS-ISs)", RFC 5120, February 2008.
>>>=20
>>>  [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>>             Engineering", RFC 5305, October 2008.
>>>=20
>>>  [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>>             "Traffic Engineering Extensions to OSPF Version 3",
>>>             RFC 5329, September 2008.
>>>=20
>>>  [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
>>>             for IPv6", RFC 5340, July 2008.
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 30]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>>  [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch,
>> "Network
>>>             Time Protocol Version 4: Protocol and Algorithms
>>>             Specification", RFC 5905, June 2010.
>>>=20
>>>  [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V.,
>> Regan,
>>>             J., and S. Amante, "Flow-Aware Transport of Pseudowires
>>>             over an MPLS Packet Switched Network", RFC 6391,
>>>             November 2011.
>>>=20
>>>  [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
>>>             L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
>>>             RFC 6790, November 2012.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 31]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 1.  Routing extensions for Timing-aware Routers
>>>=20
>>>  MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340]
>> and
>>>  IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering
>> (TE)
>>>  link information used for constraint-based routing.
>>>=20
>>>  Indeed, it is useful to advertise data plane TE router link
>>>  capabilities, such as the capability for a router to be Timing-
>> aware.
>>>  This capability MUST then be taken into account during path
>>>  computation to prefer or even require links that advertise
>> themselves
>>>  as Timing-aware.  In this way the path can ensure the entry and
>> exit
>>>  points into the LERs and, if desired, the links into the LSRs are
>>>  able to perform port based time-stamping thus minimizing their
>> impact
>>>  on the performance of the slave clock.
>>>=20
>>>  extensions are required to OSPF and IS-IS in order to advertise
>>>  Timing-aware capabilities of a link.  Such extensions are outside
>> the
>>>  scope of this document; however such extension SHOULD be able to
>>>  signal the following information per Router Link:
>>>=20
>>>  o  Capable of processing PTP, NTP or other Timing flows
>>>=20
>>>  o  Capable of performing Transparent Clock operation
>>>=20
>>>  o  Capable of performing Boundary Clock operation
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 32]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> 2.  Signaling Extensions for Creating Timing LSPs
>>>=20
>>>  RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-
>> TE
>>>  is used to setup Timing LSPs, some information that indicates that
>>>  the LSP is carrying Timing flows MUST be included in the new
>>>  Extensions to RSVP-TE:
>>>=20
>>>  The following information MAY also be included in the new
>> Extensions
>>>  to RSVP-TE:
>>>=20
>>>  o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
>>>     field
>>>=20
>>>  o  Number of VLANs in case of PW encapsulation
>>>=20
>>>  o  Timestamp field Type
>>>=20
>>>     *  Correction Field, Timestamp
>>>=20
>>>  o  Timestamp Field format
>>>=20
>>>     *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
>>>        NTP, etc.
>>>=20
>>>  Note that in case the above optional information is signaled with
>>>  RSVP-TE for a Timing LSP, all the Timing packets carried in that
>> LSP
>>>  must have the same signaled characteristics.  For example if
>>>  Timestamp format is signaled as 64-bit PTPv1, then all Timing
>> packets
>>>  must use 64-bit PTPv1 time-stamp.
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 33]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>> Authors' Addresses
>>>=20
>>>  Shahram Davari
>>>  Broadcom Corp.
>>>  San Jose, CA  95134
>>>  USA
>>>=20
>>>  Email: davari@broadcom.com
>>>=20
>>>=20
>>>  Amit Oren
>>>  Broadcom Corp.
>>>  San Jose, CA  95134
>>>  USA
>>>=20
>>>  Email: amito@broadcom.com
>>>=20
>>>=20
>>>  Manav Bhatia
>>>  Alcatel-Lucent
>>>  Bangalore,
>>>  India
>>>=20
>>>  Email: manav.bhatia@alcatel-lucent.com
>>>=20
>>>=20
>>>  Peter Roberts
>>>  Alcatel-Lucent
>>>  Kanata,
>>>  Canada
>>>=20
>>>  Email: peter.roberts@alcatel-lucent.com
>>>=20
>>>=20
>>>  Laurent Montini
>>>  Cisco Systems
>>>  San Jose CA
>>>  USA
>>>=20
>>>  Email: lmontini@cisco.com
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 34]
>=20
>>> Internet-Draft        Transporting Timing over MPLS            June
>> 2013
>>>=20
>>>=20
>>>  Luca
>>>  Cisco Systems
>>>  San Jose CA
>>>  USA
>>>=20
>>>  Email: lmartini@cisco.com
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>>=20
>>> Davari, et al.          Expires December 17, 2013              [Page
>> 35]
>=20
>>>=20
>>> --
>>> For corporate legal information go to:
>>>=20
>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>=20
>>> _______________________________________________
>>> mpls mailing list
>>> mpls@ietf.org
>>> https://www.ietf.org/mailman/listinfo/mpls
>=20
>=20
>=20
>=20


From gregimirsky@gmail.com  Fri Aug  2 09:05:12 2013
Return-Path: <gregimirsky@gmail.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 24FCB21E80AA; Fri,  2 Aug 2013 09:05:12 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.599
X-Spam-Level: 
X-Spam-Status: No, score=-2.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HTML_MESSAGE=0.001, NO_RELAYS=-0.001]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 2HTO-amJs1bu; Fri,  2 Aug 2013 09:05:03 -0700 (PDT)
Received: from mail-ve0-x234.google.com (mail-ve0-x234.google.com [IPv6:2607:f8b0:400c:c01::234]) by ietfa.amsl.com (Postfix) with ESMTP id 2DE6B11E80EC; Fri,  2 Aug 2013 09:05:03 -0700 (PDT)
Received: by mail-ve0-f180.google.com with SMTP id pb11so862794veb.39 for <multiple recipients>; Fri, 02 Aug 2013 09:04:58 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; bh=2DRL4dvauohU652vU+8NK+DlsZbcPu6JM4Z0Ix83fQg=; b=WfVI9PplvzYUYOE3UzMq8uHom7xyHhma5rcWPwM+60RV0eTWkvXQtw1vAXIz/h8Pay IDo77UIlS6dsB2mAubiwE4+ZTUaZ37+tVgtRCyddtKWzaSKP17EybK1wiNxFCEVchPm2 +/rrYXEW/f3/FalmDZMTOeoA8rCHa27qjZZenusQGz3kjts2kYNtUKxppJYM7Gc4zCSJ x0DgFOD93JBpz4cdeCBO8tjXsfwFlnu0qeiG1ERREBAXX8JtkfvLUZ2OpjLoSwGZw8o3 2fSjOLtlgs20dKXoY2fYZJFHUyiwpkQYAf2AIDI3cYDOLrUMNAqdFcbDwdJMonR6zJG8 plkA==
MIME-Version: 1.0
X-Received: by 10.52.37.232 with SMTP id b8mr1829445vdk.80.1375459498670; Fri, 02 Aug 2013 09:04:58 -0700 (PDT)
Received: by 10.220.159.143 with HTTP; Fri, 2 Aug 2013 09:04:58 -0700 (PDT)
Received: by 10.220.159.143 with HTTP; Fri, 2 Aug 2013 09:04:58 -0700 (PDT)
In-Reply-To: <BD041262-27FA-4C2B-888C-9F7DC2D99794@broadcom.com>
References: <51FB8396.9020504@cisco.com> <D2796BF4-4DDB-4816-B730-1289CAB11080@broadcom.com> <0182DEA5604B3A44A2EE61F3EE3ED69E2076CBA4@BL2PRD0510MB349.namprd05.prod.outlook.com> <BD041262-27FA-4C2B-888C-9F7DC2D99794@broadcom.com>
Date: Fri, 2 Aug 2013 18:04:58 +0200
Message-ID: <CA+RyBmX4CF+Lo1_EV8_+T-Gn8bUJnfgeDf2dLvxUmLCfpnmrZA@mail.gmail.com>
From: Greg Mirsky <gregimirsky@gmail.com>
To: Shahram Davari <davari@broadcom.com>
Content-Type: multipart/alternative; boundary=bcaec51d2c10d243ea04e2f91efd
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc-chairs@tools.ietf.org" <tictoc-chairs@tools.ietf.org>, tictoc@ietf.org, "draft-ietf-tictoc-1588overmpls@tools.ietf.org" <draft-ietf-tictoc-1588overmpls@tools.ietf.org>, "rtg-ads@tools.ietf.org" <rtg-ads@tools.ietf.org>, int-ads@tools.ietf.org, "mpls-chairs@tools.ietf.org" <mpls-chairs@tools.ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 02 Aug 2013 16:05:12 -0000

--bcaec51d2c10d243ea04e2f91efd
Content-Type: text/plain; charset=ISO-8859-1

Hi Shahram,
I think that John suggested creation of LSPs interconnecting all
1588-capable nodes. It is absolutely not what current document defines but
that reminds me of what I've suggested around November of 2011.
Regards,
Greg
On Aug 2, 2013 6:43 AM, "Shahram Davari" <davari@broadcom.com> wrote:

> Hi John
>
> In that case I fail to understand the difference between your proposal and
> the current draft.
>
> In both cases you need a special LSP to tell the HW to Timestamp the
> packet and I assume you need to use RSVPTE to sign it. That is what the
> draft does.
>
> Regards,
> Shahram
>
>
> On Aug 2, 2013, at 12:52 PM, "John E Drake" <jdrake@juniper.net> wrote:
>
> > Shahram,
> >
> > The LSPs would be between LSRs that are 1588 capable, so they could be
> one or more hops.
> >
> > Yours Irrespectively,
> >
> > John
> >
> >
> >> -----Original Message-----
> >> From: Shahram Davari [mailto:davari@broadcom.com]
> >> Sent: Friday, August 02, 2013 3:24 AM
> >> To: <stbryant@cisco.com>
> >> Cc: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
> >> ads@tools.ietf.org; rtg-ads@tools.ietf.org; draft-ietf-tictoc-
> >> 1588overmpls@tools.ietf.org; John E Drake; mpls@ietf.org;
> >> tictoc@ietf.org
> >> Subject: Re: [mpls] Comments on draft-ietf-tictoc-1588overmpls
> >>
> >> Hi Stewart
> >>
> >> We have already looked at similar approach. The issue with one hop LSP
> >> is that not all LSRs are 1588 capable. One of the requirements is for
> >> non-1588 capable routers to just switch the packet normally.
> >>
> >> Regards,
> >> Shahram
> >>
> >>
> >> On Aug 2, 2013, at 12:02 PM, "Stewart Bryant" <stbryant@cisco.com>
> >> wrote:
> >>
> >>> Talking to John Drake about this, an alternative general
> >>> model is for the LSP is that it is an LSP that timestamps
> >>> the packet and then passes it to an application associated
> >>> with the LSP at that hop.
> >>>
> >>> This has a lot of merit.
> >>>
> >>> If however we think about it some more we have a
> >>> type of network service chaining going on here and
> >>> so we don't need an LSP per say, because the path
> >>> and instruction can be in the packet.
> >>>
> >>> In other words a general solution  to the problem
> >>> is to define an LSP with the properties that the
> >>> packet is passed one hop, timestamped and delivered
> >>> to the associated application.
> >>>
> >>> How this is MPLS construct is used to support time
> >>> tranfer is entirely within the scope of TICTOC, but
> >>> at the MPLS layer we have a clean reusable network
> >>> service.
> >>>
> >>> - Stewart
> >>>
> >>>
> >>> ============
> >>> SB> This draft does not seem to provide a precise definition
> >>> SB> the properties of the new LSP type that it wishes to
> >>> SB> define, in particular it does it define the PHB of those
> >>> SB> LSPs, nor the full interaction with the MPLS
> >>> SB> architecture.
> >>> SB>
> >>> SB> I have not tracked TICTOC for a while but I thought that
> >>> SB> the original plan was to define the concept of an offset
> >>> SB> into a packet to do the correction.
> >>> SB>
> >>> SB> It is disappointing that the opportunity was not taken
> >>> SB> to define a timing shim inside the timing LSP so that
> >>> SB> a time correction could be added to any packet such that
> >>> SB> the MPLS system was isolated from the details of the
> >>> SB> complexity of the particular time transfer type.
> >>>
> >>> SB> I think that much more clarify is needed in terms of
> >>> SB> definition of the new LSP type, since it is unclear
> >>> SB> from this text how to implement one.
> >>> SB>
> >>> SB> There are a lot of other MPLS services such as
> >>> SB> LSP ping that need to be considered.
> >>> SB>
> >>> SB> Please see inline for more comments. However these
> >>> SB> comments are made in the context of the text as written
> >>> SB> whilst I have a fundamental concern that this approach
> >>> SB> lacks an MPLS architectural soundness that need
> >>> SB> greater thought with significant impact on the
> >>> SB> draft.
> >>>
> >>> - Stewart
> >>>
> >>>
> >>> TICTOC Working Group                                           S.
> >> Davari
> >>> Internet-Draft                                                   A.
> >> Oren
> >>> Intended status: Standards Track                          Broadcom
> >> Corp.
> >>> Expires: December 17, 2013                                     M.
> >> Bhatia
> >>>                                                             P.
> >> Roberts
> >>>                                                         Alcatel-
> >> Lucent
> >>>                                                             L.
> >> Montini
> >>>                                                             L.
> >> Martini
> >>>                                                          Cisco
> >> Systems
> >>>                                                          June 15,
> >> 2013
> >>>
> >>>
> >>>           Transporting Timing messages over MPLS Networks
> >>>                  draft-ietf-tictoc-1588overmpls-05
> >>>
> >>> Abstract
> >>>
> >>>  This document defines the method for transporting Timing messages
> >>>  such as PTP and NTP over an MPLS network.  The method allows for
> >> the
> >>>  easy identification of these PDUs at the port level to allow for
> >> port
> >>>
> >>> SB> What is a port
> >>>
> >>>  level processing of these PDUs in both LERs and LSRs.
> >>>
> >>>  The basic idea is to transport Timing messages inside dedicated
> >> MPLS
> >>>  LSPs.  These LSPs only carry Timing messages and possibly Control
> >> and
> >>>  Management packets, but they do not carry customer traffic.
> >>>
> >>> SB> More specifically they only carry traffic associated with the
> >>> SB> timing service and its support.
> >>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
> >>> SB> also it gets carried in a structure that causes it to get
> >>> SB> timestamped.
> >>>
> >>>  Two methods for transporting Timing messages over MPLS are defined.
> >>>
> >>> SB> Perhaps the right approach is to define the new LSP type and then
> >>> SB> seperately to define  the mapping of the various timing services
> >>> SB> over that LSP type.
> >>>
> >>>  The first method is to transport Timing messages directly over the
> >>>  dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
> >>>  MPLS networks.  The second method is to transport Timing messages
> >>>  inside a PW via Ethernet encapsulation.
> >>>
> >>> SB> I think that we should note that there are some
> >>> SB> h/w reasons for this preference. A clean sheet approach
> >>> SB> would have been to use PTP over MPLS with no intermediate
> >>> SB> layers.
> >>>
> >>> Status of this Memo
> >>>
> >>>  This Internet-Draft is submitted in full conformance with the
> >>>  provisions of BCP 78 and BCP 79.
> >>>
> >>>  Internet-Drafts are working documents of the Internet Engineering
> >>>  Task Force (IETF).  Note that other groups may also distribute
> >>>  working documents as Internet-Drafts.  The list of current
> >> Internet-
> >>>  Drafts is at http://datatracker.ietf.org/drafts/current/.
> >>>
> >>>  Internet-Drafts are draft documents valid for a maximum of six
> >> months
> >>>  and may be updated, replaced, or obsoleted by other documents at
> >> any
> >>>  time.  It is inappropriate to use Internet-Drafts as reference
> >>>  material or to cite them other than as "work in progress."
> >>>
> >>>  This Internet-Draft will expire on December 17, 2013.
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page
> >> 1]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> Copyright Notice
> >>>
> >>>  Copyright (c) 2013 IETF Trust and the persons identified as the
> >>>  document authors.  All rights reserved.
> >>>
> >>>  This document is subject to BCP 78 and the IETF Trust's Legal
> >>>  Provisions Relating to IETF Documents
> >>>  (http://trustee.ietf.org/license-info) in effect on the date of
> >>>  publication of this document.  Please review these documents
> >>>  carefully, as they describe your rights and restrictions with
> >> respect
> >>>  to this document.  Code Components extracted from this document
> >> must
> >>>  include Simplified BSD License text as described in Section 4.e of
> >>>  the Trust Legal Provisions and are provided without warranty as
> >>>  described in the Simplified BSD License.
> >>>
> >>>
> >>> Table of Contents
> >>>
> >>>  1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .
> >> 5
> >>>
> >>>  2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .
> >> 7
> >>>
> >>>  3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .
> >> 8
> >>>
> >>>  4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .
> >> 9
> >>>
> >>>  5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . .
> >> 12
> >>>
> >>>  6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . .
> >> 13
> >>>    6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . .
> >> 13
> >>>    6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . .
> >> 13
> >>>    6.3.  Other Timing Encapsulation methods . . . . . . . . . . . .
> >> 14
> >>>
> >>>  7.  Timing message Processing  . . . . . . . . . . . . . . . . . .
> >> 15
> >>>
> >>>  8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . .
> >> 16
> >>>
> >>>  9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
> >> 17
> >>>
> >>>  10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
> >> 18
> >>>
> >>>  11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . .
> >> 19
> >>>
> >>>  12. OAM, Control and Management  . . . . . . . . . . . . . . . . .
> >> 20
> >>>
> >>>  13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . .
> >> 21
> >>>
> >>>  14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . .
> >> 22
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page
> >> 2]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>>  15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . .
> >> 23
> >>>    15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . .
> >> 23
> >>>    15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . .
> >> 23
> >>>    15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . .
> >> 24
> >>>
> >>>  16. Other considerations . . . . . . . . . . . . . . . . . . . . .
> >> 25
> >>>
> >>>  17. Security Considerations  . . . . . . . . . . . . . . . . . . .
> >> 26
> >>>
> >>>  18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . .
> >> 27
> >>>
> >>>  19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . .
> >> 28
> >>>
> >>>  20. References . . . . . . . . . . . . . . . . . . . . . . . . . .
> >> 29
> >>>    20.1. Normative References . . . . . . . . . . . . . . . . . . .
> >> 29
> >>>    20.2. Informative References . . . . . . . . . . . . . . . . . .
> >> 29
> >>>
> >>>  Appendix 1.  Routing extensions for Timing-aware Routers . . . . .
> >> 32
> >>>
> >>>  Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . .
> >> 33
> >>>
> >>>  Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . .
> >> 34
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page
> >> 3]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>>  The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
> >>>  "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in
> >> this
> >>>  document are to be interpreted as described in RFC2119 [RFC2119].
> >>>
> >>>  When used in lower case, these words convey their typical use in
> >>>  common language, and are not to be interpreted as described in
> >>>  RFC2119 [RFC2119].
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page
> >> 4]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 1.  Introduction
> >>>
> >>>  The objective of Precision Time Protocol (PTP) and Network Timing
> >>>  Protocol (NTP) are to synchronize independent clocks running on
> >>>  separate nodes of a distributed system.
> >>>
> >>>  [IEEE-1588] defines PTP messages for frequency, phase and time
> >>>  synchronization.  The PTP messages include PTP PDUs over UDP/IP
> >>>  (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F
> >> of
> >>>  [IEEE-1588]).
> >>>
> >>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
> >>> SB> of other PTP mappings if they provide better optimisation.
> >>>
> >>>  This document defines mapping and transport of the PTP
> >>>  messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
> >>>  defines several clock types: ordinary clocks, boundary clocks, end-
> >>>  to-end transparent clocks, and peer-to-peer transparent clocks.
> >>>  Transparent clocks require intermediate nodes to update correction
> >>>  field inside PTP message that reflects the transit time in the
> >> node.
> >>>
> >>>  [RFC5905] defines NTP messages for clock and time synchronization.
> >>>  The PTP messages (PDUs) are transported over UDP/IP.  This document
> >>> SB> Should that be NTP messages?
> >>> SB> It needs to be made clear as soon as you introduce NTP that
> >>> SB> they use different time representations.
> >>>
> >>>  defines mapping and transport of the NTP messages defined in
> >>>  [RFC5905] over MPLS networks.
> >>>
> >>>  One key attribute of all of these Timing messages is that the Time
> >>>  stamp processing should occur as close as possible to the actual
> >>>  transmission and reception at the physical port interface.  This
> >>>  targets optimal time and/or frequency recovery by avoiding variable
> >>>  delay introduced by queues internal to the clocks.
> >>>
> >>> SB> As I recall NTP has no epoch point defined, and I am not sure
> >>> SB> where that point is in the case of PTP in this mapping
> >>> SB> Hopefully this will get defined in due course.
> >>>
> >>>  To facilitate the fast and efficient recognition of Timing messages
> >>>  at the port level when the Timing messages are carried over MPLS
> >>>  LSPs,
> >>>
> >>> SB> Over a new LSP type with time optimied characteristics
> >>>
> >>>  this document defines the specific encapsulations that should
> >>>  be used.
> >>> SB> Hopefully it will also define the PHP
> >>>
> >>>  In addition, it can be expected that there will exist LSR/
> >>>  LERs where only a subset of the physical ports will have the port-
> >>>  based Timing message processing capabilities.
> >>> SB> Do you need to clarify that this only works at base and not in
> >>> SB> a label heirarchy.
> >>>
> >>>
> >>>  In order to ensure
> >>>  that the LSPs carrying Timing packets always enter and exit ports
> >>>  with this capability, routing extensions are defined to advertise
> >>>  this capability on a port basis and to allow for the establishment
> >> of
> >>>  LSPs that only transit such ports.  While this path establishment
> >>>  restriction may be applied only at the LER Ingress and/or egress
> >>>  ports, it becomes more important when using transparent clock
> >> capable
> >>>  LSRs in the path.
> >>> SB> I do not understand the implications of the last
> >>> SB> sentences - starting ", it becomes"
> >>>
> >>>
> >>>  Port based Timing message processing involves Timing message
> >>>  recognition.  Once the Timing messages are recognized they can be
> >>>  modified based on the reception or transmission Time-stamp.
> >>>
> >>>  This document provides two methods for transporting Timing messages
> >>>  over MPLS.  One is applicable to MPLS environment and the other one
> >>>  is applicable to MPLS/MPLS-TP environment
> >>>
> >>> SB> I think the sentence is incomplete.
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page
> >> 5]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>>  The solution involves transporting Timing messages over dedicated
> >>>  LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
> >>>  carry Management and control messages, but not data plane client
> >>>  traffic.
> >>>
> >>> SB> It is not clear why this restriction applies.
> >>>
> >>>  Timing LSPs can be established statically or via signaling.
> >>> SB> s/statically/by provisioning/network management/
> >>>
> >>>  Extensions to control plane (OSPF, ISIS, etc.) is required to
> >> enable
> >>>  routers to distribute their Timing processing capabilities over
> >> MPLS
> >>>  to other routers.  However such extensions are outside the scope of
> >>>  this document.
> >>>
> >>>  When signaling is used to setup the PTP LSP, Extensions to
> >> signaling
> >>> SB> is it a PTP LSP or a Timing LSP?
> >>>
> >>>  protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
> >>>  However such extensions are outside the scope of this document.
> >>>
> >>> SB> for mpls-tp GMPLS is the signalling protocol
> >>>
> >>>  While the techniques included herein allow for the establishment of
> >>>  paths optimized to include Time-stamping capable links, the
> >>>  performance of the Slave clocks is outside the scope of this
> >>>  document.
> >>>
> >>>  At the time of publishing this specification, Transparent Clocking
> >>>  (TC) is only defined for PTP.  Therefore at this time any part of
> >>>  this specification that talks about Transparent Clocking applies
> >> only
> >>>  to PTP.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page
> >> 6]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 2.  Terminology
> >>>
> >>>  1588: The timing and synchronization as defined by IEEE 1588.
> >>>
> >>> SB> I think that there is a more formal name for the 1588 group
> >>> SB> that needs to be used here.
> >>> SB> Also do we need to talk about 1588-200? as there is
> >>> SB> an update in progress
> >>>
> >>>  NTP: The timing and synchronization protocol defined by IETF RFC-
> >> 1305
> >>>  and RFC-5905.
> >>>
> >>>  PTP: The timing and synchronization protocol used by 1588.
> >>> SB> need the proper name for 1588
> >>>
> >>>  Master Clock: The source of 1588 timing to a set of slave clocks.
> >>>
> >>>  Master Port: A port on a ordinary or boundary clock that is in
> >> Master
> >>>  state.  This is the source of timing toward slave ports.
> >>>
> >>> SB> I am not sure the reader knows what a port is
> >>>
> >>>  Slave Clock: A receiver of 1588 timing from a master clock.
> >>>
> >>>  Slave Port: A port on a boundary clock or ordinary clock that is
> >>>  receiving timing from a master clock.
> >>>
> >>>  Ordinary Clock: A device with a single PTP port.
> >>>
> >>>  Transparent Clock.  A device that measures the time taken for a PTP
> >>>  event message to transit the device and then updates the
> >>>  correctionField of the message with this transit time.
> >>>
> >>>  Boundary Clock: A device with more than one PTP port.  Generally
> >>>  boundary clocks will have one port in slave state to receive timing
> >>>  and then other ports in master state to re-distribute the timing.
> >>>
> >>>  PTP LSP: An LSP dedicated to carry PTP messages
> >>>
> >>> SB> PTP or timing?
> >>>
> >>>
> >>>  PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
> >>>  messages.
> >>>
> >>> SB> Ah I don't think that PWE3 know what one of these is
> >>>
> >>>  CW: Pseudowire Control Word
> >>>
> >>>  LAG: Link Aggregation
> >>>
> >>>  ECMP: Equal Cost Multipath
> >>>
> >>>  CF: Correction Field, a field inside certain PTP messages (message
> >>>  type 0-3)that holds the accumulative transit time inside
> >> intermediate
> >>>  switches
> >>>
> >>>  Timing messages: Timing Protocol messages that are exchanged
> >> between
> >>>  routers in order to establish a synchronized clock.
> >>>
> >>> SB> A number of these definitions look like copies of IEEE1588
> >>> SB> definitions. We need to provide references and note the
> >>> SB> priority of the IEEE base reference.
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page
> >> 7]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 3.  Problem Statement
> >>>
> >>>  [IEEE-1588] has defined methods for transporting PTP messages over
> >>>  Ethernet and IP networks.  [RFC5905] has defined the method of
> >>>  transporting NTP messages over IP networks.  There is a need to
> >>>  transport Timing messages over MPLS networks while supporting the
> >>>  Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
> >>>  functionality in the LER and LSRs in the MPLS network.
> >>>
> >>>  There are multiple ways of transporting Timing over MPLS.  However,
> >>>  there is a requirement to limit the possible encapsulation options
> >> to
> >>>  simplify the Timing message identification and processing required
> >> at
> >>>  the port level.
> >>>
> >>>  When Timing-awareness is needed, Timing messages should not be
> >>>  transported over LSPs or PWs that are carrying customer traffic
> >>>  because LSRs perform Label switching based on the top label in the
> >>>  stack.
> >>>
> >>> SB> Have you explained why?
> >>>
> >>>  To detect Timing messages inside such LSPs require special
> >>>  hardware to do deep packet inspection at line rate.  Even if such
> >>>  hardware exists, the payload can't be deterministically identified
> >> by
> >>>  LSRs because the payload type is a context of the PW label, and the
> >>>  PW label and its context are only known to the Edge routers (PEs/
> >>>  LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR,
> >> CES,
> >>>  etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
> >>>  LSRs dont have the knowledge of whether PW Control Word (CW) is
> >>>  present or not and therefore can not deterministically identify the
> >>>  payload.
> >>>
> >>>  A generic method is defined in this document that does not require
> >>>  deep packet inspection at line rate, and can deterministically
> >>>  identify Timing messages.  This method can be used to detect Timing
> >>>  Messages in both one-step and two-step clock implementations of
> >>>  ordinary, boundary and transparent clocks.
> >>>
> >>> SB> Needs a ref and I am sure many MPLS specialists will not
> >> understand
> >>> SB> the msg types.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page
> >> 8]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 4.  Timing over MPLS Architecture
> >>>
> >>>  Timing messages are exchange between Timing ports on ordinary and
> >>>
> >>> SB> Have you defined a timing port?
> >>>
> >>>  boundary clocks.  Boundary clocks terminate the Timing messages and
> >>>  act as master for other boundary clocks or for slave clocks.  End-
> >> to-
> >>>  End Transparent clocks do not terminate the Timing messages but
> >> they
> >>>  do modify the contents of the Timing messages as they transit
> >> across
> >>>  the transparent clock.
> >>>
> >>>  Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent
> >> Clock
> >>>
> >>>  (TC) could be implemented in either LERs or LSRs.
> >>>
> >>> SB> LER and LSR need to be expanded
> >>>
> >>>  An example is shown in Figure 1, where the LERs act as Ordinary
> >> Clock
> >>>  (OC) and are the initiating/terminating point for Timing messages.
> >>>  The ingress LER encapsulates the Timing messages in Timing LSP and
> >>>  the Egress LER terminates the Timing LSP.  The LSRs act as
> >>>  Transparent Clock (TC) and just update the Timing field in the
> >> Timing
> >>>  messages.
> >>>
> >>>
> >>>     +--------+     +-------+     +-------+     +-------+     +------
> >> --+
> >>>     |Switch, |     |       |     |       |     |       |
> >> |Switch, |
> >>>     | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----|
> >> Router |
> >>>     |        |     |  OC   |     |  TC   |     |  OC   |     |
> >> |
> >>>     +--------+     +-------+     +-------+     +-------+     +------
> >> --+
> >>>                    /                                 \
> >>>     +-------+     /                                   \     +-------
> >> +
> >>>     |  LER  |    /                                     \    |  LER
> >> |
> >>>     | Master|---/                                       \---| Slave
> >> |
> >>>     | Clock |                                               | Clock
> >> |
> >>>     +-------+                                               +-------
> >> +
> >>>
> >>>    Figure (1) - Deployment example 1 of timing over MPLS network
> >>>
> >>>  Another example is shown in Figure2, where LERs terminate the
> >> Timing
> >>>  messages received from switch/routers that are outside of the MPLS
> >>>  network acting as OC or BC.  In this example LERs regenerate the
> >>>  clock and initiate timing messages encapsulated in Timing LSP
> >> toward
> >>>  the MPLS network, while the LSRs act as Transparent Clock (TC) and
> >>>  just update the Timing field in the Timing messages, which are
> >>>  already encapsulated in Timing LSPs.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page
> >> 9]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>>    +--------+     +-------+     +-------+     +-------+     +-------
> >> -+
> >>>    |Switch, |     |       |     |       |     |       |     |Switch,
> >> |
> >>>    | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router
> >> |
> >>>    | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC
> >> |
> >>>    +--------+     +-------+     +-------+     +-------+     +-------
> >> -+
> >>>
> >>>    Figure (2) - Deployment example 2 of timing over MPLS network
> >>>
> >>>
> >>>  Another example is shown in Figure 3, where LERs do not terminate
> >> the
> >>>  Timing messages received from switch/routers that are outside of
> >> the
> >>>  MPLS network acting as OC, TC or BC.  The LERs act as TC and update
> >>>  the Timing field in the Timing messages as they transit the LER,
> >>>  while encapsulating them in timing LSP.  The LSRs also act as
> >>>  Transparent Clock (TC) and just update the Timing field in the
> >> Timing
> >>>  messages which are already encapsulated in Timing LSPs.
> >>>
> >>>     +--------+     +-------+     +-------+     +-------+     +------
> >> --+
> >>>     |Switch, |     |       |     |       |     |       |
> >> |Switch, |
> >>>     | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----|
> >> Router |
> >>>     |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |
> >> |OC/TC/BC|
> >>>     +--------+     +-------+     +-------+     +-------+     +------
> >> --+
> >>>
> >>>   Figure (3) - Deployment example 3 of timing over MPLS network
> >>>
> >>>  Another example is shown in Figure 4, where LERs and LSRs support
> >>>  Boundary Clocks.  A single-hop LSP is created between two adjacent
> >>>  LSRs engaged in BC operation.  Other methods such as PTP transport
> >>>  over Ethernet MAY be used for transporting timing messages if the
> >>>  link between the two routers is Ethernet.
> >>>
> >>>    +--------+     +-------+     +-------+     +-------+     +-------
> >> -+
> >>>    |Switch, |     |       |     |       |     |       |     |Switch,
> >> |
> >>>    | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router
> >> |
> >>>    | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC
> >> |
> >>>    +--------+     +-------+     +-------+     +-------+     +-------
> >> -+
> >>>
> >>>  Figure (4) - Deployment example 3 of timing over MPLS network
> >>>
> >>>  An MPLS domain MAY serve multiple customers.  In these cases the
> >> MPLS
> >>>  domain (maintained by a service provider) may provide timing
> >> services
> >>>  to multiple customers, each having their own Timing domain.
> >>>
> >>>  The Timing over MPLS architecture assumes full mesh of Timing LSPs
> >>>  between all LERs supporting this specification.
> >>>
> >>> SB> Note sure this is right - the salves surely do not need to
> >>> SB> exchange timing amongst themselves
> >>>
> >>>  It supports
> >>>  Point-to- point (VPWS) and Multipoint (VPLS) services.
> >>>
> >>> SB> What does that mean? You do not carry user data traffic?
> >>> SB> Maybe it's the ordering of the statemnets that is causing
> >>> SB> confusion.
> >>>
> >>>  This means
> >>>  that a customer may purchase a Point-to-point Timing service
> >> between
> >>>  two customer sites or a Multipoint Timing service between more than
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 10]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>>  two customer sites.
> >>>
> >>>  The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
> >>>  This means that the Timing Multicast messages such as PTP Multicast
> >>>  event messages can be transported over P2MP Timing LSP or be
> >>>  replicated and transported over many P2P Timing LSPs.
> >>>
> >>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
> >>> SB> nor a P2MP PW, although we are close.
> >>>
> >>>  Timing messages, that do not require Time stamping or Correction
> >>>  Field update MAY be transported over Timing LSPs to simplify
> >> hardware
> >>>  and software.
> >>>
> >>>  PTP Announce messages that determine the Timing LSP terminating
> >> point
> >>>  behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
> >>>  to simplify hardware and software.
> >>>
> >>> SB> have you defined and referenced PTP announce msgs?
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 11]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 5.  Dedicated LSPs for Timing messages
> >>>
> >>>  Many methods have been considered for identifying the Timing
> >> messages
> >>>  when they are encapsulated in MPLS such as using GAL/G-ACH or a new
> >>>  reserved label.  These methods were not attractive since they
> >> either
> >>>  required deep packet inspection at line rate in the intermediate
> >> LSRs
> >>>  or they required use of a scarce new reserved label.  Also one of
> >> the
> >>>  goals was to reuse existing OAM mechanisms.
> >>>
> >>> SB> RLs = SPLs are not so rare now. In any case needs a ref.
> >>>
> >>>  The method defined in this document can be used by LER and LSRs to
> >>>  identify Timing messages in MPLS tunnels by just looking at the top
> >>>  label in the MPLS label stack, which only carry Timing messages as
> >>>  well as OAM, but not data plane client traffic.
> >>>
> >>>  Compliant implementations MUST use dedicated LSPs to carry Timing
> >>>  messages over MPLS.
> >>>
> >>> SB> I think that we need a definition of the properies of these LSPs
> >>>
> >>>  These LSPs are herein referred to as "Timing
> >>>  LSPs" and the labels associated with these LSPs as "Timing LSP
> >>>  labels".  The Timing LSPs that runs between Ingress and Egress LERs
> >>>  MUST be co-routed.  Alternatively, a single bidirectional co-routed
> >>>  LSP can be used.
> >>>
> >>> SB> I though that you said you could use M2MP LSPs - these are not
> >>> SB> bidirectional.
> >>>
> >>>  Co-routing of the two directions is required to limit the
> >> difference
> >>>  in the delays in the Master clock to Slave clock direction compared
> >>>  to the Slave clock to Master clock direction.  The Timing LSP MAY
> >> be
> >>>  MPLS/MPLS-TP LSP.
> >>>
> >>>  The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
> >>>  New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
> >>>  outside the scope of this document.
> >>>
> >>>  The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such
> >> as
> >>>  BFD and LSP Ping but the LSP data plane client plane traffic MUST
> >> be
> >>>  Timing packets only.
> >>>
> >>> SB> Why?
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 12]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 6.  Timing over LSP Encapsulation
> >>>
> >>> The encapsulations is not LSP is it?
> >>>
> >>>  This document defines two methods for carrying Timing messages over
> >>>  MPLS.  The first method is carrying UDP/IP encapsulated Timing
> >>>  messages over Timing LSPs, and the second method, is carrying
> >>>  Ethernet encapsulated Timing messages over Ethernet PWs inside
> >> Timing
> >>>  LSPs.
> >>>
> >>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
> >>>
> >>>  The simplest method of transporting Timing messages over MPLS is to
> >>>  encapsulate Timing PDUs in UDP/IP and then encapsulate them in
> >> Timing
> >>>  LSP.  This format is shown in Figure 4.
> >>>
> >>>
> >>>                   +----------------------+
> >>>                   |   Timing LSP Label   |
> >>>                   +----------------------+
> >>>                   |        IPv4/6        |
> >>>                   +----------------------+
> >>>                   |         UDP          |
> >>>                   +----------------------+
> >>>                   |     Timing PDU       |
> >>>                   +----------------------+
> >>>
> >>>     Figure (4) - Timing over UDP/IP over MPLS Encapsulation
> >>>
> >>>
> >>>  This encapsulation is very simple and is useful when the network
> >>>  between Timing Master Clock and Slave Clock is MPLS network.
> >>>
> >>> SB> Simple is a judgement call
> >>>
> >>>  In order for an LER/LSR to process Timing messages, the Timing LSP
> >>>  Label must be at the top label of the label stack.  The LER/LSR
> >> MUST
> >>>  know that the Timing LSP Label is used for carrying Timing
> >> messages.
> >>>  This can be accomplished via static configuration or via RSVP-TE
> >>>  signaling.
> >>>
> >>>  The UDP/IP encapsulation of PTP MUST follow Annex D and E of
> >>>  [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
> >>>  [RFC5905].
> >>>
> >>> 6.2.  Timing over PW Encapsulation
> >>>
> >>>  Another method of transporting Timing over MPLS networks is by
> >>>  encapsulating Timing PDUs in PW which in turn is transported over
> >>>  Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
> >>>  shown in Fig 5(A) MUST be used and the Ethernet encapsulation of
> >> PTP
> >>>  MUST follow Annex F of [IEEE-1588].
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 13]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>>  The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
> >> the
> >>>  Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
> >>>  Timing over PW encapsulation MUST use the Control Word (CW) as
> >>>  specified in [RFC4448] to ensure proper detection of PTP messages
> >>>  inside the MPLS packets for Timing over LSP and Timing over PW
> >>>  encapsulation.
> >>>
> >>> SB> That needs explanation
> >>>
> >>>  The use of Sequence Number in the CW is optional.
> >>>
> >>> SB> Given that s/n are never in practice deployed, you could probably
> >>> SB> simplify things by sayig that they are not used.
> >>>
> >>>  Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over
> >> PW
> >>>  (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
> >>>
> >>>                   +----------------+  +----------------+
> >>>                    |Timing LSP Label|  |Timing LSP Label|
> >>>                    +----------------+  +----------------+
> >>>                    |    PW Label    |  |    PW Label    |
> >>>                    +----------------+  +----------------+
> >>>                    |  Control Word  |  |      IP        |
> >>>                    +----------------+  +----------------+
> >>>                    |    Ethernet    |  |      UDP       |
> >>>                    |     Header     |  +----------------+
> >>>                    +----------------+  |   Timing PDU   |
> >>>                    |S-VLAN(Optional)|  |                |
> >>>                    +----------------+  +----------------+
> >>>                    |C-VLAN(Optional)|        (B)
> >>>                    +----------------+
> >>>                    |   Timing PDU   |
> >>>                    |                |
> >>>                    +----------------+
> >>>                           (A)
> >>>
> >>>             Figure (5) - Timing over PW Encapsulations
> >>>
> >>>  In order for an LSR to process PTP messages, the top label of the
> >>>  label stack (the Tunnel Label) MUST be a Timing label.
> >>>
> >>> S> You said that before.
> >>>
> >>> 6.3.  Other Timing Encapsulation methods
> >>>
> >>>  In future other timing encapsulation methods may be introduced,
> >> such
> >>>  as a new shim header after the Bottom of Stack to carry the Timing
> >>>  information.  Such new encapsulations are outside the scope of this
> >>>  document.
> >>>
> >>>
> >>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
> >>> SB> out of the definition of the LSP
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 14]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> SB> I think we need a section on LSP processing
> >>>
> >>> 7.  Timing message Processing
> >>>
> >>>  Each Timing protocol such as PTP and NTP, define their set of
> >> Timing
> >>>  messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
> >>>  FOLLOW_UP, etc messages.
> >>>
> >>>  Some of the Timing messages require time stamping or correction
> >> field
> >>>  update at port level and some dont.  It is the job of the LER/LSR
> >> to
> >>>  parse the timing message and find out the type of the Timing
> >> message
> >>>  and decide whether and how to Time- stamp it (e.g., BC) or update
> >>>  correction field(e.g., TC).
> >>>
> >>>
> >>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
> >>> SB> function rather than the LER?
> >>>
> >>>  For example the following PTP messages (called Event messages)
> >>>  require time-stamping or correction field update:
> >>>
> >>>  o  SYNC
> >>>
> >>>  o  DELAY_REQ (Delay Request)
> >>>
> >>>  o  PDELAY_REQ (Peer Delay Request)
> >>>
> >>>  o  PDELAY_RESP (Peer Delay Response)
> >>>
> >>>  SYNC and DELAY_REQ are exchanged between Master Clock and Slave
> >> Clock
> >>>  and MUST be transported over PTP LSPs.  PDELAY_REQ and PDELAY_RESP
> >>>  are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
> >>>  Boundary, or Transparent) and SHOULD be transported over single hop
> >>>  PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
> >>>  and PDELAY_RESP_FOLLOW_UP messages MUST also be transported over
> >> the
> >>>  PTP LSPs.
> >>>
> >>>  For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
> >>>  transported over two PTP LSPs that are in opposite directions.
> >> These
> >>>  PTP LSPs, which are in opposite directions MUST be congruent and
> >> co-
> >>>  routed.  Alternatively, a single bidirectional co-routed LSP can be
> >>>  used.
> >>>
> >>>  Except as indicated above for the two-step PTP clocks, Non-Event
> >> PTP
> >>>  message types do not need to be processed by intermediate routers.
> >>>  These message types MAY be carried in PTP Tunnel LSPs.
> >>>
> >>> SB> Are you saying that a timing P router has to be msg type
> >> sensitive?
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 15]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 8.  Protection and Redundancy
> >>>
> >>>
> >>> SB> This is a bit of a jump - I don't know how the LSP itself works
> >> yet!
> >>>
> >>>  In order to ensure continuous uninterrupted operation of slave
> >>>  clocks, usually as a general practice, slave clocks (or ports)
> >> track
> >>>  redundant master clocks.
> >>>
> >>>  It is the responsibility of the network operator to ensure that
> >>>  physically disjoint Timing LSPs are established between a slave
> >> clock
> >>>  (or port) and redundant master clocks (or ports).
> >>>
> >>>  When a slave clock (or port) listens to redundant master clocks or
> >>>  ports, any prolonged Timing LSP outage will trigger the slave clock
> >>>  or port to switch to a redundant master clock or port.
> >>>
> >>>  LSP/PW protection such as Linear protection Switching (1:1, 1+1),
> >>>  Ring protection switching or MPLS Fast Reroute (FRR) generally
> >> switch
> >>>  alternative path that usually cause a change in delay, which if
> >>>  undetected by slave clock can reduce accuracy of the slave clock.
> >>>
> >>>  Therefore protection switching MAY be used, as long as phase jumps
> >>>  upon switchover due to differences in path latency are detected and
> >>>  compensated for (such compensation not being required if BCs or
> >> peer-
> >>>  peer TCs are used throughout).
> >>>
> >>>  Note that any protection or reroute mechanism that adds additional
> >>>  MPLS label to the label stack, such as Facility Backup Fast
> >> Reroute,
> >>>  MUST ensure that the pushed label is also a Timing Label to ensure
> >>>  recognition of the MPLS frame as containing Timing messages, as it
> >>>  transits the backup path.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 16]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 9.  ECMP
> >>>
> >>>  To ensure the optimal operation of slave clocks and avoid error
> >>>  introduced by forward and reverse path delay asymmetry, the
> >> physical
> >>>  path for Timing messages from master clock to slave Clock and vice
> >>>  versa must be the same for all Event Timing messages listed in
> >>>  section 7.
> >>>
> >>>  Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
> >>>  Multipath).
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 17]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 10.  PHP
> >>>
> >>>  To ensure that the label on the top of the label stack is the
> >> Timing
> >>>  LSP Label, PHP MUST not be used.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 18]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 11.  Entropy
> >>>
> >>>  To ensure all Timing messages in a Timing LSP take the same path,
> >>>  Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
> >>>  Entropy Label MUST NOT be used for the PWs that are carried inside
> >>>  Timing LSP [RFC6391].
> >>>
> >>> SB> This is incorrect - you mean that all msgs of the same timing
> >>> SB> flow need to have the same EL value.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 19]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 12.  OAM, Control and Management
> >>>
> >>>  In order to monitor Timing LSPs and their encapsulated PWs, they
> >> MUST
> >>>  be able to carry OAM and management messages.  These management
> >>>  messages MUST be differentiated from Timing messages via already
> >>>  defined IETF methods.
> >>>
> >>>  For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
> >>>  over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
> >>>  Management protocols can easily be identified by the UDP
> >> Destination
> >>>  Port number or by GAL/G-ACH respectively.
> >>>
> >>>  Also BFD, LSP-Ping and other management messages MAY run over the
> >> PWs
> >>>  encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3
> >> or
> >>>  4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is
> >> going
> >>>  to be deprecated by IETF).  In this case G-ACH, PW label (TTL=1) or
> >>>  GAL-ACH are used to identify such management messages.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 20]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 13.  QoS Considerations
> >>>
> >>>  In network deployments where not every LSR/LER is Timing-aware, it
> >> is
> >>>  important to reduce the impact of the non-Timing-aware LSR/LERs on
> >>>  the timing recovery in the slave clock.  The Timing messages are
> >> time
> >>>  critical and must be treated with the highest priority.  Therefore
> >>>  Timing over MPLS messages must be treated with the highest priority
> >>>  in the routers.  This can be achieved by proper setup of Timing
> >> LSPs.
> >>>
> >>>  It is recommended that the Timing LSPs are setup or configured
> >>>  properly to indicate EF-PHB [RFC3246]for the CoS and Green
> >> [RFC2697]
> >>>  for drop eligibility.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 21]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 14.  FCS and Checksum Recalculation
> >>>
> >>>  When time-stamp generation and timing packet adjustment is
> >> performed
> >>>  near the physical port hardware, the process MUST include
> >>>  recalculation of the Ethernet FCS.
> >>>
> >>> SB> The above is confusing - an LSR always recomputes the link layer
> >>> SB> CRC which may or may not be Ethernet.
> >>>
> >>>  Also FCS retention for the
> >>>  payload Ethernet described in [RFC4720] MUST NOT be used.
> >>>
> >>>  For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
> >>>  may be required as per UDP transport standards.
> >>>
> >>> SB> You really need to be working on getting the IPv6 C?S computation
> >>> SB> removed from PTP msgs.
> >>>
> >>>  When UDP checksum is used, each Timing-aware LER/LSR must either
> >>>  incrementally update the UDP checksum after Time stamping or
> >>>  Correction Field update or verify the UDP checksum on reception
> >> from
> >>>  upstream and recalculate the checksum completely on transmission to
> >>>  downstream node after Time stamping or Correction Field update.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 22]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 15.  Behavior of LER/LSR
> >>>
> >>>  Timing-capable/aware LERs and LSRs are routers that have one or
> >> more
> >>>
> >>> SB> You mean physical interfaces?
> >>>
> >>>  interfaces that can perform Timing operations (OC/BC/TC) on Timing
> >>>  packets and are configured to do so.  Timing-capable/aware LERs and
> >>>  LSRs can advertise their Timing-capability per-interface via
> >> control
> >>>  plane such as OSPF or IS-IS.
> >>> SB> ISIS and OSPF are routing protocols.
> >>>
> >>> The Timing-capable/aware LERs can then
> >>>  signals Timing LSPs via RSVP-TE signaling.  Alternatively the
> >> Timing
> >>>  capability of LER and LSRs may be configured in a centralized
> >>>  controller and the Timing LSP may be setup using manual
> >> configuration
> >>>  or other methods such as SDN.
> >>>
> >>> SB> it can also be configured individually rather then through
> >>> SB> a cebtral controllwe
> >>>
> >>> 15.1.  Behavior of Timing-capable/aware LER
> >>>
> >>>  When a Timing-capable/aware LER behaves as a Transparent clock and
> >>>  receives a Timing message from a Timing-capable/aware non-MPLS
> >>>  interface, the LER updates the Correction Field (CF) and
> >> encapsulates
> >>>  and forwards the timing message over previously established Timing
> >>>  LSP.
> >>>
> >>> SB> You need to call out the details so that people properly
> >>> SB> understand the definition of the new LSP.
> >>>
> >>>  Also when a Timing message is received from a Timing-capable/
> >>>  aware MPLS interface, LER updates the Correction Filed (CF) and
> >>>  decapsulates the MPLS encapsulation and forwards the timing message
> >>>  to a non-MPLS interface.
> >>>
> >>>  When a Timing-capable/aware LER behaves as a Boundary clock and
> >>>  receives a Timing message from a Timing-capable/aware non MPLS
> >>>  interface, the LER Timestamps the Timing packet and sends it to the
> >>>  LERs Boundary clock processing module.  Also when a Timing message
> >> is
> >>>  received from a Timing- capable/aware MPLS interface, the LER
> >>>  Timestamps the Timing packet and sends it to the LERs Boundary
> >> clock
> >>>  processing module.
> >>>
> >>>  When a Timing-capable/aware LER behaves as an Ordinary Clock toward
> >>>  the MPLS network, and receives a Timing message from a Timing-
> >>>  capable/aware MPLS interface, the LER Timestamps the Timing packet
> >>>  and sends it to the LERs Ordinary clock processing module.
> >>>
> >>> 15.2.  Behavior of Timing-capable/aware LSR
> >>>
> >>>  When a Timing-capable/aware LSR behaves as a Transparent clock and
> >>>  receives a Timing message from a Timing-capable/aware MPLS
> >> interface,
> >>>  The LSR updates the Correction Filed (CF) and forwards the timing
> >>>  message over another MPLS interface.
> >>>
> >>>  When a Timing-capable/aware LSR behaves as a Boundary clock and
> >>>  receives a Timing message from a Timing-capable/aware MPLS
> >> interface.
> >>>  The LSR performs the functions of a Boundary Clock in terminating
> >> the
> >>>  received Timing message and re-generating a new timing message over
> >>>  another (or the same) MPLS interface.
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 23]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 15.3.  Behavior of non-Timing-capable/aware LSR
> >>>
> >>>  It is most beneficial when all LSRs in the path of a Timing LSP be
> >>>  timing-Capable/aware LSRs.  This would ensure the highest quality
> >>>  time and clock synchronization by Timing Slave Clocks.  However,
> >> this
> >>>  specification does not mandate that all LSRs in path of a Timing
> >> LSP
> >>>  be Timing- capable/aware.
> >>>
> >>>  Non-Timing-capable/aware LSRs just switch the packets encapsulated
> >> in
> >>>  Timing LSPs and dont perform any Timing operation (TC or BC).
> >>>  However as explained in QoS section the Timing over MPLS packets
> >> MUST
> >>>  be still be treated with the highest priority based on their
> >> Traffic
> >>>  Class (TC) marking.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 24]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 16.  Other considerations
> >>>
> >>>  [IEEE-1588] defines an optional peer-to-peer Transparent clocking
> >>>  that requires peer delay measurement between two adjacent Timing-
> >>>  capable/ aware routers/switches.  Peer delay measurement messages
> >>>  need to be time stamped and terminated by the Timing-capable/aware
> >>>  routers/ switches.  This means that two adjacent LSRs may be
> >> engaged
> >>>  in a peer delay measurement.
> >>>
> >>>  For transporting such peer delay measurement messages a single-hop
> >>>  LSP SHOULD to be created between the two adjacent LSRs engaged in
> >>>  peer delay measurement to carry peer delay measurement messages.
> >>>  Other methods such as PTP transport over Ethernet MAY be used for
> >>>  transporting peer delay measurement messages if the link between
> >> the
> >>>  two routers is Ethernet.
> >>>
> >>>  In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/
> >> ware
> >>>  routers/switches MUST maintain a list of all the neighbors it needs
> >>>  to send a PDelay_Req to, where each neighbor corresponds to a
> >> timing
> >>>  LSP.
> >>>
> >>>  The use of Explicit Null Label (Label= 0 or 2) is acceptable as
> >> long
> >>>  as either the Explicit Null label is the bottom of stack label
> >>>  (applicable only to UDP/IP encapsulation) or the label below the
> >>>  Explicit Null label is a PTP label.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 25]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 17.  Security Considerations
> >>>
> >>>  MPLS PW security considerations in general are discussed in
> >> [RFC3985]
> >>>  and [RFC4447],and those considerations also apply to this document.
> >>>
> >>>  An experimental security protocol is defined in [IEEE-1588].The PTP
> >>>  security extension and protocol provides group source
> >> authentication,
> >>>  message integrity, and replay attack protection for PTP messages.
> >>>
> >>>  When the MPLS network (provider network) serves multiple customers,
> >>>  it is important to maintain and process each customers clock and
> >>>  Timing messages separately from other customers to ensure there is
> >> no
> >>>  cross- customer effect.  For example if an LER BC is synchronized
> >> to
> >>>  a specific grandmaster, belonging to customer A, then the LER MUST
> >>>  use that BC clock only for customer A to ensure that customer A
> >>>  cannot attack other customers by manipulating its time.
> >>>
> >>>  Timing messages MAY be encrypted or authenticated, provided that
> >> the
> >>>  LERs/LSRs that are Timing capable/aware can authenticate/ decrypt
> >> the
> >>>  timing messages.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 26]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 18.  Acknowledgements
> >>>
> >>>  The authors would like to thank Ron Cohen, Yaakov Stein, Tal
> >> Mizrahi,
> >>>  Stefano Ruffini, Peter Meyer, and other members of IETF for
> >> reviewing
> >>>  and providing feedback on this draft.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 27]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 19.  IANA Considerations
> >>>
> >>>  There are no IANA requirements in this specification.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 28]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 20.  References
> >>>
> >>> 20.1.  Normative References
> >>>
> >>>  [IEEE-1588]
> >>>             IEEE 1588-2008, "IEEE Standard for a Precision Clock
> >>>             Synchronization Protocol for Networked Measurement and
> >>>             Control Systems".
> >>>
> >>>  [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
> >>>             Requirement Levels", BCP 14, RFC 2119, March 1997.
> >>>
> >>>  [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
> >>>             Edge (PWE3) Architecture", RFC 3985, March 2005.
> >>>
> >>>  [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor
> >> Discovery
> >>>             Proxies (ND Proxy)", RFC 4389, April 2006.
> >>>
> >>>  [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
> >>>             Heron, "Pseudowire Setup and Maintenance Using the Label
> >>>             Distribution Protocol (LDP)", RFC 4447, April 2006.
> >>>
> >>>  [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
> >>>             "Encapsulation Methods for Transport of Ethernet over
> >> MPLS
> >>>             Networks", RFC 4448, April 2006.
> >>>
> >>>  [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
> >>>             Emulation Edge-to-Edge (PWE3) Frame Check Sequence
> >>>             Retention", RFC 4720, November 2006.
> >>>
> >>>  [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
> >>>             Connectivity Verification (VCCV): A Control Channel for
> >>>             Pseudowires", RFC 5085, December 2007.
> >>>
> >>>  [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding
> >> Detection
> >>>             (BFD)", RFC 5880, June 2010.
> >>>
> >>>  [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
> >>>             "Bidirectional Forwarding Detection (BFD) for MPLS Label
> >>>             Switched Paths (LSPs)", RFC 5884, June 2010.
> >>>
> >>> 20.2.  Informative References
> >>>
> >>>  [I-D.ietf-pwe3-fat-pw]
> >>>             Bryant, S., Filsfils, C., Drafz, U., Kompella, V.,
> >> Regan,
> >>>             J., and S. Amante, "Flow Aware Transport of Pseudowires
> >>>             over an MPLS Packet Switched Network",
> >>>             draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 29]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>>  [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
> >>>             system routeing information exchange protocol for use in
> >>>             conjunction with the Protocol for providing the
> >>>             Connectionless-mode Network Service (ISO 8473)".
> >>>
> >>>  [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
> >>>             dual environments", RFC 1195, December 1990.
> >>>
> >>>  [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
> >>>
> >>>  [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
> >>>             Marker", RFC 2697, September 1999.
> >>>
> >>>  [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le
> >> Boudec,
> >>>             J., Courtney, W., Davari, S., Firoiu, V., and D.
> >>>             Stiliadis, "An Expedited Forwarding PHB (Per-Hop
> >>>             Behavior)", RFC 3246, March 2002.
> >>>
> >>>  [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic
> >> Engineering
> >>>             (TE) Extensions to OSPF Version 2", RFC 3630,
> >>>             September 2003.
> >>>
> >>>  [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
> >>>             System (IS-IS) Extensions for Traffic Engineering (TE)",
> >>>             RFC 3784, June 2004.
> >>>
> >>>  [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
> >>>             Shaffer, "Extensions to OSPF for Advertising Optional
> >>>             Router Capabilities", RFC 4970, July 2007.
> >>>
> >>>  [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
> >>>             System to Intermediate System (IS-IS) Extensions for
> >>>             Advertising Router Information", RFC 4971, July 2007.
> >>>
> >>>  [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
> >>>             Topology (MT) Routing in Intermediate System to
> >>>             Intermediate Systems (IS-ISs)", RFC 5120, February 2008.
> >>>
> >>>  [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
> >>>             Engineering", RFC 5305, October 2008.
> >>>
> >>>  [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
> >>>             "Traffic Engineering Extensions to OSPF Version 3",
> >>>             RFC 5329, September 2008.
> >>>
> >>>  [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
> >>>             for IPv6", RFC 5340, July 2008.
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 30]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>>  [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch,
> >> "Network
> >>>             Time Protocol Version 4: Protocol and Algorithms
> >>>             Specification", RFC 5905, June 2010.
> >>>
> >>>  [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V.,
> >> Regan,
> >>>             J., and S. Amante, "Flow-Aware Transport of Pseudowires
> >>>             over an MPLS Packet Switched Network", RFC 6391,
> >>>             November 2011.
> >>>
> >>>  [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
> >>>             L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
> >>>             RFC 6790, November 2012.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 31]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 1.  Routing extensions for Timing-aware Routers
> >>>
> >>>  MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340]
> >> and
> >>>  IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering
> >> (TE)
> >>>  link information used for constraint-based routing.
> >>>
> >>>  Indeed, it is useful to advertise data plane TE router link
> >>>  capabilities, such as the capability for a router to be Timing-
> >> aware.
> >>>  This capability MUST then be taken into account during path
> >>>  computation to prefer or even require links that advertise
> >> themselves
> >>>  as Timing-aware.  In this way the path can ensure the entry and
> >> exit
> >>>  points into the LERs and, if desired, the links into the LSRs are
> >>>  able to perform port based time-stamping thus minimizing their
> >> impact
> >>>  on the performance of the slave clock.
> >>>
> >>>  extensions are required to OSPF and IS-IS in order to advertise
> >>>  Timing-aware capabilities of a link.  Such extensions are outside
> >> the
> >>>  scope of this document; however such extension SHOULD be able to
> >>>  signal the following information per Router Link:
> >>>
> >>>  o  Capable of processing PTP, NTP or other Timing flows
> >>>
> >>>  o  Capable of performing Transparent Clock operation
> >>>
> >>>  o  Capable of performing Boundary Clock operation
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 32]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> 2.  Signaling Extensions for Creating Timing LSPs
> >>>
> >>>  RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-
> >> TE
> >>>  is used to setup Timing LSPs, some information that indicates that
> >>>  the LSP is carrying Timing flows MUST be included in the new
> >>>  Extensions to RSVP-TE:
> >>>
> >>>  The following information MAY also be included in the new
> >> Extensions
> >>>  to RSVP-TE:
> >>>
> >>>  o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
> >>>     field
> >>>
> >>>  o  Number of VLANs in case of PW encapsulation
> >>>
> >>>  o  Timestamp field Type
> >>>
> >>>     *  Correction Field, Timestamp
> >>>
> >>>  o  Timestamp Field format
> >>>
> >>>     *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
> >>>        NTP, etc.
> >>>
> >>>  Note that in case the above optional information is signaled with
> >>>  RSVP-TE for a Timing LSP, all the Timing packets carried in that
> >> LSP
> >>>  must have the same signaled characteristics.  For example if
> >>>  Timestamp format is signaled as 64-bit PTPv1, then all Timing
> >> packets
> >>>  must use 64-bit PTPv1 time-stamp.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 33]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>> Authors' Addresses
> >>>
> >>>  Shahram Davari
> >>>  Broadcom Corp.
> >>>  San Jose, CA  95134
> >>>  USA
> >>>
> >>>  Email: davari@broadcom.com
> >>>
> >>>
> >>>  Amit Oren
> >>>  Broadcom Corp.
> >>>  San Jose, CA  95134
> >>>  USA
> >>>
> >>>  Email: amito@broadcom.com
> >>>
> >>>
> >>>  Manav Bhatia
> >>>  Alcatel-Lucent
> >>>  Bangalore,
> >>>  India
> >>>
> >>>  Email: manav.bhatia@alcatel-lucent.com
> >>>
> >>>
> >>>  Peter Roberts
> >>>  Alcatel-Lucent
> >>>  Kanata,
> >>>  Canada
> >>>
> >>>  Email: peter.roberts@alcatel-lucent.com
> >>>
> >>>
> >>>  Laurent Montini
> >>>  Cisco Systems
> >>>  San Jose CA
> >>>  USA
> >>>
> >>>  Email: lmontini@cisco.com
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 34]
> >
> >>> Internet-Draft        Transporting Timing over MPLS            June
> >> 2013
> >>>
> >>>
> >>>  Luca
> >>>  Cisco Systems
> >>>  San Jose CA
> >>>  USA
> >>>
> >>>  Email: lmartini@cisco.com
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page
> >> 35]
> >
> >>>
> >>> --
> >>> For corporate legal information go to:
> >>>
> >>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
> >>>
> >>> _______________________________________________
> >>> mpls mailing list
> >>> mpls@ietf.org
> >>> https://www.ietf.org/mailman/listinfo/mpls
> >
> >
> >
> >
>
> _______________________________________________
> mpls mailing list
> mpls@ietf.org
> https://www.ietf.org/mailman/listinfo/mpls
>

--bcaec51d2c10d243ea04e2f91efd
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

<p>Hi Shahram,<br>
I think that John suggested creation of LSPs interconnecting all 1588-capab=
le nodes. It is absolutely not what current document defines but that remin=
ds me of what I&#39;ve suggested around November of 2011.<br>
Regards,<br>
Greg</p>
<div class=3D"gmail_quote">On Aug 2, 2013 6:43 AM, &quot;Shahram Davari&quo=
t; &lt;<a href=3D"mailto:davari@broadcom.com">davari@broadcom.com</a>&gt; w=
rote:<br type=3D"attribution"><blockquote class=3D"gmail_quote" style=3D"ma=
rgin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Hi John<br>
<br>
In that case I fail to understand the difference between your proposal and =
the current draft.<br>
<br>
In both cases you need a special LSP to tell the HW to Timestamp the packet=
 and I assume you need to use RSVPTE to sign it. That is what the draft doe=
s.<br>
<br>
Regards,<br>
Shahram<br>
<br>
<br>
On Aug 2, 2013, at 12:52 PM, &quot;John E Drake&quot; &lt;<a href=3D"mailto=
:jdrake@juniper.net">jdrake@juniper.net</a>&gt; wrote:<br>
<br>
&gt; Shahram,<br>
&gt;<br>
&gt; The LSPs would be between LSRs that are 1588 capable, so they could be=
 one or more hops.<br>
&gt;<br>
&gt; Yours Irrespectively,<br>
&gt;<br>
&gt; John<br>
&gt;<br>
&gt;<br>
&gt;&gt; -----Original Message-----<br>
&gt;&gt; From: Shahram Davari [mailto:<a href=3D"mailto:davari@broadcom.com=
">davari@broadcom.com</a>]<br>
&gt;&gt; Sent: Friday, August 02, 2013 3:24 AM<br>
&gt;&gt; To: &lt;<a href=3D"mailto:stbryant@cisco.com">stbryant@cisco.com</=
a>&gt;<br>
&gt;&gt; Cc: <a href=3D"mailto:mpls-chairs@tools.ietf.org">mpls-chairs@tool=
s.ietf.org</a>; <a href=3D"mailto:tictoc-chairs@tools.ietf.org">tictoc-chai=
rs@tools.ietf.org</a>; int-<br>
&gt;&gt; <a href=3D"mailto:ads@tools.ietf.org">ads@tools.ietf.org</a>; <a h=
ref=3D"mailto:rtg-ads@tools.ietf.org">rtg-ads@tools.ietf.org</a>; draft-iet=
f-tictoc-<br>
&gt;&gt; <a href=3D"mailto:1588overmpls@tools.ietf.org">1588overmpls@tools.=
ietf.org</a>; John E Drake; <a href=3D"mailto:mpls@ietf.org">mpls@ietf.org<=
/a>;<br>
&gt;&gt; <a href=3D"mailto:tictoc@ietf.org">tictoc@ietf.org</a><br>
&gt;&gt; Subject: Re: [mpls] Comments on draft-ietf-tictoc-1588overmpls<br>
&gt;&gt;<br>
&gt;&gt; Hi Stewart<br>
&gt;&gt;<br>
&gt;&gt; We have already looked at similar approach. The issue with one hop=
 LSP<br>
&gt;&gt; is that not all LSRs are 1588 capable. One of the requirements is =
for<br>
&gt;&gt; non-1588 capable routers to just switch the packet normally.<br>
&gt;&gt;<br>
&gt;&gt; Regards,<br>
&gt;&gt; Shahram<br>
&gt;&gt;<br>
&gt;&gt;<br>
&gt;&gt; On Aug 2, 2013, at 12:02 PM, &quot;Stewart Bryant&quot; &lt;<a hre=
f=3D"mailto:stbryant@cisco.com">stbryant@cisco.com</a>&gt;<br>
&gt;&gt; wrote:<br>
&gt;&gt;<br>
&gt;&gt;&gt; Talking to John Drake about this, an alternative general<br>
&gt;&gt;&gt; model is for the LSP is that it is an LSP that timestamps<br>
&gt;&gt;&gt; the packet and then passes it to an application associated<br>
&gt;&gt;&gt; with the LSP at that hop.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; This has a lot of merit.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; If however we think about it some more we have a<br>
&gt;&gt;&gt; type of network service chaining going on here and<br>
&gt;&gt;&gt; so we don&#39;t need an LSP per say, because the path<br>
&gt;&gt;&gt; and instruction can be in the packet.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; In other words a general solution =A0to the problem<br>
&gt;&gt;&gt; is to define an LSP with the properties that the<br>
&gt;&gt;&gt; packet is passed one hop, timestamped and delivered<br>
&gt;&gt;&gt; to the associated application.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; How this is MPLS construct is used to support time<br>
&gt;&gt;&gt; tranfer is entirely within the scope of TICTOC, but<br>
&gt;&gt;&gt; at the MPLS layer we have a clean reusable network<br>
&gt;&gt;&gt; service.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; - Stewart<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D<br>
&gt;&gt;&gt; SB&gt; This draft does not seem to provide a precise definitio=
n<br>
&gt;&gt;&gt; SB&gt; the properties of the new LSP type that it wishes to<br=
>
&gt;&gt;&gt; SB&gt; define, in particular it does it define the PHB of thos=
e<br>
&gt;&gt;&gt; SB&gt; LSPs, nor the full interaction with the MPLS<br>
&gt;&gt;&gt; SB&gt; architecture.<br>
&gt;&gt;&gt; SB&gt;<br>
&gt;&gt;&gt; SB&gt; I have not tracked TICTOC for a while but I thought tha=
t<br>
&gt;&gt;&gt; SB&gt; the original plan was to define the concept of an offse=
t<br>
&gt;&gt;&gt; SB&gt; into a packet to do the correction.<br>
&gt;&gt;&gt; SB&gt;<br>
&gt;&gt;&gt; SB&gt; It is disappointing that the opportunity was not taken<=
br>
&gt;&gt;&gt; SB&gt; to define a timing shim inside the timing LSP so that<b=
r>
&gt;&gt;&gt; SB&gt; a time correction could be added to any packet such tha=
t<br>
&gt;&gt;&gt; SB&gt; the MPLS system was isolated from the details of the<br=
>
&gt;&gt;&gt; SB&gt; complexity of the particular time transfer type.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; I think that much more clarify is needed in terms of<br=
>
&gt;&gt;&gt; SB&gt; definition of the new LSP type, since it is unclear<br>
&gt;&gt;&gt; SB&gt; from this text how to implement one.<br>
&gt;&gt;&gt; SB&gt;<br>
&gt;&gt;&gt; SB&gt; There are a lot of other MPLS services such as<br>
&gt;&gt;&gt; SB&gt; LSP ping that need to be considered.<br>
&gt;&gt;&gt; SB&gt;<br>
&gt;&gt;&gt; SB&gt; Please see inline for more comments. However these<br>
&gt;&gt;&gt; SB&gt; comments are made in the context of the text as written=
<br>
&gt;&gt;&gt; SB&gt; whilst I have a fundamental concern that this approach<=
br>
&gt;&gt;&gt; SB&gt; lacks an MPLS architectural soundness that need<br>
&gt;&gt;&gt; SB&gt; greater thought with significant impact on the<br>
&gt;&gt;&gt; SB&gt; draft.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; - Stewart<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; TICTOC Working Group =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 S.<br>
&gt;&gt; Davari<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 A.<br>
&gt;&gt; Oren<br>
&gt;&gt;&gt; Intended status: Standards Track =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0Broadcom<br>
&gt;&gt; Corp.<br>
&gt;&gt;&gt; Expires: December 17, 2013 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 M.<br>
&gt;&gt; Bhatia<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 P.<br>
&gt;&gt; Roberts<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 Alcatel-<br>
&gt;&gt; Lucent<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 L.<br>
&gt;&gt; Montini<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 L.<br>
&gt;&gt; Martini<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0Cisco<br>
&gt;&gt; Systems<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0June 15,<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 Transporting Timing messages over MPLS Net=
works<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0draft-ietf-tictoc-1588overm=
pls-05<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Abstract<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0This document defines the method for transporting Timing me=
ssages<br>
&gt;&gt;&gt; =A0such as PTP and NTP over an MPLS network. =A0The method all=
ows for<br>
&gt;&gt; the<br>
&gt;&gt;&gt; =A0easy identification of these PDUs at the port level to allo=
w for<br>
&gt;&gt; port<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; What is a port<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0level processing of these PDUs in both LERs and LSRs.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The basic idea is to transport Timing messages inside dedic=
ated<br>
&gt;&gt; MPLS<br>
&gt;&gt;&gt; =A0LSPs. =A0These LSPs only carry Timing messages and possibly=
 Control<br>
&gt;&gt; and<br>
&gt;&gt;&gt; =A0Management packets, but they do not carry customer traffic.=
<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; More specifically they only carry traffic associated wi=
th the<br>
&gt;&gt;&gt; SB&gt; timing service and its support.<br>
&gt;&gt;&gt; SB&gt; The question arose WRT BFD - surely BFD is allowed, BUT=
 surely<br>
&gt;&gt;&gt; SB&gt; also it gets carried in a structure that causes it to g=
et<br>
&gt;&gt;&gt; SB&gt; timestamped.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Two methods for transporting Timing messages over MPLS are =
defined.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Perhaps the right approach is to define the new LSP typ=
e and then<br>
&gt;&gt;&gt; SB&gt; seperately to define =A0the mapping of the various timi=
ng services<br>
&gt;&gt;&gt; SB&gt; over that LSP type.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The first method is to transport Timing messages directly o=
ver the<br>
&gt;&gt;&gt; =A0dedicated MPLS LSP via UDP/IP encapsulation, which is suita=
ble for<br>
&gt;&gt;&gt; =A0MPLS networks. =A0The second method is to transport Timing =
messages<br>
&gt;&gt;&gt; =A0inside a PW via Ethernet encapsulation.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; I think that we should note that there are some<br>
&gt;&gt;&gt; SB&gt; h/w reasons for this preference. A clean sheet approach=
<br>
&gt;&gt;&gt; SB&gt; would have been to use PTP over MPLS with no intermedia=
te<br>
&gt;&gt;&gt; SB&gt; layers.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Status of this Memo<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0This Internet-Draft is submitted in full conformance with t=
he<br>
&gt;&gt;&gt; =A0provisions of BCP 78 and BCP 79.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Internet-Drafts are working documents of the Internet Engin=
eering<br>
&gt;&gt;&gt; =A0Task Force (IETF). =A0Note that other groups may also distr=
ibute<br>
&gt;&gt;&gt; =A0working documents as Internet-Drafts. =A0The list of curren=
t<br>
&gt;&gt; Internet-<br>
&gt;&gt;&gt; =A0Drafts is at <a href=3D"http://datatracker.ietf.org/drafts/=
current/" target=3D"_blank">http://datatracker.ietf.org/drafts/current/</a>=
.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Internet-Drafts are draft documents valid for a maximum of =
six<br>
&gt;&gt; months<br>
&gt;&gt;&gt; =A0and may be updated, replaced, or obsoleted by other documen=
ts at<br>
&gt;&gt; any<br>
&gt;&gt;&gt; =A0time. =A0It is inappropriate to use Internet-Drafts as refe=
rence<br>
&gt;&gt;&gt; =A0material or to cite them other than as &quot;work in progre=
ss.&quot;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0This Internet-Draft will expire on December 17, 2013.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 [Page<br>
&gt;&gt; 1]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Copyright Notice<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Copyright (c) 2013 IETF Trust and the persons identified as=
 the<br>
&gt;&gt;&gt; =A0document authors. =A0All rights reserved.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0This document is subject to BCP 78 and the IETF Trust&#39;s=
 Legal<br>
&gt;&gt;&gt; =A0Provisions Relating to IETF Documents<br>
&gt;&gt;&gt; =A0(<a href=3D"http://trustee.ietf.org/license-info" target=3D=
"_blank">http://trustee.ietf.org/license-info</a>) in effect on the date of=
<br>
&gt;&gt;&gt; =A0publication of this document. =A0Please review these docume=
nts<br>
&gt;&gt;&gt; =A0carefully, as they describe your rights and restrictions wi=
th<br>
&gt;&gt; respect<br>
&gt;&gt;&gt; =A0to this document. =A0Code Components extracted from this do=
cument<br>
&gt;&gt; must<br>
&gt;&gt;&gt; =A0include Simplified BSD License text as described in Section=
 4.e of<br>
&gt;&gt;&gt; =A0the Trust Legal Provisions and are provided without warrant=
y as<br>
&gt;&gt;&gt; =A0described in the Simplified BSD License.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Table of Contents<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A01. =A0Introduction . . . . . . . . . . . . . . . . . . . . =
. . . . .<br>
&gt;&gt; 5<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A02. =A0Terminology =A0. . . . . . . . . . . . . . . . . . . =
. . . . . .<br>
&gt;&gt; 7<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A03. =A0Problem Statement =A0. . . . . . . . . . . . . . . . =
. . . . . .<br>
&gt;&gt; 8<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A04. =A0Timing over MPLS Architecture =A0. . . . . . . . . . =
. . . . . .<br>
&gt;&gt; 9<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A05. =A0Dedicated LSPs for Timing messages . . . . . . . . . =
. . . . .<br>
&gt;&gt; 12<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A06. =A0Timing over LSP Encapsulation =A0. . . . . . . . . . =
. . . . . .<br>
&gt;&gt; 13<br>
&gt;&gt;&gt; =A0 =A06.1. =A0Timing over UDP/IP over MPLS Encapsulation . . =
. . . . . .<br>
&gt;&gt; 13<br>
&gt;&gt;&gt; =A0 =A06.2. =A0Timing over PW Encapsulation . . . . . . . . . =
. . . . . .<br>
&gt;&gt; 13<br>
&gt;&gt;&gt; =A0 =A06.3. =A0Other Timing Encapsulation methods . . . . . . =
. . . . . .<br>
&gt;&gt; 14<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A07. =A0Timing message Processing =A0. . . . . . . . . . . . =
. . . . . .<br>
&gt;&gt; 15<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A08. =A0Protection and Redundancy =A0. . . . . . . . . . . . =
. . . . . .<br>
&gt;&gt; 16<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A09. =A0ECMP . . . . . . . . . . . . . . . . . . . . . . . . =
. . . . .<br>
&gt;&gt; 17<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A010. PHP =A0. . . . . . . . . . . . . . . . . . . . . . . . =
. . . . .<br>
&gt;&gt; 18<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A011. Entropy =A0. . . . . . . . . . . . . . . . . . . . . . =
. . . . .<br>
&gt;&gt; 19<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A012. OAM, Control and Management =A0. . . . . . . . . . . . =
. . . . .<br>
&gt;&gt; 20<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A013. QoS Considerations . . . . . . . . . . . . . . . . . . =
. . . .<br>
&gt;&gt; 21<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A014. FCS and Checksum Recalculation . . . . . . . . . . . . =
. . . .<br>
&gt;&gt; 22<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 [Page<br>
&gt;&gt; 2]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A015. Behavior of LER/LSR =A0. . . . . . . . . . . . . . . . =
. . . . .<br>
&gt;&gt; 23<br>
&gt;&gt;&gt; =A0 =A015.1. Behavior of Timing-capable/aware LER . . . . . . =
. . . . .<br>
&gt;&gt; 23<br>
&gt;&gt;&gt; =A0 =A015.2. Behavior of Timing-capable/aware LSR . . . . . . =
. . . . .<br>
&gt;&gt; 23<br>
&gt;&gt;&gt; =A0 =A015.3. Behavior of non-Timing-capable/aware LSR . . . . =
. . . . .<br>
&gt;&gt; 24<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A016. Other considerations . . . . . . . . . . . . . . . . . =
. . . .<br>
&gt;&gt; 25<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A017. Security Considerations =A0. . . . . . . . . . . . . . =
. . . . .<br>
&gt;&gt; 26<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A018. Acknowledgements . . . . . . . . . . . . . . . . . . . =
. . . .<br>
&gt;&gt; 27<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A019. IANA Considerations =A0. . . . . . . . . . . . . . . . =
. . . . .<br>
&gt;&gt; 28<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A020. References . . . . . . . . . . . . . . . . . . . . . . =
. . . .<br>
&gt;&gt; 29<br>
&gt;&gt;&gt; =A0 =A020.1. Normative References . . . . . . . . . . . . . . =
. . . . .<br>
&gt;&gt; 29<br>
&gt;&gt;&gt; =A0 =A020.2. Informative References . . . . . . . . . . . . . =
. . . . .<br>
&gt;&gt; 29<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Appendix 1. =A0Routing extensions for Timing-aware Routers =
. . . . .<br>
&gt;&gt; 32<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Appendix 2. =A0Signaling Extensions for Creating Timing LSP=
s . . . .<br>
&gt;&gt; 33<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Authors&#39; Addresses . . . . . . . . . . . . . . . . . . =
. . . . . .<br>
&gt;&gt; 34<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 [Page<br>
&gt;&gt; 3]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The key words &quot;MUST&quot;, &quot;MUST NOT&quot;, &quot=
;REQUIRED&quot;, &quot;SHALL&quot;, &quot;SHALL NOT&quot;,<br>
&gt;&gt;&gt; =A0&quot;SHOULD&quot;, &quot;SHOULD NOT&quot;, &quot;RECOMMEND=
ED&quot;, &quot;MAY&quot;, and &quot;OPTIONAL&quot; in<br>
&gt;&gt; this<br>
&gt;&gt;&gt; =A0document are to be interpreted as described in RFC2119 [RFC=
2119].<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When used in lower case, these words convey their typical u=
se in<br>
&gt;&gt;&gt; =A0common language, and are not to be interpreted as described=
 in<br>
&gt;&gt;&gt; =A0RFC2119 [RFC2119].<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 [Page<br>
&gt;&gt; 4]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 1. =A0Introduction<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The objective of Precision Time Protocol (PTP) and Network =
Timing<br>
&gt;&gt;&gt; =A0Protocol (NTP) are to synchronize independent clocks runnin=
g on<br>
&gt;&gt;&gt; =A0separate nodes of a distributed system.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[IEEE-1588] defines PTP messages for frequency, phase and t=
ime<br>
&gt;&gt;&gt; =A0synchronization. =A0The PTP messages include PTP PDUs over =
UDP/IP<br>
&gt;&gt;&gt; =A0(Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (=
Annex F<br>
&gt;&gt; of<br>
&gt;&gt;&gt; =A0[IEEE-1588]).<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Sure BUT it is acknowledged that IEEE is open to the de=
finition<br>
&gt;&gt;&gt; SB&gt; of other PTP mappings if they provide better optimisati=
on.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0This document defines mapping and transport of the PTP<br>
&gt;&gt;&gt; =A0messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.=
 =A0PTP<br>
&gt;&gt;&gt; =A0defines several clock types: ordinary clocks, boundary cloc=
ks, end-<br>
&gt;&gt;&gt; =A0to-end transparent clocks, and peer-to-peer transparent clo=
cks.<br>
&gt;&gt;&gt; =A0Transparent clocks require intermediate nodes to update cor=
rection<br>
&gt;&gt;&gt; =A0field inside PTP message that reflects the transit time in =
the<br>
&gt;&gt; node.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC5905] defines NTP messages for clock and time synchroni=
zation.<br>
&gt;&gt;&gt; =A0The PTP messages (PDUs) are transported over UDP/IP. =A0Thi=
s document<br>
&gt;&gt;&gt; SB&gt; Should that be NTP messages?<br>
&gt;&gt;&gt; SB&gt; It needs to be made clear as soon as you introduce NTP =
that<br>
&gt;&gt;&gt; SB&gt; they use different time representations.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0defines mapping and transport of the NTP messages defined i=
n<br>
&gt;&gt;&gt; =A0[RFC5905] over MPLS networks.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0One key attribute of all of these Timing messages is that t=
he Time<br>
&gt;&gt;&gt; =A0stamp processing should occur as close as possible to the a=
ctual<br>
&gt;&gt;&gt; =A0transmission and reception at the physical port interface. =
=A0This<br>
&gt;&gt;&gt; =A0targets optimal time and/or frequency recovery by avoiding =
variable<br>
&gt;&gt;&gt; =A0delay introduced by queues internal to the clocks.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; As I recall NTP has no epoch point defined, and I am no=
t sure<br>
&gt;&gt;&gt; SB&gt; where that point is in the case of PTP in this mapping<=
br>
&gt;&gt;&gt; SB&gt; Hopefully this will get defined in due course.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0To facilitate the fast and efficient recognition of Timing =
messages<br>
&gt;&gt;&gt; =A0at the port level when the Timing messages are carried over=
 MPLS<br>
&gt;&gt;&gt; =A0LSPs,<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Over a new LSP type with time optimied characteristics<=
br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0this document defines the specific encapsulations that shou=
ld<br>
&gt;&gt;&gt; =A0be used.<br>
&gt;&gt;&gt; SB&gt; Hopefully it will also define the PHP<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0In addition, it can be expected that there will exist LSR/<=
br>
&gt;&gt;&gt; =A0LERs where only a subset of the physical ports will have th=
e port-<br>
&gt;&gt;&gt; =A0based Timing message processing capabilities.<br>
&gt;&gt;&gt; SB&gt; Do you need to clarify that this only works at base and=
 not in<br>
&gt;&gt;&gt; SB&gt; a label heirarchy.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0In order to ensure<br>
&gt;&gt;&gt; =A0that the LSPs carrying Timing packets always enter and exit=
 ports<br>
&gt;&gt;&gt; =A0with this capability, routing extensions are defined to adv=
ertise<br>
&gt;&gt;&gt; =A0this capability on a port basis and to allow for the establ=
ishment<br>
&gt;&gt; of<br>
&gt;&gt;&gt; =A0LSPs that only transit such ports. =A0While this path estab=
lishment<br>
&gt;&gt;&gt; =A0restriction may be applied only at the LER Ingress and/or e=
gress<br>
&gt;&gt;&gt; =A0ports, it becomes more important when using transparent clo=
ck<br>
&gt;&gt; capable<br>
&gt;&gt;&gt; =A0LSRs in the path.<br>
&gt;&gt;&gt; SB&gt; I do not understand the implications of the last<br>
&gt;&gt;&gt; SB&gt; sentences - starting &quot;, it becomes&quot;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Port based Timing message processing involves Timing messag=
e<br>
&gt;&gt;&gt; =A0recognition. =A0Once the Timing messages are recognized the=
y can be<br>
&gt;&gt;&gt; =A0modified based on the reception or transmission Time-stamp.=
<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0This document provides two methods for transporting Timing =
messages<br>
&gt;&gt;&gt; =A0over MPLS. =A0One is applicable to MPLS environment and the=
 other one<br>
&gt;&gt;&gt; =A0is applicable to MPLS/MPLS-TP environment<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; I think the sentence is incomplete.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 [Page<br>
&gt;&gt; 5]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The solution involves transporting Timing messages over ded=
icated<br>
&gt;&gt;&gt; =A0LSPs called Timing LSPs. =A0These LSPs carry Timing message=
s and MAY<br>
&gt;&gt;&gt; =A0carry Management and control messages, but not data plane c=
lient<br>
&gt;&gt;&gt; =A0traffic.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; It is not clear why this restriction applies.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Timing LSPs can be established statically or via signaling.=
<br>
&gt;&gt;&gt; SB&gt; s/statically/by provisioning/network management/<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Extensions to control plane (OSPF, ISIS, etc.) is required =
to<br>
&gt;&gt; enable<br>
&gt;&gt;&gt; =A0routers to distribute their Timing processing capabilities =
over<br>
&gt;&gt; MPLS<br>
&gt;&gt;&gt; =A0to other routers. =A0However such extensions are outside th=
e scope of<br>
&gt;&gt;&gt; =A0this document.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When signaling is used to setup the PTP LSP, Extensions to<=
br>
&gt;&gt; signaling<br>
&gt;&gt;&gt; SB&gt; is it a PTP LSP or a Timing LSP?<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0protocols (e.g., RSVP-TE) are required for establishing PTP=
 LSPs.<br>
&gt;&gt;&gt; =A0However such extensions are outside the scope of this docum=
ent.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; for mpls-tp GMPLS is the signalling protocol<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0While the techniques included herein allow for the establis=
hment of<br>
&gt;&gt;&gt; =A0paths optimized to include Time-stamping capable links, the=
<br>
&gt;&gt;&gt; =A0performance of the Slave clocks is outside the scope of thi=
s<br>
&gt;&gt;&gt; =A0document.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0At the time of publishing this specification, Transparent C=
locking<br>
&gt;&gt;&gt; =A0(TC) is only defined for PTP. =A0Therefore at this time any=
 part of<br>
&gt;&gt;&gt; =A0this specification that talks about Transparent Clocking ap=
plies<br>
&gt;&gt; only<br>
&gt;&gt;&gt; =A0to PTP.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 [Page<br>
&gt;&gt; 6]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 2. =A0Terminology<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A01588: The timing and synchronization as defined by IEEE 158=
8.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; I think that there is a more formal name for the 1588 g=
roup<br>
&gt;&gt;&gt; SB&gt; that needs to be used here.<br>
&gt;&gt;&gt; SB&gt; Also do we need to talk about 1588-200? as there is<br>
&gt;&gt;&gt; SB&gt; an update in progress<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0NTP: The timing and synchronization protocol defined by IET=
F RFC-<br>
&gt;&gt; 1305<br>
&gt;&gt;&gt; =A0and RFC-5905.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0PTP: The timing and synchronization protocol used by 1588.<=
br>
&gt;&gt;&gt; SB&gt; need the proper name for 1588<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Master Clock: The source of 1588 timing to a set of slave c=
locks.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Master Port: A port on a ordinary or boundary clock that is=
 in<br>
&gt;&gt; Master<br>
&gt;&gt;&gt; =A0state. =A0This is the source of timing toward slave ports.<=
br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; I am not sure the reader knows what a port is<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Slave Clock: A receiver of 1588 timing from a master clock.=
<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Slave Port: A port on a boundary clock or ordinary clock th=
at is<br>
&gt;&gt;&gt; =A0receiving timing from a master clock.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Ordinary Clock: A device with a single PTP port.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Transparent Clock. =A0A device that measures the time taken=
 for a PTP<br>
&gt;&gt;&gt; =A0event message to transit the device and then updates the<br=
>
&gt;&gt;&gt; =A0correctionField of the message with this transit time.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Boundary Clock: A device with more than one PTP port. =A0Ge=
nerally<br>
&gt;&gt;&gt; =A0boundary clocks will have one port in slave state to receiv=
e timing<br>
&gt;&gt;&gt; =A0and then other ports in master state to re-distribute the t=
iming.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0PTP LSP: An LSP dedicated to carry PTP messages<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; PTP or timing?<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0PTP PW: A PW within a PTP LSP that is dedicated to carry PT=
P<br>
&gt;&gt;&gt; =A0messages.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Ah I don&#39;t think that PWE3 know what one of these i=
s<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0CW: Pseudowire Control Word<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0LAG: Link Aggregation<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0ECMP: Equal Cost Multipath<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0CF: Correction Field, a field inside certain PTP messages (=
message<br>
&gt;&gt;&gt; =A0type 0-3)that holds the accumulative transit time inside<br=
>
&gt;&gt; intermediate<br>
&gt;&gt;&gt; =A0switches<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Timing messages: Timing Protocol messages that are exchange=
d<br>
&gt;&gt; between<br>
&gt;&gt;&gt; =A0routers in order to establish a synchronized clock.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; A number of these definitions look like copies of IEEE1=
588<br>
&gt;&gt;&gt; SB&gt; definitions. We need to provide references and note the=
<br>
&gt;&gt;&gt; SB&gt; priority of the IEEE base reference.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 [Page<br>
&gt;&gt; 7]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 3. =A0Problem Statement<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[IEEE-1588] has defined methods for transporting PTP messag=
es over<br>
&gt;&gt;&gt; =A0Ethernet and IP networks. =A0[RFC5905] has defined the meth=
od of<br>
&gt;&gt;&gt; =A0transporting NTP messages over IP networks. =A0There is a n=
eed to<br>
&gt;&gt;&gt; =A0transport Timing messages over MPLS networks while supporti=
ng the<br>
&gt;&gt;&gt; =A0Transparent Clock (TC), Boundary Clock (BC) and Ordinary Cl=
ock (OC)<br>
&gt;&gt;&gt; =A0functionality in the LER and LSRs in the MPLS network.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0There are multiple ways of transporting Timing over MPLS. =
=A0However,<br>
&gt;&gt;&gt; =A0there is a requirement to limit the possible encapsulation =
options<br>
&gt;&gt; to<br>
&gt;&gt;&gt; =A0simplify the Timing message identification and processing r=
equired<br>
&gt;&gt; at<br>
&gt;&gt;&gt; =A0the port level.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When Timing-awareness is needed, Timing messages should not=
 be<br>
&gt;&gt;&gt; =A0transported over LSPs or PWs that are carrying customer tra=
ffic<br>
&gt;&gt;&gt; =A0because LSRs perform Label switching based on the top label=
 in the<br>
&gt;&gt;&gt; =A0stack.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Have you explained why?<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0To detect Timing messages inside such LSPs require special<=
br>
&gt;&gt;&gt; =A0hardware to do deep packet inspection at line rate. =A0Even=
 if such<br>
&gt;&gt;&gt; =A0hardware exists, the payload can&#39;t be deterministically=
 identified<br>
&gt;&gt; by<br>
&gt;&gt;&gt; =A0LSRs because the payload type is a context of the PW label,=
 and the<br>
&gt;&gt;&gt; =A0PW label and its context are only known to the Edge routers=
 (PEs/<br>
&gt;&gt;&gt; =A0LERs); LSRs dont know what is a PWs payload (Ethernet, ATM,=
 FR,<br>
&gt;&gt; CES,<br>
&gt;&gt;&gt; =A0etc). =A0Even if one restricts an LSP to only carry Etherne=
t PWs, the<br>
&gt;&gt;&gt; =A0LSRs dont have the knowledge of whether PW Control Word (CW=
) is<br>
&gt;&gt;&gt; =A0present or not and therefore can not deterministically iden=
tify the<br>
&gt;&gt;&gt; =A0payload.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0A generic method is defined in this document that does not =
require<br>
&gt;&gt;&gt; =A0deep packet inspection at line rate, and can deterministica=
lly<br>
&gt;&gt;&gt; =A0identify Timing messages. =A0This method can be used to det=
ect Timing<br>
&gt;&gt;&gt; =A0Messages in both one-step and two-step clock implementation=
s of<br>
&gt;&gt;&gt; =A0ordinary, boundary and transparent clocks.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Needs a ref and I am sure many MPLS specialists will no=
t<br>
&gt;&gt; understand<br>
&gt;&gt;&gt; SB&gt; the msg types.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 [Page<br>
&gt;&gt; 8]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 4. =A0Timing over MPLS Architecture<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Timing messages are exchange between Timing ports on ordina=
ry and<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Have you defined a timing port?<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0boundary clocks. =A0Boundary clocks terminate the Timing me=
ssages and<br>
&gt;&gt;&gt; =A0act as master for other boundary clocks or for slave clocks=
. =A0End-<br>
&gt;&gt; to-<br>
&gt;&gt;&gt; =A0End Transparent clocks do not terminate the Timing messages=
 but<br>
&gt;&gt; they<br>
&gt;&gt;&gt; =A0do modify the contents of the Timing messages as they trans=
it<br>
&gt;&gt; across<br>
&gt;&gt;&gt; =A0the transparent clock.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Master/Slave clocks (OCs), Boundary Clocks (BC) and Transpa=
rent<br>
&gt;&gt; Clock<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0(TC) could be implemented in either LERs or LSRs.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; LER and LSR need to be expanded<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0An example is shown in Figure 1, where the LERs act as Ordi=
nary<br>
&gt;&gt; Clock<br>
&gt;&gt;&gt; =A0(OC) and are the initiating/terminating point for Timing me=
ssages.<br>
&gt;&gt;&gt; =A0The ingress LER encapsulates the Timing messages in Timing =
LSP and<br>
&gt;&gt;&gt; =A0the Egress LER terminates the Timing LSP. =A0The LSRs act a=
s<br>
&gt;&gt;&gt; =A0Transparent Clock (TC) and just update the Timing field in =
the<br>
&gt;&gt; Timing<br>
&gt;&gt;&gt; =A0messages.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0 +--------+ =A0 =A0 +-------+ =A0 =A0 +-------+ =A0 =A0=
 +-------+ =A0 =A0 +------<br>
&gt;&gt; --+<br>
&gt;&gt;&gt; =A0 =A0 |Switch, | =A0 =A0 | =A0 =A0 =A0 | =A0 =A0 | =A0 =A0 =
=A0 | =A0 =A0 | =A0 =A0 =A0 |<br>
&gt;&gt; |Switch, |<br>
&gt;&gt;&gt; =A0 =A0 | Router |-----| =A0LER =A0|-----| =A0LSR =A0|-----| =
=A0LER =A0|-----|<br>
&gt;&gt; Router |<br>
&gt;&gt;&gt; =A0 =A0 | =A0 =A0 =A0 =A0| =A0 =A0 | =A0OC =A0 | =A0 =A0 | =A0=
TC =A0 | =A0 =A0 | =A0OC =A0 | =A0 =A0 |<br>
&gt;&gt; |<br>
&gt;&gt;&gt; =A0 =A0 +--------+ =A0 =A0 +-------+ =A0 =A0 +-------+ =A0 =A0=
 +-------+ =A0 =A0 +------<br>
&gt;&gt; --+<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0/ =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 \<br>
&gt;&gt;&gt; =A0 =A0 +-------+ =A0 =A0 / =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 \ =A0 =A0 +-------<br>
&gt;&gt; +<br>
&gt;&gt;&gt; =A0 =A0 | =A0LER =A0| =A0 =A0/ =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 \ =A0 =A0| =A0LER<br>
&gt;&gt; |<br>
&gt;&gt;&gt; =A0 =A0 | Master|---/ =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 \---| Slave<br>
&gt;&gt; |<br>
&gt;&gt;&gt; =A0 =A0 | Clock | =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 | Clock<br>
&gt;&gt; |<br>
&gt;&gt;&gt; =A0 =A0 +-------+ =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 +-------<br>
&gt;&gt; +<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0Figure (1) - Deployment example 1 of timing over MPLS n=
etwork<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Another example is shown in Figure2, where LERs terminate t=
he<br>
&gt;&gt; Timing<br>
&gt;&gt;&gt; =A0messages received from switch/routers that are outside of t=
he MPLS<br>
&gt;&gt;&gt; =A0network acting as OC or BC. =A0In this example LERs regener=
ate the<br>
&gt;&gt;&gt; =A0clock and initiate timing messages encapsulated in Timing L=
SP<br>
&gt;&gt; toward<br>
&gt;&gt;&gt; =A0the MPLS network, while the LSRs act as Transparent Clock (=
TC) and<br>
&gt;&gt;&gt; =A0just update the Timing field in the Timing messages, which =
are<br>
&gt;&gt;&gt; =A0already encapsulated in Timing LSPs.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 [Page<br>
&gt;&gt; 9]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0+--------+ =A0 =A0 +-------+ =A0 =A0 +-------+ =A0 =A0 =
+-------+ =A0 =A0 +-------<br>
&gt;&gt; -+<br>
&gt;&gt;&gt; =A0 =A0|Switch, | =A0 =A0 | =A0 =A0 =A0 | =A0 =A0 | =A0 =A0 =
=A0 | =A0 =A0 | =A0 =A0 =A0 | =A0 =A0 |Switch,<br>
&gt;&gt; |<br>
&gt;&gt;&gt; =A0 =A0| Router |-----| =A0LER =A0|-----| =A0LSR =A0|-----| =
=A0LER =A0|-----| Router<br>
&gt;&gt; |<br>
&gt;&gt;&gt; =A0 =A0| OC/BC =A0| =A0 =A0 | =A0BC =A0 | =A0 =A0 | =A0TC =A0 =
| =A0 =A0 | =A0BC =A0 | =A0 =A0 | OC/BC<br>
&gt;&gt; |<br>
&gt;&gt;&gt; =A0 =A0+--------+ =A0 =A0 +-------+ =A0 =A0 +-------+ =A0 =A0 =
+-------+ =A0 =A0 +-------<br>
&gt;&gt; -+<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0Figure (2) - Deployment example 2 of timing over MPLS n=
etwork<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Another example is shown in Figure 3, where LERs do not ter=
minate<br>
&gt;&gt; the<br>
&gt;&gt;&gt; =A0Timing messages received from switch/routers that are outsi=
de of<br>
&gt;&gt; the<br>
&gt;&gt;&gt; =A0MPLS network acting as OC, TC or BC. =A0The LERs act as TC =
and update<br>
&gt;&gt;&gt; =A0the Timing field in the Timing messages as they transit the=
 LER,<br>
&gt;&gt;&gt; =A0while encapsulating them in timing LSP. =A0The LSRs also ac=
t as<br>
&gt;&gt;&gt; =A0Transparent Clock (TC) and just update the Timing field in =
the<br>
&gt;&gt; Timing<br>
&gt;&gt;&gt; =A0messages which are already encapsulated in Timing LSPs.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0 +--------+ =A0 =A0 +-------+ =A0 =A0 +-------+ =A0 =A0=
 +-------+ =A0 =A0 +------<br>
&gt;&gt; --+<br>
&gt;&gt;&gt; =A0 =A0 |Switch, | =A0 =A0 | =A0 =A0 =A0 | =A0 =A0 | =A0 =A0 =
=A0 | =A0 =A0 | =A0 =A0 =A0 |<br>
&gt;&gt; |Switch, |<br>
&gt;&gt;&gt; =A0 =A0 | Router |-----| =A0LER =A0|-----| =A0LSR =A0|-----| =
=A0LER =A0|-----|<br>
&gt;&gt; Router |<br>
&gt;&gt;&gt; =A0 =A0 |OC/TC/BC| =A0 =A0 | =A0TC =A0 | =A0 =A0 | =A0TC =A0 |=
 =A0 =A0 | =A0TC =A0 |<br>
&gt;&gt; |OC/TC/BC|<br>
&gt;&gt;&gt; =A0 =A0 +--------+ =A0 =A0 +-------+ =A0 =A0 +-------+ =A0 =A0=
 +-------+ =A0 =A0 +------<br>
&gt;&gt; --+<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 Figure (3) - Deployment example 3 of timing over MPLS netw=
ork<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Another example is shown in Figure 4, where LERs and LSRs s=
upport<br>
&gt;&gt;&gt; =A0Boundary Clocks. =A0A single-hop LSP is created between two=
 adjacent<br>
&gt;&gt;&gt; =A0LSRs engaged in BC operation. =A0Other methods such as PTP =
transport<br>
&gt;&gt;&gt; =A0over Ethernet MAY be used for transporting timing messages =
if the<br>
&gt;&gt;&gt; =A0link between the two routers is Ethernet.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0+--------+ =A0 =A0 +-------+ =A0 =A0 +-------+ =A0 =A0 =
+-------+ =A0 =A0 +-------<br>
&gt;&gt; -+<br>
&gt;&gt;&gt; =A0 =A0|Switch, | =A0 =A0 | =A0 =A0 =A0 | =A0 =A0 | =A0 =A0 =
=A0 | =A0 =A0 | =A0 =A0 =A0 | =A0 =A0 |Switch,<br>
&gt;&gt; |<br>
&gt;&gt;&gt; =A0 =A0| Router |-----| =A0LER =A0|-----| =A0LSR =A0|-----| =
=A0LER =A0|-----| Router<br>
&gt;&gt; |<br>
&gt;&gt;&gt; =A0 =A0| OC/BC =A0| =A0 =A0 | =A0BC =A0 | =A0 =A0 | =A0BC =A0 =
| =A0 =A0 | =A0BC =A0 | =A0 =A0 | OC/BC<br>
&gt;&gt; |<br>
&gt;&gt;&gt; =A0 =A0+--------+ =A0 =A0 +-------+ =A0 =A0 +-------+ =A0 =A0 =
+-------+ =A0 =A0 +-------<br>
&gt;&gt; -+<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Figure (4) - Deployment example 3 of timing over MPLS netwo=
rk<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0An MPLS domain MAY serve multiple customers. =A0In these ca=
ses the<br>
&gt;&gt; MPLS<br>
&gt;&gt;&gt; =A0domain (maintained by a service provider) may provide timin=
g<br>
&gt;&gt; services<br>
&gt;&gt;&gt; =A0to multiple customers, each having their own Timing domain.=
<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The Timing over MPLS architecture assumes full mesh of Timi=
ng LSPs<br>
&gt;&gt;&gt; =A0between all LERs supporting this specification.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Note sure this is right - the salves surely do not need=
 to<br>
&gt;&gt;&gt; SB&gt; exchange timing amongst themselves<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0It supports<br>
&gt;&gt;&gt; =A0Point-to- point (VPWS) and Multipoint (VPLS) services.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; What does that mean? You do not carry user data traffic=
?<br>
&gt;&gt;&gt; SB&gt; Maybe it&#39;s the ordering of the statemnets that is c=
ausing<br>
&gt;&gt;&gt; SB&gt; confusion.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0This means<br>
&gt;&gt;&gt; =A0that a customer may purchase a Point-to-point Timing servic=
e<br>
&gt;&gt; between<br>
&gt;&gt;&gt; =A0two customer sites or a Multipoint Timing service between m=
ore than<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 10]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0two customer sites.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The Timing over MPLS architecture supports P2P or P2MP Timi=
ng LSPs.<br>
&gt;&gt;&gt; =A0This means that the Timing Multicast messages such as PTP M=
ulticast<br>
&gt;&gt;&gt; =A0event messages can be transported over P2MP Timing LSP or b=
e<br>
&gt;&gt;&gt; =A0replicated and transported over many P2P Timing LSPs.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Note we do not yet have a definition of a P2MP mpls-tp =
LSP<br>
&gt;&gt;&gt; SB&gt; nor a P2MP PW, although we are close.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Timing messages, that do not require Time stamping or Corre=
ction<br>
&gt;&gt;&gt; =A0Field update MAY be transported over Timing LSPs to simplif=
y<br>
&gt;&gt; hardware<br>
&gt;&gt;&gt; =A0and software.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0PTP Announce messages that determine the Timing LSP termina=
ting<br>
&gt;&gt; point<br>
&gt;&gt;&gt; =A0behavior such as BC/OC/TC SHOULD be transported over the Ti=
ming LSP<br>
&gt;&gt;&gt; =A0to simplify hardware and software.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; have you defined and referenced PTP announce msgs?<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 11]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 5. =A0Dedicated LSPs for Timing messages<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Many methods have been considered for identifying the Timin=
g<br>
&gt;&gt; messages<br>
&gt;&gt;&gt; =A0when they are encapsulated in MPLS such as using GAL/G-ACH =
or a new<br>
&gt;&gt;&gt; =A0reserved label. =A0These methods were not attractive since =
they<br>
&gt;&gt; either<br>
&gt;&gt;&gt; =A0required deep packet inspection at line rate in the interme=
diate<br>
&gt;&gt; LSRs<br>
&gt;&gt;&gt; =A0or they required use of a scarce new reserved label. =A0Als=
o one of<br>
&gt;&gt; the<br>
&gt;&gt;&gt; =A0goals was to reuse existing OAM mechanisms.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; RLs =3D SPLs are not so rare now. In any case needs a r=
ef.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The method defined in this document can be used by LER and =
LSRs to<br>
&gt;&gt;&gt; =A0identify Timing messages in MPLS tunnels by just looking at=
 the top<br>
&gt;&gt;&gt; =A0label in the MPLS label stack, which only carry Timing mess=
ages as<br>
&gt;&gt;&gt; =A0well as OAM, but not data plane client traffic.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Compliant implementations MUST use dedicated LSPs to carry =
Timing<br>
&gt;&gt;&gt; =A0messages over MPLS.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; I think that we need a definition of the properies of t=
hese LSPs<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0These LSPs are herein referred to as &quot;Timing<br>
&gt;&gt;&gt; =A0LSPs&quot; and the labels associated with these LSPs as &qu=
ot;Timing LSP<br>
&gt;&gt;&gt; =A0labels&quot;. =A0The Timing LSPs that runs between Ingress =
and Egress LERs<br>
&gt;&gt;&gt; =A0MUST be co-routed. =A0Alternatively, a single bidirectional=
 co-routed<br>
&gt;&gt;&gt; =A0LSP can be used.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; I though that you said you could use M2MP LSPs - these =
are not<br>
&gt;&gt;&gt; SB&gt; bidirectional.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Co-routing of the two directions is required to limit the<b=
r>
&gt;&gt; difference<br>
&gt;&gt;&gt; =A0in the delays in the Master clock to Slave clock direction =
compared<br>
&gt;&gt;&gt; =A0to the Slave clock to Master clock direction. =A0The Timing=
 LSP MAY<br>
&gt;&gt; be<br>
&gt;&gt;&gt; =A0MPLS/MPLS-TP LSP.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The Timing LSPs could be configured or signaled via RSVP-TE=
/GMPLS.<br>
&gt;&gt;&gt; =A0New Extensions to RSVP-TE/GMPLS TLVs are required; however =
they are<br>
&gt;&gt;&gt; =A0outside the scope of this document.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffi=
c such<br>
&gt;&gt; as<br>
&gt;&gt;&gt; =A0BFD and LSP Ping but the LSP data plane client plane traffi=
c MUST<br>
&gt;&gt; be<br>
&gt;&gt;&gt; =A0Timing packets only.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Why?<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 12]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 6. =A0Timing over LSP Encapsulation<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; The encapsulations is not LSP is it?<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0This document defines two methods for carrying Timing messa=
ges over<br>
&gt;&gt;&gt; =A0MPLS. =A0The first method is carrying UDP/IP encapsulated T=
iming<br>
&gt;&gt;&gt; =A0messages over Timing LSPs, and the second method, is carryi=
ng<br>
&gt;&gt;&gt; =A0Ethernet encapsulated Timing messages over Ethernet PWs ins=
ide<br>
&gt;&gt; Timing<br>
&gt;&gt;&gt; =A0LSPs.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 6.1. =A0Timing over UDP/IP over MPLS Encapsulation<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The simplest method of transporting Timing messages over MP=
LS is to<br>
&gt;&gt;&gt; =A0encapsulate Timing PDUs in UDP/IP and then encapsulate them=
 in<br>
&gt;&gt; Timing<br>
&gt;&gt;&gt; =A0LSP. =A0This format is shown in Figure 4.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 +----------------------+<b=
r>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 | =A0 Timing LSP Label =A0=
 |<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 +----------------------+<b=
r>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 | =A0 =A0 =A0 =A0IPv4/6 =
=A0 =A0 =A0 =A0|<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 +----------------------+<b=
r>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 | =A0 =A0 =A0 =A0 UDP =A0 =
=A0 =A0 =A0 =A0|<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 +----------------------+<b=
r>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 | =A0 =A0 Timing PDU =A0 =
=A0 =A0 |<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 +----------------------+<b=
r>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0 Figure (4) - Timing over UDP/IP over MPLS Encapsulatio=
n<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0This encapsulation is very simple and is useful when the ne=
twork<br>
&gt;&gt;&gt; =A0between Timing Master Clock and Slave Clock is MPLS network=
.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Simple is a judgement call<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0In order for an LER/LSR to process Timing messages, the Tim=
ing LSP<br>
&gt;&gt;&gt; =A0Label must be at the top label of the label stack. =A0The L=
ER/LSR<br>
&gt;&gt; MUST<br>
&gt;&gt;&gt; =A0know that the Timing LSP Label is used for carrying Timing<=
br>
&gt;&gt; messages.<br>
&gt;&gt;&gt; =A0This can be accomplished via static configuration or via RS=
VP-TE<br>
&gt;&gt;&gt; =A0signaling.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The UDP/IP encapsulation of PTP MUST follow Annex D and E o=
f<br>
&gt;&gt;&gt; =A0[IEEE-1588]. =A0While the UDP/IP encapsulation of NTP MUST =
follow<br>
&gt;&gt;&gt; =A0[RFC5905].<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 6.2. =A0Timing over PW Encapsulation<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Another method of transporting Timing over MPLS networks is=
 by<br>
&gt;&gt;&gt; =A0encapsulating Timing PDUs in PW which in turn is transporte=
d over<br>
&gt;&gt;&gt; =A0Timing LSPs. =A0In case of PTP, Ethernet PW encapsulation [=
RFC4448],<br>
&gt;&gt;&gt; =A0shown in Fig 5(A) MUST be used and the Ethernet encapsulati=
on of<br>
&gt;&gt; PTP<br>
&gt;&gt;&gt; =A0MUST follow Annex F of [IEEE-1588].<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 13]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The RAW mode or Tagged mode defined in [RFC4448] MAY be use=
d and<br>
&gt;&gt; the<br>
&gt;&gt;&gt; =A0Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).=
 =A0The<br>
&gt;&gt;&gt; =A0Timing over PW encapsulation MUST use the Control Word (CW)=
 as<br>
&gt;&gt;&gt; =A0specified in [RFC4448] to ensure proper detection of PTP me=
ssages<br>
&gt;&gt;&gt; =A0inside the MPLS packets for Timing over LSP and Timing over=
 PW<br>
&gt;&gt;&gt; =A0encapsulation.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; That needs explanation<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The use of Sequence Number in the CW is optional.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Given that s/n are never in practice deployed, you coul=
d probably<br>
&gt;&gt;&gt; SB&gt; simplify things by sayig that they are not used.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Timing over PW encapsulation for NTP MUST use NTP over UDP/=
IP over<br>
&gt;&gt; PW<br>
&gt;&gt;&gt; =A0(the IP PW discussed in [RFC4447]) shown in Fig 5(B).<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 +----------------+ =A0+---=
-------------+<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0|Timing LSP Label| =A0|=
Timing LSP Label|<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0+----------------+ =A0+=
----------------+<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0| =A0 =A0PW Label =A0 =
=A0| =A0| =A0 =A0PW Label =A0 =A0|<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0+----------------+ =A0+=
----------------+<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0| =A0Control Word =A0| =
=A0| =A0 =A0 =A0IP =A0 =A0 =A0 =A0|<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0+----------------+ =A0+=
----------------+<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0| =A0 =A0Ethernet =A0 =
=A0| =A0| =A0 =A0 =A0UDP =A0 =A0 =A0 |<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0| =A0 =A0 Header =A0 =
=A0 | =A0+----------------+<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0+----------------+ =A0|=
 =A0 Timing PDU =A0 |<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0|S-VLAN(Optional)| =A0|=
 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0|<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0+----------------+ =A0+=
----------------+<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0|C-VLAN(Optional)| =A0 =
=A0 =A0 =A0(B)<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0+----------------+<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0| =A0 Timing PDU =A0 |<=
br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0|<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0+----------------+<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 (A)<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Figure (5) - Timing over PW Encapsulat=
ions<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0In order for an LSR to process PTP messages, the top label =
of the<br>
&gt;&gt;&gt; =A0label stack (the Tunnel Label) MUST be a Timing label.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; S&gt; You said that before.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 6.3. =A0Other Timing Encapsulation methods<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0In future other timing encapsulation methods may be introdu=
ced,<br>
&gt;&gt; such<br>
&gt;&gt;&gt; =A0as a new shim header after the Bottom of Stack to carry the=
 Timing<br>
&gt;&gt;&gt; =A0information. =A0Such new encapsulations are outside the sco=
pe of this<br>
&gt;&gt;&gt; =A0document.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Taking a pure MPLS pov, you can simplify a lot of the t=
ext<br>
&gt;&gt;&gt; SB&gt; out of the definition of the LSP<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 14]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; I think we need a section on LSP processing<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 7. =A0Timing message Processing<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Each Timing protocol such as PTP and NTP, define their set =
of<br>
&gt;&gt; Timing<br>
&gt;&gt;&gt; =A0messages. =A0For example PTP defines SYNC, DELAY_REQ, DELAY=
_RESP,<br>
&gt;&gt;&gt; =A0FOLLOW_UP, etc messages.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Some of the Timing messages require time stamping or correc=
tion<br>
&gt;&gt; field<br>
&gt;&gt;&gt; =A0update at port level and some dont. =A0It is the job of the=
 LER/LSR<br>
&gt;&gt; to<br>
&gt;&gt;&gt; =A0parse the timing message and find out the type of the Timin=
g<br>
&gt;&gt; message<br>
&gt;&gt;&gt; =A0and decide whether and how to Time- stamp it (e.g., BC) or =
update<br>
&gt;&gt;&gt; =A0correction field(e.g., TC).<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; AAAAAAAAAAAH. Surely this is the function of the PTP pr=
cessing<br>
&gt;&gt;&gt; SB&gt; function rather than the LER?<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0For example the following PTP messages (called Event messag=
es)<br>
&gt;&gt;&gt; =A0require time-stamping or correction field update:<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0o =A0SYNC<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0o =A0DELAY_REQ (Delay Request)<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0o =A0PDELAY_REQ (Peer Delay Request)<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0o =A0PDELAY_RESP (Peer Delay Response)<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0SYNC and DELAY_REQ are exchanged between Master Clock and S=
lave<br>
&gt;&gt; Clock<br>
&gt;&gt;&gt; =A0and MUST be transported over PTP LSPs. =A0PDELAY_REQ and PD=
ELAY_RESP<br>
&gt;&gt;&gt; =A0are exchanged between adjacent PTP clocks (i.e. =A0Master, =
Slave,<br>
&gt;&gt;&gt; =A0Boundary, or Transparent) and SHOULD be transported over si=
ngle hop<br>
&gt;&gt;&gt; =A0PTP LSPs. =A0If Two Step PTP clocks are present, then the F=
OLLOW_UP,<br>
&gt;&gt;&gt; =A0and PDELAY_RESP_FOLLOW_UP messages MUST also be transported=
 over<br>
&gt;&gt; the<br>
&gt;&gt;&gt; =A0PTP LSPs.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0For a given instance of 1588 protocol, SYNC and DELAY_REQ M=
UST be<br>
&gt;&gt;&gt; =A0transported over two PTP LSPs that are in opposite directio=
ns.<br>
&gt;&gt; These<br>
&gt;&gt;&gt; =A0PTP LSPs, which are in opposite directions MUST be congruen=
t and<br>
&gt;&gt; co-<br>
&gt;&gt;&gt; =A0routed. =A0Alternatively, a single bidirectional co-routed =
LSP can be<br>
&gt;&gt;&gt; =A0used.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Except as indicated above for the two-step PTP clocks, Non-=
Event<br>
&gt;&gt; PTP<br>
&gt;&gt;&gt; =A0message types do not need to be processed by intermediate r=
outers.<br>
&gt;&gt;&gt; =A0These message types MAY be carried in PTP Tunnel LSPs.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; Are you saying that a timing P router has to be msg typ=
e<br>
&gt;&gt; sensitive?<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 15]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 8. =A0Protection and Redundancy<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; This is a bit of a jump - I don&#39;t know how the LSP =
itself works<br>
&gt;&gt; yet!<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0In order to ensure continuous uninterrupted operation of sl=
ave<br>
&gt;&gt;&gt; =A0clocks, usually as a general practice, slave clocks (or por=
ts)<br>
&gt;&gt; track<br>
&gt;&gt;&gt; =A0redundant master clocks.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0It is the responsibility of the network operator to ensure =
that<br>
&gt;&gt;&gt; =A0physically disjoint Timing LSPs are established between a s=
lave<br>
&gt;&gt; clock<br>
&gt;&gt;&gt; =A0(or port) and redundant master clocks (or ports).<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When a slave clock (or port) listens to redundant master cl=
ocks or<br>
&gt;&gt;&gt; =A0ports, any prolonged Timing LSP outage will trigger the sla=
ve clock<br>
&gt;&gt;&gt; =A0or port to switch to a redundant master clock or port.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0LSP/PW protection such as Linear protection Switching (1:1,=
 1+1),<br>
&gt;&gt;&gt; =A0Ring protection switching or MPLS Fast Reroute (FRR) genera=
lly<br>
&gt;&gt; switch<br>
&gt;&gt;&gt; =A0alternative path that usually cause a change in delay, whic=
h if<br>
&gt;&gt;&gt; =A0undetected by slave clock can reduce accuracy of the slave =
clock.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Therefore protection switching MAY be used, as long as phas=
e jumps<br>
&gt;&gt;&gt; =A0upon switchover due to differences in path latency are dete=
cted and<br>
&gt;&gt;&gt; =A0compensated for (such compensation not being required if BC=
s or<br>
&gt;&gt; peer-<br>
&gt;&gt;&gt; =A0peer TCs are used throughout).<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Note that any protection or reroute mechanism that adds add=
itional<br>
&gt;&gt;&gt; =A0MPLS label to the label stack, such as Facility Backup Fast=
<br>
&gt;&gt; Reroute,<br>
&gt;&gt;&gt; =A0MUST ensure that the pushed label is also a Timing Label to=
 ensure<br>
&gt;&gt;&gt; =A0recognition of the MPLS frame as containing Timing messages=
, as it<br>
&gt;&gt;&gt; =A0transits the backup path.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 16]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 9. =A0ECMP<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0To ensure the optimal operation of slave clocks and avoid e=
rror<br>
&gt;&gt;&gt; =A0introduced by forward and reverse path delay asymmetry, the=
<br>
&gt;&gt; physical<br>
&gt;&gt;&gt; =A0path for Timing messages from master clock to slave Clock a=
nd vice<br>
&gt;&gt;&gt; =A0versa must be the same for all Event Timing messages listed=
 in<br>
&gt;&gt;&gt; =A0section 7.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Therefore the Timing LSPs MUST not be subject to ECMP (Equa=
l Cost<br>
&gt;&gt;&gt; =A0Multipath).<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 17]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 10. =A0PHP<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0To ensure that the label on the top of the label stack is t=
he<br>
&gt;&gt; Timing<br>
&gt;&gt;&gt; =A0LSP Label, PHP MUST not be used.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 18]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 11. =A0Entropy<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0To ensure all Timing messages in a Timing LSP take the same=
 path,<br>
&gt;&gt;&gt; =A0Entropy Label MUST NOT be used for the Timing LSP[RFC6790] =
and<br>
&gt;&gt;&gt; =A0Entropy Label MUST NOT be used for the PWs that are carried=
 inside<br>
&gt;&gt;&gt; =A0Timing LSP [RFC6391].<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; This is incorrect - you mean that all msgs of the same =
timing<br>
&gt;&gt;&gt; SB&gt; flow need to have the same EL value.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 19]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 12. =A0OAM, Control and Management<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0In order to monitor Timing LSPs and their encapsulated PWs,=
 they<br>
&gt;&gt; MUST<br>
&gt;&gt;&gt; =A0be able to carry OAM and management messages. =A0These mana=
gement<br>
&gt;&gt;&gt; =A0messages MUST be differentiated from Timing messages via al=
ready<br>
&gt;&gt;&gt; =A0defined IETF methods.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389]=
 MAY run<br>
&gt;&gt;&gt; =A0over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH. =
=A0These<br>
&gt;&gt;&gt; =A0Management protocols can easily be identified by the UDP<br=
>
&gt;&gt; Destination<br>
&gt;&gt;&gt; =A0Port number or by GAL/G-ACH respectively.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Also BFD, LSP-Ping and other management messages MAY run ov=
er the<br>
&gt;&gt; PWs<br>
&gt;&gt;&gt; =A0encapsulated in Timing LSP via one of the defined VCCVs (Ty=
pe 1, 3<br>
&gt;&gt; or<br>
&gt;&gt;&gt; =A04) [RFC5085] (note that VCCV Type 2 using Router Alert Labe=
l is<br>
&gt;&gt; going<br>
&gt;&gt;&gt; =A0to be deprecated by IETF). =A0In this case G-ACH, PW label =
(TTL=3D1) or<br>
&gt;&gt;&gt; =A0GAL-ACH are used to identify such management messages.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 20]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 13. =A0QoS Considerations<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0In network deployments where not every LSR/LER is Timing-aw=
are, it<br>
&gt;&gt; is<br>
&gt;&gt;&gt; =A0important to reduce the impact of the non-Timing-aware LSR/=
LERs on<br>
&gt;&gt;&gt; =A0the timing recovery in the slave clock. =A0The Timing messa=
ges are<br>
&gt;&gt; time<br>
&gt;&gt;&gt; =A0critical and must be treated with the highest priority. =A0=
Therefore<br>
&gt;&gt;&gt; =A0Timing over MPLS messages must be treated with the highest =
priority<br>
&gt;&gt;&gt; =A0in the routers. =A0This can be achieved by proper setup of =
Timing<br>
&gt;&gt; LSPs.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0It is recommended that the Timing LSPs are setup or configu=
red<br>
&gt;&gt;&gt; =A0properly to indicate EF-PHB [RFC3246]for the CoS and Green<=
br>
&gt;&gt; [RFC2697]<br>
&gt;&gt;&gt; =A0for drop eligibility.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 21]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 14. =A0FCS and Checksum Recalculation<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When time-stamp generation and timing packet adjustment is<=
br>
&gt;&gt; performed<br>
&gt;&gt;&gt; =A0near the physical port hardware, the process MUST include<b=
r>
&gt;&gt;&gt; =A0recalculation of the Ethernet FCS.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; The above is confusing - an LSR always recomputes the l=
ink layer<br>
&gt;&gt;&gt; SB&gt; CRC which may or may not be Ethernet.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Also FCS retention for the<br>
&gt;&gt;&gt; =A0payload Ethernet described in [RFC4720] MUST NOT be used.<b=
r>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0For UDP/IP encapsulation mode of Timing over MPLS, the UDP =
checksum<br>
&gt;&gt;&gt; =A0may be required as per UDP transport standards.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; You really need to be working on getting the IPv6 C?S c=
omputation<br>
&gt;&gt;&gt; SB&gt; removed from PTP msgs.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When UDP checksum is used, each Timing-aware LER/LSR must e=
ither<br>
&gt;&gt;&gt; =A0incrementally update the UDP checksum after Time stamping o=
r<br>
&gt;&gt;&gt; =A0Correction Field update or verify the UDP checksum on recep=
tion<br>
&gt;&gt; from<br>
&gt;&gt;&gt; =A0upstream and recalculate the checksum completely on transmi=
ssion to<br>
&gt;&gt;&gt; =A0downstream node after Time stamping or Correction Field upd=
ate.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 22]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 15. =A0Behavior of LER/LSR<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Timing-capable/aware LERs and LSRs are routers that have on=
e or<br>
&gt;&gt; more<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; You mean physical interfaces?<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0interfaces that can perform Timing operations (OC/BC/TC) on=
 Timing<br>
&gt;&gt;&gt; =A0packets and are configured to do so. =A0Timing-capable/awar=
e LERs and<br>
&gt;&gt;&gt; =A0LSRs can advertise their Timing-capability per-interface vi=
a<br>
&gt;&gt; control<br>
&gt;&gt;&gt; =A0plane such as OSPF or IS-IS.<br>
&gt;&gt;&gt; SB&gt; ISIS and OSPF are routing protocols.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; The Timing-capable/aware LERs can then<br>
&gt;&gt;&gt; =A0signals Timing LSPs via RSVP-TE signaling. =A0Alternatively=
 the<br>
&gt;&gt; Timing<br>
&gt;&gt;&gt; =A0capability of LER and LSRs may be configured in a centraliz=
ed<br>
&gt;&gt;&gt; =A0controller and the Timing LSP may be setup using manual<br>
&gt;&gt; configuration<br>
&gt;&gt;&gt; =A0or other methods such as SDN.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; it can also be configured individually rather then thro=
ugh<br>
&gt;&gt;&gt; SB&gt; a cebtral controllwe<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 15.1. =A0Behavior of Timing-capable/aware LER<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When a Timing-capable/aware LER behaves as a Transparent cl=
ock and<br>
&gt;&gt;&gt; =A0receives a Timing message from a Timing-capable/aware non-M=
PLS<br>
&gt;&gt;&gt; =A0interface, the LER updates the Correction Field (CF) and<br=
>
&gt;&gt; encapsulates<br>
&gt;&gt;&gt; =A0and forwards the timing message over previously established=
 Timing<br>
&gt;&gt;&gt; =A0LSP.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; SB&gt; You need to call out the details so that people properl=
y<br>
&gt;&gt;&gt; SB&gt; understand the definition of the new LSP.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Also when a Timing message is received from a Timing-capabl=
e/<br>
&gt;&gt;&gt; =A0aware MPLS interface, LER updates the Correction Filed (CF)=
 and<br>
&gt;&gt;&gt; =A0decapsulates the MPLS encapsulation and forwards the timing=
 message<br>
&gt;&gt;&gt; =A0to a non-MPLS interface.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When a Timing-capable/aware LER behaves as a Boundary clock=
 and<br>
&gt;&gt;&gt; =A0receives a Timing message from a Timing-capable/aware non M=
PLS<br>
&gt;&gt;&gt; =A0interface, the LER Timestamps the Timing packet and sends i=
t to the<br>
&gt;&gt;&gt; =A0LERs Boundary clock processing module. =A0Also when a Timin=
g message<br>
&gt;&gt; is<br>
&gt;&gt;&gt; =A0received from a Timing- capable/aware MPLS interface, the L=
ER<br>
&gt;&gt;&gt; =A0Timestamps the Timing packet and sends it to the LERs Bound=
ary<br>
&gt;&gt; clock<br>
&gt;&gt;&gt; =A0processing module.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When a Timing-capable/aware LER behaves as an Ordinary Cloc=
k toward<br>
&gt;&gt;&gt; =A0the MPLS network, and receives a Timing message from a Timi=
ng-<br>
&gt;&gt;&gt; =A0capable/aware MPLS interface, the LER Timestamps the Timing=
 packet<br>
&gt;&gt;&gt; =A0and sends it to the LERs Ordinary clock processing module.<=
br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 15.2. =A0Behavior of Timing-capable/aware LSR<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When a Timing-capable/aware LSR behaves as a Transparent cl=
ock and<br>
&gt;&gt;&gt; =A0receives a Timing message from a Timing-capable/aware MPLS<=
br>
&gt;&gt; interface,<br>
&gt;&gt;&gt; =A0The LSR updates the Correction Filed (CF) and forwards the =
timing<br>
&gt;&gt;&gt; =A0message over another MPLS interface.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When a Timing-capable/aware LSR behaves as a Boundary clock=
 and<br>
&gt;&gt;&gt; =A0receives a Timing message from a Timing-capable/aware MPLS<=
br>
&gt;&gt; interface.<br>
&gt;&gt;&gt; =A0The LSR performs the functions of a Boundary Clock in termi=
nating<br>
&gt;&gt; the<br>
&gt;&gt;&gt; =A0received Timing message and re-generating a new timing mess=
age over<br>
&gt;&gt;&gt; =A0another (or the same) MPLS interface.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 23]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 15.3. =A0Behavior of non-Timing-capable/aware LSR<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0It is most beneficial when all LSRs in the path of a Timing=
 LSP be<br>
&gt;&gt;&gt; =A0timing-Capable/aware LSRs. =A0This would ensure the highest=
 quality<br>
&gt;&gt;&gt; =A0time and clock synchronization by Timing Slave Clocks. =A0H=
owever,<br>
&gt;&gt; this<br>
&gt;&gt;&gt; =A0specification does not mandate that all LSRs in path of a T=
iming<br>
&gt;&gt; LSP<br>
&gt;&gt;&gt; =A0be Timing- capable/aware.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Non-Timing-capable/aware LSRs just switch the packets encap=
sulated<br>
&gt;&gt; in<br>
&gt;&gt;&gt; =A0Timing LSPs and dont perform any Timing operation (TC or BC=
).<br>
&gt;&gt;&gt; =A0However as explained in QoS section the Timing over MPLS pa=
ckets<br>
&gt;&gt; MUST<br>
&gt;&gt;&gt; =A0be still be treated with the highest priority based on thei=
r<br>
&gt;&gt; Traffic<br>
&gt;&gt;&gt; =A0Class (TC) marking.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 24]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 16. =A0Other considerations<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[IEEE-1588] defines an optional peer-to-peer Transparent cl=
ocking<br>
&gt;&gt;&gt; =A0that requires peer delay measurement between two adjacent T=
iming-<br>
&gt;&gt;&gt; =A0capable/ aware routers/switches. =A0Peer delay measurement =
messages<br>
&gt;&gt;&gt; =A0need to be time stamped and terminated by the Timing-capabl=
e/aware<br>
&gt;&gt;&gt; =A0routers/ switches. =A0This means that two adjacent LSRs may=
 be<br>
&gt;&gt; engaged<br>
&gt;&gt;&gt; =A0in a peer delay measurement.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0For transporting such peer delay measurement messages a sin=
gle-hop<br>
&gt;&gt;&gt; =A0LSP SHOULD to be created between the two adjacent LSRs enga=
ged in<br>
&gt;&gt;&gt; =A0peer delay measurement to carry peer delay measurement mess=
ages.<br>
&gt;&gt;&gt; =A0Other methods such as PTP transport over Ethernet MAY be us=
ed for<br>
&gt;&gt;&gt; =A0transporting peer delay measurement messages if the link be=
tween<br>
&gt;&gt; the<br>
&gt;&gt;&gt; =A0two routers is Ethernet.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0In Peer-to-peer transparent clocking (P2P TC), a Timing-cap=
able/<br>
&gt;&gt; ware<br>
&gt;&gt;&gt; =A0routers/switches MUST maintain a list of all the neighbors =
it needs<br>
&gt;&gt;&gt; =A0to send a PDelay_Req to, where each neighbor corresponds to=
 a<br>
&gt;&gt; timing<br>
&gt;&gt;&gt; =A0LSP.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The use of Explicit Null Label (Label=3D 0 or 2) is accepta=
ble as<br>
&gt;&gt; long<br>
&gt;&gt;&gt; =A0as either the Explicit Null label is the bottom of stack la=
bel<br>
&gt;&gt;&gt; =A0(applicable only to UDP/IP encapsulation) or the label belo=
w the<br>
&gt;&gt;&gt; =A0Explicit Null label is a PTP label.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 25]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 17. =A0Security Considerations<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0MPLS PW security considerations in general are discussed in=
<br>
&gt;&gt; [RFC3985]<br>
&gt;&gt;&gt; =A0and [RFC4447],and those considerations also apply to this d=
ocument.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0An experimental security protocol is defined in [IEEE-1588]=
.The PTP<br>
&gt;&gt;&gt; =A0security extension and protocol provides group source<br>
&gt;&gt; authentication,<br>
&gt;&gt;&gt; =A0message integrity, and replay attack protection for PTP mes=
sages.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0When the MPLS network (provider network) serves multiple cu=
stomers,<br>
&gt;&gt;&gt; =A0it is important to maintain and process each customers cloc=
k and<br>
&gt;&gt;&gt; =A0Timing messages separately from other customers to ensure t=
here is<br>
&gt;&gt; no<br>
&gt;&gt;&gt; =A0cross- customer effect. =A0For example if an LER BC is sync=
hronized<br>
&gt;&gt; to<br>
&gt;&gt;&gt; =A0a specific grandmaster, belonging to customer A, then the L=
ER MUST<br>
&gt;&gt;&gt; =A0use that BC clock only for customer A to ensure that custom=
er A<br>
&gt;&gt;&gt; =A0cannot attack other customers by manipulating its time.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Timing messages MAY be encrypted or authenticated, provided=
 that<br>
&gt;&gt; the<br>
&gt;&gt;&gt; =A0LERs/LSRs that are Timing capable/aware can authenticate/ d=
ecrypt<br>
&gt;&gt; the<br>
&gt;&gt;&gt; =A0timing messages.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 26]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 18. =A0Acknowledgements<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The authors would like to thank Ron Cohen, Yaakov Stein, Ta=
l<br>
&gt;&gt; Mizrahi,<br>
&gt;&gt;&gt; =A0Stefano Ruffini, Peter Meyer, and other members of IETF for=
<br>
&gt;&gt; reviewing<br>
&gt;&gt;&gt; =A0and providing feedback on this draft.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 27]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 19. =A0IANA Considerations<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0There are no IANA requirements in this specification.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 28]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 20. =A0References<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 20.1. =A0Normative References<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[IEEE-1588]<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 IEEE 1588-2008, &quot;IEEE Standard fo=
r a Precision Clock<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Synchronization Protocol for Networked=
 Measurement and<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Control Systems&quot;.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC2119] =A0Bradner, S., &quot;Key words for use in RFCs t=
o Indicate<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Requirement Levels&quot;, BCP 14, RFC =
2119, March 1997.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC3985] =A0Bryant, S. and P. Pate, &quot;Pseudo Wire Emul=
ation Edge-to-<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Edge (PWE3) Architecture&quot;, RFC 39=
85, March 2005.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC4389] =A0Thaler, D., Talwar, M., and C. Patel, &quot;Ne=
ighbor<br>
&gt;&gt; Discovery<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Proxies (ND Proxy)&quot;, RFC 4389, Ap=
ril 2006.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC4447] =A0Martini, L., Rosen, E., El-Aawar, N., Smith, T=
., and G.<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Heron, &quot;Pseudowire Setup and Main=
tenance Using the Label<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Distribution Protocol (LDP)&quot;, RFC=
 4447, April 2006.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC4448] =A0Martini, L., Rosen, E., El-Aawar, N., and G. H=
eron,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 &quot;Encapsulation Methods for Transp=
ort of Ethernet over<br>
&gt;&gt; MPLS<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Networks&quot;, RFC 4448, April 2006.<=
br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC4720] =A0Malis, A., Allan, D., and N. Del Regno, &quot;=
Pseudowire<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Emulation Edge-to-Edge (PWE3) Frame Ch=
eck Sequence<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Retention&quot;, RFC 4720, November 20=
06.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC5085] =A0Nadeau, T. and C. Pignataro, &quot;Pseudowire =
Virtual Circuit<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Connectivity Verification (VCCV): A Co=
ntrol Channel for<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Pseudowires&quot;, RFC 5085, December =
2007.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC5880] =A0Katz, D. and D. Ward, &quot;Bidirectional Forw=
arding<br>
&gt;&gt; Detection<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 (BFD)&quot;, RFC 5880, June 2010.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC5884] =A0Aggarwal, R., Kompella, K., Nadeau, T., and G.=
 Swallow,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 &quot;Bidirectional Forwarding Detecti=
on (BFD) for MPLS Label<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Switched Paths (LSPs)&quot;, RFC 5884,=
 June 2010.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 20.2. =A0Informative References<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[I-D.ietf-pwe3-fat-pw]<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Bryant, S., Filsfils, C., Drafz, U., K=
ompella, V.,<br>
&gt;&gt; Regan,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 J., and S. Amante, &quot;Flow Aware Tr=
ansport of Pseudowires<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 over an MPLS Packet Switched Network&q=
uot;,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 draft-ietf-pwe3-fat-pw-07 (work in pro=
gress), July 2011.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 29]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[ISO] =A0 =A0 =A0ISO/IEC 10589:1992, &quot;Intermediate sys=
tem to Intermediate<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 system routeing information exchange p=
rotocol for use in<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 conjunction with the Protocol for prov=
iding the<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Connectionless-mode Network Service (I=
SO 8473)&quot;.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC1195] =A0Callon, R., &quot;Use of OSI IS-IS for routing=
 in TCP/IP and<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 dual environments&quot;, RFC 1195, Dec=
ember 1990.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC2328] =A0Moy, J., &quot;OSPF Version 2&quot;, STD 54, R=
FC 2328, April 1998.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC2697] =A0Heinanen, J. and R. Guerin, &quot;A Single Rat=
e Three Color<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Marker&quot;, RFC 2697, September 1999=
.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC3246] =A0Davie, B., Charny, A., Bennet, J., Benson, K.,=
 Le<br>
&gt;&gt; Boudec,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 J., Courtney, W., Davari, S., Firoiu, =
V., and D.<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Stiliadis, &quot;An Expedited Forwardi=
ng PHB (Per-Hop<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Behavior)&quot;, RFC 3246, March 2002.=
<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC3630] =A0Katz, D., Kompella, K., and D. Yeung, &quot;Tr=
affic<br>
&gt;&gt; Engineering<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 (TE) Extensions to OSPF Version 2&quot=
;, RFC 3630,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 September 2003.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC3784] =A0Smit, H. and T. Li, &quot;Intermediate System =
to Intermediate<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 System (IS-IS) Extensions for Traffic =
Engineering (TE)&quot;,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 RFC 3784, June 2004.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC4970] =A0Lindem, A., Shen, N., Vasseur, JP., Aggarwal, =
R., and S.<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Shaffer, &quot;Extensions to OSPF for =
Advertising Optional<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Router Capabilities&quot;, RFC 4970, J=
uly 2007.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC4971] =A0Vasseur, JP., Shen, N., and R. Aggarwal, &quot=
;Intermediate<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 System to Intermediate System (IS-IS) =
Extensions for<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Advertising Router Information&quot;, =
RFC 4971, July 2007.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC5120] =A0Przygienda, T., Shen, N., and N. Sheth, &quot;=
M-ISIS: Multi<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Topology (MT) Routing in Intermediate =
System to<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Intermediate Systems (IS-ISs)&quot;, R=
FC 5120, February 2008.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC5305] =A0Li, T. and H. Smit, &quot;IS-IS Extensions for=
 Traffic<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Engineering&quot;, RFC 5305, October 2=
008.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC5329] =A0Ishiguro, K., Manral, V., Davey, A., and A. Li=
ndem,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 &quot;Traffic Engineering Extensions t=
o OSPF Version 3&quot;,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 RFC 5329, September 2008.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC5340] =A0Coltun, R., Ferguson, D., Moy, J., and A. Lind=
em, &quot;OSPF<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 for IPv6&quot;, RFC 5340, July 2008.<b=
r>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 30]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC5905] =A0Mills, D., Martin, J., Burbank, J., and W. Kas=
ch,<br>
&gt;&gt; &quot;Network<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Time Protocol Version 4: Protocol and =
Algorithms<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 Specification&quot;, RFC 5905, June 20=
10.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC6391] =A0Bryant, S., Filsfils, C., Drafz, U., Kompella,=
 V.,<br>
&gt;&gt; Regan,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 J., and S. Amante, &quot;Flow-Aware Tr=
ansport of Pseudowires<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 over an MPLS Packet Switched Network&q=
uot;, RFC 6391,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 November 2011.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0[RFC6790] =A0Kompella, K., Drake, J., Amante, S., Henderick=
x, W., and<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 L. Yong, &quot;The Use of Entropy Labe=
ls in MPLS Forwarding&quot;,<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0 =A0 =A0 RFC 6790, November 2012.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 31]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 1. =A0Routing extensions for Timing-aware Routers<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC=
5340]<br>
&gt;&gt; and<br>
&gt;&gt;&gt; =A0IS-IS [ISO] [RFC1195] in order to advertise Traffic Enginee=
ring<br>
&gt;&gt; (TE)<br>
&gt;&gt;&gt; =A0link information used for constraint-based routing.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Indeed, it is useful to advertise data plane TE router link=
<br>
&gt;&gt;&gt; =A0capabilities, such as the capability for a router to be Tim=
ing-<br>
&gt;&gt; aware.<br>
&gt;&gt;&gt; =A0This capability MUST then be taken into account during path=
<br>
&gt;&gt;&gt; =A0computation to prefer or even require links that advertise<=
br>
&gt;&gt; themselves<br>
&gt;&gt;&gt; =A0as Timing-aware. =A0In this way the path can ensure the ent=
ry and<br>
&gt;&gt; exit<br>
&gt;&gt;&gt; =A0points into the LERs and, if desired, the links into the LS=
Rs are<br>
&gt;&gt;&gt; =A0able to perform port based time-stamping thus minimizing th=
eir<br>
&gt;&gt; impact<br>
&gt;&gt;&gt; =A0on the performance of the slave clock.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0extensions are required to OSPF and IS-IS in order to adver=
tise<br>
&gt;&gt;&gt; =A0Timing-aware capabilities of a link. =A0Such extensions are=
 outside<br>
&gt;&gt; the<br>
&gt;&gt;&gt; =A0scope of this document; however such extension SHOULD be ab=
le to<br>
&gt;&gt;&gt; =A0signal the following information per Router Link:<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0o =A0Capable of processing PTP, NTP or other Timing flows<b=
r>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0o =A0Capable of performing Transparent Clock operation<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0o =A0Capable of performing Boundary Clock operation<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 32]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 2. =A0Signaling Extensions for Creating Timing LSPs<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0RSVP-TE signaling MAY be used to setup the timing LSPs. =A0=
When RSVP-<br>
&gt;&gt; TE<br>
&gt;&gt;&gt; =A0is used to setup Timing LSPs, some information that indicat=
es that<br>
&gt;&gt;&gt; =A0the LSP is carrying Timing flows MUST be included in the ne=
w<br>
&gt;&gt;&gt; =A0Extensions to RSVP-TE:<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0The following information MAY also be included in the new<b=
r>
&gt;&gt; Extensions<br>
&gt;&gt;&gt; =A0to RSVP-TE:<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0o =A0Offset from Bottom of Stack (BoS) to the start of the =
Time-stamp<br>
&gt;&gt;&gt; =A0 =A0 field<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0o =A0Number of VLANs in case of PW encapsulation<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0o =A0Timestamp field Type<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0 * =A0Correction Field, Timestamp<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0o =A0Timestamp Field format<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0 =A0 * =A064-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NT=
P, 128-bit<br>
&gt;&gt;&gt; =A0 =A0 =A0 =A0NTP, etc.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Note that in case the above optional information is signale=
d with<br>
&gt;&gt;&gt; =A0RSVP-TE for a Timing LSP, all the Timing packets carried in=
 that<br>
&gt;&gt; LSP<br>
&gt;&gt;&gt; =A0must have the same signaled characteristics. =A0For example=
 if<br>
&gt;&gt;&gt; =A0Timestamp format is signaled as 64-bit PTPv1, then all Timi=
ng<br>
&gt;&gt; packets<br>
&gt;&gt;&gt; =A0must use 64-bit PTPv1 time-stamp.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 33]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Authors&#39; Addresses<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Shahram Davari<br>
&gt;&gt;&gt; =A0Broadcom Corp.<br>
&gt;&gt;&gt; =A0San Jose, CA =A095134<br>
&gt;&gt;&gt; =A0USA<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Email: <a href=3D"mailto:davari@broadcom.com">davari@broadc=
om.com</a><br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Amit Oren<br>
&gt;&gt;&gt; =A0Broadcom Corp.<br>
&gt;&gt;&gt; =A0San Jose, CA =A095134<br>
&gt;&gt;&gt; =A0USA<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Email: <a href=3D"mailto:amito@broadcom.com">amito@broadcom=
.com</a><br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Manav Bhatia<br>
&gt;&gt;&gt; =A0Alcatel-Lucent<br>
&gt;&gt;&gt; =A0Bangalore,<br>
&gt;&gt;&gt; =A0India<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Email: <a href=3D"mailto:manav.bhatia@alcatel-lucent.com">m=
anav.bhatia@alcatel-lucent.com</a><br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Peter Roberts<br>
&gt;&gt;&gt; =A0Alcatel-Lucent<br>
&gt;&gt;&gt; =A0Kanata,<br>
&gt;&gt;&gt; =A0Canada<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Email: <a href=3D"mailto:peter.roberts@alcatel-lucent.com">=
peter.roberts@alcatel-lucent.com</a><br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Laurent Montini<br>
&gt;&gt;&gt; =A0Cisco Systems<br>
&gt;&gt;&gt; =A0San Jose CA<br>
&gt;&gt;&gt; =A0USA<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Email: <a href=3D"mailto:lmontini@cisco.com">lmontini@cisco=
.com</a><br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 34]<br>
&gt;<br>
&gt;&gt;&gt; Internet-Draft =A0 =A0 =A0 =A0Transporting Timing over MPLS =
=A0 =A0 =A0 =A0 =A0 =A0June<br>
&gt;&gt; 2013<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Luca<br>
&gt;&gt;&gt; =A0Cisco Systems<br>
&gt;&gt;&gt; =A0San Jose CA<br>
&gt;&gt;&gt; =A0USA<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; =A0Email: <a href=3D"mailto:lmartini@cisco.com">lmartini@cisco=
.com</a><br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Davari, et al. =A0 =A0 =A0 =A0 =A0Expires December 17, 2013 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0[Page<br>
&gt;&gt; 35]<br>
&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; --<br>
&gt;&gt;&gt; For corporate legal information go to:<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; <a href=3D"http://www.cisco.com/web/about/doing_business/legal=
/cri/index.html" target=3D"_blank">http://www.cisco.com/web/about/doing_bus=
iness/legal/cri/index.html</a><br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; _______________________________________________<br>
&gt;&gt;&gt; mpls mailing list<br>
&gt;&gt;&gt; <a href=3D"mailto:mpls@ietf.org">mpls@ietf.org</a><br>
&gt;&gt;&gt; <a href=3D"https://www.ietf.org/mailman/listinfo/mpls" target=
=3D"_blank">https://www.ietf.org/mailman/listinfo/mpls</a><br>
&gt;<br>
&gt;<br>
&gt;<br>
&gt;<br>
<br>
_______________________________________________<br>
mpls mailing list<br>
<a href=3D"mailto:mpls@ietf.org">mpls@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/mpls" target=3D"_blank">ht=
tps://www.ietf.org/mailman/listinfo/mpls</a><br>
</blockquote></div>

--bcaec51d2c10d243ea04e2f91efd--

From kurt@roeckx.be  Sat Aug  3 10:40:25 2013
Return-Path: <kurt@roeckx.be>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id EB82211E80AD for <tictoc@ietfa.amsl.com>; Sat,  3 Aug 2013 10:40:25 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -4.449
X-Spam-Level: 
X-Spam-Status: No, score=-4.449 tagged_above=-999 required=5 tests=[AWL=-1.850, BAYES_00=-2.599]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id M-8jVOVEr+Ui for <tictoc@ietfa.amsl.com>; Sat,  3 Aug 2013 10:40:20 -0700 (PDT)
Received: from juliette.telenet-ops.be (juliette.telenet-ops.be [195.130.137.74]) by ietfa.amsl.com (Postfix) with ESMTP id 2572C21F9F5F for <tictoc@ietf.org>; Sat,  3 Aug 2013 10:40:12 -0700 (PDT)
Received: from intrepid.roeckx.be ([94.226.199.45]) by juliette.telenet-ops.be with bizsmtp id 8Hg81m0150zFtyu06Hg8Y8; Sat, 03 Aug 2013 19:40:09 +0200
Received: by intrepid.roeckx.be (Postfix, from userid 1000) id 6C9DDEB001; Sat,  3 Aug 2013 19:40:08 +0200 (CEST)
Date: Sat, 3 Aug 2013 19:40:08 +0200
From: Kurt Roeckx <kurt@roeckx.be>
To: "David L. Mills" <mills@udel.edu>
Message-ID: <20130803174008.GA17578@roeckx.be>
References: <51FAC820.3090401@udel.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <51FAC820.3090401@udel.edu>
User-Agent: Mutt/1.5.21 (2010-09-15)
X-Mailman-Approved-At: Sat, 03 Aug 2013 12:08:30 -0700
Cc: NTP Working Group <ntpwg@lists.ntp.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [ntpwg] security ID submitted for review
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sat, 03 Aug 2013 17:40:26 -0000

On Thu, Aug 01, 2013 at 08:42:08PM +0000, David L. Mills wrote:
> Guys,
> 
> Here are some general comments on the ID.  I apologize if they are a
> little rough as I have to dictate this message instead of my usual
> polished wordsmithing.
> 
> format.
> The ID does not specify an extension field format.

The draft only seems to specify the the general idea of how it
should work and doesn't seem to go into detail how this should
work at the protocol level.  I assume he first wants to know if
the solution is something that can work and that others see
as a good solution before he works out the details.

> In particular, the private key for the message
> authentication code (MAC) , if used, must be known beforehand.  This
> is where the "cookie snatcher" issue comes up as described in
> previous messages.  In particular, an important constraint for
> normal operation is that extension fields are not used unless
> required for cookie determination.

I'm not sure I understand your remark.  As far as I know the
"cookie snatcher" you talk about is the normal "man in the middle"
attack.  As I understand we would change the verification by
replacing it by the normal X.509 verification as done for instance
SSL/TLS.

I assume you meant public key and note private key.

> Glaring vulnerability exists when using extension fields, as the key
> value is public and must be known in advance.  This invites a
> cut-and-paste attack, as well as header changes, even if
> subsequently detected by the on-wire protocol.

I've seen this cut-and-paste attack mentioned before, but I've
never really understood it.  As I understand things, the MAC
is generated over the NTP + extensions.  I don't see how you
can take the extension from 1 packet into an other and still
have a valid MAC.  Or are you saying that you do something
with a packet even when it doesn't have a valid MAC?

> certificate.
> Both RFC 5906 and the ID require the server certificate.  The server
> certificate can be conveyed, either off-line, or directly or
> indirectly via a certificate trail as described in RFC 5906 or the
> ID.  If obtained via the trail, there is a hazzard due to a
> middle-man attack and masquerade.

I really don't understand how such a man in the middle attack
can work.  Either the certificate + the chain validate, or they
don't.  It is of course important to you check that the CommonName
in the certificate matches the server you're trying to reach,
and that the root CA is in your list of trusted CAs.  But there
really isn't anything new or hard about this.

> In broadcast mode, the protocols described in RFC 5906 and the ID
> are messy and with several possible vulnerabilities.  Here is a
> suggestion that might be useful.  Define a new broadcast packet
> format as the NTP header followed by one extension field
> representing the MAC.  The extension field represents the signature
> of the header computed using the private key of the server and can
> be verified by the client using the public key in the server
> certificate.  Originally, this was considered an unacceptable
> dilution of accuracy due to the variability of the signature
> computing time and the length of the signature.  However, at least
> in modern NTP, the interleaved broadcast mode can be used to correct
> for this overhead.  This assumes that the overhead for either the
> server or client is reduced using fast modern processors.

I think that in all cases where it's no problem to keep track of
all clients you should use the interleaved mode, and we can use
something like the Diffie-Hellman key agreement for those.

But there clearly is also a need for servers that can't keep
the state of each client, and we need something else for that.
So maybe we need to define 2 ways of working.



Kurt


From Alexander.Vainshtein@ecitele.com  Sun Aug  4 01:21:58 2013
Return-Path: <Alexander.Vainshtein@ecitele.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2428321F9A6A for <tictoc@ietfa.amsl.com>; Sun,  4 Aug 2013 01:21:58 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -5.202
X-Spam-Level: 
X-Spam-Status: No, score=-5.202 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, MIME_QP_LONG_LINE=1.396, RCVD_IN_DNSWL_MED=-4,  UNPARSEABLE_RELAY=0.001]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id onhYVXNYweUH for <tictoc@ietfa.amsl.com>; Sun,  4 Aug 2013 01:21:44 -0700 (PDT)
Received: from mail1.bemta14.messagelabs.com (mail1.bemta14.messagelabs.com [193.109.254.107]) by ietfa.amsl.com (Postfix) with ESMTP id 8E93B21F9A98 for <tictoc@ietf.org>; Sun,  4 Aug 2013 01:21:33 -0700 (PDT)
Received: from [193.109.254.147:42681] by server-3.bemta-14.messagelabs.com id 23/4D-00891-C0F0EF15; Sun, 04 Aug 2013 08:21:32 +0000
X-Env-Sender: Alexander.Vainshtein@ecitele.com
X-Msg-Ref: server-3.tower-27.messagelabs.com!1375604486!2548670!3
X-Originating-IP: [147.234.242.234]
X-StarScan-Received: 
X-StarScan-Version: 6.9.11; banners=-,-,-
X-VirusChecked: Checked
Received: (qmail 26374 invoked from network); 4 Aug 2013 08:21:31 -0000
Received: from ilptbmg01-out.ecitele.com (HELO ilptbmg01-out.ecitele.com) (147.234.242.234) by server-3.tower-27.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 4 Aug 2013 08:21:31 -0000
X-AuditID: 93eaf2e7-b7f4d6d000005d50-88-51fe0f095c3b
Received: from ILPTWPVEXCA01.ecitele.com ( [172.31.244.224]) by ilptbmg01-out.ecitele.com (Symantec Messaging Gateway) with SMTP id 19.57.23888.90F0EF15; Sun,  4 Aug 2013 11:21:29 +0300 (IDT)
Received: from ILPTWPVEXMB02.ecitele.com ([fe80::5979:ca8d:419f:56df]) by ILPTWPVEXCA01.ecitele.com ([fe80::ac15:43ab:d541:dfa7%12]) with mapi id 14.03.0123.003; Sun, 4 Aug 2013 11:21:29 +0300
From: Alexander Vainshtein <Alexander.Vainshtein@ecitele.com>
To: "stbryant@cisco.com" <stbryant@cisco.com>
Thread-Topic: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
Thread-Index: AQHOj2AMghL6qRFw2EmhP8ypbGfBrJmEtKhw
Date: Sun, 4 Aug 2013 08:21:28 +0000
Message-ID: <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com>
References: <51FB7543.801@cisco.com>
In-Reply-To: <51FB7543.801@cisco.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-originating-ip: [10.4.35.10]
Content-Type: text/plain; charset="us-ascii"
content-transfer-encoding: quoted-printable
MIME-Version: 1.0
X-CFilter-Loop: Reflected
X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFupkleLIzCtJLcpLzFFi42KZ/OrTF10u/n+BBo9XiVucezqH0eJvcw+7 A5PHlN8bWT2WLPnJFMAU1cBok5iXl1+SWJKqkJJanGyrFFCUWZaYXKmkkJliq2SopFCQk5ic mpuaV2KrlFhQkJqXomTHpYABbIDKMvMUUvOS81My89JtlTyD/XUtLEwtdQ2V7NSUDY2tuUIy MosVUnVzEzNzFHJTi4sT01MVgCIJW5gzDs3sZC3o62SpuLT4KVMD46b3TF2MnBwSAiYS164d ZoewxSQu3FvP1sXIxSEkcJBR4tfXOewQzhFGial969hAqtgEbCU2rb4LZosI6ErM3nCDEcRm FvCQ+LZ8D5gtLOAg0bR3AjNEjaPEznsTWCFsI4lZt9+C1bAIqEhcONwMNodXIEDi174vQPUc QMtUJH5/swAxOQVUJVbOdwCpYAS67fupNUwQm8Qlbj2ZD3W/gMSSPeeZIWxRiZeP/7FC2HIS T56cYoGo15FYsPsTG4StLbFs4WtmiK2CEidnPmGBqJeUOLjiBssERvFZSFbMQtI+C0n7LCTt CxhZVjGKZuYUlCTlphsY6qUmZ5ak5qTqJefnbmKEJJLnOxh/zVc5xOgK9PVEZinu5HxgIsor iTc2MMDNURLnXd4Q7i8kkA5MONmpqQWpRfFFpTmpxYcYmTg4pRoYj3j78fzXEtsbes33gOci lo55R4pe6s1jm/xMq+T2gl+L9m4q++KteXFdZLxA3A/5C5JJ7ixn/BYu3z87SPcE05mHBtU8 WnHba75EZD0vOMrud/RQ8e37LY4qr+xlDjfuc80XaFsddcyFsV+9o2XKoq4jadwfq/78sriX voEv7+Cb9W2e/bv/K7EUZyQaajEXFScCAFlijRwcAwAA
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 04 Aug 2013 08:21:58 -0000

Stewart and all,
I concur with Stewart's statement that the draft does not define "full inter=
action with MPLS architecture".

E.g., one of the objectives of the draft is to provide a technique that woul=
d be backward-compatible with old LSRs that cannot provide on-path support f=
or timing distribution, while the other objective is to make every LSR on th=
e path aware that some MPLS packets are carrying timing-related messages and=
 hence require on-path support.

IMHO and FWIW, the MPLS data plane architecture allows just two methods for=
 making a transit LSR to provide special processing to a labeled packet:
- It would carry some kind of an "alert label" on top of the label stack and=
, specifically, on top of any labels used for actual forwarding
OR,
- The TTL in the top label stack entry has been set to 1.


The draft does not follow any of these approaches.

I must also admit that Section 12 "OAM, Control and Management" of the draft=
 looks somewhat in
E.g., I could not understand from the text whether BFD, LSP-Ping etc. OAM me=
ssages would be subjected to on-path support procedures for timing messages=
 or not.

My 2c,
     Sasha

> -----Original Message-----
> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf O=
f
> Stewart Bryant
> Sent: Friday, August 02, 2013 12:01 PM
> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.org; dr=
aft-
> ietf-tictoc-1588overmpls@tools.ietf.org
> Cc: mpls@ietf.org; tictoc@ietf.org
> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
> 
> SB> This draft does not seem to provide a precise definition
> SB> the properties of the new LSP type that it wishes to
> SB> define, in particular it does it define the PHB of those
> SB> LSPs, nor the full interaction with the MPLS
> SB> architecture.
> SB>
> SB> I have not tracked TICTOC for a while but I thought that
> SB> the original plan was to define the concept of an offset
> SB> into a packet to do the correction.
> SB>
> SB> It is disappointing that the opportunity was not taken
> SB> to define a timing shim inside the timing LSP so that
> SB> a time correction could be added to any packet such that
> SB> the MPLS system was isolated from the details of the
> SB> complexity of the particular time transfer type.
> 
> SB> I think that much more clarify is needed in terms of
> SB> definition of the new LSP type, since it is unclear
> SB> from this text how to implement one.
> SB>
> SB> There are a lot of other MPLS services such as
> SB> LSP ping that need to be considered.
> SB>
> SB> Please see inline for more comments. However these
> SB> comments are made in the context of the text as written
> SB> whilst I have a fundamental concern that this approach
> SB> lacks an MPLS architectural soundness that need
> SB> greater thought with significant impact on the
> SB> draft.
> 
> - Stewart
> 
> 
> TICTOC Working Group                                           S. Davari
> Internet-Draft                                                   A. Oren
> Intended status: Standards Track                          Broadcom Corp.
> Expires: December 17, 2013                                     M. Bhatia
>                                                                P. Roberts
>                                                            Alcatel-Lucent
>                                                                L. Montini
>                                                                L. Martini
>                                                             Cisco Systems
>                                                             June 15, 2013
> 
> 
>              Transporting Timing messages over MPLS Networks
>                     draft-ietf-tictoc-1588overmpls-05
> 
> Abstract
> 
>     This document defines the method for transporting Timing messages
>     such as PTP and NTP over an MPLS network.  The method allows for the
>     easy identification of these PDUs at the port level to allow for port
> 
> SB> What is a port
> 
>     level processing of these PDUs in both LERs and LSRs.
> 
>     The basic idea is to transport Timing messages inside dedicated MPLS
>     LSPs.  These LSPs only carry Timing messages and possibly Control and
>     Management packets, but they do not carry customer traffic.
> 
> SB> More specifically they only carry traffic associated with the
> SB> timing service and its support.
> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
> SB> also it gets carried in a structure that causes it to get
> SB> timestamped.
> 
>     Two methods for transporting Timing messages over MPLS are defined.
> 
> SB> Perhaps the right approach is to define the new LSP type and then
> SB> seperately to define  the mapping of the various timing services
> SB> over that LSP type.
> 
>     The first method is to transport Timing messages directly over the
>     dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
>     MPLS networks.  The second method is to transport Timing messages
>     inside a PW via Ethernet encapsulation.
> 
> SB> I think that we should note that there are some
> SB> h/w reasons for this preference. A clean sheet approach
> SB> would have been to use PTP over MPLS with no intermediate
> SB> layers.
> 
> Status of this Memo
> 
>     This Internet-Draft is submitted in full conformance with the
>     provisions of BCP 78 and BCP 79.
> 
>     Internet-Drafts are working documents of the Internet Engineering
>     Task Force (IETF).  Note that other groups may also distribute
>     working documents as Internet-Drafts.  The list of current Internet-
>     Drafts is at http://datatracker.ietf.org/drafts/current/.
> 
>     Internet-Drafts are draft documents valid for a maximum of six months
>     and may be updated, replaced, or obsoleted by other documents at any
>     time.  It is inappropriate to use Internet-Drafts as reference
>     material or to cite them other than as "work in progress."
> 
>     This Internet-Draft will expire on December 17, 2013.
> 
> 
> 
> Davari, et al.          Expires December 17, 2013               [Page 1]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> Copyright Notice
> 
>     Copyright (c) 2013 IETF Trust and the persons identified as the
>     document authors.  All rights reserved.
> 
>     This document is subject to BCP 78 and the IETF Trust's Legal
>     Provisions Relating to IETF Documents
>     (http://trustee.ietf.org/license-info) in effect on the date of
>     publication of this document.  Please review these documents
>     carefully, as they describe your rights and restrictions with respect
>     to this document.  Code Components extracted from this document must
>     include Simplified BSD License text as described in Section 4.e of
>     the Trust Legal Provisions and are provided without warranty as
>     described in the Simplified BSD License.
> 
> 
> Table of Contents
> 
>     1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  5
> 
>     2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .  7
> 
>     3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .  8
> 
>     4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .  9
> 
>     5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . . 12
> 
>     6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . . 13
>       6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . . 13
>       6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . . 13
>       6.3.  Other Timing Encapsulation methods . . . . . . . . . . . . 14
> 
>     7.  Timing message Processing  . . . . . . . . . . . . . . . . . . 15
> 
>     8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . . 16
> 
>     9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
> 
>     10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
> 
>     11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
> 
>     12. OAM, Control and Management  . . . . . . . . . . . . . . . . . 20
> 
>     13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . . 21
> 
>     14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . . 22
> 
> 
> 
> Davari, et al.          Expires December 17, 2013               [Page 2]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
>     15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . . 23
>       15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . . 23
>       15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . . 23
>       15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . . 24
> 
>     16. Other considerations . . . . . . . . . . . . . . . . . . . . . 25
> 
>     17. Security Considerations  . . . . . . . . . . . . . . . . . . . 26
> 
>     18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 27
> 
>     19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 28
> 
>     20. References . . . . . . . . . . . . . . . . . . . . . . . . . . 29
>       20.1. Normative References . . . . . . . . . . . . . . . . . . . 29
>       20.2. Informative References . . . . . . . . . . . . . . . . . . 29
> 
>     Appendix 1.  Routing extensions for Timing-aware Routers . . . . . 32
> 
>     Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . . 33
> 
>     Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 34
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013               [Page 3]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
>     The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
>     "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in
> this
>     document are to be interpreted as described in RFC2119 [RFC2119].
> 
>     When used in lower case, these words convey their typical use in
>     common language, and are not to be interpreted as described in
>     RFC2119 [RFC2119].
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013               [Page 4]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 1.  Introduction
> 
>     The objective of Precision Time Protocol (PTP) and Network Timing
>     Protocol (NTP) are to synchronize independent clocks running on
>     separate nodes of a distributed system.
> 
>     [IEEE-1588] defines PTP messages for frequency, phase and time
>     synchronization.  The PTP messages include PTP PDUs over UDP/IP
>     (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F of
>     [IEEE-1588]).
> 
> SB> Sure BUT it is acknowledged that IEEE is open to the definition
> SB> of other PTP mappings if they provide better optimisation.
> 
>     This document defines mapping and transport of the PTP
>     messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>     defines several clock types: ordinary clocks, boundary clocks, end-
>     to-end transparent clocks, and peer-to-peer transparent clocks.
>     Transparent clocks require intermediate nodes to update correction
>     field inside PTP message that reflects the transit time in the node.
> 
>     [RFC5905] defines NTP messages for clock and time synchronization.
>     The PTP messages (PDUs) are transported over UDP/IP.  This document
> SB> Should that be NTP messages?
> SB> It needs to be made clear as soon as you introduce NTP that
> SB> they use different time representations.
> 
>     defines mapping and transport of the NTP messages defined in
>     [RFC5905] over MPLS networks.
> 
>     One key attribute of all of these Timing messages is that the Time
>     stamp processing should occur as close as possible to the actual
>     transmission and reception at the physical port interface.  This
>     targets optimal time and/or frequency recovery by avoiding variable
>     delay introduced by queues internal to the clocks.
> 
> SB> As I recall NTP has no epoch point defined, and I am not sure
> SB> where that point is in the case of PTP in this mapping
> SB> Hopefully this will get defined in due course.
> 
>     To facilitate the fast and efficient recognition of Timing messages
>     at the port level when the Timing messages are carried over MPLS
>     LSPs,
> 
> SB> Over a new LSP type with time optimied characteristics
> 
>     this document defines the specific encapsulations that should
>     be used.
> SB> Hopefully it will also define the PHP
> 
>     In addition, it can be expected that there will exist LSR/
>     LERs where only a subset of the physical ports will have the port-
>     based Timing message processing capabilities.
> SB> Do you need to clarify that this only works at base and not in
> SB> a label heirarchy.
> 
> 
>     In order to ensure
>     that the LSPs carrying Timing packets always enter and exit ports
>     with this capability, routing extensions are defined to advertise
>     this capability on a port basis and to allow for the establishment of
>     LSPs that only transit such ports.  While this path establishment
>     restriction may be applied only at the LER Ingress and/or egress
>     ports, it becomes more important when using transparent clock capable
>     LSRs in the path.
> SB> I do not understand the implications of the last
> SB> sentences - starting ", it becomes"
> 
> 
>     Port based Timing message processing involves Timing message
>     recognition.  Once the Timing messages are recognized they can be
>     modified based on the reception or transmission Time-stamp.
> 
>     This document provides two methods for transporting Timing messages
>     over MPLS.  One is applicable to MPLS environment and the other one
>     is applicable to MPLS/MPLS-TP environment
> 
> SB> I think the sentence is incomplete.
> 
> 
> 
> Davari, et al.          Expires December 17, 2013               [Page 5]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
>     The solution involves transporting Timing messages over dedicated
>     LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
>     carry Management and control messages, but not data plane client
>     traffic.
> 
> SB> It is not clear why this restriction applies.
> 
>     Timing LSPs can be established statically or via signaling.
> SB> s/statically/by provisioning/network management/
> 
>     Extensions to control plane (OSPF, ISIS, etc.) is required to enable
>     routers to distribute their Timing processing capabilities over MPLS
>     to other routers.  However such extensions are outside the scope of
>     this document.
> 
>     When signaling is used to setup the PTP LSP, Extensions to signaling
> SB> is it a PTP LSP or a Timing LSP?
> 
>     protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
>     However such extensions are outside the scope of this document.
> 
> SB> for mpls-tp GMPLS is the signalling protocol
> 
>     While the techniques included herein allow for the establishment of
>     paths optimized to include Time-stamping capable links, the
>     performance of the Slave clocks is outside the scope of this
>     document.
> 
>     At the time of publishing this specification, Transparent Clocking
>     (TC) is only defined for PTP.  Therefore at this time any part of
>     this specification that talks about Transparent Clocking applies only
>     to PTP.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013               [Page 6]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 2.  Terminology
> 
>     1588: The timing and synchronization as defined by IEEE 1588.
> 
> SB> I think that there is a more formal name for the 1588 group
> SB> that needs to be used here.
> SB> Also do we need to talk about 1588-200? as there is
> SB> an update in progress
> 
>     NTP: The timing and synchronization protocol defined by IETF RFC-1305
>     and RFC-5905.
> 
>     PTP: The timing and synchronization protocol used by 1588.
> SB> need the proper name for 1588
> 
>     Master Clock: The source of 1588 timing to a set of slave clocks.
> 
>     Master Port: A port on a ordinary or boundary clock that is in Master
>     state.  This is the source of timing toward slave ports.
> 
> SB> I am not sure the reader knows what a port is
> 
>     Slave Clock: A receiver of 1588 timing from a master clock.
> 
>     Slave Port: A port on a boundary clock or ordinary clock that is
>     receiving timing from a master clock.
> 
>     Ordinary Clock: A device with a single PTP port.
> 
>     Transparent Clock.  A device that measures the time taken for a PTP
>     event message to transit the device and then updates the
>     correctionField of the message with this transit time.
> 
>     Boundary Clock: A device with more than one PTP port.  Generally
>     boundary clocks will have one port in slave state to receive timing
>     and then other ports in master state to re-distribute the timing.
> 
>     PTP LSP: An LSP dedicated to carry PTP messages
> 
> SB> PTP or timing?
> 
> 
>     PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>     messages.
> 
> SB> Ah I don't think that PWE3 know what one of these is
> 
>     CW: Pseudowire Control Word
> 
>     LAG: Link Aggregation
> 
>     ECMP: Equal Cost Multipath
> 
>     CF: Correction Field, a field inside certain PTP messages (message
>     type 0-3)that holds the accumulative transit time inside intermediate
>     switches
> 
>     Timing messages: Timing Protocol messages that are exchanged between
>     routers in order to establish a synchronized clock.
> 
> SB> A number of these definitions look like copies of IEEE1588
> SB> definitions. We need to provide references and note the
> SB> priority of the IEEE base reference.
> 
> 
> 
> Davari, et al.          Expires December 17, 2013               [Page 7]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 3.  Problem Statement
> 
>     [IEEE-1588] has defined methods for transporting PTP messages over
>     Ethernet and IP networks.  [RFC5905] has defined the method of
>     transporting NTP messages over IP networks.  There is a need to
>     transport Timing messages over MPLS networks while supporting the
>     Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
>     functionality in the LER and LSRs in the MPLS network.
> 
>     There are multiple ways of transporting Timing over MPLS.  However,
>     there is a requirement to limit the possible encapsulation options to
>     simplify the Timing message identification and processing required at
>     the port level.
> 
>     When Timing-awareness is needed, Timing messages should not be
>     transported over LSPs or PWs that are carrying customer traffic
>     because LSRs perform Label switching based on the top label in the
>     stack.
> 
> SB> Have you explained why?
> 
>     To detect Timing messages inside such LSPs require special
>     hardware to do deep packet inspection at line rate.  Even if such
>     hardware exists, the payload can't be deterministically identified by
>     LSRs because the payload type is a context of the PW label, and the
>     PW label and its context are only known to the Edge routers (PEs/
>     LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CES,
>     etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
>     LSRs dont have the knowledge of whether PW Control Word (CW) is
>     present or not and therefore can not deterministically identify the
>     payload.
> 
>     A generic method is defined in this document that does not require
>     deep packet inspection at line rate, and can deterministically
>     identify Timing messages.  This method can be used to detect Timing
>     Messages in both one-step and two-step clock implementations of
>     ordinary, boundary and transparent clocks.
> 
> SB> Needs a ref and I am sure many MPLS specialists will not understand
> SB> the msg types.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013               [Page 8]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 4.  Timing over MPLS Architecture
> 
>     Timing messages are exchange between Timing ports on ordinary and
> 
> SB> Have you defined a timing port?
> 
>     boundary clocks.  Boundary clocks terminate the Timing messages and
>     act as master for other boundary clocks or for slave clocks.  End-to-
>     End Transparent clocks do not terminate the Timing messages but they
>     do modify the contents of the Timing messages as they transit across
>     the transparent clock.
> 
>     Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Clock
> 
>     (TC) could be implemented in either LERs or LSRs.
> 
> SB> LER and LSR need to be expanded
> 
>     An example is shown in Figure 1, where the LERs act as Ordinary Clock
>     (OC) and are the initiating/terminating point for Timing messages.
>     The ingress LER encapsulates the Timing messages in Timing LSP and
>     the Egress LER terminates the Timing LSP.  The LSRs act as
>     Transparent Clock (TC) and just update the Timing field in the Timing
>     messages.
> 
> 
>        +--------+     +-------+     +-------+     +-------+     +--------+
>        |Switch, |     |       |     |       |     |       |     |Switch, |
>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>        |        |     |  OC   |     |  TC   |     |  OC   |     |        |
>        +--------+     +-------+     +-------+     +-------+     +--------+
>                       /                                 \
>        +-------+     /                                   \     +-------+
>        |  LER  |    /                                     \    |  LER  |
>        | Master|---/                                       \---| Slave |
>        | Clock |                                               | Clock |
>        +-------+                                               +-------+
> 
>       Figure (1) - Deployment example 1 of timing over MPLS network
> 
>     Another example is shown in Figure2, where LERs terminate the Timing
>     messages received from switch/routers that are outside of the MPLS
>     network acting as OC or BC.  In this example LERs regenerate the
>     clock and initiate timing messages encapsulated in Timing LSP toward
>     the MPLS network, while the LSRs act as Transparent Clock (TC) and
>     just update the Timing field in the Timing messages, which are
>     already encapsulated in Timing LSPs.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013               [Page 9]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
>       +--------+     +-------+     +-------+     +-------+     +--------+
>       |Switch, |     |       |     |       |     |       |     |Switch, |
>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>       | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC  |
>       +--------+     +-------+     +-------+     +-------+     +--------+
> 
>       Figure (2) - Deployment example 2 of timing over MPLS network
> 
> 
>     Another example is shown in Figure 3, where LERs do not terminate the
>     Timing messages received from switch/routers that are outside of the
>     MPLS network acting as OC, TC or BC.  The LERs act as TC and update
>     the Timing field in the Timing messages as they transit the LER,
>     while encapsulating them in timing LSP.  The LSRs also act as
>     Transparent Clock (TC) and just update the Timing field in the Timing
>     messages which are already encapsulated in Timing LSPs.
> 
>        +--------+     +-------+     +-------+     +-------+     +--------+
>        |Switch, |     |       |     |       |     |       |     |Switch, |
>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>        |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC/BC|
>        +--------+     +-------+     +-------+     +-------+     +--------+
> 
>      Figure (3) - Deployment example 3 of timing over MPLS network
> 
>     Another example is shown in Figure 4, where LERs and LSRs support
>     Boundary Clocks.  A single-hop LSP is created between two adjacent
>     LSRs engaged in BC operation.  Other methods such as PTP transport
>     over Ethernet MAY be used for transporting timing messages if the
>     link between the two routers is Ethernet.
> 
>       +--------+     +-------+     +-------+     +-------+     +--------+
>       |Switch, |     |       |     |       |     |       |     |Switch, |
>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>       | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC  |
>       +--------+     +-------+     +-------+     +-------+     +--------+
> 
>     Figure (4) - Deployment example 3 of timing over MPLS network
> 
>     An MPLS domain MAY serve multiple customers.  In these cases the MPLS
>     domain (maintained by a service provider) may provide timing services
>     to multiple customers, each having their own Timing domain.
> 
>     The Timing over MPLS architecture assumes full mesh of Timing LSPs
>     between all LERs supporting this specification.
> 
> SB> Note sure this is right - the salves surely do not need to
> SB> exchange timing amongst themselves
> 
>     It supports
>     Point-to- point (VPWS) and Multipoint (VPLS) services.
> 
> SB> What does that mean? You do not carry user data traffic?
> SB> Maybe it's the ordering of the statemnets that is causing
> SB> confusion.
> 
>     This means
>     that a customer may purchase a Point-to-point Timing service between
>     two customer sites or a Multipoint Timing service between more than
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 10]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
>     two customer sites.
> 
>     The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
>     This means that the Timing Multicast messages such as PTP Multicast
>     event messages can be transported over P2MP Timing LSP or be
>     replicated and transported over many P2P Timing LSPs.
> 
> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
> SB> nor a P2MP PW, although we are close.
> 
>     Timing messages, that do not require Time stamping or Correction
>     Field update MAY be transported over Timing LSPs to simplify hardware
>     and software.
> 
>     PTP Announce messages that determine the Timing LSP terminating point
>     behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
>     to simplify hardware and software.
> 
> SB> have you defined and referenced PTP announce msgs?
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 11]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 5.  Dedicated LSPs for Timing messages
> 
>     Many methods have been considered for identifying the Timing messages
>     when they are encapsulated in MPLS such as using GAL/G-ACH or a new
>     reserved label.  These methods were not attractive since they either
>     required deep packet inspection at line rate in the intermediate LSRs
>     or they required use of a scarce new reserved label.  Also one of the
>     goals was to reuse existing OAM mechanisms.
> 
> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
> 
>     The method defined in this document can be used by LER and LSRs to
>     identify Timing messages in MPLS tunnels by just looking at the top
>     label in the MPLS label stack, which only carry Timing messages as
>     well as OAM, but not data plane client traffic.
> 
>     Compliant implementations MUST use dedicated LSPs to carry Timing
>     messages over MPLS.
> 
> SB> I think that we need a definition of the properies of these LSPs
> 
>     These LSPs are herein referred to as "Timing
>     LSPs" and the labels associated with these LSPs as "Timing LSP
>     labels".  The Timing LSPs that runs between Ingress and Egress LERs
>     MUST be co-routed.  Alternatively, a single bidirectional co-routed
>     LSP can be used.
> 
> SB> I though that you said you could use M2MP LSPs - these are not
> SB> bidirectional.
> 
>     Co-routing of the two directions is required to limit the difference
>     in the delays in the Master clock to Slave clock direction compared
>     to the Slave clock to Master clock direction.  The Timing LSP MAY be
>     MPLS/MPLS-TP LSP.
> 
>     The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
>     New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
>     outside the scope of this document.
> 
>     The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such as
>     BFD and LSP Ping but the LSP data plane client plane traffic MUST be
>     Timing packets only.
> 
> SB> Why?
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 12]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 6.  Timing over LSP Encapsulation
> 
> The encapsulations is not LSP is it?
> 
>     This document defines two methods for carrying Timing messages over
>     MPLS.  The first method is carrying UDP/IP encapsulated Timing
>     messages over Timing LSPs, and the second method, is carrying
>     Ethernet encapsulated Timing messages over Ethernet PWs inside Timing
>     LSPs.
> 
> 6.1.  Timing over UDP/IP over MPLS Encapsulation
> 
>     The simplest method of transporting Timing messages over MPLS is to
>     encapsulate Timing PDUs in UDP/IP and then encapsulate them in Timing
>     LSP.  This format is shown in Figure 4.
> 
> 
>                      +----------------------+
>                      |   Timing LSP Label   |
>                      +----------------------+
>                      |        IPv4/6        |
>                      +----------------------+
>                      |         UDP          |
>                      +----------------------+
>                      |     Timing PDU       |
>                      +----------------------+
> 
>        Figure (4) - Timing over UDP/IP over MPLS Encapsulation
> 
> 
>     This encapsulation is very simple and is useful when the network
>     between Timing Master Clock and Slave Clock is MPLS network.
> 
> SB> Simple is a judgement call
> 
>     In order for an LER/LSR to process Timing messages, the Timing LSP
>     Label must be at the top label of the label stack.  The LER/LSR MUST
>     know that the Timing LSP Label is used for carrying Timing messages.
>     This can be accomplished via static configuration or via RSVP-TE
>     signaling.
> 
>     The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>     [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>     [RFC5905].
> 
> 6.2.  Timing over PW Encapsulation
> 
>     Another method of transporting Timing over MPLS networks is by
>     encapsulating Timing PDUs in PW which in turn is transported over
>     Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
>     shown in Fig 5(A) MUST be used and the Ethernet encapsulation of PTP
>     MUST follow Annex F of [IEEE-1588].
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 13]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
>     The RAW mode or Tagged mode defined in [RFC4448] MAY be used and the
>     Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>     Timing over PW encapsulation MUST use the Control Word (CW) as
>     specified in [RFC4448] to ensure proper detection of PTP messages
>     inside the MPLS packets for Timing over LSP and Timing over PW
>     encapsulation.
> 
> SB> That needs explanation
> 
>     The use of Sequence Number in the CW is optional.
> 
> SB> Given that s/n are never in practice deployed, you could probably
> SB> simplify things by sayig that they are not used.
> 
>     Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over PW
>     (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
> 
>                      +----------------+  +----------------+
>                       |Timing LSP Label|  |Timing LSP Label|
>                       +----------------+  +----------------+
>                       |    PW Label    |  |    PW Label    |
>                       +----------------+  +----------------+
>                       |  Control Word  |  |      IP        |
>                       +----------------+  +----------------+
>                       |    Ethernet    |  |      UDP       |
>                       |     Header     |  +----------------+
>                       +----------------+  |   Timing PDU   |
>                       |S-VLAN(Optional)|  |                |
>                       +----------------+  +----------------+
>                       |C-VLAN(Optional)|        (B)
>                       +----------------+
>                       |   Timing PDU   |
>                       |                |
>                       +----------------+
>                              (A)
> 
>                Figure (5) - Timing over PW Encapsulations
> 
>     In order for an LSR to process PTP messages, the top label of the
>     label stack (the Tunnel Label) MUST be a Timing label.
> 
> S> You said that before.
> 
> 6.3.  Other Timing Encapsulation methods
> 
>     In future other timing encapsulation methods may be introduced, such
>     as a new shim header after the Bottom of Stack to carry the Timing
>     information.  Such new encapsulations are outside the scope of this
>     document.
> 
> 
> SB> Taking a pure MPLS pov, you can simplify a lot of the text
> SB> out of the definition of the LSP
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 14]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> SB> I think we need a section on LSP processing
> 
> 7.  Timing message Processing
> 
>     Each Timing protocol such as PTP and NTP, define their set of Timing
>     messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>     FOLLOW_UP, etc messages.
> 
>     Some of the Timing messages require time stamping or correction field
>     update at port level and some dont.  It is the job of the LER/LSR to
>     parse the timing message and find out the type of the Timing message
>     and decide whether and how to Time- stamp it (e.g., BC) or update
>     correction field(e.g., TC).
> 
> 
> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
> SB> function rather than the LER?
> 
>     For example the following PTP messages (called Event messages)
>     require time-stamping or correction field update:
> 
>     o  SYNC
> 
>     o  DELAY_REQ (Delay Request)
> 
>     o  PDELAY_REQ (Peer Delay Request)
> 
>     o  PDELAY_RESP (Peer Delay Response)
> 
>     SYNC and DELAY_REQ are exchanged between Master Clock and Slave Clock
>     and MUST be transported over PTP LSPs.  PDELAY_REQ and PDELAY_RESP
>     are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>     Boundary, or Transparent) and SHOULD be transported over single hop
>     PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
>     and PDELAY_RESP_FOLLOW_UP messages MUST also be transported over
> the
>     PTP LSPs.
> 
>     For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
>     transported over two PTP LSPs that are in opposite directions.  These
>     PTP LSPs, which are in opposite directions MUST be congruent and co-
>     routed.  Alternatively, a single bidirectional co-routed LSP can be
>     used.
> 
>     Except as indicated above for the two-step PTP clocks, Non-Event PTP
>     message types do not need to be processed by intermediate routers.
>     These message types MAY be carried in PTP Tunnel LSPs.
> 
> SB> Are you saying that a timing P router has to be msg type sensitive?
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 15]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 8.  Protection and Redundancy
> 
> 
> SB> This is a bit of a jump - I don't know how the LSP itself works yet!
> 
>     In order to ensure continuous uninterrupted operation of slave
>     clocks, usually as a general practice, slave clocks (or ports) track
>     redundant master clocks.
> 
>     It is the responsibility of the network operator to ensure that
>     physically disjoint Timing LSPs are established between a slave clock
>     (or port) and redundant master clocks (or ports).
> 
>     When a slave clock (or port) listens to redundant master clocks or
>     ports, any prolonged Timing LSP outage will trigger the slave clock
>     or port to switch to a redundant master clock or port.
> 
>     LSP/PW protection such as Linear protection Switching (1:1, 1+1),
>     Ring protection switching or MPLS Fast Reroute (FRR) generally switch
>     alternative path that usually cause a change in delay, which if
>     undetected by slave clock can reduce accuracy of the slave clock.
> 
>     Therefore protection switching MAY be used, as long as phase jumps
>     upon switchover due to differences in path latency are detected and
>     compensated for (such compensation not being required if BCs or peer-
>     peer TCs are used throughout).
> 
>     Note that any protection or reroute mechanism that adds additional
>     MPLS label to the label stack, such as Facility Backup Fast Reroute,
>     MUST ensure that the pushed label is also a Timing Label to ensure
>     recognition of the MPLS frame as containing Timing messages, as it
>     transits the backup path.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 16]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 9.  ECMP
> 
>     To ensure the optimal operation of slave clocks and avoid error
>     introduced by forward and reverse path delay asymmetry, the physical
>     path for Timing messages from master clock to slave Clock and vice
>     versa must be the same for all Event Timing messages listed in
>     section 7.
> 
>     Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
>     Multipath).
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 17]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 10.  PHP
> 
>     To ensure that the label on the top of the label stack is the Timing
>     LSP Label, PHP MUST not be used.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 18]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 11.  Entropy
> 
>     To ensure all Timing messages in a Timing LSP take the same path,
>     Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>     Entropy Label MUST NOT be used for the PWs that are carried inside
>     Timing LSP [RFC6391].
> 
> SB> This is incorrect - you mean that all msgs of the same timing
> SB> flow need to have the same EL value.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 19]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 12.  OAM, Control and Management
> 
>     In order to monitor Timing LSPs and their encapsulated PWs, they MUST
>     be able to carry OAM and management messages.  These management
>     messages MUST be differentiated from Timing messages via already
>     defined IETF methods.
> 
>     For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
>     over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>     Management protocols can easily be identified by the UDP Destination
>     Port number or by GAL/G-ACH respectively.
> 
>     Also BFD, LSP-Ping and other management messages MAY run over the PWs
>     encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3 or
>     4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is going
>     to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1) or
>     GAL-ACH are used to identify such management messages.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 20]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 13.  QoS Considerations
> 
>     In network deployments where not every LSR/LER is Timing-aware, it is
>     important to reduce the impact of the non-Timing-aware LSR/LERs on
>     the timing recovery in the slave clock.  The Timing messages are time
>     critical and must be treated with the highest priority.  Therefore
>     Timing over MPLS messages must be treated with the highest priority
>     in the routers.  This can be achieved by proper setup of Timing LSPs.
> 
>     It is recommended that the Timing LSPs are setup or configured
>     properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC2697]
>     for drop eligibility.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 21]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 14.  FCS and Checksum Recalculation
> 
>     When time-stamp generation and timing packet adjustment is performed
>     near the physical port hardware, the process MUST include
>     recalculation of the Ethernet FCS.
> 
> SB> The above is confusing - an LSR always recomputes the link layer
> SB> CRC which may or may not be Ethernet.
> 
>     Also FCS retention for the
>     payload Ethernet described in [RFC4720] MUST NOT be used.
> 
>     For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
>     may be required as per UDP transport standards.
> 
> SB> You really need to be working on getting the IPv6 C?S computation
> SB> removed from PTP msgs.
> 
>     When UDP checksum is used, each Timing-aware LER/LSR must either
>     incrementally update the UDP checksum after Time stamping or
>     Correction Field update or verify the UDP checksum on reception from
>     upstream and recalculate the checksum completely on transmission to
>     downstream node after Time stamping or Correction Field update.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 22]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 15.  Behavior of LER/LSR
> 
>     Timing-capable/aware LERs and LSRs are routers that have one or more
> 
> SB> You mean physical interfaces?
> 
>     interfaces that can perform Timing operations (OC/BC/TC) on Timing
>     packets and are configured to do so.  Timing-capable/aware LERs and
>     LSRs can advertise their Timing-capability per-interface via control
>     plane such as OSPF or IS-IS.
> SB> ISIS and OSPF are routing protocols.
> 
>    The Timing-capable/aware LERs can then
>     signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timing
>     capability of LER and LSRs may be configured in a centralized
>     controller and the Timing LSP may be setup using manual configuration
>     or other methods such as SDN.
> 
> SB> it can also be configured individually rather then through
> SB> a cebtral controllwe
> 
> 15.1.  Behavior of Timing-capable/aware LER
> 
>     When a Timing-capable/aware LER behaves as a Transparent clock and
>     receives a Timing message from a Timing-capable/aware non-MPLS
>     interface, the LER updates the Correction Field (CF) and encapsulates
>     and forwards the timing message over previously established Timing
>     LSP.
> 
> SB> You need to call out the details so that people properly
> SB> understand the definition of the new LSP.
> 
>     Also when a Timing message is received from a Timing-capable/
>     aware MPLS interface, LER updates the Correction Filed (CF) and
>     decapsulates the MPLS encapsulation and forwards the timing message
>     to a non-MPLS interface.
> 
>     When a Timing-capable/aware LER behaves as a Boundary clock and
>     receives a Timing message from a Timing-capable/aware non MPLS
>     interface, the LER Timestamps the Timing packet and sends it to the
>     LERs Boundary clock processing module.  Also when a Timing message is
>     received from a Timing- capable/aware MPLS interface, the LER
>     Timestamps the Timing packet and sends it to the LERs Boundary clock
>     processing module.
> 
>     When a Timing-capable/aware LER behaves as an Ordinary Clock toward
>     the MPLS network, and receives a Timing message from a Timing-
>     capable/aware MPLS interface, the LER Timestamps the Timing packet
>     and sends it to the LERs Ordinary clock processing module.
> 
> 15.2.  Behavior of Timing-capable/aware LSR
> 
>     When a Timing-capable/aware LSR behaves as a Transparent clock and
>     receives a Timing message from a Timing-capable/aware MPLS interface,
>     The LSR updates the Correction Filed (CF) and forwards the timing
>     message over another MPLS interface.
> 
>     When a Timing-capable/aware LSR behaves as a Boundary clock and
>     receives a Timing message from a Timing-capable/aware MPLS interface.
>     The LSR performs the functions of a Boundary Clock in terminating the
>     received Timing message and re-generating a new timing message over
>     another (or the same) MPLS interface.
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 23]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 15.3.  Behavior of non-Timing-capable/aware LSR
> 
>     It is most beneficial when all LSRs in the path of a Timing LSP be
>     timing-Capable/aware LSRs.  This would ensure the highest quality
>     time and clock synchronization by Timing Slave Clocks.  However, this
>     specification does not mandate that all LSRs in path of a Timing LSP
>     be Timing- capable/aware.
> 
>     Non-Timing-capable/aware LSRs just switch the packets encapsulated in
>     Timing LSPs and dont perform any Timing operation (TC or BC).
>     However as explained in QoS section the Timing over MPLS packets MUST
>     be still be treated with the highest priority based on their Traffic
>     Class (TC) marking.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 24]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 16.  Other considerations
> 
>     [IEEE-1588] defines an optional peer-to-peer Transparent clocking
>     that requires peer delay measurement between two adjacent Timing-
>     capable/ aware routers/switches.  Peer delay measurement messages
>     need to be time stamped and terminated by the Timing-capable/aware
>     routers/ switches.  This means that two adjacent LSRs may be engaged
>     in a peer delay measurement.
> 
>     For transporting such peer delay measurement messages a single-hop
>     LSP SHOULD to be created between the two adjacent LSRs engaged in
>     peer delay measurement to carry peer delay measurement messages.
>     Other methods such as PTP transport over Ethernet MAY be used for
>     transporting peer delay measurement messages if the link between the
>     two routers is Ethernet.
> 
>     In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ ware
>     routers/switches MUST maintain a list of all the neighbors it needs
>     to send a PDelay_Req to, where each neighbor corresponds to a timing
>     LSP.
> 
>     The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as long
>     as either the Explicit Null label is the bottom of stack label
>     (applicable only to UDP/IP encapsulation) or the label below the
>     Explicit Null label is a PTP label.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 25]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 17.  Security Considerations
> 
>     MPLS PW security considerations in general are discussed in [RFC3985]
>     and [RFC4447],and those considerations also apply to this document.
> 
>     An experimental security protocol is defined in [IEEE-1588].The PTP
>     security extension and protocol provides group source authentication,
>     message integrity, and replay attack protection for PTP messages.
> 
>     When the MPLS network (provider network) serves multiple customers,
>     it is important to maintain and process each customers clock and
>     Timing messages separately from other customers to ensure there is no
>     cross- customer effect.  For example if an LER BC is synchronized to
>     a specific grandmaster, belonging to customer A, then the LER MUST
>     use that BC clock only for customer A to ensure that customer A
>     cannot attack other customers by manipulating its time.
> 
>     Timing messages MAY be encrypted or authenticated, provided that the
>     LERs/LSRs that are Timing capable/aware can authenticate/ decrypt the
>     timing messages.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 26]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 18.  Acknowledgements
> 
>     The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizrahi,
>     Stefano Ruffini, Peter Meyer, and other members of IETF for reviewing
>     and providing feedback on this draft.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 27]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 19.  IANA Considerations
> 
>     There are no IANA requirements in this specification.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 28]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 20.  References
> 
> 20.1.  Normative References
> 
>     [IEEE-1588]
>                IEEE 1588-2008, "IEEE Standard for a Precision Clock
>                Synchronization Protocol for Networked Measurement and
>                Control Systems".
> 
>     [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>                Requirement Levels", BCP 14, RFC 2119, March 1997.
> 
>     [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
>                Edge (PWE3) Architecture", RFC 3985, March 2005.
> 
>     [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
>                Proxies (ND Proxy)", RFC 4389, April 2006.
> 
>     [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
>                Heron, "Pseudowire Setup and Maintenance Using the Label
>                Distribution Protocol (LDP)", RFC 4447, April 2006.
> 
>     [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>                "Encapsulation Methods for Transport of Ethernet over MPLS
>                Networks", RFC 4448, April 2006.
> 
>     [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>                Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>                Retention", RFC 4720, November 2006.
> 
>     [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
>                Connectivity Verification (VCCV): A Control Channel for
>                Pseudowires", RFC 5085, December 2007.
> 
>     [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detection
>                (BFD)", RFC 5880, June 2010.
> 
>     [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
>                "Bidirectional Forwarding Detection (BFD) for MPLS Label
>                Switched Paths (LSPs)", RFC 5884, June 2010.
> 
> 20.2.  Informative References
> 
>     [I-D.ietf-pwe3-fat-pw]
>                Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>                J., and S. Amante, "Flow Aware Transport of Pseudowires
>                over an MPLS Packet Switched Network",
>                draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 29]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
>     [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
>                system routeing information exchange protocol for use in
>                conjunction with the Protocol for providing the
>                Connectionless-mode Network Service (ISO 8473)".
> 
>     [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
>                dual environments", RFC 1195, December 1990.
> 
>     [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
> 
>     [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
>                Marker", RFC 2697, September 1999.
> 
>     [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boudec,
>                J., Courtney, W., Davari, S., Firoiu, V., and D.
>                Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>                Behavior)", RFC 3246, March 2002.
> 
>     [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineering
>                (TE) Extensions to OSPF Version 2", RFC 3630,
>                September 2003.
> 
>     [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
>                System (IS-IS) Extensions for Traffic Engineering (TE)",
>                RFC 3784, June 2004.
> 
>     [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
>                Shaffer, "Extensions to OSPF for Advertising Optional
>                Router Capabilities", RFC 4970, July 2007.
> 
>     [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
>                System to Intermediate System (IS-IS) Extensions for
>                Advertising Router Information", RFC 4971, July 2007.
> 
>     [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
>                Topology (MT) Routing in Intermediate System to
>                Intermediate Systems (IS-ISs)", RFC 5120, February 2008.
> 
>     [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>                Engineering", RFC 5305, October 2008.
> 
>     [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>                "Traffic Engineering Extensions to OSPF Version 3",
>                RFC 5329, September 2008.
> 
>     [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
>                for IPv6", RFC 5340, July 2008.
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 30]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
>     [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network
>                Time Protocol Version 4: Protocol and Algorithms
>                Specification", RFC 5905, June 2010.
> 
>     [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>                J., and S. Amante, "Flow-Aware Transport of Pseudowires
>                over an MPLS Packet Switched Network", RFC 6391,
>                November 2011.
> 
>     [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
>                L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
>                RFC 6790, November 2012.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 31]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 1.  Routing extensions for Timing-aware Routers
> 
>     MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] and
>     IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (TE)
>     link information used for constraint-based routing.
> 
>     Indeed, it is useful to advertise data plane TE router link
>     capabilities, such as the capability for a router to be Timing-aware.
>     This capability MUST then be taken into account during path
>     computation to prefer or even require links that advertise themselves
>     as Timing-aware.  In this way the path can ensure the entry and exit
>     points into the LERs and, if desired, the links into the LSRs are
>     able to perform port based time-stamping thus minimizing their impact
>     on the performance of the slave clock.
> 
>     extensions are required to OSPF and IS-IS in order to advertise
>     Timing-aware capabilities of a link.  Such extensions are outside the
>     scope of this document; however such extension SHOULD be able to
>     signal the following information per Router Link:
> 
>     o  Capable of processing PTP, NTP or other Timing flows
> 
>     o  Capable of performing Transparent Clock operation
> 
>     o  Capable of performing Boundary Clock operation
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 32]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> 2.  Signaling Extensions for Creating Timing LSPs
> 
>     RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-TE
>     is used to setup Timing LSPs, some information that indicates that
>     the LSP is carrying Timing flows MUST be included in the new
>     Extensions to RSVP-TE:
> 
>     The following information MAY also be included in the new Extensions
>     to RSVP-TE:
> 
>     o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
>        field
> 
>     o  Number of VLANs in case of PW encapsulation
> 
>     o  Timestamp field Type
> 
>        *  Correction Field, Timestamp
> 
>     o  Timestamp Field format
> 
>        *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
>           NTP, etc.
> 
>     Note that in case the above optional information is signaled with
>     RSVP-TE for a Timing LSP, all the Timing packets carried in that LSP
>     must have the same signaled characteristics.  For example if
>     Timestamp format is signaled as 64-bit PTPv1, then all Timing packets
>     must use 64-bit PTPv1 time-stamp.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 33]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
> Authors' Addresses
> 
>     Shahram Davari
>     Broadcom Corp.
>     San Jose, CA  95134
>     USA
> 
>     Email: davari@broadcom.com
> 
> 
>     Amit Oren
>     Broadcom Corp.
>     San Jose, CA  95134
>     USA
> 
>     Email: amito@broadcom.com
> 
> 
>     Manav Bhatia
>     Alcatel-Lucent
>     Bangalore,
>     India
> 
>     Email: manav.bhatia@alcatel-lucent.com
> 
> 
>     Peter Roberts
>     Alcatel-Lucent
>     Kanata,
>     Canada
> 
>     Email: peter.roberts@alcatel-lucent.com
> 
> 
>     Laurent Montini
>     Cisco Systems
>     San Jose CA
>     USA
> 
>     Email: lmontini@cisco.com
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 34]
> 

> Internet-Draft        Transporting Timing over MPLS            June 2013
> 
> 
>     Luca
>     Cisco Systems
>     San Jose CA
>     USA
> 
>     Email: lmartini@cisco.com
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Davari, et al.          Expires December 17, 2013              [Page 35]
> 

> 
> --
> For corporate legal information go to:
> 
> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
> 
> _______________________________________________
> TICTOC mailing list
> TICTOC@ietf.org
> https://www.ietf.org/mailman/listinfo/tictoc


This e-mail message is intended for the recipient only and contains informat=
ion which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If yo=
u have received this transmission in error, please inform us by e-mail, phon=
e or fax, and then delete the original and all copies thereof.


From davarish@yahoo.com  Sun Aug  4 02:40:25 2013
Return-Path: <davarish@yahoo.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C2A0221F9A97 for <tictoc@ietfa.amsl.com>; Sun,  4 Aug 2013 02:40:25 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.869
X-Spam-Level: 
X-Spam-Status: No, score=-0.869 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, IP_NOT_FRIENDLY=0.334, MIME_QP_LONG_LINE=1.396]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id f6MlTjkyjdCG for <tictoc@ietfa.amsl.com>; Sun,  4 Aug 2013 02:40:21 -0700 (PDT)
Received: from nm48-vm2.bullet.mail.gq1.yahoo.com (nm48-vm2.bullet.mail.gq1.yahoo.com [67.195.87.190]) by ietfa.amsl.com (Postfix) with ESMTP id 538D621F9A7E for <tictoc@ietf.org>; Sun,  4 Aug 2013 02:40:18 -0700 (PDT)
Received: from [98.137.12.63] by nm48.bullet.mail.gq1.yahoo.com with NNFMP; 04 Aug 2013 09:40:16 -0000
Received: from [98.136.164.77] by tm8.bullet.mail.gq1.yahoo.com with NNFMP; 04 Aug 2013 09:40:16 -0000
Received: from [127.0.0.1] by smtp239.mail.gq1.yahoo.com with NNFMP; 04 Aug 2013 09:40:16 -0000
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yahoo.com; s=s1024; t=1375609216; bh=JyRw1/3tiObsThfCkXU5K0iKPNkWj626nUOoWjNweBY=; h=X-Yahoo-Newman-Id:X-Yahoo-Newman-Property:X-YMail-OSG:X-Yahoo-SMTP:X-Rocket-Received:References:Mime-Version:In-Reply-To:Content-Type:Content-Transfer-Encoding:Message-Id:Cc:X-Mailer:From:Subject:Date:To; b=D2XIbHKyq5RSyd62qy80wnlflXBkUd2GIAvz9iEtj+cekloqPfVdsMbO+Y0g3zfmTqX9Txrk5wInTDEEJE7ZM7GUm02xKlvftIGq3Q2cuoOjPFjfVY2XhYIC7uKHBw+kaofA8PN4g7SpXHHP+3+O0kTQN08O3OGDjSGjO9TSICI=
X-Yahoo-Newman-Id: 288057.58595.bm@smtp239.mail.gq1.yahoo.com
X-Yahoo-Newman-Property: ymail-3
X-YMail-OSG: SC5BptIVM1lCQhouG0U8uGHi65bxQXEUev1.flr0tBTrWtI z79mk8.GSMP6wU82FEGR7Vuwsljk2NFwmecWUrJ_d4mWOT_DKnclEClVY79G O1mqK_0uF0apW2rm2LaCPDN4WRaVJh2u6tgH2yB7gzlzT7cp03TYbrpJmw_I 39B6Z1qb1lORrhW7RNmCJ_uMOXgLT5_MPkQS8gQH9GHEOvJkq5mzxnnbjMgV KuCGSCTK5Fza3Hy00UYyOVh9n7.cv_edewfBErwwkj4uWWBfXGP8KPoo4dMr UJnk5qQ2WUGnLYSs65yTjdisamoCTxEazM_H4cJY6DyAay4NAhllOZ54fPOH NqH6x5fCodbSB3Svm149qi9OyxdjqbW2j7mNnnvvo9vmfFXC_fPCwMW0PuMr tdtlZNhJB.AK2GeffYOnUb36jrXEGK0K2t75jqKo9dS6N9DCwPnCcjIVn0HH pxHFOa2IxbD9jv_IEvezpcIR5UBrpLU6DpGZiyVJhKz5cDAZqpZh7CiaHXrd tR8hPudYClqChgLEunii0mnXiH9lkfQ4oM.hEOFI2w900ChU3iR32
X-Yahoo-SMTP: ygPrP9CswBCWPbPtKJlJyLY0KMlg
X-Rocket-Received: from [192.168.0.103] (davarish@98.248.42.143 with ) by smtp239.mail.gq1.yahoo.com with SMTP; 04 Aug 2013 09:40:16 +0000 UTC
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com>
Mime-Version: 1.0 (1.0)
In-Reply-To: <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: quoted-printable
Message-Id: <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com>
X-Mailer: iPhone Mail (10A403)
From: "S. Davari" <davarish@yahoo.com>
Date: Sun, 4 Aug 2013 02:40:13 -0700
To: Alexander Vainshtein <Alexander.Vainshtein@ecitele.com>
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls]  Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 04 Aug 2013 09:40:25 -0000

Hi Sasha

Perhaps you have not understood the draft well. The main reason that the dra=
ft chose to use a dedicated LSP that is signaled via RSVPTE indicating it is=
 a timing LSP is to be backward compatible with non-1588 nodes. Such nodes s=
imply just switch the packet.

Your proposal had been considered and was rejected since it was not backward=
 compatible. Nodes receiving alert label or TTL =3D 1 send the packet to CPU=
. You can refer to the meeting notes.

Regards,
Shahram


On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein <Alexander.Vainshtein@ecite=
le.com> wrote:

> Stewart and all,
> I concur with Stewart's statement that the draft does not define "full int=
eraction with MPLS architecture".
>=20
> E.g., one of the objectives of the draft is to provide a technique that wo=
uld be backward-compatible with old LSRs that cannot provide on-path support=
 for timing distribution, while the other objective is to make every LSR on t=
he path aware that some MPLS packets are carrying timing-related messages an=
d hence require on-path support.
>=20
> IMHO and FWIW, the MPLS data plane architecture allows just two methods fo=
r making a transit LSR to provide special processing to a labeled packet:
> - It would carry some kind of an "alert label" on top of the label stack a=
nd, specifically, on top of any labels used for actual forwarding
> OR,
> - The TTL in the top label stack entry has been set to 1.
>=20
>=20
> The draft does not follow any of these approaches.
>=20
> I must also admit that Section 12 "OAM, Control and Management" of the dra=
ft looks somewhat in
> E.g., I could not understand from the text whether BFD, LSP-Ping etc. OAM m=
essages would be subjected to on-path support procedures for timing messages=
 or not.
>=20
> My 2c,
>     Sasha
>=20
>> -----Original Message-----
>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf O=
f
>> Stewart Bryant
>> Sent: Friday, August 02, 2013 12:01 PM
>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.org; d=
raft-
>> ietf-tictoc-1588overmpls@tools.ietf.org
>> Cc: mpls@ietf.org; tictoc@ietf.org
>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
>>=20
>> SB> This draft does not seem to provide a precise definition
>> SB> the properties of the new LSP type that it wishes to
>> SB> define, in particular it does it define the PHB of those
>> SB> LSPs, nor the full interaction with the MPLS
>> SB> architecture.
>> SB>
>> SB> I have not tracked TICTOC for a while but I thought that
>> SB> the original plan was to define the concept of an offset
>> SB> into a packet to do the correction.
>> SB>
>> SB> It is disappointing that the opportunity was not taken
>> SB> to define a timing shim inside the timing LSP so that
>> SB> a time correction could be added to any packet such that
>> SB> the MPLS system was isolated from the details of the
>> SB> complexity of the particular time transfer type.
>>=20
>> SB> I think that much more clarify is needed in terms of
>> SB> definition of the new LSP type, since it is unclear
>> SB> from this text how to implement one.
>> SB>
>> SB> There are a lot of other MPLS services such as
>> SB> LSP ping that need to be considered.
>> SB>
>> SB> Please see inline for more comments. However these
>> SB> comments are made in the context of the text as written
>> SB> whilst I have a fundamental concern that this approach
>> SB> lacks an MPLS architectural soundness that need
>> SB> greater thought with significant impact on the
>> SB> draft.
>>=20
>> - Stewart
>>=20
>>=20
>> TICTOC Working Group                                           S. Davari
>> Internet-Draft                                                   A. Oren
>> Intended status: Standards Track                          Broadcom Corp.
>> Expires: December 17, 2013                                     M. Bhatia
>>                                                               P. Roberts
>>                                                           Alcatel-Lucent
>>                                                               L. Montini
>>                                                               L. Martini
>>                                                            Cisco Systems
>>                                                            June 15, 2013
>>=20
>>=20
>>             Transporting Timing messages over MPLS Networks
>>                    draft-ietf-tictoc-1588overmpls-05
>>=20
>> Abstract
>>=20
>>    This document defines the method for transporting Timing messages
>>    such as PTP and NTP over an MPLS network.  The method allows for the
>>    easy identification of these PDUs at the port level to allow for port
>>=20
>> SB> What is a port
>>=20
>>    level processing of these PDUs in both LERs and LSRs.
>>=20
>>    The basic idea is to transport Timing messages inside dedicated MPLS
>>    LSPs.  These LSPs only carry Timing messages and possibly Control and
>>    Management packets, but they do not carry customer traffic.
>>=20
>> SB> More specifically they only carry traffic associated with the
>> SB> timing service and its support.
>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
>> SB> also it gets carried in a structure that causes it to get
>> SB> timestamped.
>>=20
>>    Two methods for transporting Timing messages over MPLS are defined.
>>=20
>> SB> Perhaps the right approach is to define the new LSP type and then
>> SB> seperately to define  the mapping of the various timing services
>> SB> over that LSP type.
>>=20
>>    The first method is to transport Timing messages directly over the
>>    dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
>>    MPLS networks.  The second method is to transport Timing messages
>>    inside a PW via Ethernet encapsulation.
>>=20
>> SB> I think that we should note that there are some
>> SB> h/w reasons for this preference. A clean sheet approach
>> SB> would have been to use PTP over MPLS with no intermediate
>> SB> layers.
>>=20
>> Status of this Memo
>>=20
>>    This Internet-Draft is submitted in full conformance with the
>>    provisions of BCP 78 and BCP 79.
>>=20
>>    Internet-Drafts are working documents of the Internet Engineering
>>    Task Force (IETF).  Note that other groups may also distribute
>>    working documents as Internet-Drafts.  The list of current Internet-
>>    Drafts is at http://datatracker.ietf.org/drafts/current/.
>>=20
>>    Internet-Drafts are draft documents valid for a maximum of six months
>>    and may be updated, replaced, or obsoleted by other documents at any
>>    time.  It is inappropriate to use Internet-Drafts as reference
>>    material or to cite them other than as "work in progress."
>>=20
>>    This Internet-Draft will expire on December 17, 2013.
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013               [Page 1]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> Copyright Notice
>>=20
>>    Copyright (c) 2013 IETF Trust and the persons identified as the
>>    document authors.  All rights reserved.
>>=20
>>    This document is subject to BCP 78 and the IETF Trust's Legal
>>    Provisions Relating to IETF Documents
>>    (http://trustee.ietf.org/license-info) in effect on the date of
>>    publication of this document.  Please review these documents
>>    carefully, as they describe your rights and restrictions with respect
>>    to this document.  Code Components extracted from this document must
>>    include Simplified BSD License text as described in Section 4.e of
>>    the Trust Legal Provisions and are provided without warranty as
>>    described in the Simplified BSD License.
>>=20
>>=20
>> Table of Contents
>>=20
>>    1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  5
>>=20
>>    2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .  7
>>=20
>>    3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .  8
>>=20
>>    4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .  9
>>=20
>>    5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . . 12
>>=20
>>    6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . . 13
>>      6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . . 13
>>      6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . . 13
>>      6.3.  Other Timing Encapsulation methods . . . . . . . . . . . . 14
>>=20
>>    7.  Timing message Processing  . . . . . . . . . . . . . . . . . . 15
>>=20
>>    8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . . 16
>>=20
>>    9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
>>=20
>>    10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
>>=20
>>    11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
>>=20
>>    12. OAM, Control and Management  . . . . . . . . . . . . . . . . . 20
>>=20
>>    13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . . 21
>>=20
>>    14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . . 22
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013               [Page 2]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>>    15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . . 23
>>      15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . . 23
>>      15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . . 23
>>      15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . . 24
>>=20
>>    16. Other considerations . . . . . . . . . . . . . . . . . . . . . 25
>>=20
>>    17. Security Considerations  . . . . . . . . . . . . . . . . . . . 26
>>=20
>>    18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 27
>>=20
>>    19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 28
>>=20
>>    20. References . . . . . . . . . . . . . . . . . . . . . . . . . . 29
>>      20.1. Normative References . . . . . . . . . . . . . . . . . . . 29
>>      20.2. Informative References . . . . . . . . . . . . . . . . . . 29
>>=20
>>    Appendix 1.  Routing extensions for Timing-aware Routers . . . . . 32
>>=20
>>    Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . . 33
>>=20
>>    Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 34
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013               [Page 3]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>>    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
>>    "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in
>> this
>>    document are to be interpreted as described in RFC2119 [RFC2119].
>>=20
>>    When used in lower case, these words convey their typical use in
>>    common language, and are not to be interpreted as described in
>>    RFC2119 [RFC2119].
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013               [Page 4]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 1.  Introduction
>>=20
>>    The objective of Precision Time Protocol (PTP) and Network Timing
>>    Protocol (NTP) are to synchronize independent clocks running on
>>    separate nodes of a distributed system.
>>=20
>>    [IEEE-1588] defines PTP messages for frequency, phase and time
>>    synchronization.  The PTP messages include PTP PDUs over UDP/IP
>>    (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F of
>>    [IEEE-1588]).
>>=20
>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
>> SB> of other PTP mappings if they provide better optimisation.
>>=20
>>    This document defines mapping and transport of the PTP
>>    messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>>    defines several clock types: ordinary clocks, boundary clocks, end-
>>    to-end transparent clocks, and peer-to-peer transparent clocks.
>>    Transparent clocks require intermediate nodes to update correction
>>    field inside PTP message that reflects the transit time in the node.
>>=20
>>    [RFC5905] defines NTP messages for clock and time synchronization.
>>    The PTP messages (PDUs) are transported over UDP/IP.  This document
>> SB> Should that be NTP messages?
>> SB> It needs to be made clear as soon as you introduce NTP that
>> SB> they use different time representations.
>>=20
>>    defines mapping and transport of the NTP messages defined in
>>    [RFC5905] over MPLS networks.
>>=20
>>    One key attribute of all of these Timing messages is that the Time
>>    stamp processing should occur as close as possible to the actual
>>    transmission and reception at the physical port interface.  This
>>    targets optimal time and/or frequency recovery by avoiding variable
>>    delay introduced by queues internal to the clocks.
>>=20
>> SB> As I recall NTP has no epoch point defined, and I am not sure
>> SB> where that point is in the case of PTP in this mapping
>> SB> Hopefully this will get defined in due course.
>>=20
>>    To facilitate the fast and efficient recognition of Timing messages
>>    at the port level when the Timing messages are carried over MPLS
>>    LSPs,
>>=20
>> SB> Over a new LSP type with time optimied characteristics
>>=20
>>    this document defines the specific encapsulations that should
>>    be used.
>> SB> Hopefully it will also define the PHP
>>=20
>>    In addition, it can be expected that there will exist LSR/
>>    LERs where only a subset of the physical ports will have the port-
>>    based Timing message processing capabilities.
>> SB> Do you need to clarify that this only works at base and not in
>> SB> a label heirarchy.
>>=20
>>=20
>>    In order to ensure
>>    that the LSPs carrying Timing packets always enter and exit ports
>>    with this capability, routing extensions are defined to advertise
>>    this capability on a port basis and to allow for the establishment of
>>    LSPs that only transit such ports.  While this path establishment
>>    restriction may be applied only at the LER Ingress and/or egress
>>    ports, it becomes more important when using transparent clock capable
>>    LSRs in the path.
>> SB> I do not understand the implications of the last
>> SB> sentences - starting ", it becomes"
>>=20
>>=20
>>    Port based Timing message processing involves Timing message
>>    recognition.  Once the Timing messages are recognized they can be
>>    modified based on the reception or transmission Time-stamp.
>>=20
>>    This document provides two methods for transporting Timing messages
>>    over MPLS.  One is applicable to MPLS environment and the other one
>>    is applicable to MPLS/MPLS-TP environment
>>=20
>> SB> I think the sentence is incomplete.
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013               [Page 5]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>>    The solution involves transporting Timing messages over dedicated
>>    LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
>>    carry Management and control messages, but not data plane client
>>    traffic.
>>=20
>> SB> It is not clear why this restriction applies.
>>=20
>>    Timing LSPs can be established statically or via signaling.
>> SB> s/statically/by provisioning/network management/
>>=20
>>    Extensions to control plane (OSPF, ISIS, etc.) is required to enable
>>    routers to distribute their Timing processing capabilities over MPLS
>>    to other routers.  However such extensions are outside the scope of
>>    this document.
>>=20
>>    When signaling is used to setup the PTP LSP, Extensions to signaling
>> SB> is it a PTP LSP or a Timing LSP?
>>=20
>>    protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
>>    However such extensions are outside the scope of this document.
>>=20
>> SB> for mpls-tp GMPLS is the signalling protocol
>>=20
>>    While the techniques included herein allow for the establishment of
>>    paths optimized to include Time-stamping capable links, the
>>    performance of the Slave clocks is outside the scope of this
>>    document.
>>=20
>>    At the time of publishing this specification, Transparent Clocking
>>    (TC) is only defined for PTP.  Therefore at this time any part of
>>    this specification that talks about Transparent Clocking applies only
>>    to PTP.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013               [Page 6]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 2.  Terminology
>>=20
>>    1588: The timing and synchronization as defined by IEEE 1588.
>>=20
>> SB> I think that there is a more formal name for the 1588 group
>> SB> that needs to be used here.
>> SB> Also do we need to talk about 1588-200? as there is
>> SB> an update in progress
>>=20
>>    NTP: The timing and synchronization protocol defined by IETF RFC-1305
>>    and RFC-5905.
>>=20
>>    PTP: The timing and synchronization protocol used by 1588.
>> SB> need the proper name for 1588
>>=20
>>    Master Clock: The source of 1588 timing to a set of slave clocks.
>>=20
>>    Master Port: A port on a ordinary or boundary clock that is in Master
>>    state.  This is the source of timing toward slave ports.
>>=20
>> SB> I am not sure the reader knows what a port is
>>=20
>>    Slave Clock: A receiver of 1588 timing from a master clock.
>>=20
>>    Slave Port: A port on a boundary clock or ordinary clock that is
>>    receiving timing from a master clock.
>>=20
>>    Ordinary Clock: A device with a single PTP port.
>>=20
>>    Transparent Clock.  A device that measures the time taken for a PTP
>>    event message to transit the device and then updates the
>>    correctionField of the message with this transit time.
>>=20
>>    Boundary Clock: A device with more than one PTP port.  Generally
>>    boundary clocks will have one port in slave state to receive timing
>>    and then other ports in master state to re-distribute the timing.
>>=20
>>    PTP LSP: An LSP dedicated to carry PTP messages
>>=20
>> SB> PTP or timing?
>>=20
>>=20
>>    PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>    messages.
>>=20
>> SB> Ah I don't think that PWE3 know what one of these is
>>=20
>>    CW: Pseudowire Control Word
>>=20
>>    LAG: Link Aggregation
>>=20
>>    ECMP: Equal Cost Multipath
>>=20
>>    CF: Correction Field, a field inside certain PTP messages (message
>>    type 0-3)that holds the accumulative transit time inside intermediate
>>    switches
>>=20
>>    Timing messages: Timing Protocol messages that are exchanged between
>>    routers in order to establish a synchronized clock.
>>=20
>> SB> A number of these definitions look like copies of IEEE1588
>> SB> definitions. We need to provide references and note the
>> SB> priority of the IEEE base reference.
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013               [Page 7]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 3.  Problem Statement
>>=20
>>    [IEEE-1588] has defined methods for transporting PTP messages over
>>    Ethernet and IP networks.  [RFC5905] has defined the method of
>>    transporting NTP messages over IP networks.  There is a need to
>>    transport Timing messages over MPLS networks while supporting the
>>    Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
>>    functionality in the LER and LSRs in the MPLS network.
>>=20
>>    There are multiple ways of transporting Timing over MPLS.  However,
>>    there is a requirement to limit the possible encapsulation options to
>>    simplify the Timing message identification and processing required at
>>    the port level.
>>=20
>>    When Timing-awareness is needed, Timing messages should not be
>>    transported over LSPs or PWs that are carrying customer traffic
>>    because LSRs perform Label switching based on the top label in the
>>    stack.
>>=20
>> SB> Have you explained why?
>>=20
>>    To detect Timing messages inside such LSPs require special
>>    hardware to do deep packet inspection at line rate.  Even if such
>>    hardware exists, the payload can't be deterministically identified by
>>    LSRs because the payload type is a context of the PW label, and the
>>    PW label and its context are only known to the Edge routers (PEs/
>>    LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CES,
>>    etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
>>    LSRs dont have the knowledge of whether PW Control Word (CW) is
>>    present or not and therefore can not deterministically identify the
>>    payload.
>>=20
>>    A generic method is defined in this document that does not require
>>    deep packet inspection at line rate, and can deterministically
>>    identify Timing messages.  This method can be used to detect Timing
>>    Messages in both one-step and two-step clock implementations of
>>    ordinary, boundary and transparent clocks.
>>=20
>> SB> Needs a ref and I am sure many MPLS specialists will not understand
>> SB> the msg types.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013               [Page 8]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 4.  Timing over MPLS Architecture
>>=20
>>    Timing messages are exchange between Timing ports on ordinary and
>>=20
>> SB> Have you defined a timing port?
>>=20
>>    boundary clocks.  Boundary clocks terminate the Timing messages and
>>    act as master for other boundary clocks or for slave clocks.  End-to-
>>    End Transparent clocks do not terminate the Timing messages but they
>>    do modify the contents of the Timing messages as they transit across
>>    the transparent clock.
>>=20
>>    Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Clock
>>=20
>>    (TC) could be implemented in either LERs or LSRs.
>>=20
>> SB> LER and LSR need to be expanded
>>=20
>>    An example is shown in Figure 1, where the LERs act as Ordinary Clock
>>    (OC) and are the initiating/terminating point for Timing messages.
>>    The ingress LER encapsulates the Timing messages in Timing LSP and
>>    the Egress LER terminates the Timing LSP.  The LSRs act as
>>    Transparent Clock (TC) and just update the Timing field in the Timing
>>    messages.
>>=20
>>=20
>>       +--------+     +-------+     +-------+     +-------+     +--------+=

>>       |Switch, |     |       |     |       |     |       |     |Switch, |=

>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |=

>>       |        |     |  OC   |     |  TC   |     |  OC   |     |        |=

>>       +--------+     +-------+     +-------+     +-------+     +--------+=

>>                      /                                 \
>>       +-------+     /                                   \     +-------+
>>       |  LER  |    /                                     \    |  LER  |
>>       | Master|---/                                       \---| Slave |
>>       | Clock |                                               | Clock |
>>       +-------+                                               +-------+
>>=20
>>      Figure (1) - Deployment example 1 of timing over MPLS network
>>=20
>>    Another example is shown in Figure2, where LERs terminate the Timing
>>    messages received from switch/routers that are outside of the MPLS
>>    network acting as OC or BC.  In this example LERs regenerate the
>>    clock and initiate timing messages encapsulated in Timing LSP toward
>>    the MPLS network, while the LSRs act as Transparent Clock (TC) and
>>    just update the Timing field in the Timing messages, which are
>>    already encapsulated in Timing LSPs.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013               [Page 9]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>>      +--------+     +-------+     +-------+     +-------+     +--------+
>>      |Switch, |     |       |     |       |     |       |     |Switch, |
>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>      | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC  |
>>      +--------+     +-------+     +-------+     +-------+     +--------+
>>=20
>>      Figure (2) - Deployment example 2 of timing over MPLS network
>>=20
>>=20
>>    Another example is shown in Figure 3, where LERs do not terminate the
>>    Timing messages received from switch/routers that are outside of the
>>    MPLS network acting as OC, TC or BC.  The LERs act as TC and update
>>    the Timing field in the Timing messages as they transit the LER,
>>    while encapsulating them in timing LSP.  The LSRs also act as
>>    Transparent Clock (TC) and just update the Timing field in the Timing
>>    messages which are already encapsulated in Timing LSPs.
>>=20
>>       +--------+     +-------+     +-------+     +-------+     +--------+=

>>       |Switch, |     |       |     |       |     |       |     |Switch, |=

>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |=

>>       |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC/BC|=

>>       +--------+     +-------+     +-------+     +-------+     +--------+=

>>=20
>>     Figure (3) - Deployment example 3 of timing over MPLS network
>>=20
>>    Another example is shown in Figure 4, where LERs and LSRs support
>>    Boundary Clocks.  A single-hop LSP is created between two adjacent
>>    LSRs engaged in BC operation.  Other methods such as PTP transport
>>    over Ethernet MAY be used for transporting timing messages if the
>>    link between the two routers is Ethernet.
>>=20
>>      +--------+     +-------+     +-------+     +-------+     +--------+
>>      |Switch, |     |       |     |       |     |       |     |Switch, |
>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>      | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC  |
>>      +--------+     +-------+     +-------+     +-------+     +--------+
>>=20
>>    Figure (4) - Deployment example 3 of timing over MPLS network
>>=20
>>    An MPLS domain MAY serve multiple customers.  In these cases the MPLS
>>    domain (maintained by a service provider) may provide timing services
>>    to multiple customers, each having their own Timing domain.
>>=20
>>    The Timing over MPLS architecture assumes full mesh of Timing LSPs
>>    between all LERs supporting this specification.
>>=20
>> SB> Note sure this is right - the salves surely do not need to
>> SB> exchange timing amongst themselves
>>=20
>>    It supports
>>    Point-to- point (VPWS) and Multipoint (VPLS) services.
>>=20
>> SB> What does that mean? You do not carry user data traffic?
>> SB> Maybe it's the ordering of the statemnets that is causing
>> SB> confusion.
>>=20
>>    This means
>>    that a customer may purchase a Point-to-point Timing service between
>>    two customer sites or a Multipoint Timing service between more than
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 10]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>>    two customer sites.
>>=20
>>    The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
>>    This means that the Timing Multicast messages such as PTP Multicast
>>    event messages can be transported over P2MP Timing LSP or be
>>    replicated and transported over many P2P Timing LSPs.
>>=20
>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>> SB> nor a P2MP PW, although we are close.
>>=20
>>    Timing messages, that do not require Time stamping or Correction
>>    Field update MAY be transported over Timing LSPs to simplify hardware
>>    and software.
>>=20
>>    PTP Announce messages that determine the Timing LSP terminating point
>>    behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
>>    to simplify hardware and software.
>>=20
>> SB> have you defined and referenced PTP announce msgs?
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 11]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 5.  Dedicated LSPs for Timing messages
>>=20
>>    Many methods have been considered for identifying the Timing messages
>>    when they are encapsulated in MPLS such as using GAL/G-ACH or a new
>>    reserved label.  These methods were not attractive since they either
>>    required deep packet inspection at line rate in the intermediate LSRs
>>    or they required use of a scarce new reserved label.  Also one of the
>>    goals was to reuse existing OAM mechanisms.
>>=20
>> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
>>=20
>>    The method defined in this document can be used by LER and LSRs to
>>    identify Timing messages in MPLS tunnels by just looking at the top
>>    label in the MPLS label stack, which only carry Timing messages as
>>    well as OAM, but not data plane client traffic.
>>=20
>>    Compliant implementations MUST use dedicated LSPs to carry Timing
>>    messages over MPLS.
>>=20
>> SB> I think that we need a definition of the properies of these LSPs
>>=20
>>    These LSPs are herein referred to as "Timing
>>    LSPs" and the labels associated with these LSPs as "Timing LSP
>>    labels".  The Timing LSPs that runs between Ingress and Egress LERs
>>    MUST be co-routed.  Alternatively, a single bidirectional co-routed
>>    LSP can be used.
>>=20
>> SB> I though that you said you could use M2MP LSPs - these are not
>> SB> bidirectional.
>>=20
>>    Co-routing of the two directions is required to limit the difference
>>    in the delays in the Master clock to Slave clock direction compared
>>    to the Slave clock to Master clock direction.  The Timing LSP MAY be
>>    MPLS/MPLS-TP LSP.
>>=20
>>    The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
>>    New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
>>    outside the scope of this document.
>>=20
>>    The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such as
>>    BFD and LSP Ping but the LSP data plane client plane traffic MUST be
>>    Timing packets only.
>>=20
>> SB> Why?
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 12]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 6.  Timing over LSP Encapsulation
>>=20
>> The encapsulations is not LSP is it?
>>=20
>>    This document defines two methods for carrying Timing messages over
>>    MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>    messages over Timing LSPs, and the second method, is carrying
>>    Ethernet encapsulated Timing messages over Ethernet PWs inside Timing
>>    LSPs.
>>=20
>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>=20
>>    The simplest method of transporting Timing messages over MPLS is to
>>    encapsulate Timing PDUs in UDP/IP and then encapsulate them in Timing
>>    LSP.  This format is shown in Figure 4.
>>=20
>>=20
>>                     +----------------------+
>>                     |   Timing LSP Label   |
>>                     +----------------------+
>>                     |        IPv4/6        |
>>                     +----------------------+
>>                     |         UDP          |
>>                     +----------------------+
>>                     |     Timing PDU       |
>>                     +----------------------+
>>=20
>>       Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>=20
>>=20
>>    This encapsulation is very simple and is useful when the network
>>    between Timing Master Clock and Slave Clock is MPLS network.
>>=20
>> SB> Simple is a judgement call
>>=20
>>    In order for an LER/LSR to process Timing messages, the Timing LSP
>>    Label must be at the top label of the label stack.  The LER/LSR MUST
>>    know that the Timing LSP Label is used for carrying Timing messages.
>>    This can be accomplished via static configuration or via RSVP-TE
>>    signaling.
>>=20
>>    The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>    [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>>    [RFC5905].
>>=20
>> 6.2.  Timing over PW Encapsulation
>>=20
>>    Another method of transporting Timing over MPLS networks is by
>>    encapsulating Timing PDUs in PW which in turn is transported over
>>    Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
>>    shown in Fig 5(A) MUST be used and the Ethernet encapsulation of PTP
>>    MUST follow Annex F of [IEEE-1588].
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 13]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>>    The RAW mode or Tagged mode defined in [RFC4448] MAY be used and the
>>    Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>>    Timing over PW encapsulation MUST use the Control Word (CW) as
>>    specified in [RFC4448] to ensure proper detection of PTP messages
>>    inside the MPLS packets for Timing over LSP and Timing over PW
>>    encapsulation.
>>=20
>> SB> That needs explanation
>>=20
>>    The use of Sequence Number in the CW is optional.
>>=20
>> SB> Given that s/n are never in practice deployed, you could probably
>> SB> simplify things by sayig that they are not used.
>>=20
>>    Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over PW
>>    (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>=20
>>                     +----------------+  +----------------+
>>                      |Timing LSP Label|  |Timing LSP Label|
>>                      +----------------+  +----------------+
>>                      |    PW Label    |  |    PW Label    |
>>                      +----------------+  +----------------+
>>                      |  Control Word  |  |      IP        |
>>                      +----------------+  +----------------+
>>                      |    Ethernet    |  |      UDP       |
>>                      |     Header     |  +----------------+
>>                      +----------------+  |   Timing PDU   |
>>                      |S-VLAN(Optional)|  |                |
>>                      +----------------+  +----------------+
>>                      |C-VLAN(Optional)|        (B)
>>                      +----------------+
>>                      |   Timing PDU   |
>>                      |                |
>>                      +----------------+
>>                             (A)
>>=20
>>               Figure (5) - Timing over PW Encapsulations
>>=20
>>    In order for an LSR to process PTP messages, the top label of the
>>    label stack (the Tunnel Label) MUST be a Timing label.
>>=20
>> S> You said that before.
>>=20
>> 6.3.  Other Timing Encapsulation methods
>>=20
>>    In future other timing encapsulation methods may be introduced, such
>>    as a new shim header after the Bottom of Stack to carry the Timing
>>    information.  Such new encapsulations are outside the scope of this
>>    document.
>>=20
>>=20
>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>> SB> out of the definition of the LSP
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 14]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> SB> I think we need a section on LSP processing
>>=20
>> 7.  Timing message Processing
>>=20
>>    Each Timing protocol such as PTP and NTP, define their set of Timing
>>    messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>>    FOLLOW_UP, etc messages.
>>=20
>>    Some of the Timing messages require time stamping or correction field
>>    update at port level and some dont.  It is the job of the LER/LSR to
>>    parse the timing message and find out the type of the Timing message
>>    and decide whether and how to Time- stamp it (e.g., BC) or update
>>    correction field(e.g., TC).
>>=20
>>=20
>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
>> SB> function rather than the LER?
>>=20
>>    For example the following PTP messages (called Event messages)
>>    require time-stamping or correction field update:
>>=20
>>    o  SYNC
>>=20
>>    o  DELAY_REQ (Delay Request)
>>=20
>>    o  PDELAY_REQ (Peer Delay Request)
>>=20
>>    o  PDELAY_RESP (Peer Delay Response)
>>=20
>>    SYNC and DELAY_REQ are exchanged between Master Clock and Slave Clock
>>    and MUST be transported over PTP LSPs.  PDELAY_REQ and PDELAY_RESP
>>    are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>>    Boundary, or Transparent) and SHOULD be transported over single hop
>>    PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
>>    and PDELAY_RESP_FOLLOW_UP messages MUST also be transported over
>> the
>>    PTP LSPs.
>>=20
>>    For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
>>    transported over two PTP LSPs that are in opposite directions.  These
>>    PTP LSPs, which are in opposite directions MUST be congruent and co-
>>    routed.  Alternatively, a single bidirectional co-routed LSP can be
>>    used.
>>=20
>>    Except as indicated above for the two-step PTP clocks, Non-Event PTP
>>    message types do not need to be processed by intermediate routers.
>>    These message types MAY be carried in PTP Tunnel LSPs.
>>=20
>> SB> Are you saying that a timing P router has to be msg type sensitive?
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 15]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 8.  Protection and Redundancy
>>=20
>>=20
>> SB> This is a bit of a jump - I don't know how the LSP itself works yet!
>>=20
>>    In order to ensure continuous uninterrupted operation of slave
>>    clocks, usually as a general practice, slave clocks (or ports) track
>>    redundant master clocks.
>>=20
>>    It is the responsibility of the network operator to ensure that
>>    physically disjoint Timing LSPs are established between a slave clock
>>    (or port) and redundant master clocks (or ports).
>>=20
>>    When a slave clock (or port) listens to redundant master clocks or
>>    ports, any prolonged Timing LSP outage will trigger the slave clock
>>    or port to switch to a redundant master clock or port.
>>=20
>>    LSP/PW protection such as Linear protection Switching (1:1, 1+1),
>>    Ring protection switching or MPLS Fast Reroute (FRR) generally switch
>>    alternative path that usually cause a change in delay, which if
>>    undetected by slave clock can reduce accuracy of the slave clock.
>>=20
>>    Therefore protection switching MAY be used, as long as phase jumps
>>    upon switchover due to differences in path latency are detected and
>>    compensated for (such compensation not being required if BCs or peer-
>>    peer TCs are used throughout).
>>=20
>>    Note that any protection or reroute mechanism that adds additional
>>    MPLS label to the label stack, such as Facility Backup Fast Reroute,
>>    MUST ensure that the pushed label is also a Timing Label to ensure
>>    recognition of the MPLS frame as containing Timing messages, as it
>>    transits the backup path.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 16]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 9.  ECMP
>>=20
>>    To ensure the optimal operation of slave clocks and avoid error
>>    introduced by forward and reverse path delay asymmetry, the physical
>>    path for Timing messages from master clock to slave Clock and vice
>>    versa must be the same for all Event Timing messages listed in
>>    section 7.
>>=20
>>    Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
>>    Multipath).
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 17]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 10.  PHP
>>=20
>>    To ensure that the label on the top of the label stack is the Timing
>>    LSP Label, PHP MUST not be used.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 18]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 11.  Entropy
>>=20
>>    To ensure all Timing messages in a Timing LSP take the same path,
>>    Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>    Entropy Label MUST NOT be used for the PWs that are carried inside
>>    Timing LSP [RFC6391].
>>=20
>> SB> This is incorrect - you mean that all msgs of the same timing
>> SB> flow need to have the same EL value.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 19]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 12.  OAM, Control and Management
>>=20
>>    In order to monitor Timing LSPs and their encapsulated PWs, they MUST
>>    be able to carry OAM and management messages.  These management
>>    messages MUST be differentiated from Timing messages via already
>>    defined IETF methods.
>>=20
>>    For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
>>    over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>>    Management protocols can easily be identified by the UDP Destination
>>    Port number or by GAL/G-ACH respectively.
>>=20
>>    Also BFD, LSP-Ping and other management messages MAY run over the PWs
>>    encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3 or
>>    4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is going
>>    to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1) or
>>    GAL-ACH are used to identify such management messages.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 20]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 13.  QoS Considerations
>>=20
>>    In network deployments where not every LSR/LER is Timing-aware, it is
>>    important to reduce the impact of the non-Timing-aware LSR/LERs on
>>    the timing recovery in the slave clock.  The Timing messages are time
>>    critical and must be treated with the highest priority.  Therefore
>>    Timing over MPLS messages must be treated with the highest priority
>>    in the routers.  This can be achieved by proper setup of Timing LSPs.
>>=20
>>    It is recommended that the Timing LSPs are setup or configured
>>    properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC2697]
>>    for drop eligibility.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 21]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 14.  FCS and Checksum Recalculation
>>=20
>>    When time-stamp generation and timing packet adjustment is performed
>>    near the physical port hardware, the process MUST include
>>    recalculation of the Ethernet FCS.
>>=20
>> SB> The above is confusing - an LSR always recomputes the link layer
>> SB> CRC which may or may not be Ethernet.
>>=20
>>    Also FCS retention for the
>>    payload Ethernet described in [RFC4720] MUST NOT be used.
>>=20
>>    For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
>>    may be required as per UDP transport standards.
>>=20
>> SB> You really need to be working on getting the IPv6 C?S computation
>> SB> removed from PTP msgs.
>>=20
>>    When UDP checksum is used, each Timing-aware LER/LSR must either
>>    incrementally update the UDP checksum after Time stamping or
>>    Correction Field update or verify the UDP checksum on reception from
>>    upstream and recalculate the checksum completely on transmission to
>>    downstream node after Time stamping or Correction Field update.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 22]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 15.  Behavior of LER/LSR
>>=20
>>    Timing-capable/aware LERs and LSRs are routers that have one or more
>>=20
>> SB> You mean physical interfaces?
>>=20
>>    interfaces that can perform Timing operations (OC/BC/TC) on Timing
>>    packets and are configured to do so.  Timing-capable/aware LERs and
>>    LSRs can advertise their Timing-capability per-interface via control
>>    plane such as OSPF or IS-IS.
>> SB> ISIS and OSPF are routing protocols.
>>=20
>>   The Timing-capable/aware LERs can then
>>    signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timing
>>    capability of LER and LSRs may be configured in a centralized
>>    controller and the Timing LSP may be setup using manual configuration
>>    or other methods such as SDN.
>>=20
>> SB> it can also be configured individually rather then through
>> SB> a cebtral controllwe
>>=20
>> 15.1.  Behavior of Timing-capable/aware LER
>>=20
>>    When a Timing-capable/aware LER behaves as a Transparent clock and
>>    receives a Timing message from a Timing-capable/aware non-MPLS
>>    interface, the LER updates the Correction Field (CF) and encapsulates
>>    and forwards the timing message over previously established Timing
>>    LSP.
>>=20
>> SB> You need to call out the details so that people properly
>> SB> understand the definition of the new LSP.
>>=20
>>    Also when a Timing message is received from a Timing-capable/
>>    aware MPLS interface, LER updates the Correction Filed (CF) and
>>    decapsulates the MPLS encapsulation and forwards the timing message
>>    to a non-MPLS interface.
>>=20
>>    When a Timing-capable/aware LER behaves as a Boundary clock and
>>    receives a Timing message from a Timing-capable/aware non MPLS
>>    interface, the LER Timestamps the Timing packet and sends it to the
>>    LERs Boundary clock processing module.  Also when a Timing message is
>>    received from a Timing- capable/aware MPLS interface, the LER
>>    Timestamps the Timing packet and sends it to the LERs Boundary clock
>>    processing module.
>>=20
>>    When a Timing-capable/aware LER behaves as an Ordinary Clock toward
>>    the MPLS network, and receives a Timing message from a Timing-
>>    capable/aware MPLS interface, the LER Timestamps the Timing packet
>>    and sends it to the LERs Ordinary clock processing module.
>>=20
>> 15.2.  Behavior of Timing-capable/aware LSR
>>=20
>>    When a Timing-capable/aware LSR behaves as a Transparent clock and
>>    receives a Timing message from a Timing-capable/aware MPLS interface,
>>    The LSR updates the Correction Filed (CF) and forwards the timing
>>    message over another MPLS interface.
>>=20
>>    When a Timing-capable/aware LSR behaves as a Boundary clock and
>>    receives a Timing message from a Timing-capable/aware MPLS interface.
>>    The LSR performs the functions of a Boundary Clock in terminating the
>>    received Timing message and re-generating a new timing message over
>>    another (or the same) MPLS interface.
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 23]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>=20
>>    It is most beneficial when all LSRs in the path of a Timing LSP be
>>    timing-Capable/aware LSRs.  This would ensure the highest quality
>>    time and clock synchronization by Timing Slave Clocks.  However, this
>>    specification does not mandate that all LSRs in path of a Timing LSP
>>    be Timing- capable/aware.
>>=20
>>    Non-Timing-capable/aware LSRs just switch the packets encapsulated in
>>    Timing LSPs and dont perform any Timing operation (TC or BC).
>>    However as explained in QoS section the Timing over MPLS packets MUST
>>    be still be treated with the highest priority based on their Traffic
>>    Class (TC) marking.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 24]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 16.  Other considerations
>>=20
>>    [IEEE-1588] defines an optional peer-to-peer Transparent clocking
>>    that requires peer delay measurement between two adjacent Timing-
>>    capable/ aware routers/switches.  Peer delay measurement messages
>>    need to be time stamped and terminated by the Timing-capable/aware
>>    routers/ switches.  This means that two adjacent LSRs may be engaged
>>    in a peer delay measurement.
>>=20
>>    For transporting such peer delay measurement messages a single-hop
>>    LSP SHOULD to be created between the two adjacent LSRs engaged in
>>    peer delay measurement to carry peer delay measurement messages.
>>    Other methods such as PTP transport over Ethernet MAY be used for
>>    transporting peer delay measurement messages if the link between the
>>    two routers is Ethernet.
>>=20
>>    In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ ware
>>    routers/switches MUST maintain a list of all the neighbors it needs
>>    to send a PDelay_Req to, where each neighbor corresponds to a timing
>>    LSP.
>>=20
>>    The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as long=

>>    as either the Explicit Null label is the bottom of stack label
>>    (applicable only to UDP/IP encapsulation) or the label below the
>>    Explicit Null label is a PTP label.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 25]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 17.  Security Considerations
>>=20
>>    MPLS PW security considerations in general are discussed in [RFC3985]
>>    and [RFC4447],and those considerations also apply to this document.
>>=20
>>    An experimental security protocol is defined in [IEEE-1588].The PTP
>>    security extension and protocol provides group source authentication,
>>    message integrity, and replay attack protection for PTP messages.
>>=20
>>    When the MPLS network (provider network) serves multiple customers,
>>    it is important to maintain and process each customers clock and
>>    Timing messages separately from other customers to ensure there is no
>>    cross- customer effect.  For example if an LER BC is synchronized to
>>    a specific grandmaster, belonging to customer A, then the LER MUST
>>    use that BC clock only for customer A to ensure that customer A
>>    cannot attack other customers by manipulating its time.
>>=20
>>    Timing messages MAY be encrypted or authenticated, provided that the
>>    LERs/LSRs that are Timing capable/aware can authenticate/ decrypt the
>>    timing messages.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 26]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 18.  Acknowledgements
>>=20
>>    The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizrahi,
>>    Stefano Ruffini, Peter Meyer, and other members of IETF for reviewing
>>    and providing feedback on this draft.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 27]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 19.  IANA Considerations
>>=20
>>    There are no IANA requirements in this specification.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 28]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 20.  References
>>=20
>> 20.1.  Normative References
>>=20
>>    [IEEE-1588]
>>               IEEE 1588-2008, "IEEE Standard for a Precision Clock
>>               Synchronization Protocol for Networked Measurement and
>>               Control Systems".
>>=20
>>    [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>               Requirement Levels", BCP 14, RFC 2119, March 1997.
>>=20
>>    [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
>>               Edge (PWE3) Architecture", RFC 3985, March 2005.
>>=20
>>    [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
>>               Proxies (ND Proxy)", RFC 4389, April 2006.
>>=20
>>    [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
>>               Heron, "Pseudowire Setup and Maintenance Using the Label
>>               Distribution Protocol (LDP)", RFC 4447, April 2006.
>>=20
>>    [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>               "Encapsulation Methods for Transport of Ethernet over MPLS
>>               Networks", RFC 4448, April 2006.
>>=20
>>    [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>               Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>               Retention", RFC 4720, November 2006.
>>=20
>>    [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
>>               Connectivity Verification (VCCV): A Control Channel for
>>               Pseudowires", RFC 5085, December 2007.
>>=20
>>    [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detection
>>               (BFD)", RFC 5880, June 2010.
>>=20
>>    [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
>>               "Bidirectional Forwarding Detection (BFD) for MPLS Label
>>               Switched Paths (LSPs)", RFC 5884, June 2010.
>>=20
>> 20.2.  Informative References
>>=20
>>    [I-D.ietf-pwe3-fat-pw]
>>               Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>>               J., and S. Amante, "Flow Aware Transport of Pseudowires
>>               over an MPLS Packet Switched Network",
>>               draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 29]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>>    [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
>>               system routeing information exchange protocol for use in
>>               conjunction with the Protocol for providing the
>>               Connectionless-mode Network Service (ISO 8473)".
>>=20
>>    [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
>>               dual environments", RFC 1195, December 1990.
>>=20
>>    [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
>>=20
>>    [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
>>               Marker", RFC 2697, September 1999.
>>=20
>>    [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boudec,
>>               J., Courtney, W., Davari, S., Firoiu, V., and D.
>>               Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>               Behavior)", RFC 3246, March 2002.
>>=20
>>    [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineering
>>               (TE) Extensions to OSPF Version 2", RFC 3630,
>>               September 2003.
>>=20
>>    [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
>>               System (IS-IS) Extensions for Traffic Engineering (TE)",
>>               RFC 3784, June 2004.
>>=20
>>    [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
>>               Shaffer, "Extensions to OSPF for Advertising Optional
>>               Router Capabilities", RFC 4970, July 2007.
>>=20
>>    [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
>>               System to Intermediate System (IS-IS) Extensions for
>>               Advertising Router Information", RFC 4971, July 2007.
>>=20
>>    [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
>>               Topology (MT) Routing in Intermediate System to
>>               Intermediate Systems (IS-ISs)", RFC 5120, February 2008.
>>=20
>>    [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>               Engineering", RFC 5305, October 2008.
>>=20
>>    [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>               "Traffic Engineering Extensions to OSPF Version 3",
>>               RFC 5329, September 2008.
>>=20
>>    [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
>>               for IPv6", RFC 5340, July 2008.
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 30]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>>    [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network
>>               Time Protocol Version 4: Protocol and Algorithms
>>               Specification", RFC 5905, June 2010.
>>=20
>>    [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>>               J., and S. Amante, "Flow-Aware Transport of Pseudowires
>>               over an MPLS Packet Switched Network", RFC 6391,
>>               November 2011.
>>=20
>>    [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
>>               L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
>>               RFC 6790, November 2012.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 31]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 1.  Routing extensions for Timing-aware Routers
>>=20
>>    MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] and
>>    IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (TE)
>>    link information used for constraint-based routing.
>>=20
>>    Indeed, it is useful to advertise data plane TE router link
>>    capabilities, such as the capability for a router to be Timing-aware.
>>    This capability MUST then be taken into account during path
>>    computation to prefer or even require links that advertise themselves
>>    as Timing-aware.  In this way the path can ensure the entry and exit
>>    points into the LERs and, if desired, the links into the LSRs are
>>    able to perform port based time-stamping thus minimizing their impact
>>    on the performance of the slave clock.
>>=20
>>    extensions are required to OSPF and IS-IS in order to advertise
>>    Timing-aware capabilities of a link.  Such extensions are outside the
>>    scope of this document; however such extension SHOULD be able to
>>    signal the following information per Router Link:
>>=20
>>    o  Capable of processing PTP, NTP or other Timing flows
>>=20
>>    o  Capable of performing Transparent Clock operation
>>=20
>>    o  Capable of performing Boundary Clock operation
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 32]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> 2.  Signaling Extensions for Creating Timing LSPs
>>=20
>>    RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-TE
>>    is used to setup Timing LSPs, some information that indicates that
>>    the LSP is carrying Timing flows MUST be included in the new
>>    Extensions to RSVP-TE:
>>=20
>>    The following information MAY also be included in the new Extensions
>>    to RSVP-TE:
>>=20
>>    o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
>>       field
>>=20
>>    o  Number of VLANs in case of PW encapsulation
>>=20
>>    o  Timestamp field Type
>>=20
>>       *  Correction Field, Timestamp
>>=20
>>    o  Timestamp Field format
>>=20
>>       *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
>>          NTP, etc.
>>=20
>>    Note that in case the above optional information is signaled with
>>    RSVP-TE for a Timing LSP, all the Timing packets carried in that LSP
>>    must have the same signaled characteristics.  For example if
>>    Timestamp format is signaled as 64-bit PTPv1, then all Timing packets
>>    must use 64-bit PTPv1 time-stamp.
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 33]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>> Authors' Addresses
>>=20
>>    Shahram Davari
>>    Broadcom Corp.
>>    San Jose, CA  95134
>>    USA
>>=20
>>    Email: davari@broadcom.com
>>=20
>>=20
>>    Amit Oren
>>    Broadcom Corp.
>>    San Jose, CA  95134
>>    USA
>>=20
>>    Email: amito@broadcom.com
>>=20
>>=20
>>    Manav Bhatia
>>    Alcatel-Lucent
>>    Bangalore,
>>    India
>>=20
>>    Email: manav.bhatia@alcatel-lucent.com
>>=20
>>=20
>>    Peter Roberts
>>    Alcatel-Lucent
>>    Kanata,
>>    Canada
>>=20
>>    Email: peter.roberts@alcatel-lucent.com
>>=20
>>=20
>>    Laurent Montini
>>    Cisco Systems
>>    San Jose CA
>>    USA
>>=20
>>    Email: lmontini@cisco.com
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 34]
>=20
>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>=20
>>=20
>>    Luca
>>    Cisco Systems
>>    San Jose CA
>>    USA
>>=20
>>    Email: lmartini@cisco.com
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>>=20
>> Davari, et al.          Expires December 17, 2013              [Page 35]
>=20
>>=20
>> --
>> For corporate legal information go to:
>>=20
>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>=20
>> _______________________________________________
>> TICTOC mailing list
>> TICTOC@ietf.org
>> https://www.ietf.org/mailman/listinfo/tictoc
>=20
>=20
> This e-mail message is intended for the recipient only and contains inform=
ation which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If y=
ou have received this transmission in error, please inform us by e-mail, pho=
ne or fax, and then delete the original and all copies thereof.
>=20
> _______________________________________________
> mpls mailing list
> mpls@ietf.org
> https://www.ietf.org/mailman/listinfo/mpls

From Alexander.Vainshtein@ecitele.com  Sun Aug  4 02:47:38 2013
Return-Path: <Alexander.Vainshtein@ecitele.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C8C4E21F8488 for <tictoc@ietfa.amsl.com>; Sun,  4 Aug 2013 02:47:38 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -4.702
X-Spam-Level: 
X-Spam-Status: No, score=-4.702 tagged_above=-999 required=5 tests=[AWL=0.500,  BAYES_00=-2.599, MIME_QP_LONG_LINE=1.396, RCVD_IN_DNSWL_MED=-4,  UNPARSEABLE_RELAY=0.001]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 7+7lryavB6Hb for <tictoc@ietfa.amsl.com>; Sun,  4 Aug 2013 02:47:34 -0700 (PDT)
Received: from mail1.bemta3.messagelabs.com (mail1.bemta3.messagelabs.com [195.245.230.163]) by ietfa.amsl.com (Postfix) with ESMTP id BE4B221F999E for <tictoc@ietf.org>; Sun,  4 Aug 2013 02:47:32 -0700 (PDT)
Received: from [85.158.138.51:34536] by server-3.bemta-3.messagelabs.com id AF/31-08372-4332EF15; Sun, 04 Aug 2013 09:47:32 +0000
X-Env-Sender: Alexander.Vainshtein@ecitele.com
X-Msg-Ref: server-2.tower-174.messagelabs.com!1375609648!30031696!2
X-Originating-IP: [147.234.242.234]
X-StarScan-Received: 
X-StarScan-Version: 6.9.11; banners=-,-,-
X-VirusChecked: Checked
Received: (qmail 21221 invoked from network); 4 Aug 2013 09:47:30 -0000
Received: from ilptbmg01-out.ecitele.com (HELO ilptbmg01-out.ecitele.com) (147.234.242.234) by server-2.tower-174.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 4 Aug 2013 09:47:30 -0000
X-AuditID: 93eaf2e7-b7f4d6d000005d50-d5-51fe232f39bd
Received: from ILPTWPVEXCA02.ecitele.com ( [172.31.244.232]) by ilptbmg01-out.ecitele.com (Symantec Messaging Gateway) with SMTP id 86.89.23888.F232EF15; Sun,  4 Aug 2013 12:47:28 +0300 (IDT)
Received: from ILPTWPVEXMB02.ecitele.com ([fe80::5979:ca8d:419f:56df]) by ILPTWPVEXCA02.ecitele.com ([fe80::c473:490d:3a7e:e34a%12]) with mapi id 14.03.0123.003; Sun, 4 Aug 2013 12:47:27 +0300
From: Alexander Vainshtein <Alexander.Vainshtein@ecitele.com>
To: "S. Davari" <davarish@yahoo.com>
Thread-Topic: [mpls] [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
Thread-Index: AQHOj2AMghL6qRFw2EmhP8ypbGfBrJmEtKhw///niYCAADPcwA==
Date: Sun, 4 Aug 2013 09:47:26 +0000
Message-ID: <F9336571731ADE42A5397FC831CEAA0215125037@ILPTWPVEXMB02.ecitele.com>
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com>
In-Reply-To: <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-originating-ip: [10.4.35.10]
Content-Type: text/plain; charset="us-ascii"
content-transfer-encoding: quoted-printable
MIME-Version: 1.0
X-CFilter-Loop: Reflected
X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFuphleLIzCtJLcpLzFFi42KZ/OrTF10D5X+BBh8va1mcezqH0eJvcw+7 A5PHlN8bWT2WLPnJFMAU1cBok5iXl1+SWJKqkJJanGyrFFCUWZaYXKmkkJliq2SopFCQk5ic mpuaV2KrlFhQkJqXomTHpYABbIDKMvMUUvOS81My89JtlTyD/XUtLEwtdQ2V7NSUDY2tuUIy MosVUnVzEzNzFHJTi4sT01MVgCIJW5gz+p+tZi84tYal4tCsq8wNjK2NzF2MnBwSAiYSD6ct gLLFJC7cW8/WxcjFISRwkFFib99ZRgjnCKPEm8MvwKrYBGwlNq2+ywZiiwioSCzdvpQFxGYW KJU4f2czI4gtLOAu0f9iC1SNh0TvrnuMELaTxJvbn1lBbBag3t3f/oLZvAIBEj/blrJCLJvH KNHQfxqsgRNo2ae+d2ALGIHO+35qDRPEMnGJW0/mM0GcLSCxZM95qBdEJV4+/scKYctJPHly Cuo4HYkFuz+xQdjaEssWvmaGWCwocXLmExaIekmJgytusExgFJ+FZMUsJO2zkLTPQtK+gJFl FaNoZk5BSVJuuoGhXmpyZklqTqpecn7uJkZIOnm+g/HXfJVDjK5Aj09kluJOzgemo7ySeGMD A9wcJXHe5Q3h/kIC6cC0k52aWpBaFF9UmpNafIiRiYNTqoHxWqvld9WK00yHUls2rtuZLjtn x1HGw745fEs3Oh3v0LnL7Ffxh4tvY+u0PtV2pmlO9xTqLn1MPV9/NKGseF0M69wXNw03r7aT zl0vvveJT//+TRb7Mv/JnAn+tGR+6d6YFacn/Mg0Yq+2f9py58MKBiWNx9y5+y5vUEs8dKby OO8bz7V5Hc4NSizFGYmGWsxFxYkA4e+vkR8DAAA=
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls]  Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 04 Aug 2013 09:47:38 -0000

Shahram hi,
I think that I have understood the objectives of the draft.
I have been just trying to say that, IMHO and FWIW, these objectives CANNOT=
 BE MET within the scope of the MPLS data plane architecture.

Regards,
     Sasha

> -----Original Message-----
> From: S. Davari [mailto:davarish@yahoo.com]
> Sent: Sunday, August 04, 2013 12:40 PM
> To: Alexander Vainshtein
> Cc: stbryant@cisco.com; mpls@ietf.org; tictoc@ietf.org
> Subject: Re: [mpls] [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
> 
> Hi Sasha
> 
> Perhaps you have not understood the draft well. The main reason that the d=
raft
> chose to use a dedicated LSP that is signaled via RSVPTE indicating it is=
 a
> timing LSP is to be backward compatible with non-1588 nodes. Such nodes
> simply just switch the packet.
> 
> Your proposal had been considered and was rejected since it was not backwa=
rd
> compatible. Nodes receiving alert label or TTL =3D 1 send the packet to CP=
U. You
> can refer to the meeting notes.
> 
> Regards,
> Shahram
> 
> 
> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein
> <Alexander.Vainshtein@ecitele.com> wrote:
> 
> > Stewart and all,
> > I concur with Stewart's statement that the draft does not define "full
> interaction with MPLS architecture".
> >
> > E.g., one of the objectives of the draft is to provide a technique that=
 would be
> backward-compatible with old LSRs that cannot provide on-path support for
> timing distribution, while the other objective is to make every LSR on the=
 path
> aware that some MPLS packets are carrying timing-related messages and henc=
e
> require on-path support.
> >
> > IMHO and FWIW, the MPLS data plane architecture allows just two methods
> for making a transit LSR to provide special processing to a labeled packet=
:
> > - It would carry some kind of an "alert label" on top of the label stack=
 and,
> specifically, on top of any labels used for actual forwarding
> > OR,
> > - The TTL in the top label stack entry has been set to 1.
> >
> >
> > The draft does not follow any of these approaches.
> >
> > I must also admit that Section 12 "OAM, Control and Management" of the
> draft looks somewhat in
> > E.g., I could not understand from the text whether BFD, LSP-Ping etc. OA=
M
> messages would be subjected to on-path support procedures for timing
> messages or not.
> >
> > My 2c,
> >     Sasha
> >
> >> -----Original Message-----
> >> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behal=
f
> Of
> >> Stewart Bryant
> >> Sent: Friday, August 02, 2013 12:01 PM
> >> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
> >> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.org;=
 draft-
> >> ietf-tictoc-1588overmpls@tools.ietf.org
> >> Cc: mpls@ietf.org; tictoc@ietf.org
> >> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
> >>
> >> SB> This draft does not seem to provide a precise definition
> >> SB> the properties of the new LSP type that it wishes to
> >> SB> define, in particular it does it define the PHB of those
> >> SB> LSPs, nor the full interaction with the MPLS
> >> SB> architecture.
> >> SB>
> >> SB> I have not tracked TICTOC for a while but I thought that
> >> SB> the original plan was to define the concept of an offset
> >> SB> into a packet to do the correction.
> >> SB>
> >> SB> It is disappointing that the opportunity was not taken
> >> SB> to define a timing shim inside the timing LSP so that
> >> SB> a time correction could be added to any packet such that
> >> SB> the MPLS system was isolated from the details of the
> >> SB> complexity of the particular time transfer type.
> >>
> >> SB> I think that much more clarify is needed in terms of
> >> SB> definition of the new LSP type, since it is unclear
> >> SB> from this text how to implement one.
> >> SB>
> >> SB> There are a lot of other MPLS services such as
> >> SB> LSP ping that need to be considered.
> >> SB>
> >> SB> Please see inline for more comments. However these
> >> SB> comments are made in the context of the text as written
> >> SB> whilst I have a fundamental concern that this approach
> >> SB> lacks an MPLS architectural soundness that need
> >> SB> greater thought with significant impact on the
> >> SB> draft.
> >>
> >> - Stewart
> >>
> >>
> >> TICTOC Working Group                                           S. Davar=
i
> >> Internet-Draft                                                   A. Ore=
n
> >> Intended status: Standards Track                          Broadcom Corp=
.
> >> Expires: December 17, 2013                                     M. Bhati=
a
> >>                                                               P. Robert=
s
> >>                                                           Alcatel-Lucen=
t
> >>                                                               L. Montin=
i
> >>                                                               L. Martin=
i
> >>                                                            Cisco System=
s
> >>                                                            June 15, 201=
3
> >>
> >>
> >>             Transporting Timing messages over MPLS Networks
> >>                    draft-ietf-tictoc-1588overmpls-05
> >>
> >> Abstract
> >>
> >>    This document defines the method for transporting Timing messages
> >>    such as PTP and NTP over an MPLS network.  The method allows for the
> >>    easy identification of these PDUs at the port level to allow for por=
t
> >>
> >> SB> What is a port
> >>
> >>    level processing of these PDUs in both LERs and LSRs.
> >>
> >>    The basic idea is to transport Timing messages inside dedicated MPLS
> >>    LSPs.  These LSPs only carry Timing messages and possibly Control an=
d
> >>    Management packets, but they do not carry customer traffic.
> >>
> >> SB> More specifically they only carry traffic associated with the
> >> SB> timing service and its support.
> >> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
> >> SB> also it gets carried in a structure that causes it to get
> >> SB> timestamped.
> >>
> >>    Two methods for transporting Timing messages over MPLS are defined.
> >>
> >> SB> Perhaps the right approach is to define the new LSP type and then
> >> SB> seperately to define  the mapping of the various timing services
> >> SB> over that LSP type.
> >>
> >>    The first method is to transport Timing messages directly over the
> >>    dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
> >>    MPLS networks.  The second method is to transport Timing messages
> >>    inside a PW via Ethernet encapsulation.
> >>
> >> SB> I think that we should note that there are some
> >> SB> h/w reasons for this preference. A clean sheet approach
> >> SB> would have been to use PTP over MPLS with no intermediate
> >> SB> layers.
> >>
> >> Status of this Memo
> >>
> >>    This Internet-Draft is submitted in full conformance with the
> >>    provisions of BCP 78 and BCP 79.
> >>
> >>    Internet-Drafts are working documents of the Internet Engineering
> >>    Task Force (IETF).  Note that other groups may also distribute
> >>    working documents as Internet-Drafts.  The list of current Internet-
> >>    Drafts is at http://datatracker.ietf.org/drafts/current/.
> >>
> >>    Internet-Drafts are draft documents valid for a maximum of six month=
s
> >>    and may be updated, replaced, or obsoleted by other documents at any
> >>    time.  It is inappropriate to use Internet-Drafts as reference
> >>    material or to cite them other than as "work in progress."
> >>
> >>    This Internet-Draft will expire on December 17, 2013.
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013               [Page 1=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> Copyright Notice
> >>
> >>    Copyright (c) 2013 IETF Trust and the persons identified as the
> >>    document authors.  All rights reserved.
> >>
> >>    This document is subject to BCP 78 and the IETF Trust's Legal
> >>    Provisions Relating to IETF Documents
> >>    (http://trustee.ietf.org/license-info) in effect on the date of
> >>    publication of this document.  Please review these documents
> >>    carefully, as they describe your rights and restrictions with respec=
t
> >>    to this document.  Code Components extracted from this document must
> >>    include Simplified BSD License text as described in Section 4.e of
> >>    the Trust Legal Provisions and are provided without warranty as
> >>    described in the Simplified BSD License.
> >>
> >>
> >> Table of Contents
> >>
> >>    1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . . =
 5
> >>
> >>    2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . . =
 7
> >>
> >>    3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . . =
 8
> >>
> >>    4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . . =
 9
> >>
> >>    5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . . 1=
2
> >>
> >>    6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . . 1=
3
> >>      6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . . 1=
3
> >>      6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . . 1=
3
> >>      6.3.  Other Timing Encapsulation methods . . . . . . . . . . . . 1=
4
> >>
> >>    7.  Timing message Processing  . . . . . . . . . . . . . . . . . . 1=
5
> >>
> >>    8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . . 1=
6
> >>
> >>    9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1=
7
> >>
> >>    10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1=
8
> >>
> >>    11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . . 1=
9
> >>
> >>    12. OAM, Control and Management  . . . . . . . . . . . . . . . . . 2=
0
> >>
> >>    13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . . 2=
1
> >>
> >>    14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . . 2=
2
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013               [Page 2=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >>    15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . . 2=
3
> >>      15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . . 2=
3
> >>      15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . . 2=
3
> >>      15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . . 2=
4
> >>
> >>    16. Other considerations . . . . . . . . . . . . . . . . . . . . . 2=
5
> >>
> >>    17. Security Considerations  . . . . . . . . . . . . . . . . . . . 2=
6
> >>
> >>    18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 2=
7
> >>
> >>    19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 2=
8
> >>
> >>    20. References . . . . . . . . . . . . . . . . . . . . . . . . . . 2=
9
> >>      20.1. Normative References . . . . . . . . . . . . . . . . . . . 2=
9
> >>      20.2. Informative References . . . . . . . . . . . . . . . . . . 2=
9
> >>
> >>    Appendix 1.  Routing extensions for Timing-aware Routers . . . . . 3=
2
> >>
> >>    Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . . 3=
3
> >>
> >>    Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 3=
4
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013               [Page 3=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >>    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
> >>    "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in
> >> this
> >>    document are to be interpreted as described in RFC2119 [RFC2119].
> >>
> >>    When used in lower case, these words convey their typical use in
> >>    common language, and are not to be interpreted as described in
> >>    RFC2119 [RFC2119].
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013               [Page 4=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 1.  Introduction
> >>
> >>    The objective of Precision Time Protocol (PTP) and Network Timing
> >>    Protocol (NTP) are to synchronize independent clocks running on
> >>    separate nodes of a distributed system.
> >>
> >>    [IEEE-1588] defines PTP messages for frequency, phase and time
> >>    synchronization.  The PTP messages include PTP PDUs over UDP/IP
> >>    (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F o=
f
> >>    [IEEE-1588]).
> >>
> >> SB> Sure BUT it is acknowledged that IEEE is open to the definition
> >> SB> of other PTP mappings if they provide better optimisation.
> >>
> >>    This document defines mapping and transport of the PTP
> >>    messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
> >>    defines several clock types: ordinary clocks, boundary clocks, end-
> >>    to-end transparent clocks, and peer-to-peer transparent clocks.
> >>    Transparent clocks require intermediate nodes to update correction
> >>    field inside PTP message that reflects the transit time in the node.
> >>
> >>    [RFC5905] defines NTP messages for clock and time synchronization.
> >>    The PTP messages (PDUs) are transported over UDP/IP.  This document
> >> SB> Should that be NTP messages?
> >> SB> It needs to be made clear as soon as you introduce NTP that
> >> SB> they use different time representations.
> >>
> >>    defines mapping and transport of the NTP messages defined in
> >>    [RFC5905] over MPLS networks.
> >>
> >>    One key attribute of all of these Timing messages is that the Time
> >>    stamp processing should occur as close as possible to the actual
> >>    transmission and reception at the physical port interface.  This
> >>    targets optimal time and/or frequency recovery by avoiding variable
> >>    delay introduced by queues internal to the clocks.
> >>
> >> SB> As I recall NTP has no epoch point defined, and I am not sure
> >> SB> where that point is in the case of PTP in this mapping
> >> SB> Hopefully this will get defined in due course.
> >>
> >>    To facilitate the fast and efficient recognition of Timing messages
> >>    at the port level when the Timing messages are carried over MPLS
> >>    LSPs,
> >>
> >> SB> Over a new LSP type with time optimied characteristics
> >>
> >>    this document defines the specific encapsulations that should
> >>    be used.
> >> SB> Hopefully it will also define the PHP
> >>
> >>    In addition, it can be expected that there will exist LSR/
> >>    LERs where only a subset of the physical ports will have the port-
> >>    based Timing message processing capabilities.
> >> SB> Do you need to clarify that this only works at base and not in
> >> SB> a label heirarchy.
> >>
> >>
> >>    In order to ensure
> >>    that the LSPs carrying Timing packets always enter and exit ports
> >>    with this capability, routing extensions are defined to advertise
> >>    this capability on a port basis and to allow for the establishment o=
f
> >>    LSPs that only transit such ports.  While this path establishment
> >>    restriction may be applied only at the LER Ingress and/or egress
> >>    ports, it becomes more important when using transparent clock capabl=
e
> >>    LSRs in the path.
> >> SB> I do not understand the implications of the last
> >> SB> sentences - starting ", it becomes"
> >>
> >>
> >>    Port based Timing message processing involves Timing message
> >>    recognition.  Once the Timing messages are recognized they can be
> >>    modified based on the reception or transmission Time-stamp.
> >>
> >>    This document provides two methods for transporting Timing messages
> >>    over MPLS.  One is applicable to MPLS environment and the other one
> >>    is applicable to MPLS/MPLS-TP environment
> >>
> >> SB> I think the sentence is incomplete.
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013               [Page 5=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >>    The solution involves transporting Timing messages over dedicated
> >>    LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
> >>    carry Management and control messages, but not data plane client
> >>    traffic.
> >>
> >> SB> It is not clear why this restriction applies.
> >>
> >>    Timing LSPs can be established statically or via signaling.
> >> SB> s/statically/by provisioning/network management/
> >>
> >>    Extensions to control plane (OSPF, ISIS, etc.) is required to enable
> >>    routers to distribute their Timing processing capabilities over MPLS
> >>    to other routers.  However such extensions are outside the scope of
> >>    this document.
> >>
> >>    When signaling is used to setup the PTP LSP, Extensions to signaling
> >> SB> is it a PTP LSP or a Timing LSP?
> >>
> >>    protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
> >>    However such extensions are outside the scope of this document.
> >>
> >> SB> for mpls-tp GMPLS is the signalling protocol
> >>
> >>    While the techniques included herein allow for the establishment of
> >>    paths optimized to include Time-stamping capable links, the
> >>    performance of the Slave clocks is outside the scope of this
> >>    document.
> >>
> >>    At the time of publishing this specification, Transparent Clocking
> >>    (TC) is only defined for PTP.  Therefore at this time any part of
> >>    this specification that talks about Transparent Clocking applies onl=
y
> >>    to PTP.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013               [Page 6=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 2.  Terminology
> >>
> >>    1588: The timing and synchronization as defined by IEEE 1588.
> >>
> >> SB> I think that there is a more formal name for the 1588 group
> >> SB> that needs to be used here.
> >> SB> Also do we need to talk about 1588-200? as there is
> >> SB> an update in progress
> >>
> >>    NTP: The timing and synchronization protocol defined by IETF RFC-130=
5
> >>    and RFC-5905.
> >>
> >>    PTP: The timing and synchronization protocol used by 1588.
> >> SB> need the proper name for 1588
> >>
> >>    Master Clock: The source of 1588 timing to a set of slave clocks.
> >>
> >>    Master Port: A port on a ordinary or boundary clock that is in Maste=
r
> >>    state.  This is the source of timing toward slave ports.
> >>
> >> SB> I am not sure the reader knows what a port is
> >>
> >>    Slave Clock: A receiver of 1588 timing from a master clock.
> >>
> >>    Slave Port: A port on a boundary clock or ordinary clock that is
> >>    receiving timing from a master clock.
> >>
> >>    Ordinary Clock: A device with a single PTP port.
> >>
> >>    Transparent Clock.  A device that measures the time taken for a PTP
> >>    event message to transit the device and then updates the
> >>    correctionField of the message with this transit time.
> >>
> >>    Boundary Clock: A device with more than one PTP port.  Generally
> >>    boundary clocks will have one port in slave state to receive timing
> >>    and then other ports in master state to re-distribute the timing.
> >>
> >>    PTP LSP: An LSP dedicated to carry PTP messages
> >>
> >> SB> PTP or timing?
> >>
> >>
> >>    PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
> >>    messages.
> >>
> >> SB> Ah I don't think that PWE3 know what one of these is
> >>
> >>    CW: Pseudowire Control Word
> >>
> >>    LAG: Link Aggregation
> >>
> >>    ECMP: Equal Cost Multipath
> >>
> >>    CF: Correction Field, a field inside certain PTP messages (message
> >>    type 0-3)that holds the accumulative transit time inside intermediat=
e
> >>    switches
> >>
> >>    Timing messages: Timing Protocol messages that are exchanged between
> >>    routers in order to establish a synchronized clock.
> >>
> >> SB> A number of these definitions look like copies of IEEE1588
> >> SB> definitions. We need to provide references and note the
> >> SB> priority of the IEEE base reference.
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013               [Page 7=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 3.  Problem Statement
> >>
> >>    [IEEE-1588] has defined methods for transporting PTP messages over
> >>    Ethernet and IP networks.  [RFC5905] has defined the method of
> >>    transporting NTP messages over IP networks.  There is a need to
> >>    transport Timing messages over MPLS networks while supporting the
> >>    Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
> >>    functionality in the LER and LSRs in the MPLS network.
> >>
> >>    There are multiple ways of transporting Timing over MPLS.  However,
> >>    there is a requirement to limit the possible encapsulation options t=
o
> >>    simplify the Timing message identification and processing required a=
t
> >>    the port level.
> >>
> >>    When Timing-awareness is needed, Timing messages should not be
> >>    transported over LSPs or PWs that are carrying customer traffic
> >>    because LSRs perform Label switching based on the top label in the
> >>    stack.
> >>
> >> SB> Have you explained why?
> >>
> >>    To detect Timing messages inside such LSPs require special
> >>    hardware to do deep packet inspection at line rate.  Even if such
> >>    hardware exists, the payload can't be deterministically identified b=
y
> >>    LSRs because the payload type is a context of the PW label, and the
> >>    PW label and its context are only known to the Edge routers (PEs/
> >>    LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CES,
> >>    etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
> >>    LSRs dont have the knowledge of whether PW Control Word (CW) is
> >>    present or not and therefore can not deterministically identify the
> >>    payload.
> >>
> >>    A generic method is defined in this document that does not require
> >>    deep packet inspection at line rate, and can deterministically
> >>    identify Timing messages.  This method can be used to detect Timing
> >>    Messages in both one-step and two-step clock implementations of
> >>    ordinary, boundary and transparent clocks.
> >>
> >> SB> Needs a ref and I am sure many MPLS specialists will not understand
> >> SB> the msg types.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013               [Page 8=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 4.  Timing over MPLS Architecture
> >>
> >>    Timing messages are exchange between Timing ports on ordinary and
> >>
> >> SB> Have you defined a timing port?
> >>
> >>    boundary clocks.  Boundary clocks terminate the Timing messages and
> >>    act as master for other boundary clocks or for slave clocks.  End-to=
-
> >>    End Transparent clocks do not terminate the Timing messages but they
> >>    do modify the contents of the Timing messages as they transit across
> >>    the transparent clock.
> >>
> >>    Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Cloc=
k
> >>
> >>    (TC) could be implemented in either LERs or LSRs.
> >>
> >> SB> LER and LSR need to be expanded
> >>
> >>    An example is shown in Figure 1, where the LERs act as Ordinary Cloc=
k
> >>    (OC) and are the initiating/terminating point for Timing messages.
> >>    The ingress LER encapsulates the Timing messages in Timing LSP and
> >>    the Egress LER terminates the Timing LSP.  The LSRs act as
> >>    Transparent Clock (TC) and just update the Timing field in the Timin=
g
> >>    messages.
> >>
> >>
> >>       +--------+     +-------+     +-------+     +-------+     +-------=
-+
> >>       |Switch, |     |       |     |       |     |       |     |Switch,=
 |
> >>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router=
 |
> >>       |        |     |  OC   |     |  TC   |     |  OC   |     |      =
  |
> >>       +--------+     +-------+     +-------+     +-------+     +-------=
-+
> >>                      /                                 \
> >>       +-------+     /                                   \     +-------+
> >>       |  LER  |    /                                     \    |  LER  |
> >>       | Master|---/                                       \---| Slave |
> >>       | Clock |                                               | Clock |
> >>       +-------+                                               +-------+
> >>
> >>      Figure (1) - Deployment example 1 of timing over MPLS network
> >>
> >>    Another example is shown in Figure2, where LERs terminate the Timing
> >>    messages received from switch/routers that are outside of the MPLS
> >>    network acting as OC or BC.  In this example LERs regenerate the
> >>    clock and initiate timing messages encapsulated in Timing LSP toward
> >>    the MPLS network, while the LSRs act as Transparent Clock (TC) and
> >>    just update the Timing field in the Timing messages, which are
> >>    already encapsulated in Timing LSPs.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013               [Page 9=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >>      +--------+     +-------+     +-------+     +-------+     +--------=
+
> >>      |Switch, |     |       |     |       |     |       |     |Switch,=
 |
> >>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router=
 |
> >>      | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC =
 |
> >>      +--------+     +-------+     +-------+     +-------+     +--------=
+
> >>
> >>      Figure (2) - Deployment example 2 of timing over MPLS network
> >>
> >>
> >>    Another example is shown in Figure 3, where LERs do not terminate th=
e
> >>    Timing messages received from switch/routers that are outside of the
> >>    MPLS network acting as OC, TC or BC.  The LERs act as TC and update
> >>    the Timing field in the Timing messages as they transit the LER,
> >>    while encapsulating them in timing LSP.  The LSRs also act as
> >>    Transparent Clock (TC) and just update the Timing field in the Timin=
g
> >>    messages which are already encapsulated in Timing LSPs.
> >>
> >>       +--------+     +-------+     +-------+     +-------+     +-------=
-+
> >>       |Switch, |     |       |     |       |     |       |     |Switch,=
 |
> >>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router=
 |
> >>       |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC/B=
C|
> >>       +--------+     +-------+     +-------+     +-------+     +-------=
-+
> >>
> >>     Figure (3) - Deployment example 3 of timing over MPLS network
> >>
> >>    Another example is shown in Figure 4, where LERs and LSRs support
> >>    Boundary Clocks.  A single-hop LSP is created between two adjacent
> >>    LSRs engaged in BC operation.  Other methods such as PTP transport
> >>    over Ethernet MAY be used for transporting timing messages if the
> >>    link between the two routers is Ethernet.
> >>
> >>      +--------+     +-------+     +-------+     +-------+     +--------=
+
> >>      |Switch, |     |       |     |       |     |       |     |Switch,=
 |
> >>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router=
 |
> >>      | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC =
 |
> >>      +--------+     +-------+     +-------+     +-------+     +--------=
+
> >>
> >>    Figure (4) - Deployment example 3 of timing over MPLS network
> >>
> >>    An MPLS domain MAY serve multiple customers.  In these cases the MPL=
S
> >>    domain (maintained by a service provider) may provide timing service=
s
> >>    to multiple customers, each having their own Timing domain.
> >>
> >>    The Timing over MPLS architecture assumes full mesh of Timing LSPs
> >>    between all LERs supporting this specification.
> >>
> >> SB> Note sure this is right - the salves surely do not need to
> >> SB> exchange timing amongst themselves
> >>
> >>    It supports
> >>    Point-to- point (VPWS) and Multipoint (VPLS) services.
> >>
> >> SB> What does that mean? You do not carry user data traffic?
> >> SB> Maybe it's the ordering of the statemnets that is causing
> >> SB> confusion.
> >>
> >>    This means
> >>    that a customer may purchase a Point-to-point Timing service between
> >>    two customer sites or a Multipoint Timing service between more than
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 10=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >>    two customer sites.
> >>
> >>    The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
> >>    This means that the Timing Multicast messages such as PTP Multicast
> >>    event messages can be transported over P2MP Timing LSP or be
> >>    replicated and transported over many P2P Timing LSPs.
> >>
> >> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
> >> SB> nor a P2MP PW, although we are close.
> >>
> >>    Timing messages, that do not require Time stamping or Correction
> >>    Field update MAY be transported over Timing LSPs to simplify hardwar=
e
> >>    and software.
> >>
> >>    PTP Announce messages that determine the Timing LSP terminating poin=
t
> >>    behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
> >>    to simplify hardware and software.
> >>
> >> SB> have you defined and referenced PTP announce msgs?
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 11=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 5.  Dedicated LSPs for Timing messages
> >>
> >>    Many methods have been considered for identifying the Timing message=
s
> >>    when they are encapsulated in MPLS such as using GAL/G-ACH or a new
> >>    reserved label.  These methods were not attractive since they either
> >>    required deep packet inspection at line rate in the intermediate LSR=
s
> >>    or they required use of a scarce new reserved label.  Also one of th=
e
> >>    goals was to reuse existing OAM mechanisms.
> >>
> >> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
> >>
> >>    The method defined in this document can be used by LER and LSRs to
> >>    identify Timing messages in MPLS tunnels by just looking at the top
> >>    label in the MPLS label stack, which only carry Timing messages as
> >>    well as OAM, but not data plane client traffic.
> >>
> >>    Compliant implementations MUST use dedicated LSPs to carry Timing
> >>    messages over MPLS.
> >>
> >> SB> I think that we need a definition of the properies of these LSPs
> >>
> >>    These LSPs are herein referred to as "Timing
> >>    LSPs" and the labels associated with these LSPs as "Timing LSP
> >>    labels".  The Timing LSPs that runs between Ingress and Egress LERs
> >>    MUST be co-routed.  Alternatively, a single bidirectional co-routed
> >>    LSP can be used.
> >>
> >> SB> I though that you said you could use M2MP LSPs - these are not
> >> SB> bidirectional.
> >>
> >>    Co-routing of the two directions is required to limit the difference
> >>    in the delays in the Master clock to Slave clock direction compared
> >>    to the Slave clock to Master clock direction.  The Timing LSP MAY be
> >>    MPLS/MPLS-TP LSP.
> >>
> >>    The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
> >>    New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
> >>    outside the scope of this document.
> >>
> >>    The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such as
> >>    BFD and LSP Ping but the LSP data plane client plane traffic MUST be
> >>    Timing packets only.
> >>
> >> SB> Why?
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 12=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 6.  Timing over LSP Encapsulation
> >>
> >> The encapsulations is not LSP is it?
> >>
> >>    This document defines two methods for carrying Timing messages over
> >>    MPLS.  The first method is carrying UDP/IP encapsulated Timing
> >>    messages over Timing LSPs, and the second method, is carrying
> >>    Ethernet encapsulated Timing messages over Ethernet PWs inside Timin=
g
> >>    LSPs.
> >>
> >> 6.1.  Timing over UDP/IP over MPLS Encapsulation
> >>
> >>    The simplest method of transporting Timing messages over MPLS is to
> >>    encapsulate Timing PDUs in UDP/IP and then encapsulate them in Timin=
g
> >>    LSP.  This format is shown in Figure 4.
> >>
> >>
> >>                     +----------------------+
> >>                     |   Timing LSP Label   |
> >>                     +----------------------+
> >>                     |        IPv4/6        |
> >>                     +----------------------+
> >>                     |         UDP          |
> >>                     +----------------------+
> >>                     |     Timing PDU       |
> >>                     +----------------------+
> >>
> >>       Figure (4) - Timing over UDP/IP over MPLS Encapsulation
> >>
> >>
> >>    This encapsulation is very simple and is useful when the network
> >>    between Timing Master Clock and Slave Clock is MPLS network.
> >>
> >> SB> Simple is a judgement call
> >>
> >>    In order for an LER/LSR to process Timing messages, the Timing LSP
> >>    Label must be at the top label of the label stack.  The LER/LSR MUST
> >>    know that the Timing LSP Label is used for carrying Timing messages.
> >>    This can be accomplished via static configuration or via RSVP-TE
> >>    signaling.
> >>
> >>    The UDP/IP encapsulation of PTP MUST follow Annex D and E of
> >>    [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
> >>    [RFC5905].
> >>
> >> 6.2.  Timing over PW Encapsulation
> >>
> >>    Another method of transporting Timing over MPLS networks is by
> >>    encapsulating Timing PDUs in PW which in turn is transported over
> >>    Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
> >>    shown in Fig 5(A) MUST be used and the Ethernet encapsulation of PTP
> >>    MUST follow Annex F of [IEEE-1588].
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 13=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >>    The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
> the
> >>    Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
> >>    Timing over PW encapsulation MUST use the Control Word (CW) as
> >>    specified in [RFC4448] to ensure proper detection of PTP messages
> >>    inside the MPLS packets for Timing over LSP and Timing over PW
> >>    encapsulation.
> >>
> >> SB> That needs explanation
> >>
> >>    The use of Sequence Number in the CW is optional.
> >>
> >> SB> Given that s/n are never in practice deployed, you could probably
> >> SB> simplify things by sayig that they are not used.
> >>
> >>    Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over
> PW
> >>    (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
> >>
> >>                     +----------------+  +----------------+
> >>                      |Timing LSP Label|  |Timing LSP Label|
> >>                      +----------------+  +----------------+
> >>                      |    PW Label    |  |    PW Label    |
> >>                      +----------------+  +----------------+
> >>                      |  Control Word  |  |      IP        |
> >>                      +----------------+  +----------------+
> >>                      |    Ethernet    |  |      UDP       |
> >>                      |     Header     |  +----------------+
> >>                      +----------------+  |   Timing PDU   |
> >>                      |S-VLAN(Optional)|  |                |
> >>                      +----------------+  +----------------+
> >>                      |C-VLAN(Optional)|        (B)
> >>                      +----------------+
> >>                      |   Timing PDU   |
> >>                      |                |
> >>                      +----------------+
> >>                             (A)
> >>
> >>               Figure (5) - Timing over PW Encapsulations
> >>
> >>    In order for an LSR to process PTP messages, the top label of the
> >>    label stack (the Tunnel Label) MUST be a Timing label.
> >>
> >> S> You said that before.
> >>
> >> 6.3.  Other Timing Encapsulation methods
> >>
> >>    In future other timing encapsulation methods may be introduced, such
> >>    as a new shim header after the Bottom of Stack to carry the Timing
> >>    information.  Such new encapsulations are outside the scope of this
> >>    document.
> >>
> >>
> >> SB> Taking a pure MPLS pov, you can simplify a lot of the text
> >> SB> out of the definition of the LSP
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 14=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> SB> I think we need a section on LSP processing
> >>
> >> 7.  Timing message Processing
> >>
> >>    Each Timing protocol such as PTP and NTP, define their set of Timing
> >>    messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
> >>    FOLLOW_UP, etc messages.
> >>
> >>    Some of the Timing messages require time stamping or correction fiel=
d
> >>    update at port level and some dont.  It is the job of the LER/LSR to
> >>    parse the timing message and find out the type of the Timing message
> >>    and decide whether and how to Time- stamp it (e.g., BC) or update
> >>    correction field(e.g., TC).
> >>
> >>
> >> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
> >> SB> function rather than the LER?
> >>
> >>    For example the following PTP messages (called Event messages)
> >>    require time-stamping or correction field update:
> >>
> >>    o  SYNC
> >>
> >>    o  DELAY_REQ (Delay Request)
> >>
> >>    o  PDELAY_REQ (Peer Delay Request)
> >>
> >>    o  PDELAY_RESP (Peer Delay Response)
> >>
> >>    SYNC and DELAY_REQ are exchanged between Master Clock and Slave
> Clock
> >>    and MUST be transported over PTP LSPs.  PDELAY_REQ and PDELAY_RESP
> >>    are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
> >>    Boundary, or Transparent) and SHOULD be transported over single hop
> >>    PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
> >>    and PDELAY_RESP_FOLLOW_UP messages MUST also be transported over
> >> the
> >>    PTP LSPs.
> >>
> >>    For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
> >>    transported over two PTP LSPs that are in opposite directions.  Thes=
e
> >>    PTP LSPs, which are in opposite directions MUST be congruent and co-
> >>    routed.  Alternatively, a single bidirectional co-routed LSP can be
> >>    used.
> >>
> >>    Except as indicated above for the two-step PTP clocks, Non-Event PTP
> >>    message types do not need to be processed by intermediate routers.
> >>    These message types MAY be carried in PTP Tunnel LSPs.
> >>
> >> SB> Are you saying that a timing P router has to be msg type sensitive?
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 15=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 8.  Protection and Redundancy
> >>
> >>
> >> SB> This is a bit of a jump - I don't know how the LSP itself works yet=
!
> >>
> >>    In order to ensure continuous uninterrupted operation of slave
> >>    clocks, usually as a general practice, slave clocks (or ports) track
> >>    redundant master clocks.
> >>
> >>    It is the responsibility of the network operator to ensure that
> >>    physically disjoint Timing LSPs are established between a slave cloc=
k
> >>    (or port) and redundant master clocks (or ports).
> >>
> >>    When a slave clock (or port) listens to redundant master clocks or
> >>    ports, any prolonged Timing LSP outage will trigger the slave clock
> >>    or port to switch to a redundant master clock or port.
> >>
> >>    LSP/PW protection such as Linear protection Switching (1:1, 1+1),
> >>    Ring protection switching or MPLS Fast Reroute (FRR) generally switc=
h
> >>    alternative path that usually cause a change in delay, which if
> >>    undetected by slave clock can reduce accuracy of the slave clock.
> >>
> >>    Therefore protection switching MAY be used, as long as phase jumps
> >>    upon switchover due to differences in path latency are detected and
> >>    compensated for (such compensation not being required if BCs or peer=
-
> >>    peer TCs are used throughout).
> >>
> >>    Note that any protection or reroute mechanism that adds additional
> >>    MPLS label to the label stack, such as Facility Backup Fast Reroute,
> >>    MUST ensure that the pushed label is also a Timing Label to ensure
> >>    recognition of the MPLS frame as containing Timing messages, as it
> >>    transits the backup path.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 16=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 9.  ECMP
> >>
> >>    To ensure the optimal operation of slave clocks and avoid error
> >>    introduced by forward and reverse path delay asymmetry, the physical
> >>    path for Timing messages from master clock to slave Clock and vice
> >>    versa must be the same for all Event Timing messages listed in
> >>    section 7.
> >>
> >>    Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
> >>    Multipath).
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 17=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 10.  PHP
> >>
> >>    To ensure that the label on the top of the label stack is the Timing
> >>    LSP Label, PHP MUST not be used.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 18=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 11.  Entropy
> >>
> >>    To ensure all Timing messages in a Timing LSP take the same path,
> >>    Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
> >>    Entropy Label MUST NOT be used for the PWs that are carried inside
> >>    Timing LSP [RFC6391].
> >>
> >> SB> This is incorrect - you mean that all msgs of the same timing
> >> SB> flow need to have the same EL value.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 19=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 12.  OAM, Control and Management
> >>
> >>    In order to monitor Timing LSPs and their encapsulated PWs, they MUS=
T
> >>    be able to carry OAM and management messages.  These management
> >>    messages MUST be differentiated from Timing messages via already
> >>    defined IETF methods.
> >>
> >>    For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
> >>    over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
> >>    Management protocols can easily be identified by the UDP Destination
> >>    Port number or by GAL/G-ACH respectively.
> >>
> >>    Also BFD, LSP-Ping and other management messages MAY run over the
> PWs
> >>    encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3 o=
r
> >>    4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is goin=
g
> >>    to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1) o=
r
> >>    GAL-ACH are used to identify such management messages.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 20=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 13.  QoS Considerations
> >>
> >>    In network deployments where not every LSR/LER is Timing-aware, it i=
s
> >>    important to reduce the impact of the non-Timing-aware LSR/LERs on
> >>    the timing recovery in the slave clock.  The Timing messages are tim=
e
> >>    critical and must be treated with the highest priority.  Therefore
> >>    Timing over MPLS messages must be treated with the highest priority
> >>    in the routers.  This can be achieved by proper setup of Timing LSPs=
.
> >>
> >>    It is recommended that the Timing LSPs are setup or configured
> >>    properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC2697]
> >>    for drop eligibility.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 21=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 14.  FCS and Checksum Recalculation
> >>
> >>    When time-stamp generation and timing packet adjustment is performed
> >>    near the physical port hardware, the process MUST include
> >>    recalculation of the Ethernet FCS.
> >>
> >> SB> The above is confusing - an LSR always recomputes the link layer
> >> SB> CRC which may or may not be Ethernet.
> >>
> >>    Also FCS retention for the
> >>    payload Ethernet described in [RFC4720] MUST NOT be used.
> >>
> >>    For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
> >>    may be required as per UDP transport standards.
> >>
> >> SB> You really need to be working on getting the IPv6 C?S computation
> >> SB> removed from PTP msgs.
> >>
> >>    When UDP checksum is used, each Timing-aware LER/LSR must either
> >>    incrementally update the UDP checksum after Time stamping or
> >>    Correction Field update or verify the UDP checksum on reception from
> >>    upstream and recalculate the checksum completely on transmission to
> >>    downstream node after Time stamping or Correction Field update.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 22=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 15.  Behavior of LER/LSR
> >>
> >>    Timing-capable/aware LERs and LSRs are routers that have one or more
> >>
> >> SB> You mean physical interfaces?
> >>
> >>    interfaces that can perform Timing operations (OC/BC/TC) on Timing
> >>    packets and are configured to do so.  Timing-capable/aware LERs and
> >>    LSRs can advertise their Timing-capability per-interface via control
> >>    plane such as OSPF or IS-IS.
> >> SB> ISIS and OSPF are routing protocols.
> >>
> >>   The Timing-capable/aware LERs can then
> >>    signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timing
> >>    capability of LER and LSRs may be configured in a centralized
> >>    controller and the Timing LSP may be setup using manual configuratio=
n
> >>    or other methods such as SDN.
> >>
> >> SB> it can also be configured individually rather then through
> >> SB> a cebtral controllwe
> >>
> >> 15.1.  Behavior of Timing-capable/aware LER
> >>
> >>    When a Timing-capable/aware LER behaves as a Transparent clock and
> >>    receives a Timing message from a Timing-capable/aware non-MPLS
> >>    interface, the LER updates the Correction Field (CF) and encapsulate=
s
> >>    and forwards the timing message over previously established Timing
> >>    LSP.
> >>
> >> SB> You need to call out the details so that people properly
> >> SB> understand the definition of the new LSP.
> >>
> >>    Also when a Timing message is received from a Timing-capable/
> >>    aware MPLS interface, LER updates the Correction Filed (CF) and
> >>    decapsulates the MPLS encapsulation and forwards the timing message
> >>    to a non-MPLS interface.
> >>
> >>    When a Timing-capable/aware LER behaves as a Boundary clock and
> >>    receives a Timing message from a Timing-capable/aware non MPLS
> >>    interface, the LER Timestamps the Timing packet and sends it to the
> >>    LERs Boundary clock processing module.  Also when a Timing message i=
s
> >>    received from a Timing- capable/aware MPLS interface, the LER
> >>    Timestamps the Timing packet and sends it to the LERs Boundary clock
> >>    processing module.
> >>
> >>    When a Timing-capable/aware LER behaves as an Ordinary Clock toward
> >>    the MPLS network, and receives a Timing message from a Timing-
> >>    capable/aware MPLS interface, the LER Timestamps the Timing packet
> >>    and sends it to the LERs Ordinary clock processing module.
> >>
> >> 15.2.  Behavior of Timing-capable/aware LSR
> >>
> >>    When a Timing-capable/aware LSR behaves as a Transparent clock and
> >>    receives a Timing message from a Timing-capable/aware MPLS interface=
,
> >>    The LSR updates the Correction Filed (CF) and forwards the timing
> >>    message over another MPLS interface.
> >>
> >>    When a Timing-capable/aware LSR behaves as a Boundary clock and
> >>    receives a Timing message from a Timing-capable/aware MPLS interface=
.
> >>    The LSR performs the functions of a Boundary Clock in terminating th=
e
> >>    received Timing message and re-generating a new timing message over
> >>    another (or the same) MPLS interface.
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 23=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 15.3.  Behavior of non-Timing-capable/aware LSR
> >>
> >>    It is most beneficial when all LSRs in the path of a Timing LSP be
> >>    timing-Capable/aware LSRs.  This would ensure the highest quality
> >>    time and clock synchronization by Timing Slave Clocks.  However, thi=
s
> >>    specification does not mandate that all LSRs in path of a Timing LSP
> >>    be Timing- capable/aware.
> >>
> >>    Non-Timing-capable/aware LSRs just switch the packets encapsulated i=
n
> >>    Timing LSPs and dont perform any Timing operation (TC or BC).
> >>    However as explained in QoS section the Timing over MPLS packets MUS=
T
> >>    be still be treated with the highest priority based on their Traffic
> >>    Class (TC) marking.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 24=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 16.  Other considerations
> >>
> >>    [IEEE-1588] defines an optional peer-to-peer Transparent clocking
> >>    that requires peer delay measurement between two adjacent Timing-
> >>    capable/ aware routers/switches.  Peer delay measurement messages
> >>    need to be time stamped and terminated by the Timing-capable/aware
> >>    routers/ switches.  This means that two adjacent LSRs may be engaged
> >>    in a peer delay measurement.
> >>
> >>    For transporting such peer delay measurement messages a single-hop
> >>    LSP SHOULD to be created between the two adjacent LSRs engaged in
> >>    peer delay measurement to carry peer delay measurement messages.
> >>    Other methods such as PTP transport over Ethernet MAY be used for
> >>    transporting peer delay measurement messages if the link between the
> >>    two routers is Ethernet.
> >>
> >>    In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ war=
e
> >>    routers/switches MUST maintain a list of all the neighbors it needs
> >>    to send a PDelay_Req to, where each neighbor corresponds to a timing
> >>    LSP.
> >>
> >>    The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as lo=
ng
> >>    as either the Explicit Null label is the bottom of stack label
> >>    (applicable only to UDP/IP encapsulation) or the label below the
> >>    Explicit Null label is a PTP label.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 25=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 17.  Security Considerations
> >>
> >>    MPLS PW security considerations in general are discussed in [RFC3985=
]
> >>    and [RFC4447],and those considerations also apply to this document.
> >>
> >>    An experimental security protocol is defined in [IEEE-1588].The PTP
> >>    security extension and protocol provides group source authentication=
,
> >>    message integrity, and replay attack protection for PTP messages.
> >>
> >>    When the MPLS network (provider network) serves multiple customers,
> >>    it is important to maintain and process each customers clock and
> >>    Timing messages separately from other customers to ensure there is n=
o
> >>    cross- customer effect.  For example if an LER BC is synchronized to
> >>    a specific grandmaster, belonging to customer A, then the LER MUST
> >>    use that BC clock only for customer A to ensure that customer A
> >>    cannot attack other customers by manipulating its time.
> >>
> >>    Timing messages MAY be encrypted or authenticated, provided that the
> >>    LERs/LSRs that are Timing capable/aware can authenticate/ decrypt th=
e
> >>    timing messages.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 26=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 18.  Acknowledgements
> >>
> >>    The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizrahi=
,
> >>    Stefano Ruffini, Peter Meyer, and other members of IETF for reviewin=
g
> >>    and providing feedback on this draft.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 27=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 19.  IANA Considerations
> >>
> >>    There are no IANA requirements in this specification.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 28=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 20.  References
> >>
> >> 20.1.  Normative References
> >>
> >>    [IEEE-1588]
> >>               IEEE 1588-2008, "IEEE Standard for a Precision Clock
> >>               Synchronization Protocol for Networked Measurement and
> >>               Control Systems".
> >>
> >>    [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
> >>               Requirement Levels", BCP 14, RFC 2119, March 1997.
> >>
> >>    [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
> >>               Edge (PWE3) Architecture", RFC 3985, March 2005.
> >>
> >>    [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
> >>               Proxies (ND Proxy)", RFC 4389, April 2006.
> >>
> >>    [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
> >>               Heron, "Pseudowire Setup and Maintenance Using the Label
> >>               Distribution Protocol (LDP)", RFC 4447, April 2006.
> >>
> >>    [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
> >>               "Encapsulation Methods for Transport of Ethernet over MPL=
S
> >>               Networks", RFC 4448, April 2006.
> >>
> >>    [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
> >>               Emulation Edge-to-Edge (PWE3) Frame Check Sequence
> >>               Retention", RFC 4720, November 2006.
> >>
> >>    [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
> >>               Connectivity Verification (VCCV): A Control Channel for
> >>               Pseudowires", RFC 5085, December 2007.
> >>
> >>    [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detection
> >>               (BFD)", RFC 5880, June 2010.
> >>
> >>    [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
> >>               "Bidirectional Forwarding Detection (BFD) for MPLS Label
> >>               Switched Paths (LSPs)", RFC 5884, June 2010.
> >>
> >> 20.2.  Informative References
> >>
> >>    [I-D.ietf-pwe3-fat-pw]
> >>               Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
> >>               J., and S. Amante, "Flow Aware Transport of Pseudowires
> >>               over an MPLS Packet Switched Network",
> >>               draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 29=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >>    [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
> >>               system routeing information exchange protocol for use in
> >>               conjunction with the Protocol for providing the
> >>               Connectionless-mode Network Service (ISO 8473)".
> >>
> >>    [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
> >>               dual environments", RFC 1195, December 1990.
> >>
> >>    [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
> >>
> >>    [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
> >>               Marker", RFC 2697, September 1999.
> >>
> >>    [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boudec,
> >>               J., Courtney, W., Davari, S., Firoiu, V., and D.
> >>               Stiliadis, "An Expedited Forwarding PHB (Per-Hop
> >>               Behavior)", RFC 3246, March 2002.
> >>
> >>    [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineerin=
g
> >>               (TE) Extensions to OSPF Version 2", RFC 3630,
> >>               September 2003.
> >>
> >>    [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
> >>               System (IS-IS) Extensions for Traffic Engineering (TE)",
> >>               RFC 3784, June 2004.
> >>
> >>    [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
> >>               Shaffer, "Extensions to OSPF for Advertising Optional
> >>               Router Capabilities", RFC 4970, July 2007.
> >>
> >>    [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
> >>               System to Intermediate System (IS-IS) Extensions for
> >>               Advertising Router Information", RFC 4971, July 2007.
> >>
> >>    [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
> >>               Topology (MT) Routing in Intermediate System to
> >>               Intermediate Systems (IS-ISs)", RFC 5120, February 2008.
> >>
> >>    [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
> >>               Engineering", RFC 5305, October 2008.
> >>
> >>    [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
> >>               "Traffic Engineering Extensions to OSPF Version 3",
> >>               RFC 5329, September 2008.
> >>
> >>    [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
> >>               for IPv6", RFC 5340, July 2008.
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 30=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >>    [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Networ=
k
> >>               Time Protocol Version 4: Protocol and Algorithms
> >>               Specification", RFC 5905, June 2010.
> >>
> >>    [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
> >>               J., and S. Amante, "Flow-Aware Transport of Pseudowires
> >>               over an MPLS Packet Switched Network", RFC 6391,
> >>               November 2011.
> >>
> >>    [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
> >>               L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
> >>               RFC 6790, November 2012.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 31=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 1.  Routing extensions for Timing-aware Routers
> >>
> >>    MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] and
> >>    IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (TE)
> >>    link information used for constraint-based routing.
> >>
> >>    Indeed, it is useful to advertise data plane TE router link
> >>    capabilities, such as the capability for a router to be Timing-aware=
.
> >>    This capability MUST then be taken into account during path
> >>    computation to prefer or even require links that advertise themselve=
s
> >>    as Timing-aware.  In this way the path can ensure the entry and exit
> >>    points into the LERs and, if desired, the links into the LSRs are
> >>    able to perform port based time-stamping thus minimizing their impac=
t
> >>    on the performance of the slave clock.
> >>
> >>    extensions are required to OSPF and IS-IS in order to advertise
> >>    Timing-aware capabilities of a link.  Such extensions are outside th=
e
> >>    scope of this document; however such extension SHOULD be able to
> >>    signal the following information per Router Link:
> >>
> >>    o  Capable of processing PTP, NTP or other Timing flows
> >>
> >>    o  Capable of performing Transparent Clock operation
> >>
> >>    o  Capable of performing Boundary Clock operation
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 32=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> 2.  Signaling Extensions for Creating Timing LSPs
> >>
> >>    RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-T=
E
> >>    is used to setup Timing LSPs, some information that indicates that
> >>    the LSP is carrying Timing flows MUST be included in the new
> >>    Extensions to RSVP-TE:
> >>
> >>    The following information MAY also be included in the new Extensions
> >>    to RSVP-TE:
> >>
> >>    o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
> >>       field
> >>
> >>    o  Number of VLANs in case of PW encapsulation
> >>
> >>    o  Timestamp field Type
> >>
> >>       *  Correction Field, Timestamp
> >>
> >>    o  Timestamp Field format
> >>
> >>       *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
> >>          NTP, etc.
> >>
> >>    Note that in case the above optional information is signaled with
> >>    RSVP-TE for a Timing LSP, all the Timing packets carried in that LSP
> >>    must have the same signaled characteristics.  For example if
> >>    Timestamp format is signaled as 64-bit PTPv1, then all Timing packet=
s
> >>    must use 64-bit PTPv1 time-stamp.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 33=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >> Authors' Addresses
> >>
> >>    Shahram Davari
> >>    Broadcom Corp.
> >>    San Jose, CA  95134
> >>    USA
> >>
> >>    Email: davari@broadcom.com
> >>
> >>
> >>    Amit Oren
> >>    Broadcom Corp.
> >>    San Jose, CA  95134
> >>    USA
> >>
> >>    Email: amito@broadcom.com
> >>
> >>
> >>    Manav Bhatia
> >>    Alcatel-Lucent
> >>    Bangalore,
> >>    India
> >>
> >>    Email: manav.bhatia@alcatel-lucent.com
> >>
> >>
> >>    Peter Roberts
> >>    Alcatel-Lucent
> >>    Kanata,
> >>    Canada
> >>
> >>    Email: peter.roberts@alcatel-lucent.com
> >>
> >>
> >>    Laurent Montini
> >>    Cisco Systems
> >>    San Jose CA
> >>    USA
> >>
> >>    Email: lmontini@cisco.com
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 34=
]
> >
> >> Internet-Draft        Transporting Timing over MPLS            June 201=
3
> >>
> >>
> >>    Luca
> >>    Cisco Systems
> >>    San Jose CA
> >>    USA
> >>
> >>    Email: lmartini@cisco.com
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> Davari, et al.          Expires December 17, 2013              [Page 35=
]
> >
> >>
> >> --
> >> For corporate legal information go to:
> >>
> >> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
> >>
> >> _______________________________________________
> >> TICTOC mailing list
> >> TICTOC@ietf.org
> >> https://www.ietf.org/mailman/listinfo/tictoc
> >
> >
> > This e-mail message is intended for the recipient only and contains
> information which is CONFIDENTIAL and which may be proprietary to ECI
> Telecom. If you have received this transmission in error, please inform us=
 by e-
> mail, phone or fax, and then delete the original and all copies thereof.
> >
> > _______________________________________________
> > mpls mailing list
> > mpls@ietf.org
> > https://www.ietf.org/mailman/listinfo/mpls


This e-mail message is intended for the recipient only and contains informat=
ion which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If yo=
u have received this transmission in error, please inform us by e-mail, phon=
e or fax, and then delete the original and all copies thereof.


From stbryant@cisco.com  Mon Aug  5 02:30:27 2013
Return-Path: <stbryant@cisco.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 700F221F8EB3; Mon,  5 Aug 2013 02:30:19 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -110.599
X-Spam-Level: 
X-Spam-Status: No, score=-110.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Kj6u+2UzcgHs; Mon,  5 Aug 2013 02:30:09 -0700 (PDT)
Received: from ams-iport-1.cisco.com (ams-iport-1.cisco.com [144.254.224.140]) by ietfa.amsl.com (Postfix) with ESMTP id 8779C21F97C7; Mon,  5 Aug 2013 02:29:23 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=69317; q=dns/txt; s=iport; t=1375694963; x=1376904563; h=message-id:date:from:reply-to:mime-version:to:cc:subject: references:in-reply-to:content-transfer-encoding; bh=Pf82lB9bGTRKsubYizE4CBG/tR2TdaUDf3oMUbE1+Lg=; b=fOVuzPULB4lczKII/ui3tPYTDyiHuSwIoFnpFEIjy9A47b0rRn4fp8j6 ws/DbEeuf469/CrKAokdXQcx8I24FsWZIZQiLLXGPZId31d0DppacLdLj 07ADeemj1z11341OX00Gdz+0oIF1ZAA8zSoY6va+QT8rESD32CHvNDBTU c=;
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AgUFAEFv/1GQ/khN/2dsb2JhbABQCoMGNb9JgR4WdIIkAQEBAwEBAQEXAR0uAQcKAQwECxEBAwEBAQkMCggHCQMCAQIBFR8DBggGCgMBBQIBAReHbwYMtG2ORgIJBgUFgRYiBwYEhAMDjEGHB0GDV4EqhW2KOIMYgWcBBQMX
X-IronPort-AV: E=Sophos;i="4.89,817,1367971200"; d="scan'208";a="158008151"
Received: from ams-core-4.cisco.com ([144.254.72.77]) by ams-iport-1.cisco.com with ESMTP; 05 Aug 2013 09:29:20 +0000
Received: from cisco.com (mrwint.cisco.com [64.103.70.36]) by ams-core-4.cisco.com (8.14.5/8.14.5) with ESMTP id r759TIbR011511 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Mon, 5 Aug 2013 09:29:18 GMT
Received: from [IPv6:::1] (localhost [127.0.0.1]) by cisco.com (8.14.4+Sun/8.8.8) with ESMTP id r759TC9Z028298; Mon, 5 Aug 2013 10:29:12 +0100 (BST)
Message-ID: <51FF7068.9080600@cisco.com>
Date: Mon, 05 Aug 2013 10:29:12 +0100
From: Stewart Bryant <stbryant@cisco.com>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:17.0) Gecko/20130620 Thunderbird/17.0.7
MIME-Version: 1.0
To: "S. Davari" <davarish@yahoo.com>
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com>
In-Reply-To: <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls]  Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
Reply-To: stbryant@cisco.com
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 05 Aug 2013 09:30:27 -0000

My concern is that we are creating a new LSP type in MPLS
(which is a architectural change and thus needs to be very carefully
considered) without asking the question "how can I do this
in the most general way, to maximize flexibility/reuse"

I suggested an LSP type that has the properties "timestamp
and pass to application", but I think that Sasha raises a good
point about using router alert. However RA has no implicit
timestamp, so maybe we need a new type of RA that has the
properties "timestamp, and pass top application indicated by
the next label". That would be quite useful in a number
of OAM applications. There is possibly some GAL variant
of that design that should also be considered.

The application could worry about the path and where it
was necessary to skip some hops a hierarchical LSP would
accomplish that, with the specific benefit that the application
would consciously do this, and may be able to apply some
form of compensation within the network.

- Stewart

On 04/08/2013 10:40, S. Davari wrote:
> Hi Sasha
>
> Perhaps you have not understood the draft well. The main reason that the draft chose to use a dedicated LSP that is signaled via RSVPTE indicating it is a timing LSP is to be backward compatible with non-1588 nodes. Such nodes simply just switch the packet.
>
> Your proposal had been considered and was rejected since it was not backward compatible. Nodes receiving alert label or TTL = 1 send the packet to CPU. You can refer to the meeting notes.
>
> Regards,
> Shahram
>
>
> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein <Alexander.Vainshtein@ecitele.com> wrote:
>
>> Stewart and all,
>> I concur with Stewart's statement that the draft does not define "full interaction with MPLS architecture".
>>
>> E.g., one of the objectives of the draft is to provide a technique that would be backward-compatible with old LSRs that cannot provide on-path support for timing distribution, while the other objective is to make every LSR on the path aware that some MPLS packets are carrying timing-related messages and hence require on-path support.
>>
>> IMHO and FWIW, the MPLS data plane architecture allows just two methods for making a transit LSR to provide special processing to a labeled packet:
>> - It would carry some kind of an "alert label" on top of the label stack and, specifically, on top of any labels used for actual forwarding
>> OR,
>> - The TTL in the top label stack entry has been set to 1.
>>
>>
>> The draft does not follow any of these approaches.
>>
>> I must also admit that Section 12 "OAM, Control and Management" of the draft looks somewhat in
>> E.g., I could not understand from the text whether BFD, LSP-Ping etc. OAM messages would be subjected to on-path support procedures for timing messages or not.
>>
>> My 2c,
>>      Sasha
>>
>>> -----Original Message-----
>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf Of
>>> Stewart Bryant
>>> Sent: Friday, August 02, 2013 12:01 PM
>>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
>>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.org; draft-
>>> ietf-tictoc-1588overmpls@tools.ietf.org
>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
>>>
>>> SB> This draft does not seem to provide a precise definition
>>> SB> the properties of the new LSP type that it wishes to
>>> SB> define, in particular it does it define the PHB of those
>>> SB> LSPs, nor the full interaction with the MPLS
>>> SB> architecture.
>>> SB>
>>> SB> I have not tracked TICTOC for a while but I thought that
>>> SB> the original plan was to define the concept of an offset
>>> SB> into a packet to do the correction.
>>> SB>
>>> SB> It is disappointing that the opportunity was not taken
>>> SB> to define a timing shim inside the timing LSP so that
>>> SB> a time correction could be added to any packet such that
>>> SB> the MPLS system was isolated from the details of the
>>> SB> complexity of the particular time transfer type.
>>>
>>> SB> I think that much more clarify is needed in terms of
>>> SB> definition of the new LSP type, since it is unclear
>>> SB> from this text how to implement one.
>>> SB>
>>> SB> There are a lot of other MPLS services such as
>>> SB> LSP ping that need to be considered.
>>> SB>
>>> SB> Please see inline for more comments. However these
>>> SB> comments are made in the context of the text as written
>>> SB> whilst I have a fundamental concern that this approach
>>> SB> lacks an MPLS architectural soundness that need
>>> SB> greater thought with significant impact on the
>>> SB> draft.
>>>
>>> - Stewart
>>>
>>>
>>> TICTOC Working Group                                           S. Davari
>>> Internet-Draft                                                   A. Oren
>>> Intended status: Standards Track                          Broadcom Corp.
>>> Expires: December 17, 2013                                     M. Bhatia
>>>                                                                P. Roberts
>>>                                                            Alcatel-Lucent
>>>                                                                L. Montini
>>>                                                                L. Martini
>>>                                                             Cisco Systems
>>>                                                             June 15, 2013
>>>
>>>
>>>              Transporting Timing messages over MPLS Networks
>>>                     draft-ietf-tictoc-1588overmpls-05
>>>
>>> Abstract
>>>
>>>     This document defines the method for transporting Timing messages
>>>     such as PTP and NTP over an MPLS network.  The method allows for the
>>>     easy identification of these PDUs at the port level to allow for port
>>>
>>> SB> What is a port
>>>
>>>     level processing of these PDUs in both LERs and LSRs.
>>>
>>>     The basic idea is to transport Timing messages inside dedicated MPLS
>>>     LSPs.  These LSPs only carry Timing messages and possibly Control and
>>>     Management packets, but they do not carry customer traffic.
>>>
>>> SB> More specifically they only carry traffic associated with the
>>> SB> timing service and its support.
>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
>>> SB> also it gets carried in a structure that causes it to get
>>> SB> timestamped.
>>>
>>>     Two methods for transporting Timing messages over MPLS are defined.
>>>
>>> SB> Perhaps the right approach is to define the new LSP type and then
>>> SB> seperately to define  the mapping of the various timing services
>>> SB> over that LSP type.
>>>
>>>     The first method is to transport Timing messages directly over the
>>>     dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
>>>     MPLS networks.  The second method is to transport Timing messages
>>>     inside a PW via Ethernet encapsulation.
>>>
>>> SB> I think that we should note that there are some
>>> SB> h/w reasons for this preference. A clean sheet approach
>>> SB> would have been to use PTP over MPLS with no intermediate
>>> SB> layers.
>>>
>>> Status of this Memo
>>>
>>>     This Internet-Draft is submitted in full conformance with the
>>>     provisions of BCP 78 and BCP 79.
>>>
>>>     Internet-Drafts are working documents of the Internet Engineering
>>>     Task Force (IETF).  Note that other groups may also distribute
>>>     working documents as Internet-Drafts.  The list of current Internet-
>>>     Drafts is at http://datatracker.ietf.org/drafts/current/.
>>>
>>>     Internet-Drafts are draft documents valid for a maximum of six months
>>>     and may be updated, replaced, or obsoleted by other documents at any
>>>     time.  It is inappropriate to use Internet-Drafts as reference
>>>     material or to cite them other than as "work in progress."
>>>
>>>     This Internet-Draft will expire on December 17, 2013.
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 1]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> Copyright Notice
>>>
>>>     Copyright (c) 2013 IETF Trust and the persons identified as the
>>>     document authors.  All rights reserved.
>>>
>>>     This document is subject to BCP 78 and the IETF Trust's Legal
>>>     Provisions Relating to IETF Documents
>>>     (http://trustee.ietf.org/license-info) in effect on the date of
>>>     publication of this document.  Please review these documents
>>>     carefully, as they describe your rights and restrictions with respect
>>>     to this document.  Code Components extracted from this document must
>>>     include Simplified BSD License text as described in Section 4.e of
>>>     the Trust Legal Provisions and are provided without warranty as
>>>     described in the Simplified BSD License.
>>>
>>>
>>> Table of Contents
>>>
>>>     1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  5
>>>
>>>     2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .  7
>>>
>>>     3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .  8
>>>
>>>     4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .  9
>>>
>>>     5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . . 12
>>>
>>>     6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . . 13
>>>       6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . . 13
>>>       6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . . 13
>>>       6.3.  Other Timing Encapsulation methods . . . . . . . . . . . . 14
>>>
>>>     7.  Timing message Processing  . . . . . . . . . . . . . . . . . . 15
>>>
>>>     8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . . 16
>>>
>>>     9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
>>>
>>>     10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
>>>
>>>     11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
>>>
>>>     12. OAM, Control and Management  . . . . . . . . . . . . . . . . . 20
>>>
>>>     13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . . 21
>>>
>>>     14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . . 22
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 2]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>>     15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . . 23
>>>       15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . . 23
>>>       15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . . 23
>>>       15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . . 24
>>>
>>>     16. Other considerations . . . . . . . . . . . . . . . . . . . . . 25
>>>
>>>     17. Security Considerations  . . . . . . . . . . . . . . . . . . . 26
>>>
>>>     18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 27
>>>
>>>     19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 28
>>>
>>>     20. References . . . . . . . . . . . . . . . . . . . . . . . . . . 29
>>>       20.1. Normative References . . . . . . . . . . . . . . . . . . . 29
>>>       20.2. Informative References . . . . . . . . . . . . . . . . . . 29
>>>
>>>     Appendix 1.  Routing extensions for Timing-aware Routers . . . . . 32
>>>
>>>     Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . . 33
>>>
>>>     Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 34
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 3]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>>     The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
>>>     "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in
>>> this
>>>     document are to be interpreted as described in RFC2119 [RFC2119].
>>>
>>>     When used in lower case, these words convey their typical use in
>>>     common language, and are not to be interpreted as described in
>>>     RFC2119 [RFC2119].
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 4]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 1.  Introduction
>>>
>>>     The objective of Precision Time Protocol (PTP) and Network Timing
>>>     Protocol (NTP) are to synchronize independent clocks running on
>>>     separate nodes of a distributed system.
>>>
>>>     [IEEE-1588] defines PTP messages for frequency, phase and time
>>>     synchronization.  The PTP messages include PTP PDUs over UDP/IP
>>>     (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F of
>>>     [IEEE-1588]).
>>>
>>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
>>> SB> of other PTP mappings if they provide better optimisation.
>>>
>>>     This document defines mapping and transport of the PTP
>>>     messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>>>     defines several clock types: ordinary clocks, boundary clocks, end-
>>>     to-end transparent clocks, and peer-to-peer transparent clocks.
>>>     Transparent clocks require intermediate nodes to update correction
>>>     field inside PTP message that reflects the transit time in the node.
>>>
>>>     [RFC5905] defines NTP messages for clock and time synchronization.
>>>     The PTP messages (PDUs) are transported over UDP/IP.  This document
>>> SB> Should that be NTP messages?
>>> SB> It needs to be made clear as soon as you introduce NTP that
>>> SB> they use different time representations.
>>>
>>>     defines mapping and transport of the NTP messages defined in
>>>     [RFC5905] over MPLS networks.
>>>
>>>     One key attribute of all of these Timing messages is that the Time
>>>     stamp processing should occur as close as possible to the actual
>>>     transmission and reception at the physical port interface.  This
>>>     targets optimal time and/or frequency recovery by avoiding variable
>>>     delay introduced by queues internal to the clocks.
>>>
>>> SB> As I recall NTP has no epoch point defined, and I am not sure
>>> SB> where that point is in the case of PTP in this mapping
>>> SB> Hopefully this will get defined in due course.
>>>
>>>     To facilitate the fast and efficient recognition of Timing messages
>>>     at the port level when the Timing messages are carried over MPLS
>>>     LSPs,
>>>
>>> SB> Over a new LSP type with time optimied characteristics
>>>
>>>     this document defines the specific encapsulations that should
>>>     be used.
>>> SB> Hopefully it will also define the PHP
>>>
>>>     In addition, it can be expected that there will exist LSR/
>>>     LERs where only a subset of the physical ports will have the port-
>>>     based Timing message processing capabilities.
>>> SB> Do you need to clarify that this only works at base and not in
>>> SB> a label heirarchy.
>>>
>>>
>>>     In order to ensure
>>>     that the LSPs carrying Timing packets always enter and exit ports
>>>     with this capability, routing extensions are defined to advertise
>>>     this capability on a port basis and to allow for the establishment of
>>>     LSPs that only transit such ports.  While this path establishment
>>>     restriction may be applied only at the LER Ingress and/or egress
>>>     ports, it becomes more important when using transparent clock capable
>>>     LSRs in the path.
>>> SB> I do not understand the implications of the last
>>> SB> sentences - starting ", it becomes"
>>>
>>>
>>>     Port based Timing message processing involves Timing message
>>>     recognition.  Once the Timing messages are recognized they can be
>>>     modified based on the reception or transmission Time-stamp.
>>>
>>>     This document provides two methods for transporting Timing messages
>>>     over MPLS.  One is applicable to MPLS environment and the other one
>>>     is applicable to MPLS/MPLS-TP environment
>>>
>>> SB> I think the sentence is incomplete.
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 5]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>>     The solution involves transporting Timing messages over dedicated
>>>     LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
>>>     carry Management and control messages, but not data plane client
>>>     traffic.
>>>
>>> SB> It is not clear why this restriction applies.
>>>
>>>     Timing LSPs can be established statically or via signaling.
>>> SB> s/statically/by provisioning/network management/
>>>
>>>     Extensions to control plane (OSPF, ISIS, etc.) is required to enable
>>>     routers to distribute their Timing processing capabilities over MPLS
>>>     to other routers.  However such extensions are outside the scope of
>>>     this document.
>>>
>>>     When signaling is used to setup the PTP LSP, Extensions to signaling
>>> SB> is it a PTP LSP or a Timing LSP?
>>>
>>>     protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
>>>     However such extensions are outside the scope of this document.
>>>
>>> SB> for mpls-tp GMPLS is the signalling protocol
>>>
>>>     While the techniques included herein allow for the establishment of
>>>     paths optimized to include Time-stamping capable links, the
>>>     performance of the Slave clocks is outside the scope of this
>>>     document.
>>>
>>>     At the time of publishing this specification, Transparent Clocking
>>>     (TC) is only defined for PTP.  Therefore at this time any part of
>>>     this specification that talks about Transparent Clocking applies only
>>>     to PTP.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 6]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 2.  Terminology
>>>
>>>     1588: The timing and synchronization as defined by IEEE 1588.
>>>
>>> SB> I think that there is a more formal name for the 1588 group
>>> SB> that needs to be used here.
>>> SB> Also do we need to talk about 1588-200? as there is
>>> SB> an update in progress
>>>
>>>     NTP: The timing and synchronization protocol defined by IETF RFC-1305
>>>     and RFC-5905.
>>>
>>>     PTP: The timing and synchronization protocol used by 1588.
>>> SB> need the proper name for 1588
>>>
>>>     Master Clock: The source of 1588 timing to a set of slave clocks.
>>>
>>>     Master Port: A port on a ordinary or boundary clock that is in Master
>>>     state.  This is the source of timing toward slave ports.
>>>
>>> SB> I am not sure the reader knows what a port is
>>>
>>>     Slave Clock: A receiver of 1588 timing from a master clock.
>>>
>>>     Slave Port: A port on a boundary clock or ordinary clock that is
>>>     receiving timing from a master clock.
>>>
>>>     Ordinary Clock: A device with a single PTP port.
>>>
>>>     Transparent Clock.  A device that measures the time taken for a PTP
>>>     event message to transit the device and then updates the
>>>     correctionField of the message with this transit time.
>>>
>>>     Boundary Clock: A device with more than one PTP port.  Generally
>>>     boundary clocks will have one port in slave state to receive timing
>>>     and then other ports in master state to re-distribute the timing.
>>>
>>>     PTP LSP: An LSP dedicated to carry PTP messages
>>>
>>> SB> PTP or timing?
>>>
>>>
>>>     PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>>     messages.
>>>
>>> SB> Ah I don't think that PWE3 know what one of these is
>>>
>>>     CW: Pseudowire Control Word
>>>
>>>     LAG: Link Aggregation
>>>
>>>     ECMP: Equal Cost Multipath
>>>
>>>     CF: Correction Field, a field inside certain PTP messages (message
>>>     type 0-3)that holds the accumulative transit time inside intermediate
>>>     switches
>>>
>>>     Timing messages: Timing Protocol messages that are exchanged between
>>>     routers in order to establish a synchronized clock.
>>>
>>> SB> A number of these definitions look like copies of IEEE1588
>>> SB> definitions. We need to provide references and note the
>>> SB> priority of the IEEE base reference.
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 7]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 3.  Problem Statement
>>>
>>>     [IEEE-1588] has defined methods for transporting PTP messages over
>>>     Ethernet and IP networks.  [RFC5905] has defined the method of
>>>     transporting NTP messages over IP networks.  There is a need to
>>>     transport Timing messages over MPLS networks while supporting the
>>>     Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
>>>     functionality in the LER and LSRs in the MPLS network.
>>>
>>>     There are multiple ways of transporting Timing over MPLS.  However,
>>>     there is a requirement to limit the possible encapsulation options to
>>>     simplify the Timing message identification and processing required at
>>>     the port level.
>>>
>>>     When Timing-awareness is needed, Timing messages should not be
>>>     transported over LSPs or PWs that are carrying customer traffic
>>>     because LSRs perform Label switching based on the top label in the
>>>     stack.
>>>
>>> SB> Have you explained why?
>>>
>>>     To detect Timing messages inside such LSPs require special
>>>     hardware to do deep packet inspection at line rate.  Even if such
>>>     hardware exists, the payload can't be deterministically identified by
>>>     LSRs because the payload type is a context of the PW label, and the
>>>     PW label and its context are only known to the Edge routers (PEs/
>>>     LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CES,
>>>     etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
>>>     LSRs dont have the knowledge of whether PW Control Word (CW) is
>>>     present or not and therefore can not deterministically identify the
>>>     payload.
>>>
>>>     A generic method is defined in this document that does not require
>>>     deep packet inspection at line rate, and can deterministically
>>>     identify Timing messages.  This method can be used to detect Timing
>>>     Messages in both one-step and two-step clock implementations of
>>>     ordinary, boundary and transparent clocks.
>>>
>>> SB> Needs a ref and I am sure many MPLS specialists will not understand
>>> SB> the msg types.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 8]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 4.  Timing over MPLS Architecture
>>>
>>>     Timing messages are exchange between Timing ports on ordinary and
>>>
>>> SB> Have you defined a timing port?
>>>
>>>     boundary clocks.  Boundary clocks terminate the Timing messages and
>>>     act as master for other boundary clocks or for slave clocks.  End-to-
>>>     End Transparent clocks do not terminate the Timing messages but they
>>>     do modify the contents of the Timing messages as they transit across
>>>     the transparent clock.
>>>
>>>     Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Clock
>>>
>>>     (TC) could be implemented in either LERs or LSRs.
>>>
>>> SB> LER and LSR need to be expanded
>>>
>>>     An example is shown in Figure 1, where the LERs act as Ordinary Clock
>>>     (OC) and are the initiating/terminating point for Timing messages.
>>>     The ingress LER encapsulates the Timing messages in Timing LSP and
>>>     the Egress LER terminates the Timing LSP.  The LSRs act as
>>>     Transparent Clock (TC) and just update the Timing field in the Timing
>>>     messages.
>>>
>>>
>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>        |Switch, |     |       |     |       |     |       |     |Switch, |
>>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>        |        |     |  OC   |     |  TC   |     |  OC   |     |        |
>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>                       /                                 \
>>>        +-------+     /                                   \     +-------+
>>>        |  LER  |    /                                     \    |  LER  |
>>>        | Master|---/                                       \---| Slave |
>>>        | Clock |                                               | Clock |
>>>        +-------+                                               +-------+
>>>
>>>       Figure (1) - Deployment example 1 of timing over MPLS network
>>>
>>>     Another example is shown in Figure2, where LERs terminate the Timing
>>>     messages received from switch/routers that are outside of the MPLS
>>>     network acting as OC or BC.  In this example LERs regenerate the
>>>     clock and initiate timing messages encapsulated in Timing LSP toward
>>>     the MPLS network, while the LSRs act as Transparent Clock (TC) and
>>>     just update the Timing field in the Timing messages, which are
>>>     already encapsulated in Timing LSPs.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 9]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>       |Switch, |     |       |     |       |     |       |     |Switch, |
>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>       | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC  |
>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>
>>>       Figure (2) - Deployment example 2 of timing over MPLS network
>>>
>>>
>>>     Another example is shown in Figure 3, where LERs do not terminate the
>>>     Timing messages received from switch/routers that are outside of the
>>>     MPLS network acting as OC, TC or BC.  The LERs act as TC and update
>>>     the Timing field in the Timing messages as they transit the LER,
>>>     while encapsulating them in timing LSP.  The LSRs also act as
>>>     Transparent Clock (TC) and just update the Timing field in the Timing
>>>     messages which are already encapsulated in Timing LSPs.
>>>
>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>        |Switch, |     |       |     |       |     |       |     |Switch, |
>>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>        |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC/BC|
>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>
>>>      Figure (3) - Deployment example 3 of timing over MPLS network
>>>
>>>     Another example is shown in Figure 4, where LERs and LSRs support
>>>     Boundary Clocks.  A single-hop LSP is created between two adjacent
>>>     LSRs engaged in BC operation.  Other methods such as PTP transport
>>>     over Ethernet MAY be used for transporting timing messages if the
>>>     link between the two routers is Ethernet.
>>>
>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>       |Switch, |     |       |     |       |     |       |     |Switch, |
>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>       | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC  |
>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>
>>>     Figure (4) - Deployment example 3 of timing over MPLS network
>>>
>>>     An MPLS domain MAY serve multiple customers.  In these cases the MPLS
>>>     domain (maintained by a service provider) may provide timing services
>>>     to multiple customers, each having their own Timing domain.
>>>
>>>     The Timing over MPLS architecture assumes full mesh of Timing LSPs
>>>     between all LERs supporting this specification.
>>>
>>> SB> Note sure this is right - the salves surely do not need to
>>> SB> exchange timing amongst themselves
>>>
>>>     It supports
>>>     Point-to- point (VPWS) and Multipoint (VPLS) services.
>>>
>>> SB> What does that mean? You do not carry user data traffic?
>>> SB> Maybe it's the ordering of the statemnets that is causing
>>> SB> confusion.
>>>
>>>     This means
>>>     that a customer may purchase a Point-to-point Timing service between
>>>     two customer sites or a Multipoint Timing service between more than
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 10]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>>     two customer sites.
>>>
>>>     The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
>>>     This means that the Timing Multicast messages such as PTP Multicast
>>>     event messages can be transported over P2MP Timing LSP or be
>>>     replicated and transported over many P2P Timing LSPs.
>>>
>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>>> SB> nor a P2MP PW, although we are close.
>>>
>>>     Timing messages, that do not require Time stamping or Correction
>>>     Field update MAY be transported over Timing LSPs to simplify hardware
>>>     and software.
>>>
>>>     PTP Announce messages that determine the Timing LSP terminating point
>>>     behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
>>>     to simplify hardware and software.
>>>
>>> SB> have you defined and referenced PTP announce msgs?
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 11]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 5.  Dedicated LSPs for Timing messages
>>>
>>>     Many methods have been considered for identifying the Timing messages
>>>     when they are encapsulated in MPLS such as using GAL/G-ACH or a new
>>>     reserved label.  These methods were not attractive since they either
>>>     required deep packet inspection at line rate in the intermediate LSRs
>>>     or they required use of a scarce new reserved label.  Also one of the
>>>     goals was to reuse existing OAM mechanisms.
>>>
>>> SB> RLs = SPLs are not so rare now. In any case needs a ref.
>>>
>>>     The method defined in this document can be used by LER and LSRs to
>>>     identify Timing messages in MPLS tunnels by just looking at the top
>>>     label in the MPLS label stack, which only carry Timing messages as
>>>     well as OAM, but not data plane client traffic.
>>>
>>>     Compliant implementations MUST use dedicated LSPs to carry Timing
>>>     messages over MPLS.
>>>
>>> SB> I think that we need a definition of the properies of these LSPs
>>>
>>>     These LSPs are herein referred to as "Timing
>>>     LSPs" and the labels associated with these LSPs as "Timing LSP
>>>     labels".  The Timing LSPs that runs between Ingress and Egress LERs
>>>     MUST be co-routed.  Alternatively, a single bidirectional co-routed
>>>     LSP can be used.
>>>
>>> SB> I though that you said you could use M2MP LSPs - these are not
>>> SB> bidirectional.
>>>
>>>     Co-routing of the two directions is required to limit the difference
>>>     in the delays in the Master clock to Slave clock direction compared
>>>     to the Slave clock to Master clock direction.  The Timing LSP MAY be
>>>     MPLS/MPLS-TP LSP.
>>>
>>>     The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
>>>     New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
>>>     outside the scope of this document.
>>>
>>>     The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such as
>>>     BFD and LSP Ping but the LSP data plane client plane traffic MUST be
>>>     Timing packets only.
>>>
>>> SB> Why?
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 12]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 6.  Timing over LSP Encapsulation
>>>
>>> The encapsulations is not LSP is it?
>>>
>>>     This document defines two methods for carrying Timing messages over
>>>     MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>>     messages over Timing LSPs, and the second method, is carrying
>>>     Ethernet encapsulated Timing messages over Ethernet PWs inside Timing
>>>     LSPs.
>>>
>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>>
>>>     The simplest method of transporting Timing messages over MPLS is to
>>>     encapsulate Timing PDUs in UDP/IP and then encapsulate them in Timing
>>>     LSP.  This format is shown in Figure 4.
>>>
>>>
>>>                      +----------------------+
>>>                      |   Timing LSP Label   |
>>>                      +----------------------+
>>>                      |        IPv4/6        |
>>>                      +----------------------+
>>>                      |         UDP          |
>>>                      +----------------------+
>>>                      |     Timing PDU       |
>>>                      +----------------------+
>>>
>>>        Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>>
>>>
>>>     This encapsulation is very simple and is useful when the network
>>>     between Timing Master Clock and Slave Clock is MPLS network.
>>>
>>> SB> Simple is a judgement call
>>>
>>>     In order for an LER/LSR to process Timing messages, the Timing LSP
>>>     Label must be at the top label of the label stack.  The LER/LSR MUST
>>>     know that the Timing LSP Label is used for carrying Timing messages.
>>>     This can be accomplished via static configuration or via RSVP-TE
>>>     signaling.
>>>
>>>     The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>>     [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>>>     [RFC5905].
>>>
>>> 6.2.  Timing over PW Encapsulation
>>>
>>>     Another method of transporting Timing over MPLS networks is by
>>>     encapsulating Timing PDUs in PW which in turn is transported over
>>>     Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
>>>     shown in Fig 5(A) MUST be used and the Ethernet encapsulation of PTP
>>>     MUST follow Annex F of [IEEE-1588].
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 13]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>>     The RAW mode or Tagged mode defined in [RFC4448] MAY be used and the
>>>     Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>>>     Timing over PW encapsulation MUST use the Control Word (CW) as
>>>     specified in [RFC4448] to ensure proper detection of PTP messages
>>>     inside the MPLS packets for Timing over LSP and Timing over PW
>>>     encapsulation.
>>>
>>> SB> That needs explanation
>>>
>>>     The use of Sequence Number in the CW is optional.
>>>
>>> SB> Given that s/n are never in practice deployed, you could probably
>>> SB> simplify things by sayig that they are not used.
>>>
>>>     Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over PW
>>>     (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>>
>>>                      +----------------+  +----------------+
>>>                       |Timing LSP Label|  |Timing LSP Label|
>>>                       +----------------+  +----------------+
>>>                       |    PW Label    |  |    PW Label    |
>>>                       +----------------+  +----------------+
>>>                       |  Control Word  |  |      IP        |
>>>                       +----------------+  +----------------+
>>>                       |    Ethernet    |  |      UDP       |
>>>                       |     Header     |  +----------------+
>>>                       +----------------+  |   Timing PDU   |
>>>                       |S-VLAN(Optional)|  |                |
>>>                       +----------------+  +----------------+
>>>                       |C-VLAN(Optional)|        (B)
>>>                       +----------------+
>>>                       |   Timing PDU   |
>>>                       |                |
>>>                       +----------------+
>>>                              (A)
>>>
>>>                Figure (5) - Timing over PW Encapsulations
>>>
>>>     In order for an LSR to process PTP messages, the top label of the
>>>     label stack (the Tunnel Label) MUST be a Timing label.
>>>
>>> S> You said that before.
>>>
>>> 6.3.  Other Timing Encapsulation methods
>>>
>>>     In future other timing encapsulation methods may be introduced, such
>>>     as a new shim header after the Bottom of Stack to carry the Timing
>>>     information.  Such new encapsulations are outside the scope of this
>>>     document.
>>>
>>>
>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>>> SB> out of the definition of the LSP
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 14]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> SB> I think we need a section on LSP processing
>>>
>>> 7.  Timing message Processing
>>>
>>>     Each Timing protocol such as PTP and NTP, define their set of Timing
>>>     messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>>>     FOLLOW_UP, etc messages.
>>>
>>>     Some of the Timing messages require time stamping or correction field
>>>     update at port level and some dont.  It is the job of the LER/LSR to
>>>     parse the timing message and find out the type of the Timing message
>>>     and decide whether and how to Time- stamp it (e.g., BC) or update
>>>     correction field(e.g., TC).
>>>
>>>
>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
>>> SB> function rather than the LER?
>>>
>>>     For example the following PTP messages (called Event messages)
>>>     require time-stamping or correction field update:
>>>
>>>     o  SYNC
>>>
>>>     o  DELAY_REQ (Delay Request)
>>>
>>>     o  PDELAY_REQ (Peer Delay Request)
>>>
>>>     o  PDELAY_RESP (Peer Delay Response)
>>>
>>>     SYNC and DELAY_REQ are exchanged between Master Clock and Slave Clock
>>>     and MUST be transported over PTP LSPs.  PDELAY_REQ and PDELAY_RESP
>>>     are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>>>     Boundary, or Transparent) and SHOULD be transported over single hop
>>>     PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
>>>     and PDELAY_RESP_FOLLOW_UP messages MUST also be transported over
>>> the
>>>     PTP LSPs.
>>>
>>>     For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
>>>     transported over two PTP LSPs that are in opposite directions.  These
>>>     PTP LSPs, which are in opposite directions MUST be congruent and co-
>>>     routed.  Alternatively, a single bidirectional co-routed LSP can be
>>>     used.
>>>
>>>     Except as indicated above for the two-step PTP clocks, Non-Event PTP
>>>     message types do not need to be processed by intermediate routers.
>>>     These message types MAY be carried in PTP Tunnel LSPs.
>>>
>>> SB> Are you saying that a timing P router has to be msg type sensitive?
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 15]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 8.  Protection and Redundancy
>>>
>>>
>>> SB> This is a bit of a jump - I don't know how the LSP itself works yet!
>>>
>>>     In order to ensure continuous uninterrupted operation of slave
>>>     clocks, usually as a general practice, slave clocks (or ports) track
>>>     redundant master clocks.
>>>
>>>     It is the responsibility of the network operator to ensure that
>>>     physically disjoint Timing LSPs are established between a slave clock
>>>     (or port) and redundant master clocks (or ports).
>>>
>>>     When a slave clock (or port) listens to redundant master clocks or
>>>     ports, any prolonged Timing LSP outage will trigger the slave clock
>>>     or port to switch to a redundant master clock or port.
>>>
>>>     LSP/PW protection such as Linear protection Switching (1:1, 1+1),
>>>     Ring protection switching or MPLS Fast Reroute (FRR) generally switch
>>>     alternative path that usually cause a change in delay, which if
>>>     undetected by slave clock can reduce accuracy of the slave clock.
>>>
>>>     Therefore protection switching MAY be used, as long as phase jumps
>>>     upon switchover due to differences in path latency are detected and
>>>     compensated for (such compensation not being required if BCs or peer-
>>>     peer TCs are used throughout).
>>>
>>>     Note that any protection or reroute mechanism that adds additional
>>>     MPLS label to the label stack, such as Facility Backup Fast Reroute,
>>>     MUST ensure that the pushed label is also a Timing Label to ensure
>>>     recognition of the MPLS frame as containing Timing messages, as it
>>>     transits the backup path.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 16]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 9.  ECMP
>>>
>>>     To ensure the optimal operation of slave clocks and avoid error
>>>     introduced by forward and reverse path delay asymmetry, the physical
>>>     path for Timing messages from master clock to slave Clock and vice
>>>     versa must be the same for all Event Timing messages listed in
>>>     section 7.
>>>
>>>     Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
>>>     Multipath).
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 17]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 10.  PHP
>>>
>>>     To ensure that the label on the top of the label stack is the Timing
>>>     LSP Label, PHP MUST not be used.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 18]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 11.  Entropy
>>>
>>>     To ensure all Timing messages in a Timing LSP take the same path,
>>>     Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>>     Entropy Label MUST NOT be used for the PWs that are carried inside
>>>     Timing LSP [RFC6391].
>>>
>>> SB> This is incorrect - you mean that all msgs of the same timing
>>> SB> flow need to have the same EL value.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 19]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 12.  OAM, Control and Management
>>>
>>>     In order to monitor Timing LSPs and their encapsulated PWs, they MUST
>>>     be able to carry OAM and management messages.  These management
>>>     messages MUST be differentiated from Timing messages via already
>>>     defined IETF methods.
>>>
>>>     For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
>>>     over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>>>     Management protocols can easily be identified by the UDP Destination
>>>     Port number or by GAL/G-ACH respectively.
>>>
>>>     Also BFD, LSP-Ping and other management messages MAY run over the PWs
>>>     encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3 or
>>>     4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is going
>>>     to be deprecated by IETF).  In this case G-ACH, PW label (TTL=1) or
>>>     GAL-ACH are used to identify such management messages.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 20]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 13.  QoS Considerations
>>>
>>>     In network deployments where not every LSR/LER is Timing-aware, it is
>>>     important to reduce the impact of the non-Timing-aware LSR/LERs on
>>>     the timing recovery in the slave clock.  The Timing messages are time
>>>     critical and must be treated with the highest priority.  Therefore
>>>     Timing over MPLS messages must be treated with the highest priority
>>>     in the routers.  This can be achieved by proper setup of Timing LSPs.
>>>
>>>     It is recommended that the Timing LSPs are setup or configured
>>>     properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC2697]
>>>     for drop eligibility.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 21]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 14.  FCS and Checksum Recalculation
>>>
>>>     When time-stamp generation and timing packet adjustment is performed
>>>     near the physical port hardware, the process MUST include
>>>     recalculation of the Ethernet FCS.
>>>
>>> SB> The above is confusing - an LSR always recomputes the link layer
>>> SB> CRC which may or may not be Ethernet.
>>>
>>>     Also FCS retention for the
>>>     payload Ethernet described in [RFC4720] MUST NOT be used.
>>>
>>>     For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
>>>     may be required as per UDP transport standards.
>>>
>>> SB> You really need to be working on getting the IPv6 C?S computation
>>> SB> removed from PTP msgs.
>>>
>>>     When UDP checksum is used, each Timing-aware LER/LSR must either
>>>     incrementally update the UDP checksum after Time stamping or
>>>     Correction Field update or verify the UDP checksum on reception from
>>>     upstream and recalculate the checksum completely on transmission to
>>>     downstream node after Time stamping or Correction Field update.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 22]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 15.  Behavior of LER/LSR
>>>
>>>     Timing-capable/aware LERs and LSRs are routers that have one or more
>>>
>>> SB> You mean physical interfaces?
>>>
>>>     interfaces that can perform Timing operations (OC/BC/TC) on Timing
>>>     packets and are configured to do so.  Timing-capable/aware LERs and
>>>     LSRs can advertise their Timing-capability per-interface via control
>>>     plane such as OSPF or IS-IS.
>>> SB> ISIS and OSPF are routing protocols.
>>>
>>>    The Timing-capable/aware LERs can then
>>>     signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timing
>>>     capability of LER and LSRs may be configured in a centralized
>>>     controller and the Timing LSP may be setup using manual configuration
>>>     or other methods such as SDN.
>>>
>>> SB> it can also be configured individually rather then through
>>> SB> a cebtral controllwe
>>>
>>> 15.1.  Behavior of Timing-capable/aware LER
>>>
>>>     When a Timing-capable/aware LER behaves as a Transparent clock and
>>>     receives a Timing message from a Timing-capable/aware non-MPLS
>>>     interface, the LER updates the Correction Field (CF) and encapsulates
>>>     and forwards the timing message over previously established Timing
>>>     LSP.
>>>
>>> SB> You need to call out the details so that people properly
>>> SB> understand the definition of the new LSP.
>>>
>>>     Also when a Timing message is received from a Timing-capable/
>>>     aware MPLS interface, LER updates the Correction Filed (CF) and
>>>     decapsulates the MPLS encapsulation and forwards the timing message
>>>     to a non-MPLS interface.
>>>
>>>     When a Timing-capable/aware LER behaves as a Boundary clock and
>>>     receives a Timing message from a Timing-capable/aware non MPLS
>>>     interface, the LER Timestamps the Timing packet and sends it to the
>>>     LERs Boundary clock processing module.  Also when a Timing message is
>>>     received from a Timing- capable/aware MPLS interface, the LER
>>>     Timestamps the Timing packet and sends it to the LERs Boundary clock
>>>     processing module.
>>>
>>>     When a Timing-capable/aware LER behaves as an Ordinary Clock toward
>>>     the MPLS network, and receives a Timing message from a Timing-
>>>     capable/aware MPLS interface, the LER Timestamps the Timing packet
>>>     and sends it to the LERs Ordinary clock processing module.
>>>
>>> 15.2.  Behavior of Timing-capable/aware LSR
>>>
>>>     When a Timing-capable/aware LSR behaves as a Transparent clock and
>>>     receives a Timing message from a Timing-capable/aware MPLS interface,
>>>     The LSR updates the Correction Filed (CF) and forwards the timing
>>>     message over another MPLS interface.
>>>
>>>     When a Timing-capable/aware LSR behaves as a Boundary clock and
>>>     receives a Timing message from a Timing-capable/aware MPLS interface.
>>>     The LSR performs the functions of a Boundary Clock in terminating the
>>>     received Timing message and re-generating a new timing message over
>>>     another (or the same) MPLS interface.
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 23]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>>
>>>     It is most beneficial when all LSRs in the path of a Timing LSP be
>>>     timing-Capable/aware LSRs.  This would ensure the highest quality
>>>     time and clock synchronization by Timing Slave Clocks.  However, this
>>>     specification does not mandate that all LSRs in path of a Timing LSP
>>>     be Timing- capable/aware.
>>>
>>>     Non-Timing-capable/aware LSRs just switch the packets encapsulated in
>>>     Timing LSPs and dont perform any Timing operation (TC or BC).
>>>     However as explained in QoS section the Timing over MPLS packets MUST
>>>     be still be treated with the highest priority based on their Traffic
>>>     Class (TC) marking.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 24]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 16.  Other considerations
>>>
>>>     [IEEE-1588] defines an optional peer-to-peer Transparent clocking
>>>     that requires peer delay measurement between two adjacent Timing-
>>>     capable/ aware routers/switches.  Peer delay measurement messages
>>>     need to be time stamped and terminated by the Timing-capable/aware
>>>     routers/ switches.  This means that two adjacent LSRs may be engaged
>>>     in a peer delay measurement.
>>>
>>>     For transporting such peer delay measurement messages a single-hop
>>>     LSP SHOULD to be created between the two adjacent LSRs engaged in
>>>     peer delay measurement to carry peer delay measurement messages.
>>>     Other methods such as PTP transport over Ethernet MAY be used for
>>>     transporting peer delay measurement messages if the link between the
>>>     two routers is Ethernet.
>>>
>>>     In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ ware
>>>     routers/switches MUST maintain a list of all the neighbors it needs
>>>     to send a PDelay_Req to, where each neighbor corresponds to a timing
>>>     LSP.
>>>
>>>     The use of Explicit Null Label (Label= 0 or 2) is acceptable as long
>>>     as either the Explicit Null label is the bottom of stack label
>>>     (applicable only to UDP/IP encapsulation) or the label below the
>>>     Explicit Null label is a PTP label.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 25]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 17.  Security Considerations
>>>
>>>     MPLS PW security considerations in general are discussed in [RFC3985]
>>>     and [RFC4447],and those considerations also apply to this document.
>>>
>>>     An experimental security protocol is defined in [IEEE-1588].The PTP
>>>     security extension and protocol provides group source authentication,
>>>     message integrity, and replay attack protection for PTP messages.
>>>
>>>     When the MPLS network (provider network) serves multiple customers,
>>>     it is important to maintain and process each customers clock and
>>>     Timing messages separately from other customers to ensure there is no
>>>     cross- customer effect.  For example if an LER BC is synchronized to
>>>     a specific grandmaster, belonging to customer A, then the LER MUST
>>>     use that BC clock only for customer A to ensure that customer A
>>>     cannot attack other customers by manipulating its time.
>>>
>>>     Timing messages MAY be encrypted or authenticated, provided that the
>>>     LERs/LSRs that are Timing capable/aware can authenticate/ decrypt the
>>>     timing messages.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 26]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 18.  Acknowledgements
>>>
>>>     The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizrahi,
>>>     Stefano Ruffini, Peter Meyer, and other members of IETF for reviewing
>>>     and providing feedback on this draft.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 27]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 19.  IANA Considerations
>>>
>>>     There are no IANA requirements in this specification.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 28]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 20.  References
>>>
>>> 20.1.  Normative References
>>>
>>>     [IEEE-1588]
>>>                IEEE 1588-2008, "IEEE Standard for a Precision Clock
>>>                Synchronization Protocol for Networked Measurement and
>>>                Control Systems".
>>>
>>>     [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>>                Requirement Levels", BCP 14, RFC 2119, March 1997.
>>>
>>>     [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
>>>                Edge (PWE3) Architecture", RFC 3985, March 2005.
>>>
>>>     [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
>>>                Proxies (ND Proxy)", RFC 4389, April 2006.
>>>
>>>     [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
>>>                Heron, "Pseudowire Setup and Maintenance Using the Label
>>>                Distribution Protocol (LDP)", RFC 4447, April 2006.
>>>
>>>     [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>>                "Encapsulation Methods for Transport of Ethernet over MPLS
>>>                Networks", RFC 4448, April 2006.
>>>
>>>     [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>>                Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>>                Retention", RFC 4720, November 2006.
>>>
>>>     [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
>>>                Connectivity Verification (VCCV): A Control Channel for
>>>                Pseudowires", RFC 5085, December 2007.
>>>
>>>     [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detection
>>>                (BFD)", RFC 5880, June 2010.
>>>
>>>     [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
>>>                "Bidirectional Forwarding Detection (BFD) for MPLS Label
>>>                Switched Paths (LSPs)", RFC 5884, June 2010.
>>>
>>> 20.2.  Informative References
>>>
>>>     [I-D.ietf-pwe3-fat-pw]
>>>                Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>>>                J., and S. Amante, "Flow Aware Transport of Pseudowires
>>>                over an MPLS Packet Switched Network",
>>>                draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 29]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>>     [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
>>>                system routeing information exchange protocol for use in
>>>                conjunction with the Protocol for providing the
>>>                Connectionless-mode Network Service (ISO 8473)".
>>>
>>>     [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
>>>                dual environments", RFC 1195, December 1990.
>>>
>>>     [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
>>>
>>>     [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
>>>                Marker", RFC 2697, September 1999.
>>>
>>>     [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boudec,
>>>                J., Courtney, W., Davari, S., Firoiu, V., and D.
>>>                Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>>                Behavior)", RFC 3246, March 2002.
>>>
>>>     [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineering
>>>                (TE) Extensions to OSPF Version 2", RFC 3630,
>>>                September 2003.
>>>
>>>     [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
>>>                System (IS-IS) Extensions for Traffic Engineering (TE)",
>>>                RFC 3784, June 2004.
>>>
>>>     [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
>>>                Shaffer, "Extensions to OSPF for Advertising Optional
>>>                Router Capabilities", RFC 4970, July 2007.
>>>
>>>     [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
>>>                System to Intermediate System (IS-IS) Extensions for
>>>                Advertising Router Information", RFC 4971, July 2007.
>>>
>>>     [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
>>>                Topology (MT) Routing in Intermediate System to
>>>                Intermediate Systems (IS-ISs)", RFC 5120, February 2008.
>>>
>>>     [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>>                Engineering", RFC 5305, October 2008.
>>>
>>>     [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>>                "Traffic Engineering Extensions to OSPF Version 3",
>>>                RFC 5329, September 2008.
>>>
>>>     [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
>>>                for IPv6", RFC 5340, July 2008.
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 30]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>>     [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network
>>>                Time Protocol Version 4: Protocol and Algorithms
>>>                Specification", RFC 5905, June 2010.
>>>
>>>     [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>>>                J., and S. Amante, "Flow-Aware Transport of Pseudowires
>>>                over an MPLS Packet Switched Network", RFC 6391,
>>>                November 2011.
>>>
>>>     [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
>>>                L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
>>>                RFC 6790, November 2012.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 31]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 1.  Routing extensions for Timing-aware Routers
>>>
>>>     MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] and
>>>     IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (TE)
>>>     link information used for constraint-based routing.
>>>
>>>     Indeed, it is useful to advertise data plane TE router link
>>>     capabilities, such as the capability for a router to be Timing-aware.
>>>     This capability MUST then be taken into account during path
>>>     computation to prefer or even require links that advertise themselves
>>>     as Timing-aware.  In this way the path can ensure the entry and exit
>>>     points into the LERs and, if desired, the links into the LSRs are
>>>     able to perform port based time-stamping thus minimizing their impact
>>>     on the performance of the slave clock.
>>>
>>>     extensions are required to OSPF and IS-IS in order to advertise
>>>     Timing-aware capabilities of a link.  Such extensions are outside the
>>>     scope of this document; however such extension SHOULD be able to
>>>     signal the following information per Router Link:
>>>
>>>     o  Capable of processing PTP, NTP or other Timing flows
>>>
>>>     o  Capable of performing Transparent Clock operation
>>>
>>>     o  Capable of performing Boundary Clock operation
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 32]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> 2.  Signaling Extensions for Creating Timing LSPs
>>>
>>>     RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-TE
>>>     is used to setup Timing LSPs, some information that indicates that
>>>     the LSP is carrying Timing flows MUST be included in the new
>>>     Extensions to RSVP-TE:
>>>
>>>     The following information MAY also be included in the new Extensions
>>>     to RSVP-TE:
>>>
>>>     o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
>>>        field
>>>
>>>     o  Number of VLANs in case of PW encapsulation
>>>
>>>     o  Timestamp field Type
>>>
>>>        *  Correction Field, Timestamp
>>>
>>>     o  Timestamp Field format
>>>
>>>        *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
>>>           NTP, etc.
>>>
>>>     Note that in case the above optional information is signaled with
>>>     RSVP-TE for a Timing LSP, all the Timing packets carried in that LSP
>>>     must have the same signaled characteristics.  For example if
>>>     Timestamp format is signaled as 64-bit PTPv1, then all Timing packets
>>>     must use 64-bit PTPv1 time-stamp.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 33]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>> Authors' Addresses
>>>
>>>     Shahram Davari
>>>     Broadcom Corp.
>>>     San Jose, CA  95134
>>>     USA
>>>
>>>     Email: davari@broadcom.com
>>>
>>>
>>>     Amit Oren
>>>     Broadcom Corp.
>>>     San Jose, CA  95134
>>>     USA
>>>
>>>     Email: amito@broadcom.com
>>>
>>>
>>>     Manav Bhatia
>>>     Alcatel-Lucent
>>>     Bangalore,
>>>     India
>>>
>>>     Email: manav.bhatia@alcatel-lucent.com
>>>
>>>
>>>     Peter Roberts
>>>     Alcatel-Lucent
>>>     Kanata,
>>>     Canada
>>>
>>>     Email: peter.roberts@alcatel-lucent.com
>>>
>>>
>>>     Laurent Montini
>>>     Cisco Systems
>>>     San Jose CA
>>>     USA
>>>
>>>     Email: lmontini@cisco.com
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 34]
>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>
>>>
>>>     Luca
>>>     Cisco Systems
>>>     San Jose CA
>>>     USA
>>>
>>>     Email: lmartini@cisco.com
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 35]
>>> --
>>> For corporate legal information go to:
>>>
>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>
>>> _______________________________________________
>>> TICTOC mailing list
>>> TICTOC@ietf.org
>>> https://www.ietf.org/mailman/listinfo/tictoc
>>
>> This e-mail message is intended for the recipient only and contains information which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If you have received this transmission in error, please inform us by e-mail, phone or fax, and then delete the original and all copies thereof.
>>
>> _______________________________________________
>> mpls mailing list
>> mpls@ietf.org
>> https://www.ietf.org/mailman/listinfo/mpls
> .
>


-- 
For corporate legal information go to:

http://www.cisco.com/web/about/doing_business/legal/cri/index.html


From davari@broadcom.com  Mon Aug  5 13:29:40 2013
Return-Path: <davari@broadcom.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 729F321F8421; Mon,  5 Aug 2013 13:29:40 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.599
X-Spam-Level: 
X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id HiLtiy-VpSMx; Mon,  5 Aug 2013 13:29:27 -0700 (PDT)
Received: from mms1.broadcom.com (mms1.broadcom.com [216.31.210.17]) by ietfa.amsl.com (Postfix) with ESMTP id 8335921F9F2D; Mon,  5 Aug 2013 13:29:26 -0700 (PDT)
Received: from [10.9.208.53] by mms1.broadcom.com with ESMTP (Broadcom SMTP Relay (Email Firewall v6.5)); Mon, 05 Aug 2013 13:25:27 -0700
X-Server-Uuid: 06151B78-6688-425E-9DE2-57CB27892261
Received: from SJEXCHCAS07.corp.ad.broadcom.com (10.16.203.16) by IRVEXCHCAS06.corp.ad.broadcom.com (10.9.208.53) with Microsoft SMTP Server (TLS) id 14.1.438.0; Mon, 5 Aug 2013 13:29:17 -0700
Received: from SJEXCHMB12.corp.ad.broadcom.com ( [fe80::bc15:c1e1:c29a:36f7]) by SJEXCHCAS07.corp.ad.broadcom.com ( [::1]) with mapi id 14.01.0438.000; Mon, 5 Aug 2013 13:29:16 -0700
From: "Shahram Davari" <davari@broadcom.com>
To: "stbryant@cisco.com" <stbryant@cisco.com>, "S. Davari" <davarish@yahoo.com>
Thread-Topic: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
Thread-Index: AQHOkPa4jpilABu4V0eu/dc/FPmvd5mGz+gAgABBHaA=
Date: Mon, 5 Aug 2013 20:29:16 +0000
Message-ID: <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com>
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com>
In-Reply-To: <51FF7068.9080600@cisco.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-originating-ip: [10.16.203.100]
MIME-Version: 1.0
X-WSS-ID: 7E1ED5BD31W80904505-01-01
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: quoted-printable
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 05 Aug 2013 20:29:40 -0000

Hi Stewart,

Your suggestion of " I suggested an LSP type that has the properties "times=
tamp and pass to application", is a subset of the existing draft and should=
 work. It basically limits the time stamping and correction field update to=
 LERs (while the draft supports time stamping at LER and LSR).

However Sasha's suggestion of using a reserved label (RAL, GAL or any other=
 reserved label) does not satisfy one of the major requirements, which is b=
ackward compatibility. Routers that don't understand this reserved label wi=
ll drop or copy to CPU such packets.=20

Regards,
Shahram

-----Original Message-----
From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf Of=
 Stewart Bryant
Sent: Monday, August 05, 2013 2:29 AM
To: S. Davari
Cc: mpls@ietf.org; tictoc@ietf.org
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls

My concern is that we are creating a new LSP type in MPLS
(which is a architectural change and thus needs to be very carefully
considered) without asking the question "how can I do this
in the most general way, to maximize flexibility/reuse"

I suggested an LSP type that has the properties "timestamp
and pass to application", but I think that Sasha raises a good
point about using router alert. However RA has no implicit
timestamp, so maybe we need a new type of RA that has the
properties "timestamp, and pass top application indicated by
the next label". That would be quite useful in a number
of OAM applications. There is possibly some GAL variant
of that design that should also be considered.

The application could worry about the path and where it
was necessary to skip some hops a hierarchical LSP would
accomplish that, with the specific benefit that the application
would consciously do this, and may be able to apply some
form of compensation within the network.

- Stewart

On 04/08/2013 10:40, S. Davari wrote:
> Hi Sasha
>
> Perhaps you have not understood the draft well. The main reason that the =
draft chose to use a dedicated LSP that is signaled via RSVPTE indicating i=
t is a timing LSP is to be backward compatible with non-1588 nodes. Such no=
des simply just switch the packet.
>
> Your proposal had been considered and was rejected since it was not backw=
ard compatible. Nodes receiving alert label or TTL =3D 1 send the packet to=
 CPU. You can refer to the meeting notes.
>
> Regards,
> Shahram
>
>
> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein <Alexander.Vainshtein@ec=
itele.com> wrote:
>
>> Stewart and all,
>> I concur with Stewart's statement that the draft does not define "full i=
nteraction with MPLS architecture".
>>
>> E.g., one of the objectives of the draft is to provide a technique that =
would be backward-compatible with old LSRs that cannot provide on-path supp=
ort for timing distribution, while the other objective is to make every LSR=
 on the path aware that some MPLS packets are carrying timing-related messa=
ges and hence require on-path support.
>>
>> IMHO and FWIW, the MPLS data plane architecture allows just two methods =
for making a transit LSR to provide special processing to a labeled packet:
>> - It would carry some kind of an "alert label" on top of the label stack=
 and, specifically, on top of any labels used for actual forwarding
>> OR,
>> - The TTL in the top label stack entry has been set to 1.
>>
>>
>> The draft does not follow any of these approaches.
>>
>> I must also admit that Section 12 "OAM, Control and Management" of the d=
raft looks somewhat in
>> E.g., I could not understand from the text whether BFD, LSP-Ping etc. OA=
M messages would be subjected to on-path support procedures for timing mess=
ages or not.
>>
>> My 2c,
>>      Sasha
>>
>>> -----Original Message-----
>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behal=
f Of
>>> Stewart Bryant
>>> Sent: Friday, August 02, 2013 12:01 PM
>>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
>>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.org;=
 draft-
>>> ietf-tictoc-1588overmpls@tools.ietf.org
>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
>>>
>>> SB> This draft does not seem to provide a precise definition
>>> SB> the properties of the new LSP type that it wishes to
>>> SB> define, in particular it does it define the PHB of those
>>> SB> LSPs, nor the full interaction with the MPLS
>>> SB> architecture.
>>> SB>
>>> SB> I have not tracked TICTOC for a while but I thought that
>>> SB> the original plan was to define the concept of an offset
>>> SB> into a packet to do the correction.
>>> SB>
>>> SB> It is disappointing that the opportunity was not taken
>>> SB> to define a timing shim inside the timing LSP so that
>>> SB> a time correction could be added to any packet such that
>>> SB> the MPLS system was isolated from the details of the
>>> SB> complexity of the particular time transfer type.
>>>
>>> SB> I think that much more clarify is needed in terms of
>>> SB> definition of the new LSP type, since it is unclear
>>> SB> from this text how to implement one.
>>> SB>
>>> SB> There are a lot of other MPLS services such as
>>> SB> LSP ping that need to be considered.
>>> SB>
>>> SB> Please see inline for more comments. However these
>>> SB> comments are made in the context of the text as written
>>> SB> whilst I have a fundamental concern that this approach
>>> SB> lacks an MPLS architectural soundness that need
>>> SB> greater thought with significant impact on the
>>> SB> draft.
>>>
>>> - Stewart
>>>
>>>
>>> TICTOC Working Group                                           S. Davar=
i
>>> Internet-Draft                                                   A. Ore=
n
>>> Intended status: Standards Track                          Broadcom Corp=
.
>>> Expires: December 17, 2013                                     M. Bhati=
a
>>>                                                                P. Rober=
ts
>>>                                                            Alcatel-Luce=
nt
>>>                                                                L. Monti=
ni
>>>                                                                L. Marti=
ni
>>>                                                             Cisco Syste=
ms
>>>                                                             June 15, 20=
13
>>>
>>>
>>>              Transporting Timing messages over MPLS Networks
>>>                     draft-ietf-tictoc-1588overmpls-05
>>>
>>> Abstract
>>>
>>>     This document defines the method for transporting Timing messages
>>>     such as PTP and NTP over an MPLS network.  The method allows for th=
e
>>>     easy identification of these PDUs at the port level to allow for po=
rt
>>>
>>> SB> What is a port
>>>
>>>     level processing of these PDUs in both LERs and LSRs.
>>>
>>>     The basic idea is to transport Timing messages inside dedicated MPL=
S
>>>     LSPs.  These LSPs only carry Timing messages and possibly Control a=
nd
>>>     Management packets, but they do not carry customer traffic.
>>>
>>> SB> More specifically they only carry traffic associated with the
>>> SB> timing service and its support.
>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
>>> SB> also it gets carried in a structure that causes it to get
>>> SB> timestamped.
>>>
>>>     Two methods for transporting Timing messages over MPLS are defined.
>>>
>>> SB> Perhaps the right approach is to define the new LSP type and then
>>> SB> seperately to define  the mapping of the various timing services
>>> SB> over that LSP type.
>>>
>>>     The first method is to transport Timing messages directly over the
>>>     dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
>>>     MPLS networks.  The second method is to transport Timing messages
>>>     inside a PW via Ethernet encapsulation.
>>>
>>> SB> I think that we should note that there are some
>>> SB> h/w reasons for this preference. A clean sheet approach
>>> SB> would have been to use PTP over MPLS with no intermediate
>>> SB> layers.
>>>
>>> Status of this Memo
>>>
>>>     This Internet-Draft is submitted in full conformance with the
>>>     provisions of BCP 78 and BCP 79.
>>>
>>>     Internet-Drafts are working documents of the Internet Engineering
>>>     Task Force (IETF).  Note that other groups may also distribute
>>>     working documents as Internet-Drafts.  The list of current Internet=
-
>>>     Drafts is at http://datatracker.ietf.org/drafts/current/.
>>>
>>>     Internet-Drafts are draft documents valid for a maximum of six mont=
hs
>>>     and may be updated, replaced, or obsoleted by other documents at an=
y
>>>     time.  It is inappropriate to use Internet-Drafts as reference
>>>     material or to cite them other than as "work in progress."
>>>
>>>     This Internet-Draft will expire on December 17, 2013.
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 1=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> Copyright Notice
>>>
>>>     Copyright (c) 2013 IETF Trust and the persons identified as the
>>>     document authors.  All rights reserved.
>>>
>>>     This document is subject to BCP 78 and the IETF Trust's Legal
>>>     Provisions Relating to IETF Documents
>>>     (http://trustee.ietf.org/license-info) in effect on the date of
>>>     publication of this document.  Please review these documents
>>>     carefully, as they describe your rights and restrictions with respe=
ct
>>>     to this document.  Code Components extracted from this document mus=
t
>>>     include Simplified BSD License text as described in Section 4.e of
>>>     the Trust Legal Provisions and are provided without warranty as
>>>     described in the Simplified BSD License.
>>>
>>>
>>> Table of Contents
>>>
>>>     1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . . =
 5
>>>
>>>     2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . . =
 7
>>>
>>>     3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . . =
 8
>>>
>>>     4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . . =
 9
>>>
>>>     5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . . =
12
>>>
>>>     6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . . =
13
>>>       6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . . =
13
>>>       6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . . =
13
>>>       6.3.  Other Timing Encapsulation methods . . . . . . . . . . . . =
14
>>>
>>>     7.  Timing message Processing  . . . . . . . . . . . . . . . . . . =
15
>>>
>>>     8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . . =
16
>>>
>>>     9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . =
17
>>>
>>>     10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . =
18
>>>
>>>     11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . . =
19
>>>
>>>     12. OAM, Control and Management  . . . . . . . . . . . . . . . . . =
20
>>>
>>>     13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . . =
21
>>>
>>>     14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . . =
22
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 2=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>>     15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . . =
23
>>>       15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . . =
23
>>>       15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . . =
23
>>>       15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . . =
24
>>>
>>>     16. Other considerations . . . . . . . . . . . . . . . . . . . . . =
25
>>>
>>>     17. Security Considerations  . . . . . . . . . . . . . . . . . . . =
26
>>>
>>>     18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . =
27
>>>
>>>     19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . =
28
>>>
>>>     20. References . . . . . . . . . . . . . . . . . . . . . . . . . . =
29
>>>       20.1. Normative References . . . . . . . . . . . . . . . . . . . =
29
>>>       20.2. Informative References . . . . . . . . . . . . . . . . . . =
29
>>>
>>>     Appendix 1.  Routing extensions for Timing-aware Routers . . . . . =
32
>>>
>>>     Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . . =
33
>>>
>>>     Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . =
34
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 3=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>>     The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
>>>     "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in
>>> this
>>>     document are to be interpreted as described in RFC2119 [RFC2119].
>>>
>>>     When used in lower case, these words convey their typical use in
>>>     common language, and are not to be interpreted as described in
>>>     RFC2119 [RFC2119].
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 4=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 1.  Introduction
>>>
>>>     The objective of Precision Time Protocol (PTP) and Network Timing
>>>     Protocol (NTP) are to synchronize independent clocks running on
>>>     separate nodes of a distributed system.
>>>
>>>     [IEEE-1588] defines PTP messages for frequency, phase and time
>>>     synchronization.  The PTP messages include PTP PDUs over UDP/IP
>>>     (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F =
of
>>>     [IEEE-1588]).
>>>
>>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
>>> SB> of other PTP mappings if they provide better optimisation.
>>>
>>>     This document defines mapping and transport of the PTP
>>>     messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>>>     defines several clock types: ordinary clocks, boundary clocks, end-
>>>     to-end transparent clocks, and peer-to-peer transparent clocks.
>>>     Transparent clocks require intermediate nodes to update correction
>>>     field inside PTP message that reflects the transit time in the node=
.
>>>
>>>     [RFC5905] defines NTP messages for clock and time synchronization.
>>>     The PTP messages (PDUs) are transported over UDP/IP.  This document
>>> SB> Should that be NTP messages?
>>> SB> It needs to be made clear as soon as you introduce NTP that
>>> SB> they use different time representations.
>>>
>>>     defines mapping and transport of the NTP messages defined in
>>>     [RFC5905] over MPLS networks.
>>>
>>>     One key attribute of all of these Timing messages is that the Time
>>>     stamp processing should occur as close as possible to the actual
>>>     transmission and reception at the physical port interface.  This
>>>     targets optimal time and/or frequency recovery by avoiding variable
>>>     delay introduced by queues internal to the clocks.
>>>
>>> SB> As I recall NTP has no epoch point defined, and I am not sure
>>> SB> where that point is in the case of PTP in this mapping
>>> SB> Hopefully this will get defined in due course.
>>>
>>>     To facilitate the fast and efficient recognition of Timing messages
>>>     at the port level when the Timing messages are carried over MPLS
>>>     LSPs,
>>>
>>> SB> Over a new LSP type with time optimied characteristics
>>>
>>>     this document defines the specific encapsulations that should
>>>     be used.
>>> SB> Hopefully it will also define the PHP
>>>
>>>     In addition, it can be expected that there will exist LSR/
>>>     LERs where only a subset of the physical ports will have the port-
>>>     based Timing message processing capabilities.
>>> SB> Do you need to clarify that this only works at base and not in
>>> SB> a label heirarchy.
>>>
>>>
>>>     In order to ensure
>>>     that the LSPs carrying Timing packets always enter and exit ports
>>>     with this capability, routing extensions are defined to advertise
>>>     this capability on a port basis and to allow for the establishment =
of
>>>     LSPs that only transit such ports.  While this path establishment
>>>     restriction may be applied only at the LER Ingress and/or egress
>>>     ports, it becomes more important when using transparent clock capab=
le
>>>     LSRs in the path.
>>> SB> I do not understand the implications of the last
>>> SB> sentences - starting ", it becomes"
>>>
>>>
>>>     Port based Timing message processing involves Timing message
>>>     recognition.  Once the Timing messages are recognized they can be
>>>     modified based on the reception or transmission Time-stamp.
>>>
>>>     This document provides two methods for transporting Timing messages
>>>     over MPLS.  One is applicable to MPLS environment and the other one
>>>     is applicable to MPLS/MPLS-TP environment
>>>
>>> SB> I think the sentence is incomplete.
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 5=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>>     The solution involves transporting Timing messages over dedicated
>>>     LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
>>>     carry Management and control messages, but not data plane client
>>>     traffic.
>>>
>>> SB> It is not clear why this restriction applies.
>>>
>>>     Timing LSPs can be established statically or via signaling.
>>> SB> s/statically/by provisioning/network management/
>>>
>>>     Extensions to control plane (OSPF, ISIS, etc.) is required to enabl=
e
>>>     routers to distribute their Timing processing capabilities over MPL=
S
>>>     to other routers.  However such extensions are outside the scope of
>>>     this document.
>>>
>>>     When signaling is used to setup the PTP LSP, Extensions to signalin=
g
>>> SB> is it a PTP LSP or a Timing LSP?
>>>
>>>     protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
>>>     However such extensions are outside the scope of this document.
>>>
>>> SB> for mpls-tp GMPLS is the signalling protocol
>>>
>>>     While the techniques included herein allow for the establishment of
>>>     paths optimized to include Time-stamping capable links, the
>>>     performance of the Slave clocks is outside the scope of this
>>>     document.
>>>
>>>     At the time of publishing this specification, Transparent Clocking
>>>     (TC) is only defined for PTP.  Therefore at this time any part of
>>>     this specification that talks about Transparent Clocking applies on=
ly
>>>     to PTP.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 6=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 2.  Terminology
>>>
>>>     1588: The timing and synchronization as defined by IEEE 1588.
>>>
>>> SB> I think that there is a more formal name for the 1588 group
>>> SB> that needs to be used here.
>>> SB> Also do we need to talk about 1588-200? as there is
>>> SB> an update in progress
>>>
>>>     NTP: The timing and synchronization protocol defined by IETF RFC-13=
05
>>>     and RFC-5905.
>>>
>>>     PTP: The timing and synchronization protocol used by 1588.
>>> SB> need the proper name for 1588
>>>
>>>     Master Clock: The source of 1588 timing to a set of slave clocks.
>>>
>>>     Master Port: A port on a ordinary or boundary clock that is in Mast=
er
>>>     state.  This is the source of timing toward slave ports.
>>>
>>> SB> I am not sure the reader knows what a port is
>>>
>>>     Slave Clock: A receiver of 1588 timing from a master clock.
>>>
>>>     Slave Port: A port on a boundary clock or ordinary clock that is
>>>     receiving timing from a master clock.
>>>
>>>     Ordinary Clock: A device with a single PTP port.
>>>
>>>     Transparent Clock.  A device that measures the time taken for a PTP
>>>     event message to transit the device and then updates the
>>>     correctionField of the message with this transit time.
>>>
>>>     Boundary Clock: A device with more than one PTP port.  Generally
>>>     boundary clocks will have one port in slave state to receive timing
>>>     and then other ports in master state to re-distribute the timing.
>>>
>>>     PTP LSP: An LSP dedicated to carry PTP messages
>>>
>>> SB> PTP or timing?
>>>
>>>
>>>     PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>>     messages.
>>>
>>> SB> Ah I don't think that PWE3 know what one of these is
>>>
>>>     CW: Pseudowire Control Word
>>>
>>>     LAG: Link Aggregation
>>>
>>>     ECMP: Equal Cost Multipath
>>>
>>>     CF: Correction Field, a field inside certain PTP messages (message
>>>     type 0-3)that holds the accumulative transit time inside intermedia=
te
>>>     switches
>>>
>>>     Timing messages: Timing Protocol messages that are exchanged betwee=
n
>>>     routers in order to establish a synchronized clock.
>>>
>>> SB> A number of these definitions look like copies of IEEE1588
>>> SB> definitions. We need to provide references and note the
>>> SB> priority of the IEEE base reference.
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 7=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 3.  Problem Statement
>>>
>>>     [IEEE-1588] has defined methods for transporting PTP messages over
>>>     Ethernet and IP networks.  [RFC5905] has defined the method of
>>>     transporting NTP messages over IP networks.  There is a need to
>>>     transport Timing messages over MPLS networks while supporting the
>>>     Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
>>>     functionality in the LER and LSRs in the MPLS network.
>>>
>>>     There are multiple ways of transporting Timing over MPLS.  However,
>>>     there is a requirement to limit the possible encapsulation options =
to
>>>     simplify the Timing message identification and processing required =
at
>>>     the port level.
>>>
>>>     When Timing-awareness is needed, Timing messages should not be
>>>     transported over LSPs or PWs that are carrying customer traffic
>>>     because LSRs perform Label switching based on the top label in the
>>>     stack.
>>>
>>> SB> Have you explained why?
>>>
>>>     To detect Timing messages inside such LSPs require special
>>>     hardware to do deep packet inspection at line rate.  Even if such
>>>     hardware exists, the payload can't be deterministically identified =
by
>>>     LSRs because the payload type is a context of the PW label, and the
>>>     PW label and its context are only known to the Edge routers (PEs/
>>>     LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CES=
,
>>>     etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
>>>     LSRs dont have the knowledge of whether PW Control Word (CW) is
>>>     present or not and therefore can not deterministically identify the
>>>     payload.
>>>
>>>     A generic method is defined in this document that does not require
>>>     deep packet inspection at line rate, and can deterministically
>>>     identify Timing messages.  This method can be used to detect Timing
>>>     Messages in both one-step and two-step clock implementations of
>>>     ordinary, boundary and transparent clocks.
>>>
>>> SB> Needs a ref and I am sure many MPLS specialists will not understand
>>> SB> the msg types.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 8=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 4.  Timing over MPLS Architecture
>>>
>>>     Timing messages are exchange between Timing ports on ordinary and
>>>
>>> SB> Have you defined a timing port?
>>>
>>>     boundary clocks.  Boundary clocks terminate the Timing messages and
>>>     act as master for other boundary clocks or for slave clocks.  End-t=
o-
>>>     End Transparent clocks do not terminate the Timing messages but the=
y
>>>     do modify the contents of the Timing messages as they transit acros=
s
>>>     the transparent clock.
>>>
>>>     Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Clo=
ck
>>>
>>>     (TC) could be implemented in either LERs or LSRs.
>>>
>>> SB> LER and LSR need to be expanded
>>>
>>>     An example is shown in Figure 1, where the LERs act as Ordinary Clo=
ck
>>>     (OC) and are the initiating/terminating point for Timing messages.
>>>     The ingress LER encapsulates the Timing messages in Timing LSP and
>>>     the Egress LER terminates the Timing LSP.  The LSRs act as
>>>     Transparent Clock (TC) and just update the Timing field in the Timi=
ng
>>>     messages.
>>>
>>>
>>>        +--------+     +-------+     +-------+     +-------+     +------=
--+
>>>        |Switch, |     |       |     |       |     |       |     |Switch=
, |
>>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Route=
r |
>>>        |        |     |  OC   |     |  TC   |     |  OC   |     |      =
  |
>>>        +--------+     +-------+     +-------+     +-------+     +------=
--+
>>>                       /                                 \
>>>        +-------+     /                                   \     +-------=
+
>>>        |  LER  |    /                                     \    |  LER  =
|
>>>        | Master|---/                                       \---| Slave =
|
>>>        | Clock |                                               | Clock =
|
>>>        +-------+                                               +-------=
+
>>>
>>>       Figure (1) - Deployment example 1 of timing over MPLS network
>>>
>>>     Another example is shown in Figure2, where LERs terminate the Timin=
g
>>>     messages received from switch/routers that are outside of the MPLS
>>>     network acting as OC or BC.  In this example LERs regenerate the
>>>     clock and initiate timing messages encapsulated in Timing LSP towar=
d
>>>     the MPLS network, while the LSRs act as Transparent Clock (TC) and
>>>     just update the Timing field in the Timing messages, which are
>>>     already encapsulated in Timing LSPs.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013               [Page 9=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>>       +--------+     +-------+     +-------+     +-------+     +-------=
-+
>>>       |Switch, |     |       |     |       |     |       |     |Switch,=
 |
>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router=
 |
>>>       | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC =
 |
>>>       +--------+     +-------+     +-------+     +-------+     +-------=
-+
>>>
>>>       Figure (2) - Deployment example 2 of timing over MPLS network
>>>
>>>
>>>     Another example is shown in Figure 3, where LERs do not terminate t=
he
>>>     Timing messages received from switch/routers that are outside of th=
e
>>>     MPLS network acting as OC, TC or BC.  The LERs act as TC and update
>>>     the Timing field in the Timing messages as they transit the LER,
>>>     while encapsulating them in timing LSP.  The LSRs also act as
>>>     Transparent Clock (TC) and just update the Timing field in the Timi=
ng
>>>     messages which are already encapsulated in Timing LSPs.
>>>
>>>        +--------+     +-------+     +-------+     +-------+     +------=
--+
>>>        |Switch, |     |       |     |       |     |       |     |Switch=
, |
>>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Route=
r |
>>>        |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC/=
BC|
>>>        +--------+     +-------+     +-------+     +-------+     +------=
--+
>>>
>>>      Figure (3) - Deployment example 3 of timing over MPLS network
>>>
>>>     Another example is shown in Figure 4, where LERs and LSRs support
>>>     Boundary Clocks.  A single-hop LSP is created between two adjacent
>>>     LSRs engaged in BC operation.  Other methods such as PTP transport
>>>     over Ethernet MAY be used for transporting timing messages if the
>>>     link between the two routers is Ethernet.
>>>
>>>       +--------+     +-------+     +-------+     +-------+     +-------=
-+
>>>       |Switch, |     |       |     |       |     |       |     |Switch,=
 |
>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router=
 |
>>>       | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC =
 |
>>>       +--------+     +-------+     +-------+     +-------+     +-------=
-+
>>>
>>>     Figure (4) - Deployment example 3 of timing over MPLS network
>>>
>>>     An MPLS domain MAY serve multiple customers.  In these cases the MP=
LS
>>>     domain (maintained by a service provider) may provide timing servic=
es
>>>     to multiple customers, each having their own Timing domain.
>>>
>>>     The Timing over MPLS architecture assumes full mesh of Timing LSPs
>>>     between all LERs supporting this specification.
>>>
>>> SB> Note sure this is right - the salves surely do not need to
>>> SB> exchange timing amongst themselves
>>>
>>>     It supports
>>>     Point-to- point (VPWS) and Multipoint (VPLS) services.
>>>
>>> SB> What does that mean? You do not carry user data traffic?
>>> SB> Maybe it's the ordering of the statemnets that is causing
>>> SB> confusion.
>>>
>>>     This means
>>>     that a customer may purchase a Point-to-point Timing service betwee=
n
>>>     two customer sites or a Multipoint Timing service between more than
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 10=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>>     two customer sites.
>>>
>>>     The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
>>>     This means that the Timing Multicast messages such as PTP Multicast
>>>     event messages can be transported over P2MP Timing LSP or be
>>>     replicated and transported over many P2P Timing LSPs.
>>>
>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>>> SB> nor a P2MP PW, although we are close.
>>>
>>>     Timing messages, that do not require Time stamping or Correction
>>>     Field update MAY be transported over Timing LSPs to simplify hardwa=
re
>>>     and software.
>>>
>>>     PTP Announce messages that determine the Timing LSP terminating poi=
nt
>>>     behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
>>>     to simplify hardware and software.
>>>
>>> SB> have you defined and referenced PTP announce msgs?
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 11=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 5.  Dedicated LSPs for Timing messages
>>>
>>>     Many methods have been considered for identifying the Timing messag=
es
>>>     when they are encapsulated in MPLS such as using GAL/G-ACH or a new
>>>     reserved label.  These methods were not attractive since they eithe=
r
>>>     required deep packet inspection at line rate in the intermediate LS=
Rs
>>>     or they required use of a scarce new reserved label.  Also one of t=
he
>>>     goals was to reuse existing OAM mechanisms.
>>>
>>> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
>>>
>>>     The method defined in this document can be used by LER and LSRs to
>>>     identify Timing messages in MPLS tunnels by just looking at the top
>>>     label in the MPLS label stack, which only carry Timing messages as
>>>     well as OAM, but not data plane client traffic.
>>>
>>>     Compliant implementations MUST use dedicated LSPs to carry Timing
>>>     messages over MPLS.
>>>
>>> SB> I think that we need a definition of the properies of these LSPs
>>>
>>>     These LSPs are herein referred to as "Timing
>>>     LSPs" and the labels associated with these LSPs as "Timing LSP
>>>     labels".  The Timing LSPs that runs between Ingress and Egress LERs
>>>     MUST be co-routed.  Alternatively, a single bidirectional co-routed
>>>     LSP can be used.
>>>
>>> SB> I though that you said you could use M2MP LSPs - these are not
>>> SB> bidirectional.
>>>
>>>     Co-routing of the two directions is required to limit the differenc=
e
>>>     in the delays in the Master clock to Slave clock direction compared
>>>     to the Slave clock to Master clock direction.  The Timing LSP MAY b=
e
>>>     MPLS/MPLS-TP LSP.
>>>
>>>     The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
>>>     New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
>>>     outside the scope of this document.
>>>
>>>     The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such a=
s
>>>     BFD and LSP Ping but the LSP data plane client plane traffic MUST b=
e
>>>     Timing packets only.
>>>
>>> SB> Why?
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 12=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 6.  Timing over LSP Encapsulation
>>>
>>> The encapsulations is not LSP is it?
>>>
>>>     This document defines two methods for carrying Timing messages over
>>>     MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>>     messages over Timing LSPs, and the second method, is carrying
>>>     Ethernet encapsulated Timing messages over Ethernet PWs inside Timi=
ng
>>>     LSPs.
>>>
>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>>
>>>     The simplest method of transporting Timing messages over MPLS is to
>>>     encapsulate Timing PDUs in UDP/IP and then encapsulate them in Timi=
ng
>>>     LSP.  This format is shown in Figure 4.
>>>
>>>
>>>                      +----------------------+
>>>                      |   Timing LSP Label   |
>>>                      +----------------------+
>>>                      |        IPv4/6        |
>>>                      +----------------------+
>>>                      |         UDP          |
>>>                      +----------------------+
>>>                      |     Timing PDU       |
>>>                      +----------------------+
>>>
>>>        Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>>
>>>
>>>     This encapsulation is very simple and is useful when the network
>>>     between Timing Master Clock and Slave Clock is MPLS network.
>>>
>>> SB> Simple is a judgement call
>>>
>>>     In order for an LER/LSR to process Timing messages, the Timing LSP
>>>     Label must be at the top label of the label stack.  The LER/LSR MUS=
T
>>>     know that the Timing LSP Label is used for carrying Timing messages=
.
>>>     This can be accomplished via static configuration or via RSVP-TE
>>>     signaling.
>>>
>>>     The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>>     [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>>>     [RFC5905].
>>>
>>> 6.2.  Timing over PW Encapsulation
>>>
>>>     Another method of transporting Timing over MPLS networks is by
>>>     encapsulating Timing PDUs in PW which in turn is transported over
>>>     Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
>>>     shown in Fig 5(A) MUST be used and the Ethernet encapsulation of PT=
P
>>>     MUST follow Annex F of [IEEE-1588].
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 13=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>>     The RAW mode or Tagged mode defined in [RFC4448] MAY be used and th=
e
>>>     Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>>>     Timing over PW encapsulation MUST use the Control Word (CW) as
>>>     specified in [RFC4448] to ensure proper detection of PTP messages
>>>     inside the MPLS packets for Timing over LSP and Timing over PW
>>>     encapsulation.
>>>
>>> SB> That needs explanation
>>>
>>>     The use of Sequence Number in the CW is optional.
>>>
>>> SB> Given that s/n are never in practice deployed, you could probably
>>> SB> simplify things by sayig that they are not used.
>>>
>>>     Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over =
PW
>>>     (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>>
>>>                      +----------------+  +----------------+
>>>                       |Timing LSP Label|  |Timing LSP Label|
>>>                       +----------------+  +----------------+
>>>                       |    PW Label    |  |    PW Label    |
>>>                       +----------------+  +----------------+
>>>                       |  Control Word  |  |      IP        |
>>>                       +----------------+  +----------------+
>>>                       |    Ethernet    |  |      UDP       |
>>>                       |     Header     |  +----------------+
>>>                       +----------------+  |   Timing PDU   |
>>>                       |S-VLAN(Optional)|  |                |
>>>                       +----------------+  +----------------+
>>>                       |C-VLAN(Optional)|        (B)
>>>                       +----------------+
>>>                       |   Timing PDU   |
>>>                       |                |
>>>                       +----------------+
>>>                              (A)
>>>
>>>                Figure (5) - Timing over PW Encapsulations
>>>
>>>     In order for an LSR to process PTP messages, the top label of the
>>>     label stack (the Tunnel Label) MUST be a Timing label.
>>>
>>> S> You said that before.
>>>
>>> 6.3.  Other Timing Encapsulation methods
>>>
>>>     In future other timing encapsulation methods may be introduced, suc=
h
>>>     as a new shim header after the Bottom of Stack to carry the Timing
>>>     information.  Such new encapsulations are outside the scope of this
>>>     document.
>>>
>>>
>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>>> SB> out of the definition of the LSP
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 14=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> SB> I think we need a section on LSP processing
>>>
>>> 7.  Timing message Processing
>>>
>>>     Each Timing protocol such as PTP and NTP, define their set of Timin=
g
>>>     messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>>>     FOLLOW_UP, etc messages.
>>>
>>>     Some of the Timing messages require time stamping or correction fie=
ld
>>>     update at port level and some dont.  It is the job of the LER/LSR t=
o
>>>     parse the timing message and find out the type of the Timing messag=
e
>>>     and decide whether and how to Time- stamp it (e.g., BC) or update
>>>     correction field(e.g., TC).
>>>
>>>
>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
>>> SB> function rather than the LER?
>>>
>>>     For example the following PTP messages (called Event messages)
>>>     require time-stamping or correction field update:
>>>
>>>     o  SYNC
>>>
>>>     o  DELAY_REQ (Delay Request)
>>>
>>>     o  PDELAY_REQ (Peer Delay Request)
>>>
>>>     o  PDELAY_RESP (Peer Delay Response)
>>>
>>>     SYNC and DELAY_REQ are exchanged between Master Clock and Slave Clo=
ck
>>>     and MUST be transported over PTP LSPs.  PDELAY_REQ and PDELAY_RESP
>>>     are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>>>     Boundary, or Transparent) and SHOULD be transported over single hop
>>>     PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
>>>     and PDELAY_RESP_FOLLOW_UP messages MUST also be transported over
>>> the
>>>     PTP LSPs.
>>>
>>>     For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
>>>     transported over two PTP LSPs that are in opposite directions.  The=
se
>>>     PTP LSPs, which are in opposite directions MUST be congruent and co=
-
>>>     routed.  Alternatively, a single bidirectional co-routed LSP can be
>>>     used.
>>>
>>>     Except as indicated above for the two-step PTP clocks, Non-Event PT=
P
>>>     message types do not need to be processed by intermediate routers.
>>>     These message types MAY be carried in PTP Tunnel LSPs.
>>>
>>> SB> Are you saying that a timing P router has to be msg type sensitive?
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 15=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 8.  Protection and Redundancy
>>>
>>>
>>> SB> This is a bit of a jump - I don't know how the LSP itself works yet=
!
>>>
>>>     In order to ensure continuous uninterrupted operation of slave
>>>     clocks, usually as a general practice, slave clocks (or ports) trac=
k
>>>     redundant master clocks.
>>>
>>>     It is the responsibility of the network operator to ensure that
>>>     physically disjoint Timing LSPs are established between a slave clo=
ck
>>>     (or port) and redundant master clocks (or ports).
>>>
>>>     When a slave clock (or port) listens to redundant master clocks or
>>>     ports, any prolonged Timing LSP outage will trigger the slave clock
>>>     or port to switch to a redundant master clock or port.
>>>
>>>     LSP/PW protection such as Linear protection Switching (1:1, 1+1),
>>>     Ring protection switching or MPLS Fast Reroute (FRR) generally swit=
ch
>>>     alternative path that usually cause a change in delay, which if
>>>     undetected by slave clock can reduce accuracy of the slave clock.
>>>
>>>     Therefore protection switching MAY be used, as long as phase jumps
>>>     upon switchover due to differences in path latency are detected and
>>>     compensated for (such compensation not being required if BCs or pee=
r-
>>>     peer TCs are used throughout).
>>>
>>>     Note that any protection or reroute mechanism that adds additional
>>>     MPLS label to the label stack, such as Facility Backup Fast Reroute=
,
>>>     MUST ensure that the pushed label is also a Timing Label to ensure
>>>     recognition of the MPLS frame as containing Timing messages, as it
>>>     transits the backup path.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 16=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 9.  ECMP
>>>
>>>     To ensure the optimal operation of slave clocks and avoid error
>>>     introduced by forward and reverse path delay asymmetry, the physica=
l
>>>     path for Timing messages from master clock to slave Clock and vice
>>>     versa must be the same for all Event Timing messages listed in
>>>     section 7.
>>>
>>>     Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
>>>     Multipath).
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 17=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 10.  PHP
>>>
>>>     To ensure that the label on the top of the label stack is the Timin=
g
>>>     LSP Label, PHP MUST not be used.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 18=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 11.  Entropy
>>>
>>>     To ensure all Timing messages in a Timing LSP take the same path,
>>>     Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>>     Entropy Label MUST NOT be used for the PWs that are carried inside
>>>     Timing LSP [RFC6391].
>>>
>>> SB> This is incorrect - you mean that all msgs of the same timing
>>> SB> flow need to have the same EL value.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 19=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 12.  OAM, Control and Management
>>>
>>>     In order to monitor Timing LSPs and their encapsulated PWs, they MU=
ST
>>>     be able to carry OAM and management messages.  These management
>>>     messages MUST be differentiated from Timing messages via already
>>>     defined IETF methods.
>>>
>>>     For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
>>>     over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>>>     Management protocols can easily be identified by the UDP Destinatio=
n
>>>     Port number or by GAL/G-ACH respectively.
>>>
>>>     Also BFD, LSP-Ping and other management messages MAY run over the P=
Ws
>>>     encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3 =
or
>>>     4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is goi=
ng
>>>     to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1) =
or
>>>     GAL-ACH are used to identify such management messages.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 20=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 13.  QoS Considerations
>>>
>>>     In network deployments where not every LSR/LER is Timing-aware, it =
is
>>>     important to reduce the impact of the non-Timing-aware LSR/LERs on
>>>     the timing recovery in the slave clock.  The Timing messages are ti=
me
>>>     critical and must be treated with the highest priority.  Therefore
>>>     Timing over MPLS messages must be treated with the highest priority
>>>     in the routers.  This can be achieved by proper setup of Timing LSP=
s.
>>>
>>>     It is recommended that the Timing LSPs are setup or configured
>>>     properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC2697=
]
>>>     for drop eligibility.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 21=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 14.  FCS and Checksum Recalculation
>>>
>>>     When time-stamp generation and timing packet adjustment is performe=
d
>>>     near the physical port hardware, the process MUST include
>>>     recalculation of the Ethernet FCS.
>>>
>>> SB> The above is confusing - an LSR always recomputes the link layer
>>> SB> CRC which may or may not be Ethernet.
>>>
>>>     Also FCS retention for the
>>>     payload Ethernet described in [RFC4720] MUST NOT be used.
>>>
>>>     For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
>>>     may be required as per UDP transport standards.
>>>
>>> SB> You really need to be working on getting the IPv6 C?S computation
>>> SB> removed from PTP msgs.
>>>
>>>     When UDP checksum is used, each Timing-aware LER/LSR must either
>>>     incrementally update the UDP checksum after Time stamping or
>>>     Correction Field update or verify the UDP checksum on reception fro=
m
>>>     upstream and recalculate the checksum completely on transmission to
>>>     downstream node after Time stamping or Correction Field update.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 22=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 15.  Behavior of LER/LSR
>>>
>>>     Timing-capable/aware LERs and LSRs are routers that have one or mor=
e
>>>
>>> SB> You mean physical interfaces?
>>>
>>>     interfaces that can perform Timing operations (OC/BC/TC) on Timing
>>>     packets and are configured to do so.  Timing-capable/aware LERs and
>>>     LSRs can advertise their Timing-capability per-interface via contro=
l
>>>     plane such as OSPF or IS-IS.
>>> SB> ISIS and OSPF are routing protocols.
>>>
>>>    The Timing-capable/aware LERs can then
>>>     signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timin=
g
>>>     capability of LER and LSRs may be configured in a centralized
>>>     controller and the Timing LSP may be setup using manual configurati=
on
>>>     or other methods such as SDN.
>>>
>>> SB> it can also be configured individually rather then through
>>> SB> a cebtral controllwe
>>>
>>> 15.1.  Behavior of Timing-capable/aware LER
>>>
>>>     When a Timing-capable/aware LER behaves as a Transparent clock and
>>>     receives a Timing message from a Timing-capable/aware non-MPLS
>>>     interface, the LER updates the Correction Field (CF) and encapsulat=
es
>>>     and forwards the timing message over previously established Timing
>>>     LSP.
>>>
>>> SB> You need to call out the details so that people properly
>>> SB> understand the definition of the new LSP.
>>>
>>>     Also when a Timing message is received from a Timing-capable/
>>>     aware MPLS interface, LER updates the Correction Filed (CF) and
>>>     decapsulates the MPLS encapsulation and forwards the timing message
>>>     to a non-MPLS interface.
>>>
>>>     When a Timing-capable/aware LER behaves as a Boundary clock and
>>>     receives a Timing message from a Timing-capable/aware non MPLS
>>>     interface, the LER Timestamps the Timing packet and sends it to the
>>>     LERs Boundary clock processing module.  Also when a Timing message =
is
>>>     received from a Timing- capable/aware MPLS interface, the LER
>>>     Timestamps the Timing packet and sends it to the LERs Boundary cloc=
k
>>>     processing module.
>>>
>>>     When a Timing-capable/aware LER behaves as an Ordinary Clock toward
>>>     the MPLS network, and receives a Timing message from a Timing-
>>>     capable/aware MPLS interface, the LER Timestamps the Timing packet
>>>     and sends it to the LERs Ordinary clock processing module.
>>>
>>> 15.2.  Behavior of Timing-capable/aware LSR
>>>
>>>     When a Timing-capable/aware LSR behaves as a Transparent clock and
>>>     receives a Timing message from a Timing-capable/aware MPLS interfac=
e,
>>>     The LSR updates the Correction Filed (CF) and forwards the timing
>>>     message over another MPLS interface.
>>>
>>>     When a Timing-capable/aware LSR behaves as a Boundary clock and
>>>     receives a Timing message from a Timing-capable/aware MPLS interfac=
e.
>>>     The LSR performs the functions of a Boundary Clock in terminating t=
he
>>>     received Timing message and re-generating a new timing message over
>>>     another (or the same) MPLS interface.
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 23=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>>
>>>     It is most beneficial when all LSRs in the path of a Timing LSP be
>>>     timing-Capable/aware LSRs.  This would ensure the highest quality
>>>     time and clock synchronization by Timing Slave Clocks.  However, th=
is
>>>     specification does not mandate that all LSRs in path of a Timing LS=
P
>>>     be Timing- capable/aware.
>>>
>>>     Non-Timing-capable/aware LSRs just switch the packets encapsulated =
in
>>>     Timing LSPs and dont perform any Timing operation (TC or BC).
>>>     However as explained in QoS section the Timing over MPLS packets MU=
ST
>>>     be still be treated with the highest priority based on their Traffi=
c
>>>     Class (TC) marking.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 24=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 16.  Other considerations
>>>
>>>     [IEEE-1588] defines an optional peer-to-peer Transparent clocking
>>>     that requires peer delay measurement between two adjacent Timing-
>>>     capable/ aware routers/switches.  Peer delay measurement messages
>>>     need to be time stamped and terminated by the Timing-capable/aware
>>>     routers/ switches.  This means that two adjacent LSRs may be engage=
d
>>>     in a peer delay measurement.
>>>
>>>     For transporting such peer delay measurement messages a single-hop
>>>     LSP SHOULD to be created between the two adjacent LSRs engaged in
>>>     peer delay measurement to carry peer delay measurement messages.
>>>     Other methods such as PTP transport over Ethernet MAY be used for
>>>     transporting peer delay measurement messages if the link between th=
e
>>>     two routers is Ethernet.
>>>
>>>     In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ wa=
re
>>>     routers/switches MUST maintain a list of all the neighbors it needs
>>>     to send a PDelay_Req to, where each neighbor corresponds to a timin=
g
>>>     LSP.
>>>
>>>     The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as l=
ong
>>>     as either the Explicit Null label is the bottom of stack label
>>>     (applicable only to UDP/IP encapsulation) or the label below the
>>>     Explicit Null label is a PTP label.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 25=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 17.  Security Considerations
>>>
>>>     MPLS PW security considerations in general are discussed in [RFC398=
5]
>>>     and [RFC4447],and those considerations also apply to this document.
>>>
>>>     An experimental security protocol is defined in [IEEE-1588].The PTP
>>>     security extension and protocol provides group source authenticatio=
n,
>>>     message integrity, and replay attack protection for PTP messages.
>>>
>>>     When the MPLS network (provider network) serves multiple customers,
>>>     it is important to maintain and process each customers clock and
>>>     Timing messages separately from other customers to ensure there is =
no
>>>     cross- customer effect.  For example if an LER BC is synchronized t=
o
>>>     a specific grandmaster, belonging to customer A, then the LER MUST
>>>     use that BC clock only for customer A to ensure that customer A
>>>     cannot attack other customers by manipulating its time.
>>>
>>>     Timing messages MAY be encrypted or authenticated, provided that th=
e
>>>     LERs/LSRs that are Timing capable/aware can authenticate/ decrypt t=
he
>>>     timing messages.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 26=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 18.  Acknowledgements
>>>
>>>     The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizrah=
i,
>>>     Stefano Ruffini, Peter Meyer, and other members of IETF for reviewi=
ng
>>>     and providing feedback on this draft.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 27=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 19.  IANA Considerations
>>>
>>>     There are no IANA requirements in this specification.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 28=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 20.  References
>>>
>>> 20.1.  Normative References
>>>
>>>     [IEEE-1588]
>>>                IEEE 1588-2008, "IEEE Standard for a Precision Clock
>>>                Synchronization Protocol for Networked Measurement and
>>>                Control Systems".
>>>
>>>     [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>>                Requirement Levels", BCP 14, RFC 2119, March 1997.
>>>
>>>     [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
>>>                Edge (PWE3) Architecture", RFC 3985, March 2005.
>>>
>>>     [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discover=
y
>>>                Proxies (ND Proxy)", RFC 4389, April 2006.
>>>
>>>     [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
>>>                Heron, "Pseudowire Setup and Maintenance Using the Label
>>>                Distribution Protocol (LDP)", RFC 4447, April 2006.
>>>
>>>     [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>>                "Encapsulation Methods for Transport of Ethernet over MP=
LS
>>>                Networks", RFC 4448, April 2006.
>>>
>>>     [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>>                Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>>                Retention", RFC 4720, November 2006.
>>>
>>>     [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
>>>                Connectivity Verification (VCCV): A Control Channel for
>>>                Pseudowires", RFC 5085, December 2007.
>>>
>>>     [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detectio=
n
>>>                (BFD)", RFC 5880, June 2010.
>>>
>>>     [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
>>>                "Bidirectional Forwarding Detection (BFD) for MPLS Label
>>>                Switched Paths (LSPs)", RFC 5884, June 2010.
>>>
>>> 20.2.  Informative References
>>>
>>>     [I-D.ietf-pwe3-fat-pw]
>>>                Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan=
,
>>>                J., and S. Amante, "Flow Aware Transport of Pseudowires
>>>                over an MPLS Packet Switched Network",
>>>                draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 29=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>>     [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
>>>                system routeing information exchange protocol for use in
>>>                conjunction with the Protocol for providing the
>>>                Connectionless-mode Network Service (ISO 8473)".
>>>
>>>     [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
>>>                dual environments", RFC 1195, December 1990.
>>>
>>>     [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
>>>
>>>     [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
>>>                Marker", RFC 2697, September 1999.
>>>
>>>     [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boudec=
,
>>>                J., Courtney, W., Davari, S., Firoiu, V., and D.
>>>                Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>>                Behavior)", RFC 3246, March 2002.
>>>
>>>     [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineeri=
ng
>>>                (TE) Extensions to OSPF Version 2", RFC 3630,
>>>                September 2003.
>>>
>>>     [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
>>>                System (IS-IS) Extensions for Traffic Engineering (TE)",
>>>                RFC 3784, June 2004.
>>>
>>>     [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
>>>                Shaffer, "Extensions to OSPF for Advertising Optional
>>>                Router Capabilities", RFC 4970, July 2007.
>>>
>>>     [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
>>>                System to Intermediate System (IS-IS) Extensions for
>>>                Advertising Router Information", RFC 4971, July 2007.
>>>
>>>     [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
>>>                Topology (MT) Routing in Intermediate System to
>>>                Intermediate Systems (IS-ISs)", RFC 5120, February 2008.
>>>
>>>     [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>>                Engineering", RFC 5305, October 2008.
>>>
>>>     [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>>                "Traffic Engineering Extensions to OSPF Version 3",
>>>                RFC 5329, September 2008.
>>>
>>>     [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
>>>                for IPv6", RFC 5340, July 2008.
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 30=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>>     [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Netwo=
rk
>>>                Time Protocol Version 4: Protocol and Algorithms
>>>                Specification", RFC 5905, June 2010.
>>>
>>>     [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan=
,
>>>                J., and S. Amante, "Flow-Aware Transport of Pseudowires
>>>                over an MPLS Packet Switched Network", RFC 6391,
>>>                November 2011.
>>>
>>>     [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
>>>                L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
>>>                RFC 6790, November 2012.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 31=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 1.  Routing extensions for Timing-aware Routers
>>>
>>>     MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] an=
d
>>>     IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (TE=
)
>>>     link information used for constraint-based routing.
>>>
>>>     Indeed, it is useful to advertise data plane TE router link
>>>     capabilities, such as the capability for a router to be Timing-awar=
e.
>>>     This capability MUST then be taken into account during path
>>>     computation to prefer or even require links that advertise themselv=
es
>>>     as Timing-aware.  In this way the path can ensure the entry and exi=
t
>>>     points into the LERs and, if desired, the links into the LSRs are
>>>     able to perform port based time-stamping thus minimizing their impa=
ct
>>>     on the performance of the slave clock.
>>>
>>>     extensions are required to OSPF and IS-IS in order to advertise
>>>     Timing-aware capabilities of a link.  Such extensions are outside t=
he
>>>     scope of this document; however such extension SHOULD be able to
>>>     signal the following information per Router Link:
>>>
>>>     o  Capable of processing PTP, NTP or other Timing flows
>>>
>>>     o  Capable of performing Transparent Clock operation
>>>
>>>     o  Capable of performing Boundary Clock operation
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 32=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> 2.  Signaling Extensions for Creating Timing LSPs
>>>
>>>     RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-=
TE
>>>     is used to setup Timing LSPs, some information that indicates that
>>>     the LSP is carrying Timing flows MUST be included in the new
>>>     Extensions to RSVP-TE:
>>>
>>>     The following information MAY also be included in the new Extension=
s
>>>     to RSVP-TE:
>>>
>>>     o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
>>>        field
>>>
>>>     o  Number of VLANs in case of PW encapsulation
>>>
>>>     o  Timestamp field Type
>>>
>>>        *  Correction Field, Timestamp
>>>
>>>     o  Timestamp Field format
>>>
>>>        *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
>>>           NTP, etc.
>>>
>>>     Note that in case the above optional information is signaled with
>>>     RSVP-TE for a Timing LSP, all the Timing packets carried in that LS=
P
>>>     must have the same signaled characteristics.  For example if
>>>     Timestamp format is signaled as 64-bit PTPv1, then all Timing packe=
ts
>>>     must use 64-bit PTPv1 time-stamp.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 33=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>> Authors' Addresses
>>>
>>>     Shahram Davari
>>>     Broadcom Corp.
>>>     San Jose, CA  95134
>>>     USA
>>>
>>>     Email: davari@broadcom.com
>>>
>>>
>>>     Amit Oren
>>>     Broadcom Corp.
>>>     San Jose, CA  95134
>>>     USA
>>>
>>>     Email: amito@broadcom.com
>>>
>>>
>>>     Manav Bhatia
>>>     Alcatel-Lucent
>>>     Bangalore,
>>>     India
>>>
>>>     Email: manav.bhatia@alcatel-lucent.com
>>>
>>>
>>>     Peter Roberts
>>>     Alcatel-Lucent
>>>     Kanata,
>>>     Canada
>>>
>>>     Email: peter.roberts@alcatel-lucent.com
>>>
>>>
>>>     Laurent Montini
>>>     Cisco Systems
>>>     San Jose CA
>>>     USA
>>>
>>>     Email: lmontini@cisco.com
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 34=
]
>>> Internet-Draft        Transporting Timing over MPLS            June 201=
3
>>>
>>>
>>>     Luca
>>>     Cisco Systems
>>>     San Jose CA
>>>     USA
>>>
>>>     Email: lmartini@cisco.com
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> Davari, et al.          Expires December 17, 2013              [Page 35=
]
>>> --
>>> For corporate legal information go to:
>>>
>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>
>>> _______________________________________________
>>> TICTOC mailing list
>>> TICTOC@ietf.org
>>> https://www.ietf.org/mailman/listinfo/tictoc
>>
>> This e-mail message is intended for the recipient only and contains info=
rmation which is CONFIDENTIAL and which may be proprietary to ECI Telecom. =
If you have received this transmission in error, please inform us by e-mail=
, phone or fax, and then delete the original and all copies thereof.
>>
>> _______________________________________________
>> mpls mailing list
>> mpls@ietf.org
>> https://www.ietf.org/mailman/listinfo/mpls
> .
>


--=20
For corporate legal information go to:

http://www.cisco.com/web/about/doing_business/legal/cri/index.html

_______________________________________________
TICTOC mailing list
TICTOC@ietf.org
https://www.ietf.org/mailman/listinfo/tictoc



From Alexander.Vainshtein@ecitele.com  Tue Aug  6 00:38:55 2013
Return-Path: <Alexander.Vainshtein@ecitele.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 986D111E80E8 for <tictoc@ietfa.amsl.com>; Tue,  6 Aug 2013 00:38:55 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.102
X-Spam-Level: 
X-Spam-Status: No, score=-3.102 tagged_above=-999 required=5 tests=[AWL=-0.900, BAYES_00=-2.599, MIME_QP_LONG_LINE=1.396, RCVD_IN_DNSWL_LOW=-1, UNPARSEABLE_RELAY=0.001]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id q8nyVXg5UmBy for <tictoc@ietfa.amsl.com>; Tue,  6 Aug 2013 00:38:51 -0700 (PDT)
Received: from mail1.bemta3.messagelabs.com (mail1.bemta3.messagelabs.com [195.245.230.168]) by ietfa.amsl.com (Postfix) with ESMTP id B765921F9E26 for <tictoc@ietf.org>; Tue,  6 Aug 2013 00:38:46 -0700 (PDT)
Received: from [85.158.138.51:65372] by server-8.bemta-3.messagelabs.com id 61/E6-18692-508A0025; Tue, 06 Aug 2013 07:38:45 +0000
X-Env-Sender: Alexander.Vainshtein@ecitele.com
X-Msg-Ref: server-15.tower-174.messagelabs.com!1375774719!28438247!4
X-Originating-IP: [147.234.242.234]
X-StarScan-Received: 
X-StarScan-Version: 6.9.11; banners=-,-,-
X-VirusChecked: Checked
Received: (qmail 1351 invoked from network); 6 Aug 2013 07:38:44 -0000
Received: from ilptbmg01-out.ecitele.com (HELO ilptbmg01-out.ecitele.com) (147.234.242.234) by server-15.tower-174.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 6 Aug 2013 07:38:44 -0000
X-AuditID: 93eaf2e7-b7ef66d00000140d-77-5200a802b343
Received: from ILPTWPVEXCA01.ecitele.com ( [172.31.244.224]) by ilptbmg01-out.ecitele.com (Symantec Messaging Gateway) with SMTP id 01.18.05133.208A0025; Tue,  6 Aug 2013 10:38:42 +0300 (IDT)
Received: from ILPTWPVEXMB01.ecitele.com ([fe80::f152:8eaf:8fb0:a5da]) by ILPTWPVEXCA01.ecitele.com ([fe80::ac15:43ab:d541:dfa7%12]) with mapi id 14.03.0123.003; Tue, 6 Aug 2013 10:38:41 +0300
From: Alexander Vainshtein <Alexander.Vainshtein@ecitele.com>
To: Shahram Davari <davari@broadcom.com>
Thread-Topic: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
Thread-Index: AQHOkhqJ+XHALQ/MzkWnZzr9BEl1M5mHyTyg
Date: Tue, 6 Aug 2013 07:38:40 +0000
Message-ID: <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com>
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com>
In-Reply-To: <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-originating-ip: [10.4.35.10]
Content-Type: text/plain; charset="us-ascii"
content-transfer-encoding: quoted-printable
MIME-Version: 1.0
X-CFilter-Loop: Reflected
X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFprCKsWRmVeSWpSXmKPExsUy+dWnL7pMKxiCDBZ90bVY3+tpce7pHEaL v8097A7MHrPun2XzmPJ7I6vHkiU/mQKYoxoYbRLz8vJLEktSFVJSi5NtlQKKMssSkyuVFDJT bJUMlRQKchKTU3NT80pslRILClLzUpTsuBQwgA1QWWaeQmpecn5KZl66rZJnsL+uhYWppa6h kp2asqGxNVdIRmaxQqpubmJmjkJuanFxYnqqAlAkYQtzxrrnvewFu+6yVPz9/JO5gXH3MuYu Rk4OCQETifmPtjJC2GISF+6tZ+ti5OIQEjjIKHF6US8zhHOEUeLQmVMsIFVsArYSm1bfZQOx RQQ0JA7eugI2iVlgOqPE24cWXYwcHMIC7hKHf/hBlHhIdE49zwhhG0m0HXkPZrMIqEicfnEG bCSvQIDEhjW/mCB2dTFJHP8zD6yIUyBc4viO62C7GIGu+35qDRPELnGJW0/mM0FcLSCxZM95 qG9EJV4+/scKYctJPHkCcTOzgI7Egt2f2CBsbYllC18zQywWlDg58wkLRL2kxMEVN1gmMIrP QrJiFpL2WUjaZyFpX8DIsopRNDOnoCQpN93AUC81ObMkNSdVLzk/dxMjJL0838H4a77KIUZX oMcnMktxJ+cD01NeSbyxgQFujpI47/KGcH8hgXRg0slOTS1ILYovKs1JLT7EyMTBKdXAyMLG p6Ehbb/umcny2PyDNz+eOb+kaN2eSub63H3vc/PeXUz0/Sie95Pphq3Nva1u+s3vQqq5605w HZzYp7M1adW5hW8/dT1IX6x65cRNG5sHrB27Za7OdfptZm73UO6+0uuA2O9X99/JX5pSqCvF 9KhC4PWT8p/93nvyzSdfC62z2lpQYsUur8RSnJFoqMVcVJwIAIX9DkInAwAA
Cc: "mpls@ietf.org" <mpls@ietf.org>, "S. Davari" <davarish@yahoo.com>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 06 Aug 2013 07:38:55 -0000

Shahram,
I agree with you that the routers that do not recognize a Timing Alert label=
 would send the packet to the CPU. (This would be easy to arrange since we a=
re speaking about a reserved label.)

I do not, however, see this as a serious issue, because the SW running on th=
is CPU would (hopefully) be upgradable to handle the packet correctly i.e.,=
 to forward it to where it should be forwarded and, if it is forwarded as a=
 labeled packet, to prepend the Timing Alert Label on top of its label stack=
. It could even record the residence time (as observed by the CPU in the pro=
per place in the packet. This would mean that introducing additional error t=
o whatever timing information is associated with this packet - but this is w=
hat you should anyway expect if there are non-compliant routers on your path=
, right?

Regards,
     Sasha

> -----Original Message-----
> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf O=
f
> Shahram Davari
> Sent: Monday, August 05, 2013 11:29 PM
> To: stbryant@cisco.com; S. Davari
> Cc: mpls@ietf.org; tictoc@ietf.org
> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
> 
> Hi Stewart,
> 
> Your suggestion of " I suggested an LSP type that has the properties "time=
stamp
> and pass to application", is a subset of the existing draft and should wor=
k. It
> basically limits the time stamping and correction field update to LERs (wh=
ile the
> draft supports time stamping at LER and LSR).
> 
> However Sasha's suggestion of using a reserved label (RAL, GAL or any othe=
r
> reserved label) does not satisfy one of the major requirements, which is
> backward compatibility. Routers that don't understand this reserved label=
 will
> drop or copy to CPU such packets.
> 
> Regards,
> Shahram
> 
> -----Original Message-----
> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf O=
f
> Stewart Bryant
> Sent: Monday, August 05, 2013 2:29 AM
> To: S. Davari
> Cc: mpls@ietf.org; tictoc@ietf.org
> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
> 
> My concern is that we are creating a new LSP type in MPLS
> (which is a architectural change and thus needs to be very carefully
> considered) without asking the question "how can I do this
> in the most general way, to maximize flexibility/reuse"
> 
> I suggested an LSP type that has the properties "timestamp
> and pass to application", but I think that Sasha raises a good
> point about using router alert. However RA has no implicit
> timestamp, so maybe we need a new type of RA that has the
> properties "timestamp, and pass top application indicated by
> the next label". That would be quite useful in a number
> of OAM applications. There is possibly some GAL variant
> of that design that should also be considered.
> 
> The application could worry about the path and where it
> was necessary to skip some hops a hierarchical LSP would
> accomplish that, with the specific benefit that the application
> would consciously do this, and may be able to apply some
> form of compensation within the network.
> 
> - Stewart
> 
> On 04/08/2013 10:40, S. Davari wrote:
> > Hi Sasha
> >
> > Perhaps you have not understood the draft well. The main reason that the
> draft chose to use a dedicated LSP that is signaled via RSVPTE indicating=
 it is a
> timing LSP is to be backward compatible with non-1588 nodes. Such nodes
> simply just switch the packet.
> >
> > Your proposal had been considered and was rejected since it was not
> backward compatible. Nodes receiving alert label or TTL =3D 1 send the pac=
ket to
> CPU. You can refer to the meeting notes.
> >
> > Regards,
> > Shahram
> >
> >
> > On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein
> <Alexander.Vainshtein@ecitele.com> wrote:
> >
> >> Stewart and all,
> >> I concur with Stewart's statement that the draft does not define "full
> interaction with MPLS architecture".
> >>
> >> E.g., one of the objectives of the draft is to provide a technique that=
 would
> be backward-compatible with old LSRs that cannot provide on-path support f=
or
> timing distribution, while the other objective is to make every LSR on the=
 path
> aware that some MPLS packets are carrying timing-related messages and henc=
e
> require on-path support.
> >>
> >> IMHO and FWIW, the MPLS data plane architecture allows just two methods
> for making a transit LSR to provide special processing to a labeled packet=
:
> >> - It would carry some kind of an "alert label" on top of the label stac=
k and,
> specifically, on top of any labels used for actual forwarding
> >> OR,
> >> - The TTL in the top label stack entry has been set to 1.
> >>
> >>
> >> The draft does not follow any of these approaches.
> >>
> >> I must also admit that Section 12 "OAM, Control and Management" of the
> draft looks somewhat in
> >> E.g., I could not understand from the text whether BFD, LSP-Ping etc. O=
AM
> messages would be subjected to on-path support procedures for timing
> messages or not.
> >>
> >> My 2c,
> >>      Sasha
> >>
> >>> -----Original Message-----
> >>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Beha=
lf
> Of
> >>> Stewart Bryant
> >>> Sent: Friday, August 02, 2013 12:01 PM
> >>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
> >>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.org=
;
> draft-
> >>> ietf-tictoc-1588overmpls@tools.ietf.org
> >>> Cc: mpls@ietf.org; tictoc@ietf.org
> >>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
> >>>
> >>> SB> This draft does not seem to provide a precise definition
> >>> SB> the properties of the new LSP type that it wishes to
> >>> SB> define, in particular it does it define the PHB of those
> >>> SB> LSPs, nor the full interaction with the MPLS
> >>> SB> architecture.
> >>> SB>
> >>> SB> I have not tracked TICTOC for a while but I thought that
> >>> SB> the original plan was to define the concept of an offset
> >>> SB> into a packet to do the correction.
> >>> SB>
> >>> SB> It is disappointing that the opportunity was not taken
> >>> SB> to define a timing shim inside the timing LSP so that
> >>> SB> a time correction could be added to any packet such that
> >>> SB> the MPLS system was isolated from the details of the
> >>> SB> complexity of the particular time transfer type.
> >>>
> >>> SB> I think that much more clarify is needed in terms of
> >>> SB> definition of the new LSP type, since it is unclear
> >>> SB> from this text how to implement one.
> >>> SB>
> >>> SB> There are a lot of other MPLS services such as
> >>> SB> LSP ping that need to be considered.
> >>> SB>
> >>> SB> Please see inline for more comments. However these
> >>> SB> comments are made in the context of the text as written
> >>> SB> whilst I have a fundamental concern that this approach
> >>> SB> lacks an MPLS architectural soundness that need
> >>> SB> greater thought with significant impact on the
> >>> SB> draft.
> >>>
> >>> - Stewart
> >>>
> >>>
> >>> TICTOC Working Group                                           S. Dava=
ri
> >>> Internet-Draft                                                   A. Or=
en
> >>> Intended status: Standards Track                          Broadcom Cor=
p.
> >>> Expires: December 17, 2013                                     M. Bhat=
ia
> >>>                                                                P. Robe=
rts
> >>>                                                            Alcatel-Luc=
ent
> >>>                                                                L. Mont=
ini
> >>>                                                                L. Mart=
ini
> >>>                                                             Cisco Syst=
ems
> >>>                                                             June 15, 2=
013
> >>>
> >>>
> >>>              Transporting Timing messages over MPLS Networks
> >>>                     draft-ietf-tictoc-1588overmpls-05
> >>>
> >>> Abstract
> >>>
> >>>     This document defines the method for transporting Timing messages
> >>>     such as PTP and NTP over an MPLS network.  The method allows for t=
he
> >>>     easy identification of these PDUs at the port level to allow for p=
ort
> >>>
> >>> SB> What is a port
> >>>
> >>>     level processing of these PDUs in both LERs and LSRs.
> >>>
> >>>     The basic idea is to transport Timing messages inside dedicated MP=
LS
> >>>     LSPs.  These LSPs only carry Timing messages and possibly Control=
 and
> >>>     Management packets, but they do not carry customer traffic.
> >>>
> >>> SB> More specifically they only carry traffic associated with the
> >>> SB> timing service and its support.
> >>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
> >>> SB> also it gets carried in a structure that causes it to get
> >>> SB> timestamped.
> >>>
> >>>     Two methods for transporting Timing messages over MPLS are defined=
.
> >>>
> >>> SB> Perhaps the right approach is to define the new LSP type and then
> >>> SB> seperately to define  the mapping of the various timing services
> >>> SB> over that LSP type.
> >>>
> >>>     The first method is to transport Timing messages directly over the
> >>>     dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
> >>>     MPLS networks.  The second method is to transport Timing messages
> >>>     inside a PW via Ethernet encapsulation.
> >>>
> >>> SB> I think that we should note that there are some
> >>> SB> h/w reasons for this preference. A clean sheet approach
> >>> SB> would have been to use PTP over MPLS with no intermediate
> >>> SB> layers.
> >>>
> >>> Status of this Memo
> >>>
> >>>     This Internet-Draft is submitted in full conformance with the
> >>>     provisions of BCP 78 and BCP 79.
> >>>
> >>>     Internet-Drafts are working documents of the Internet Engineering
> >>>     Task Force (IETF).  Note that other groups may also distribute
> >>>     working documents as Internet-Drafts.  The list of current Interne=
t-
> >>>     Drafts is at http://datatracker.ietf.org/drafts/current/.
> >>>
> >>>     Internet-Drafts are draft documents valid for a maximum of six mon=
ths
> >>>     and may be updated, replaced, or obsoleted by other documents at a=
ny
> >>>     time.  It is inappropriate to use Internet-Drafts as reference
> >>>     material or to cite them other than as "work in progress."
> >>>
> >>>     This Internet-Draft will expire on December 17, 2013.
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page=
 1]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> Copyright Notice
> >>>
> >>>     Copyright (c) 2013 IETF Trust and the persons identified as the
> >>>     document authors.  All rights reserved.
> >>>
> >>>     This document is subject to BCP 78 and the IETF Trust's Legal
> >>>     Provisions Relating to IETF Documents
> >>>     (http://trustee.ietf.org/license-info) in effect on the date of
> >>>     publication of this document.  Please review these documents
> >>>     carefully, as they describe your rights and restrictions with resp=
ect
> >>>     to this document.  Code Components extracted from this document mu=
st
> >>>     include Simplified BSD License text as described in Section 4.e of
> >>>     the Trust Legal Provisions and are provided without warranty as
> >>>     described in the Simplified BSD License.
> >>>
> >>>
> >>> Table of Contents
> >>>
> >>>     1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .=
  5
> >>>
> >>>     2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .=
  7
> >>>
> >>>     3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .=
  8
> >>>
> >>>     4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .=
  9
> >>>
> >>>     5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . .=
 12
> >>>
> >>>     6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . .=
 13
> >>>       6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . .=
 13
> >>>       6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . .=
 13
> >>>       6.3.  Other Timing Encapsulation methods . . . . . . . . . . . .=
 14
> >>>
> >>>     7.  Timing message Processing  . . . . . . . . . . . . . . . . . .=
 15
> >>>
> >>>     8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . .=
 16
> >>>
> >>>     9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 17
> >>>
> >>>     10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 18
> >>>
> >>>     11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 19
> >>>
> >>>     12. OAM, Control and Management  . . . . . . . . . . . . . . . . .=
 20
> >>>
> >>>     13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . .=
 21
> >>>
> >>>     14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . .=
 22
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page=
 2]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>>     15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . .=
 23
> >>>       15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . .=
 23
> >>>       15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . .=
 23
> >>>       15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . .=
 24
> >>>
> >>>     16. Other considerations . . . . . . . . . . . . . . . . . . . . .=
 25
> >>>
> >>>     17. Security Considerations  . . . . . . . . . . . . . . . . . . .=
 26
> >>>
> >>>     18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . .=
 27
> >>>
> >>>     19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . .=
 28
> >>>
> >>>     20. References . . . . . . . . . . . . . . . . . . . . . . . . . .=
 29
> >>>       20.1. Normative References . . . . . . . . . . . . . . . . . . .=
 29
> >>>       20.2. Informative References . . . . . . . . . . . . . . . . . .=
 29
> >>>
> >>>     Appendix 1.  Routing extensions for Timing-aware Routers . . . . .=
 32
> >>>
> >>>     Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . .=
 33
> >>>
> >>>     Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . .=
 34
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page=
 3]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>>     The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
> NOT",
> >>>     "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL"
> in
> >>> this
> >>>     document are to be interpreted as described in RFC2119 [RFC2119].
> >>>
> >>>     When used in lower case, these words convey their typical use in
> >>>     common language, and are not to be interpreted as described in
> >>>     RFC2119 [RFC2119].
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page=
 4]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 1.  Introduction
> >>>
> >>>     The objective of Precision Time Protocol (PTP) and Network Timing
> >>>     Protocol (NTP) are to synchronize independent clocks running on
> >>>     separate nodes of a distributed system.
> >>>
> >>>     [IEEE-1588] defines PTP messages for frequency, phase and time
> >>>     synchronization.  The PTP messages include PTP PDUs over UDP/IP
> >>>     (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F=
 of
> >>>     [IEEE-1588]).
> >>>
> >>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
> >>> SB> of other PTP mappings if they provide better optimisation.
> >>>
> >>>     This document defines mapping and transport of the PTP
> >>>     messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
> >>>     defines several clock types: ordinary clocks, boundary clocks, end=
-
> >>>     to-end transparent clocks, and peer-to-peer transparent clocks.
> >>>     Transparent clocks require intermediate nodes to update correction
> >>>     field inside PTP message that reflects the transit time in the nod=
e.
> >>>
> >>>     [RFC5905] defines NTP messages for clock and time synchronization.
> >>>     The PTP messages (PDUs) are transported over UDP/IP.  This documen=
t
> >>> SB> Should that be NTP messages?
> >>> SB> It needs to be made clear as soon as you introduce NTP that
> >>> SB> they use different time representations.
> >>>
> >>>     defines mapping and transport of the NTP messages defined in
> >>>     [RFC5905] over MPLS networks.
> >>>
> >>>     One key attribute of all of these Timing messages is that the Time
> >>>     stamp processing should occur as close as possible to the actual
> >>>     transmission and reception at the physical port interface.  This
> >>>     targets optimal time and/or frequency recovery by avoiding variabl=
e
> >>>     delay introduced by queues internal to the clocks.
> >>>
> >>> SB> As I recall NTP has no epoch point defined, and I am not sure
> >>> SB> where that point is in the case of PTP in this mapping
> >>> SB> Hopefully this will get defined in due course.
> >>>
> >>>     To facilitate the fast and efficient recognition of Timing message=
s
> >>>     at the port level when the Timing messages are carried over MPLS
> >>>     LSPs,
> >>>
> >>> SB> Over a new LSP type with time optimied characteristics
> >>>
> >>>     this document defines the specific encapsulations that should
> >>>     be used.
> >>> SB> Hopefully it will also define the PHP
> >>>
> >>>     In addition, it can be expected that there will exist LSR/
> >>>     LERs where only a subset of the physical ports will have the port-
> >>>     based Timing message processing capabilities.
> >>> SB> Do you need to clarify that this only works at base and not in
> >>> SB> a label heirarchy.
> >>>
> >>>
> >>>     In order to ensure
> >>>     that the LSPs carrying Timing packets always enter and exit ports
> >>>     with this capability, routing extensions are defined to advertise
> >>>     this capability on a port basis and to allow for the establishment=
 of
> >>>     LSPs that only transit such ports.  While this path establishment
> >>>     restriction may be applied only at the LER Ingress and/or egress
> >>>     ports, it becomes more important when using transparent clock capa=
ble
> >>>     LSRs in the path.
> >>> SB> I do not understand the implications of the last
> >>> SB> sentences - starting ", it becomes"
> >>>
> >>>
> >>>     Port based Timing message processing involves Timing message
> >>>     recognition.  Once the Timing messages are recognized they can be
> >>>     modified based on the reception or transmission Time-stamp.
> >>>
> >>>     This document provides two methods for transporting Timing message=
s
> >>>     over MPLS.  One is applicable to MPLS environment and the other on=
e
> >>>     is applicable to MPLS/MPLS-TP environment
> >>>
> >>> SB> I think the sentence is incomplete.
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page=
 5]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>>     The solution involves transporting Timing messages over dedicated
> >>>     LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
> >>>     carry Management and control messages, but not data plane client
> >>>     traffic.
> >>>
> >>> SB> It is not clear why this restriction applies.
> >>>
> >>>     Timing LSPs can be established statically or via signaling.
> >>> SB> s/statically/by provisioning/network management/
> >>>
> >>>     Extensions to control plane (OSPF, ISIS, etc.) is required to enab=
le
> >>>     routers to distribute their Timing processing capabilities over MP=
LS
> >>>     to other routers.  However such extensions are outside the scope o=
f
> >>>     this document.
> >>>
> >>>     When signaling is used to setup the PTP LSP, Extensions to signali=
ng
> >>> SB> is it a PTP LSP or a Timing LSP?
> >>>
> >>>     protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
> >>>     However such extensions are outside the scope of this document.
> >>>
> >>> SB> for mpls-tp GMPLS is the signalling protocol
> >>>
> >>>     While the techniques included herein allow for the establishment o=
f
> >>>     paths optimized to include Time-stamping capable links, the
> >>>     performance of the Slave clocks is outside the scope of this
> >>>     document.
> >>>
> >>>     At the time of publishing this specification, Transparent Clocking
> >>>     (TC) is only defined for PTP.  Therefore at this time any part of
> >>>     this specification that talks about Transparent Clocking applies o=
nly
> >>>     to PTP.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page=
 6]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 2.  Terminology
> >>>
> >>>     1588: The timing and synchronization as defined by IEEE 1588.
> >>>
> >>> SB> I think that there is a more formal name for the 1588 group
> >>> SB> that needs to be used here.
> >>> SB> Also do we need to talk about 1588-200? as there is
> >>> SB> an update in progress
> >>>
> >>>     NTP: The timing and synchronization protocol defined by IETF RFC-1=
305
> >>>     and RFC-5905.
> >>>
> >>>     PTP: The timing and synchronization protocol used by 1588.
> >>> SB> need the proper name for 1588
> >>>
> >>>     Master Clock: The source of 1588 timing to a set of slave clocks.
> >>>
> >>>     Master Port: A port on a ordinary or boundary clock that is in Mas=
ter
> >>>     state.  This is the source of timing toward slave ports.
> >>>
> >>> SB> I am not sure the reader knows what a port is
> >>>
> >>>     Slave Clock: A receiver of 1588 timing from a master clock.
> >>>
> >>>     Slave Port: A port on a boundary clock or ordinary clock that is
> >>>     receiving timing from a master clock.
> >>>
> >>>     Ordinary Clock: A device with a single PTP port.
> >>>
> >>>     Transparent Clock.  A device that measures the time taken for a PT=
P
> >>>     event message to transit the device and then updates the
> >>>     correctionField of the message with this transit time.
> >>>
> >>>     Boundary Clock: A device with more than one PTP port.  Generally
> >>>     boundary clocks will have one port in slave state to receive timin=
g
> >>>     and then other ports in master state to re-distribute the timing.
> >>>
> >>>     PTP LSP: An LSP dedicated to carry PTP messages
> >>>
> >>> SB> PTP or timing?
> >>>
> >>>
> >>>     PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
> >>>     messages.
> >>>
> >>> SB> Ah I don't think that PWE3 know what one of these is
> >>>
> >>>     CW: Pseudowire Control Word
> >>>
> >>>     LAG: Link Aggregation
> >>>
> >>>     ECMP: Equal Cost Multipath
> >>>
> >>>     CF: Correction Field, a field inside certain PTP messages (message
> >>>     type 0-3)that holds the accumulative transit time inside intermedi=
ate
> >>>     switches
> >>>
> >>>     Timing messages: Timing Protocol messages that are exchanged
> between
> >>>     routers in order to establish a synchronized clock.
> >>>
> >>> SB> A number of these definitions look like copies of IEEE1588
> >>> SB> definitions. We need to provide references and note the
> >>> SB> priority of the IEEE base reference.
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page=
 7]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 3.  Problem Statement
> >>>
> >>>     [IEEE-1588] has defined methods for transporting PTP messages over
> >>>     Ethernet and IP networks.  [RFC5905] has defined the method of
> >>>     transporting NTP messages over IP networks.  There is a need to
> >>>     transport Timing messages over MPLS networks while supporting the
> >>>     Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC=
)
> >>>     functionality in the LER and LSRs in the MPLS network.
> >>>
> >>>     There are multiple ways of transporting Timing over MPLS.  However=
,
> >>>     there is a requirement to limit the possible encapsulation options=
 to
> >>>     simplify the Timing message identification and processing required=
 at
> >>>     the port level.
> >>>
> >>>     When Timing-awareness is needed, Timing messages should not be
> >>>     transported over LSPs or PWs that are carrying customer traffic
> >>>     because LSRs perform Label switching based on the top label in the
> >>>     stack.
> >>>
> >>> SB> Have you explained why?
> >>>
> >>>     To detect Timing messages inside such LSPs require special
> >>>     hardware to do deep packet inspection at line rate.  Even if such
> >>>     hardware exists, the payload can't be deterministically identified=
 by
> >>>     LSRs because the payload type is a context of the PW label, and th=
e
> >>>     PW label and its context are only known to the Edge routers (PEs/
> >>>     LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CE=
S,
> >>>     etc).  Even if one restricts an LSP to only carry Ethernet PWs, th=
e
> >>>     LSRs dont have the knowledge of whether PW Control Word (CW) is
> >>>     present or not and therefore can not deterministically identify th=
e
> >>>     payload.
> >>>
> >>>     A generic method is defined in this document that does not require
> >>>     deep packet inspection at line rate, and can deterministically
> >>>     identify Timing messages.  This method can be used to detect Timin=
g
> >>>     Messages in both one-step and two-step clock implementations of
> >>>     ordinary, boundary and transparent clocks.
> >>>
> >>> SB> Needs a ref and I am sure many MPLS specialists will not understan=
d
> >>> SB> the msg types.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page=
 8]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 4.  Timing over MPLS Architecture
> >>>
> >>>     Timing messages are exchange between Timing ports on ordinary and
> >>>
> >>> SB> Have you defined a timing port?
> >>>
> >>>     boundary clocks.  Boundary clocks terminate the Timing messages an=
d
> >>>     act as master for other boundary clocks or for slave clocks.  End-=
to-
> >>>     End Transparent clocks do not terminate the Timing messages but th=
ey
> >>>     do modify the contents of the Timing messages as they transit acro=
ss
> >>>     the transparent clock.
> >>>
> >>>     Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Cl=
ock
> >>>
> >>>     (TC) could be implemented in either LERs or LSRs.
> >>>
> >>> SB> LER and LSR need to be expanded
> >>>
> >>>     An example is shown in Figure 1, where the LERs act as Ordinary Cl=
ock
> >>>     (OC) and are the initiating/terminating point for Timing messages.
> >>>     The ingress LER encapsulates the Timing messages in Timing LSP and
> >>>     the Egress LER terminates the Timing LSP.  The LSRs act as
> >>>     Transparent Clock (TC) and just update the Timing field in the Tim=
ing
> >>>     messages.
> >>>
> >>>
> >>>        +--------+     +-------+     +-------+     +-------+     +-----=
---+
> >>>        |Switch, |     |       |     |       |     |       |     |Switc=
h, |
> >>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
> >>>        |        |     |  OC   |     |  TC   |     |  OC   |     |    =
    |
> >>>        +--------+     +-------+     +-------+     +-------+     +-----=
---+
> >>>                       /                                 \
> >>>        +-------+     /                                   \     +------=
-+
> >>>        |  LER  |    /                                     \    |  LER=
  |
> >>>        | Master|---/                                       \---| Slave=
 |
> >>>        | Clock |                                               | Clock=
 |
> >>>        +-------+                                               +------=
-+
> >>>
> >>>       Figure (1) - Deployment example 1 of timing over MPLS network
> >>>
> >>>     Another example is shown in Figure2, where LERs terminate the Timi=
ng
> >>>     messages received from switch/routers that are outside of the MPLS
> >>>     network acting as OC or BC.  In this example LERs regenerate the
> >>>     clock and initiate timing messages encapsulated in Timing LSP towa=
rd
> >>>     the MPLS network, while the LSRs act as Transparent Clock (TC) and
> >>>     just update the Timing field in the Timing messages, which are
> >>>     already encapsulated in Timing LSPs.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013               [Page=
 9]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>>       +--------+     +-------+     +-------+     +-------+     +------=
--+
> >>>       |Switch, |     |       |     |       |     |       |     |Switch=
, |
> >>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Route=
r |
> >>>       | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC=
  |
> >>>       +--------+     +-------+     +-------+     +-------+     +------=
--+
> >>>
> >>>       Figure (2) - Deployment example 2 of timing over MPLS network
> >>>
> >>>
> >>>     Another example is shown in Figure 3, where LERs do not terminate=
 the
> >>>     Timing messages received from switch/routers that are outside of t=
he
> >>>     MPLS network acting as OC, TC or BC.  The LERs act as TC and updat=
e
> >>>     the Timing field in the Timing messages as they transit the LER,
> >>>     while encapsulating them in timing LSP.  The LSRs also act as
> >>>     Transparent Clock (TC) and just update the Timing field in the Tim=
ing
> >>>     messages which are already encapsulated in Timing LSPs.
> >>>
> >>>        +--------+     +-------+     +-------+     +-------+     +-----=
---+
> >>>        |Switch, |     |       |     |       |     |       |     |Switc=
h, |
> >>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
> >>>        |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC=
/BC|
> >>>        +--------+     +-------+     +-------+     +-------+     +-----=
---+
> >>>
> >>>      Figure (3) - Deployment example 3 of timing over MPLS network
> >>>
> >>>     Another example is shown in Figure 4, where LERs and LSRs support
> >>>     Boundary Clocks.  A single-hop LSP is created between two adjacent
> >>>     LSRs engaged in BC operation.  Other methods such as PTP transport
> >>>     over Ethernet MAY be used for transporting timing messages if the
> >>>     link between the two routers is Ethernet.
> >>>
> >>>       +--------+     +-------+     +-------+     +-------+     +------=
--+
> >>>       |Switch, |     |       |     |       |     |       |     |Switch=
, |
> >>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Route=
r |
> >>>       | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC=
  |
> >>>       +--------+     +-------+     +-------+     +-------+     +------=
--+
> >>>
> >>>     Figure (4) - Deployment example 3 of timing over MPLS network
> >>>
> >>>     An MPLS domain MAY serve multiple customers.  In these cases the
> MPLS
> >>>     domain (maintained by a service provider) may provide timing servi=
ces
> >>>     to multiple customers, each having their own Timing domain.
> >>>
> >>>     The Timing over MPLS architecture assumes full mesh of Timing LSPs
> >>>     between all LERs supporting this specification.
> >>>
> >>> SB> Note sure this is right - the salves surely do not need to
> >>> SB> exchange timing amongst themselves
> >>>
> >>>     It supports
> >>>     Point-to- point (VPWS) and Multipoint (VPLS) services.
> >>>
> >>> SB> What does that mean? You do not carry user data traffic?
> >>> SB> Maybe it's the ordering of the statemnets that is causing
> >>> SB> confusion.
> >>>
> >>>     This means
> >>>     that a customer may purchase a Point-to-point Timing service betwe=
en
> >>>     two customer sites or a Multipoint Timing service between more tha=
n
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 1=
0]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>>     two customer sites.
> >>>
> >>>     The Timing over MPLS architecture supports P2P or P2MP Timing LSPs=
.
> >>>     This means that the Timing Multicast messages such as PTP Multicas=
t
> >>>     event messages can be transported over P2MP Timing LSP or be
> >>>     replicated and transported over many P2P Timing LSPs.
> >>>
> >>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
> >>> SB> nor a P2MP PW, although we are close.
> >>>
> >>>     Timing messages, that do not require Time stamping or Correction
> >>>     Field update MAY be transported over Timing LSPs to simplify hardw=
are
> >>>     and software.
> >>>
> >>>     PTP Announce messages that determine the Timing LSP terminating
> point
> >>>     behavior such as BC/OC/TC SHOULD be transported over the Timing LS=
P
> >>>     to simplify hardware and software.
> >>>
> >>> SB> have you defined and referenced PTP announce msgs?
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 1=
1]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 5.  Dedicated LSPs for Timing messages
> >>>
> >>>     Many methods have been considered for identifying the Timing
> messages
> >>>     when they are encapsulated in MPLS such as using GAL/G-ACH or a ne=
w
> >>>     reserved label.  These methods were not attractive since they eith=
er
> >>>     required deep packet inspection at line rate in the intermediate L=
SRs
> >>>     or they required use of a scarce new reserved label.  Also one of=
 the
> >>>     goals was to reuse existing OAM mechanisms.
> >>>
> >>> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
> >>>
> >>>     The method defined in this document can be used by LER and LSRs to
> >>>     identify Timing messages in MPLS tunnels by just looking at the to=
p
> >>>     label in the MPLS label stack, which only carry Timing messages as
> >>>     well as OAM, but not data plane client traffic.
> >>>
> >>>     Compliant implementations MUST use dedicated LSPs to carry Timing
> >>>     messages over MPLS.
> >>>
> >>> SB> I think that we need a definition of the properies of these LSPs
> >>>
> >>>     These LSPs are herein referred to as "Timing
> >>>     LSPs" and the labels associated with these LSPs as "Timing LSP
> >>>     labels".  The Timing LSPs that runs between Ingress and Egress LER=
s
> >>>     MUST be co-routed.  Alternatively, a single bidirectional co-route=
d
> >>>     LSP can be used.
> >>>
> >>> SB> I though that you said you could use M2MP LSPs - these are not
> >>> SB> bidirectional.
> >>>
> >>>     Co-routing of the two directions is required to limit the differen=
ce
> >>>     in the delays in the Master clock to Slave clock direction compare=
d
> >>>     to the Slave clock to Master clock direction.  The Timing LSP MAY=
 be
> >>>     MPLS/MPLS-TP LSP.
> >>>
> >>>     The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
> >>>     New Extensions to RSVP-TE/GMPLS TLVs are required; however they ar=
e
> >>>     outside the scope of this document.
> >>>
> >>>     The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such=
 as
> >>>     BFD and LSP Ping but the LSP data plane client plane traffic MUST=
 be
> >>>     Timing packets only.
> >>>
> >>> SB> Why?
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 1=
2]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 6.  Timing over LSP Encapsulation
> >>>
> >>> The encapsulations is not LSP is it?
> >>>
> >>>     This document defines two methods for carrying Timing messages ove=
r
> >>>     MPLS.  The first method is carrying UDP/IP encapsulated Timing
> >>>     messages over Timing LSPs, and the second method, is carrying
> >>>     Ethernet encapsulated Timing messages over Ethernet PWs inside
> Timing
> >>>     LSPs.
> >>>
> >>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
> >>>
> >>>     The simplest method of transporting Timing messages over MPLS is t=
o
> >>>     encapsulate Timing PDUs in UDP/IP and then encapsulate them in
> Timing
> >>>     LSP.  This format is shown in Figure 4.
> >>>
> >>>
> >>>                      +----------------------+
> >>>                      |   Timing LSP Label   |
> >>>                      +----------------------+
> >>>                      |        IPv4/6        |
> >>>                      +----------------------+
> >>>                      |         UDP          |
> >>>                      +----------------------+
> >>>                      |     Timing PDU       |
> >>>                      +----------------------+
> >>>
> >>>        Figure (4) - Timing over UDP/IP over MPLS Encapsulation
> >>>
> >>>
> >>>     This encapsulation is very simple and is useful when the network
> >>>     between Timing Master Clock and Slave Clock is MPLS network.
> >>>
> >>> SB> Simple is a judgement call
> >>>
> >>>     In order for an LER/LSR to process Timing messages, the Timing LSP
> >>>     Label must be at the top label of the label stack.  The LER/LSR MU=
ST
> >>>     know that the Timing LSP Label is used for carrying Timing message=
s.
> >>>     This can be accomplished via static configuration or via RSVP-TE
> >>>     signaling.
> >>>
> >>>     The UDP/IP encapsulation of PTP MUST follow Annex D and E of
> >>>     [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
> >>>     [RFC5905].
> >>>
> >>> 6.2.  Timing over PW Encapsulation
> >>>
> >>>     Another method of transporting Timing over MPLS networks is by
> >>>     encapsulating Timing PDUs in PW which in turn is transported over
> >>>     Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
> >>>     shown in Fig 5(A) MUST be used and the Ethernet encapsulation of P=
TP
> >>>     MUST follow Annex F of [IEEE-1588].
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 1=
3]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>>     The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
> the
> >>>     Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
> >>>     Timing over PW encapsulation MUST use the Control Word (CW) as
> >>>     specified in [RFC4448] to ensure proper detection of PTP messages
> >>>     inside the MPLS packets for Timing over LSP and Timing over PW
> >>>     encapsulation.
> >>>
> >>> SB> That needs explanation
> >>>
> >>>     The use of Sequence Number in the CW is optional.
> >>>
> >>> SB> Given that s/n are never in practice deployed, you could probably
> >>> SB> simplify things by sayig that they are not used.
> >>>
> >>>     Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over
> PW
> >>>     (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
> >>>
> >>>                      +----------------+  +----------------+
> >>>                       |Timing LSP Label|  |Timing LSP Label|
> >>>                       +----------------+  +----------------+
> >>>                       |    PW Label    |  |    PW Label    |
> >>>                       +----------------+  +----------------+
> >>>                       |  Control Word  |  |      IP        |
> >>>                       +----------------+  +----------------+
> >>>                       |    Ethernet    |  |      UDP       |
> >>>                       |     Header     |  +----------------+
> >>>                       +----------------+  |   Timing PDU   |
> >>>                       |S-VLAN(Optional)|  |                |
> >>>                       +----------------+  +----------------+
> >>>                       |C-VLAN(Optional)|        (B)
> >>>                       +----------------+
> >>>                       |   Timing PDU   |
> >>>                       |                |
> >>>                       +----------------+
> >>>                              (A)
> >>>
> >>>                Figure (5) - Timing over PW Encapsulations
> >>>
> >>>     In order for an LSR to process PTP messages, the top label of the
> >>>     label stack (the Tunnel Label) MUST be a Timing label.
> >>>
> >>> S> You said that before.
> >>>
> >>> 6.3.  Other Timing Encapsulation methods
> >>>
> >>>     In future other timing encapsulation methods may be introduced, su=
ch
> >>>     as a new shim header after the Bottom of Stack to carry the Timing
> >>>     information.  Such new encapsulations are outside the scope of thi=
s
> >>>     document.
> >>>
> >>>
> >>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
> >>> SB> out of the definition of the LSP
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 1=
4]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> SB> I think we need a section on LSP processing
> >>>
> >>> 7.  Timing message Processing
> >>>
> >>>     Each Timing protocol such as PTP and NTP, define their set of Timi=
ng
> >>>     messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
> >>>     FOLLOW_UP, etc messages.
> >>>
> >>>     Some of the Timing messages require time stamping or correction fi=
eld
> >>>     update at port level and some dont.  It is the job of the LER/LSR=
 to
> >>>     parse the timing message and find out the type of the Timing messa=
ge
> >>>     and decide whether and how to Time- stamp it (e.g., BC) or update
> >>>     correction field(e.g., TC).
> >>>
> >>>
> >>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
> >>> SB> function rather than the LER?
> >>>
> >>>     For example the following PTP messages (called Event messages)
> >>>     require time-stamping or correction field update:
> >>>
> >>>     o  SYNC
> >>>
> >>>     o  DELAY_REQ (Delay Request)
> >>>
> >>>     o  PDELAY_REQ (Peer Delay Request)
> >>>
> >>>     o  PDELAY_RESP (Peer Delay Response)
> >>>
> >>>     SYNC and DELAY_REQ are exchanged between Master Clock and Slave
> Clock
> >>>     and MUST be transported over PTP LSPs.  PDELAY_REQ and
> PDELAY_RESP
> >>>     are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
> >>>     Boundary, or Transparent) and SHOULD be transported over single ho=
p
> >>>     PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
> >>>     and PDELAY_RESP_FOLLOW_UP messages MUST also be transported
> over
> >>> the
> >>>     PTP LSPs.
> >>>
> >>>     For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
> >>>     transported over two PTP LSPs that are in opposite directions.  Th=
ese
> >>>     PTP LSPs, which are in opposite directions MUST be congruent and c=
o-
> >>>     routed.  Alternatively, a single bidirectional co-routed LSP can b=
e
> >>>     used.
> >>>
> >>>     Except as indicated above for the two-step PTP clocks, Non-Event P=
TP
> >>>     message types do not need to be processed by intermediate routers.
> >>>     These message types MAY be carried in PTP Tunnel LSPs.
> >>>
> >>> SB> Are you saying that a timing P router has to be msg type sensitive=
?
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 1=
5]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 8.  Protection and Redundancy
> >>>
> >>>
> >>> SB> This is a bit of a jump - I don't know how the LSP itself works ye=
t!
> >>>
> >>>     In order to ensure continuous uninterrupted operation of slave
> >>>     clocks, usually as a general practice, slave clocks (or ports) tra=
ck
> >>>     redundant master clocks.
> >>>
> >>>     It is the responsibility of the network operator to ensure that
> >>>     physically disjoint Timing LSPs are established between a slave cl=
ock
> >>>     (or port) and redundant master clocks (or ports).
> >>>
> >>>     When a slave clock (or port) listens to redundant master clocks or
> >>>     ports, any prolonged Timing LSP outage will trigger the slave cloc=
k
> >>>     or port to switch to a redundant master clock or port.
> >>>
> >>>     LSP/PW protection such as Linear protection Switching (1:1, 1+1),
> >>>     Ring protection switching or MPLS Fast Reroute (FRR) generally swi=
tch
> >>>     alternative path that usually cause a change in delay, which if
> >>>     undetected by slave clock can reduce accuracy of the slave clock.
> >>>
> >>>     Therefore protection switching MAY be used, as long as phase jumps
> >>>     upon switchover due to differences in path latency are detected an=
d
> >>>     compensated for (such compensation not being required if BCs or pe=
er-
> >>>     peer TCs are used throughout).
> >>>
> >>>     Note that any protection or reroute mechanism that adds additional
> >>>     MPLS label to the label stack, such as Facility Backup Fast Rerout=
e,
> >>>     MUST ensure that the pushed label is also a Timing Label to ensure
> >>>     recognition of the MPLS frame as containing Timing messages, as it
> >>>     transits the backup path.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 1=
6]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 9.  ECMP
> >>>
> >>>     To ensure the optimal operation of slave clocks and avoid error
> >>>     introduced by forward and reverse path delay asymmetry, the physic=
al
> >>>     path for Timing messages from master clock to slave Clock and vice
> >>>     versa must be the same for all Event Timing messages listed in
> >>>     section 7.
> >>>
> >>>     Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
> >>>     Multipath).
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 1=
7]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 10.  PHP
> >>>
> >>>     To ensure that the label on the top of the label stack is the Timi=
ng
> >>>     LSP Label, PHP MUST not be used.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 1=
8]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 11.  Entropy
> >>>
> >>>     To ensure all Timing messages in a Timing LSP take the same path,
> >>>     Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
> >>>     Entropy Label MUST NOT be used for the PWs that are carried inside
> >>>     Timing LSP [RFC6391].
> >>>
> >>> SB> This is incorrect - you mean that all msgs of the same timing
> >>> SB> flow need to have the same EL value.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 1=
9]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 12.  OAM, Control and Management
> >>>
> >>>     In order to monitor Timing LSPs and their encapsulated PWs, they M=
UST
> >>>     be able to carry OAM and management messages.  These management
> >>>     messages MUST be differentiated from Timing messages via already
> >>>     defined IETF methods.
> >>>
> >>>     For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY ru=
n
> >>>     over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
> >>>     Management protocols can easily be identified by the UDP Destinati=
on
> >>>     Port number or by GAL/G-ACH respectively.
> >>>
> >>>     Also BFD, LSP-Ping and other management messages MAY run over the
> PWs
> >>>     encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3=
 or
> >>>     4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is go=
ing
> >>>     to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1)=
 or
> >>>     GAL-ACH are used to identify such management messages.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 2=
0]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 13.  QoS Considerations
> >>>
> >>>     In network deployments where not every LSR/LER is Timing-aware, it=
 is
> >>>     important to reduce the impact of the non-Timing-aware LSR/LERs on
> >>>     the timing recovery in the slave clock.  The Timing messages are t=
ime
> >>>     critical and must be treated with the highest priority.  Therefore
> >>>     Timing over MPLS messages must be treated with the highest priorit=
y
> >>>     in the routers.  This can be achieved by proper setup of Timing LS=
Ps.
> >>>
> >>>     It is recommended that the Timing LSPs are setup or configured
> >>>     properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC269=
7]
> >>>     for drop eligibility.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 2=
1]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 14.  FCS and Checksum Recalculation
> >>>
> >>>     When time-stamp generation and timing packet adjustment is perform=
ed
> >>>     near the physical port hardware, the process MUST include
> >>>     recalculation of the Ethernet FCS.
> >>>
> >>> SB> The above is confusing - an LSR always recomputes the link layer
> >>> SB> CRC which may or may not be Ethernet.
> >>>
> >>>     Also FCS retention for the
> >>>     payload Ethernet described in [RFC4720] MUST NOT be used.
> >>>
> >>>     For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksu=
m
> >>>     may be required as per UDP transport standards.
> >>>
> >>> SB> You really need to be working on getting the IPv6 C?S computation
> >>> SB> removed from PTP msgs.
> >>>
> >>>     When UDP checksum is used, each Timing-aware LER/LSR must either
> >>>     incrementally update the UDP checksum after Time stamping or
> >>>     Correction Field update or verify the UDP checksum on reception fr=
om
> >>>     upstream and recalculate the checksum completely on transmission t=
o
> >>>     downstream node after Time stamping or Correction Field update.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 2=
2]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 15.  Behavior of LER/LSR
> >>>
> >>>     Timing-capable/aware LERs and LSRs are routers that have one or mo=
re
> >>>
> >>> SB> You mean physical interfaces?
> >>>
> >>>     interfaces that can perform Timing operations (OC/BC/TC) on Timing
> >>>     packets and are configured to do so.  Timing-capable/aware LERs an=
d
> >>>     LSRs can advertise their Timing-capability per-interface via contr=
ol
> >>>     plane such as OSPF or IS-IS.
> >>> SB> ISIS and OSPF are routing protocols.
> >>>
> >>>    The Timing-capable/aware LERs can then
> >>>     signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timi=
ng
> >>>     capability of LER and LSRs may be configured in a centralized
> >>>     controller and the Timing LSP may be setup using manual configurat=
ion
> >>>     or other methods such as SDN.
> >>>
> >>> SB> it can also be configured individually rather then through
> >>> SB> a cebtral controllwe
> >>>
> >>> 15.1.  Behavior of Timing-capable/aware LER
> >>>
> >>>     When a Timing-capable/aware LER behaves as a Transparent clock and
> >>>     receives a Timing message from a Timing-capable/aware non-MPLS
> >>>     interface, the LER updates the Correction Field (CF) and encapsula=
tes
> >>>     and forwards the timing message over previously established Timing
> >>>     LSP.
> >>>
> >>> SB> You need to call out the details so that people properly
> >>> SB> understand the definition of the new LSP.
> >>>
> >>>     Also when a Timing message is received from a Timing-capable/
> >>>     aware MPLS interface, LER updates the Correction Filed (CF) and
> >>>     decapsulates the MPLS encapsulation and forwards the timing messag=
e
> >>>     to a non-MPLS interface.
> >>>
> >>>     When a Timing-capable/aware LER behaves as a Boundary clock and
> >>>     receives a Timing message from a Timing-capable/aware non MPLS
> >>>     interface, the LER Timestamps the Timing packet and sends it to th=
e
> >>>     LERs Boundary clock processing module.  Also when a Timing message=
 is
> >>>     received from a Timing- capable/aware MPLS interface, the LER
> >>>     Timestamps the Timing packet and sends it to the LERs Boundary clo=
ck
> >>>     processing module.
> >>>
> >>>     When a Timing-capable/aware LER behaves as an Ordinary Clock towar=
d
> >>>     the MPLS network, and receives a Timing message from a Timing-
> >>>     capable/aware MPLS interface, the LER Timestamps the Timing packet
> >>>     and sends it to the LERs Ordinary clock processing module.
> >>>
> >>> 15.2.  Behavior of Timing-capable/aware LSR
> >>>
> >>>     When a Timing-capable/aware LSR behaves as a Transparent clock and
> >>>     receives a Timing message from a Timing-capable/aware MPLS
> interface,
> >>>     The LSR updates the Correction Filed (CF) and forwards the timing
> >>>     message over another MPLS interface.
> >>>
> >>>     When a Timing-capable/aware LSR behaves as a Boundary clock and
> >>>     receives a Timing message from a Timing-capable/aware MPLS
> interface.
> >>>     The LSR performs the functions of a Boundary Clock in terminating=
 the
> >>>     received Timing message and re-generating a new timing message ove=
r
> >>>     another (or the same) MPLS interface.
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 2=
3]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 15.3.  Behavior of non-Timing-capable/aware LSR
> >>>
> >>>     It is most beneficial when all LSRs in the path of a Timing LSP be
> >>>     timing-Capable/aware LSRs.  This would ensure the highest quality
> >>>     time and clock synchronization by Timing Slave Clocks.  However, t=
his
> >>>     specification does not mandate that all LSRs in path of a Timing L=
SP
> >>>     be Timing- capable/aware.
> >>>
> >>>     Non-Timing-capable/aware LSRs just switch the packets encapsulated=
 in
> >>>     Timing LSPs and dont perform any Timing operation (TC or BC).
> >>>     However as explained in QoS section the Timing over MPLS packets
> MUST
> >>>     be still be treated with the highest priority based on their Traff=
ic
> >>>     Class (TC) marking.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 2=
4]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 16.  Other considerations
> >>>
> >>>     [IEEE-1588] defines an optional peer-to-peer Transparent clocking
> >>>     that requires peer delay measurement between two adjacent Timing-
> >>>     capable/ aware routers/switches.  Peer delay measurement messages
> >>>     need to be time stamped and terminated by the Timing-capable/aware
> >>>     routers/ switches.  This means that two adjacent LSRs may be engag=
ed
> >>>     in a peer delay measurement.
> >>>
> >>>     For transporting such peer delay measurement messages a single-hop
> >>>     LSP SHOULD to be created between the two adjacent LSRs engaged in
> >>>     peer delay measurement to carry peer delay measurement messages.
> >>>     Other methods such as PTP transport over Ethernet MAY be used for
> >>>     transporting peer delay measurement messages if the link between t=
he
> >>>     two routers is Ethernet.
> >>>
> >>>     In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ w=
are
> >>>     routers/switches MUST maintain a list of all the neighbors it need=
s
> >>>     to send a PDelay_Req to, where each neighbor corresponds to a timi=
ng
> >>>     LSP.
> >>>
> >>>     The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as=
 long
> >>>     as either the Explicit Null label is the bottom of stack label
> >>>     (applicable only to UDP/IP encapsulation) or the label below the
> >>>     Explicit Null label is a PTP label.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 2=
5]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 17.  Security Considerations
> >>>
> >>>     MPLS PW security considerations in general are discussed in [RFC39=
85]
> >>>     and [RFC4447],and those considerations also apply to this document=
.
> >>>
> >>>     An experimental security protocol is defined in [IEEE-1588].The PT=
P
> >>>     security extension and protocol provides group source authenticati=
on,
> >>>     message integrity, and replay attack protection for PTP messages.
> >>>
> >>>     When the MPLS network (provider network) serves multiple customers=
,
> >>>     it is important to maintain and process each customers clock and
> >>>     Timing messages separately from other customers to ensure there is=
 no
> >>>     cross- customer effect.  For example if an LER BC is synchronized=
 to
> >>>     a specific grandmaster, belonging to customer A, then the LER MUST
> >>>     use that BC clock only for customer A to ensure that customer A
> >>>     cannot attack other customers by manipulating its time.
> >>>
> >>>     Timing messages MAY be encrypted or authenticated, provided that t=
he
> >>>     LERs/LSRs that are Timing capable/aware can authenticate/ decrypt=
 the
> >>>     timing messages.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 2=
6]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 18.  Acknowledgements
> >>>
> >>>     The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizra=
hi,
> >>>     Stefano Ruffini, Peter Meyer, and other members of IETF for review=
ing
> >>>     and providing feedback on this draft.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 2=
7]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 19.  IANA Considerations
> >>>
> >>>     There are no IANA requirements in this specification.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 2=
8]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 20.  References
> >>>
> >>> 20.1.  Normative References
> >>>
> >>>     [IEEE-1588]
> >>>                IEEE 1588-2008, "IEEE Standard for a Precision Clock
> >>>                Synchronization Protocol for Networked Measurement and
> >>>                Control Systems".
> >>>
> >>>     [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
> >>>                Requirement Levels", BCP 14, RFC 2119, March 1997.
> >>>
> >>>     [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
> >>>                Edge (PWE3) Architecture", RFC 3985, March 2005.
> >>>
> >>>     [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discove=
ry
> >>>                Proxies (ND Proxy)", RFC 4389, April 2006.
> >>>
> >>>     [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
> >>>                Heron, "Pseudowire Setup and Maintenance Using the Labe=
l
> >>>                Distribution Protocol (LDP)", RFC 4447, April 2006.
> >>>
> >>>     [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
> >>>                "Encapsulation Methods for Transport of Ethernet over M=
PLS
> >>>                Networks", RFC 4448, April 2006.
> >>>
> >>>     [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
> >>>                Emulation Edge-to-Edge (PWE3) Frame Check Sequence
> >>>                Retention", RFC 4720, November 2006.
> >>>
> >>>     [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circui=
t
> >>>                Connectivity Verification (VCCV): A Control Channel for
> >>>                Pseudowires", RFC 5085, December 2007.
> >>>
> >>>     [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detecti=
on
> >>>                (BFD)", RFC 5880, June 2010.
> >>>
> >>>     [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
> >>>                "Bidirectional Forwarding Detection (BFD) for MPLS Labe=
l
> >>>                Switched Paths (LSPs)", RFC 5884, June 2010.
> >>>
> >>> 20.2.  Informative References
> >>>
> >>>     [I-D.ietf-pwe3-fat-pw]
> >>>                Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Rega=
n,
> >>>                J., and S. Amante, "Flow Aware Transport of Pseudowires
> >>>                over an MPLS Packet Switched Network",
> >>>                draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011=
.
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 2=
9]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>>     [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediat=
e
> >>>                system routeing information exchange protocol for use i=
n
> >>>                conjunction with the Protocol for providing the
> >>>                Connectionless-mode Network Service (ISO 8473)".
> >>>
> >>>     [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
> >>>                dual environments", RFC 1195, December 1990.
> >>>
> >>>     [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998=
.
> >>>
> >>>     [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
> >>>                Marker", RFC 2697, September 1999.
> >>>
> >>>     [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boude=
c,
> >>>                J., Courtney, W., Davari, S., Firoiu, V., and D.
> >>>                Stiliadis, "An Expedited Forwarding PHB (Per-Hop
> >>>                Behavior)", RFC 3246, March 2002.
> >>>
> >>>     [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineer=
ing
> >>>                (TE) Extensions to OSPF Version 2", RFC 3630,
> >>>                September 2003.
> >>>
> >>>     [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediat=
e
> >>>                System (IS-IS) Extensions for Traffic Engineering (TE)"=
,
> >>>                RFC 3784, June 2004.
> >>>
> >>>     [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S=
.
> >>>                Shaffer, "Extensions to OSPF for Advertising Optional
> >>>                Router Capabilities", RFC 4970, July 2007.
> >>>
> >>>     [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
> >>>                System to Intermediate System (IS-IS) Extensions for
> >>>                Advertising Router Information", RFC 4971, July 2007.
> >>>
> >>>     [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
> >>>                Topology (MT) Routing in Intermediate System to
> >>>                Intermediate Systems (IS-ISs)", RFC 5120, February 2008=
.
> >>>
> >>>     [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
> >>>                Engineering", RFC 5305, October 2008.
> >>>
> >>>     [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
> >>>                "Traffic Engineering Extensions to OSPF Version 3",
> >>>                RFC 5329, September 2008.
> >>>
> >>>     [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
> >>>                for IPv6", RFC 5340, July 2008.
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 3=
0]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>>     [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Netw=
ork
> >>>                Time Protocol Version 4: Protocol and Algorithms
> >>>                Specification", RFC 5905, June 2010.
> >>>
> >>>     [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Rega=
n,
> >>>                J., and S. Amante, "Flow-Aware Transport of Pseudowires
> >>>                over an MPLS Packet Switched Network", RFC 6391,
> >>>                November 2011.
> >>>
> >>>     [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., an=
d
> >>>                L. Yong, "The Use of Entropy Labels in MPLS Forwarding"=
,
> >>>                RFC 6790, November 2012.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 3=
1]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 1.  Routing extensions for Timing-aware Routers
> >>>
> >>>     MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] a=
nd
> >>>     IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (T=
E)
> >>>     link information used for constraint-based routing.
> >>>
> >>>     Indeed, it is useful to advertise data plane TE router link
> >>>     capabilities, such as the capability for a router to be Timing-awa=
re.
> >>>     This capability MUST then be taken into account during path
> >>>     computation to prefer or even require links that advertise themsel=
ves
> >>>     as Timing-aware.  In this way the path can ensure the entry and ex=
it
> >>>     points into the LERs and, if desired, the links into the LSRs are
> >>>     able to perform port based time-stamping thus minimizing their imp=
act
> >>>     on the performance of the slave clock.
> >>>
> >>>     extensions are required to OSPF and IS-IS in order to advertise
> >>>     Timing-aware capabilities of a link.  Such extensions are outside=
 the
> >>>     scope of this document; however such extension SHOULD be able to
> >>>     signal the following information per Router Link:
> >>>
> >>>     o  Capable of processing PTP, NTP or other Timing flows
> >>>
> >>>     o  Capable of performing Transparent Clock operation
> >>>
> >>>     o  Capable of performing Boundary Clock operation
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 3=
2]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> 2.  Signaling Extensions for Creating Timing LSPs
> >>>
> >>>     RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP=
-TE
> >>>     is used to setup Timing LSPs, some information that indicates that
> >>>     the LSP is carrying Timing flows MUST be included in the new
> >>>     Extensions to RSVP-TE:
> >>>
> >>>     The following information MAY also be included in the new Extensio=
ns
> >>>     to RSVP-TE:
> >>>
> >>>     o  Offset from Bottom of Stack (BoS) to the start of the Time-stam=
p
> >>>        field
> >>>
> >>>     o  Number of VLANs in case of PW encapsulation
> >>>
> >>>     o  Timestamp field Type
> >>>
> >>>        *  Correction Field, Timestamp
> >>>
> >>>     o  Timestamp Field format
> >>>
> >>>        *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
> >>>           NTP, etc.
> >>>
> >>>     Note that in case the above optional information is signaled with
> >>>     RSVP-TE for a Timing LSP, all the Timing packets carried in that L=
SP
> >>>     must have the same signaled characteristics.  For example if
> >>>     Timestamp format is signaled as 64-bit PTPv1, then all Timing pack=
ets
> >>>     must use 64-bit PTPv1 time-stamp.
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 3=
3]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>> Authors' Addresses
> >>>
> >>>     Shahram Davari
> >>>     Broadcom Corp.
> >>>     San Jose, CA  95134
> >>>     USA
> >>>
> >>>     Email: davari@broadcom.com
> >>>
> >>>
> >>>     Amit Oren
> >>>     Broadcom Corp.
> >>>     San Jose, CA  95134
> >>>     USA
> >>>
> >>>     Email: amito@broadcom.com
> >>>
> >>>
> >>>     Manav Bhatia
> >>>     Alcatel-Lucent
> >>>     Bangalore,
> >>>     India
> >>>
> >>>     Email: manav.bhatia@alcatel-lucent.com
> >>>
> >>>
> >>>     Peter Roberts
> >>>     Alcatel-Lucent
> >>>     Kanata,
> >>>     Canada
> >>>
> >>>     Email: peter.roberts@alcatel-lucent.com
> >>>
> >>>
> >>>     Laurent Montini
> >>>     Cisco Systems
> >>>     San Jose CA
> >>>     USA
> >>>
> >>>     Email: lmontini@cisco.com
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 3=
4]
> >>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
> >>>
> >>>
> >>>     Luca
> >>>     Cisco Systems
> >>>     San Jose CA
> >>>     USA
> >>>
> >>>     Email: lmartini@cisco.com
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> Davari, et al.          Expires December 17, 2013              [Page 3=
5]
> >>> --
> >>> For corporate legal information go to:
> >>>
> >>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
> >>>
> >>> _______________________________________________
> >>> TICTOC mailing list
> >>> TICTOC@ietf.org
> >>> https://www.ietf.org/mailman/listinfo/tictoc
> >>
> >> This e-mail message is intended for the recipient only and contains
> information which is CONFIDENTIAL and which may be proprietary to ECI
> Telecom. If you have received this transmission in error, please inform us=
 by e-
> mail, phone or fax, and then delete the original and all copies thereof.
> >>
> >> _______________________________________________
> >> mpls mailing list
> >> mpls@ietf.org
> >> https://www.ietf.org/mailman/listinfo/mpls
> > .
> >
> 
> 
> --
> For corporate legal information go to:
> 
> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
> 
> _______________________________________________
> TICTOC mailing list
> TICTOC@ietf.org
> https://www.ietf.org/mailman/listinfo/tictoc
> 
> 
> _______________________________________________
> TICTOC mailing list
> TICTOC@ietf.org
> https://www.ietf.org/mailman/listinfo/tictoc


This e-mail message is intended for the recipient only and contains informat=
ion which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If yo=
u have received this transmission in error, please inform us by e-mail, phon=
e or fax, and then delete the original and all copies thereof.


From davarish@yahoo.com  Tue Aug  6 06:35:32 2013
Return-Path: <davarish@yahoo.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D44AB21F9F80 for <tictoc@ietfa.amsl.com>; Tue,  6 Aug 2013 06:35:32 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.203
X-Spam-Level: 
X-Spam-Status: No, score=-1.203 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, MIME_QP_LONG_LINE=1.396]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id nhbgFvguexDB for <tictoc@ietfa.amsl.com>; Tue,  6 Aug 2013 06:35:28 -0700 (PDT)
Received: from nm15.bullet.mail.bf1.yahoo.com (nm15.bullet.mail.bf1.yahoo.com [98.139.212.174]) by ietfa.amsl.com (Postfix) with ESMTP id C10C921F9CCC for <tictoc@ietf.org>; Tue,  6 Aug 2013 06:35:24 -0700 (PDT)
Received: from [98.139.212.147] by nm15.bullet.mail.bf1.yahoo.com with NNFMP; 06 Aug 2013 13:35:22 -0000
Received: from [98.139.211.199] by tm4.bullet.mail.bf1.yahoo.com with NNFMP; 06 Aug 2013 13:35:22 -0000
Received: from [127.0.0.1] by smtp208.mail.bf1.yahoo.com with NNFMP; 06 Aug 2013 13:35:22 -0000
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yahoo.com; s=s1024; t=1375796122; bh=GyRmXObHO/J+l9hhJ5xWBh9mIjV1vt/Eq75FXXAxScc=; h=X-Yahoo-Newman-Id:X-Yahoo-Newman-Property:X-YMail-OSG:X-Yahoo-SMTP:X-Rocket-Received:References:Mime-Version:In-Reply-To:Content-Type:Content-Transfer-Encoding:Message-Id:Cc:X-Mailer:From:Subject:Date:To; b=WQvCjf+nMPOHoxrdw+X/EzR6ecX/caNXqbasmtDzbjjQeRx5g1Fpq9XIhat84AlO3Z0J7j/DnLQuBdswksZcgvKrPK2XXPz3YriXKKCc/jPmRvq2mGA6hc7sW8+tRfVGNIwLBpVmUYtWWKd1Lnofsh1UfN7Ti7lptU9GadGuSQo=
X-Yahoo-Newman-Id: 482266.10068.bm@smtp208.mail.bf1.yahoo.com
X-Yahoo-Newman-Property: ymail-3
X-YMail-OSG: bF6jTbAVM1ksMcJmHbO4rXWaSwGE4M6qiD46htd5rfB9YVV k7YG39zJ.Yjq93nPvQ8Bhxd2Ime.HAmvHMpppbUE2MLY2hlPk9S3ZVjf7kf_ qZeBo0nHYt1KEQhkWUA27uFQpjTYws0sX970GCCeYKELL4sF36qxDXZyY2_o 3CNwA1J5s0VEB.7kThR.I45gxslG026vN5JztMyS3y_1S5BH4SXbFzN8met0 ymldn.dQ57WWh1rSwQu4wJVLdxvY266qQNnI_mGuI8bY2WHlsqCjpkO41q4k hA2gpMLLZ1.acImjzMtacit55NUmzew5Mb9kpoZ3TonYZXuO9tAxcl.BuWPi BOeoCSK9jBpH5nl0vJe6drTdw.ee1RXeOAAwy4027TYd7.HUuzf0GUiYHaTy 7h8Z1Tc9R5VKduBLVBKGfHPgZoIYLjpGO8ZeA1Tus1miGUUfzg5RVVZnDDwq t9L.cm3gFQP5DUIcW3fDhd4cIJ6RXDr6Q0riMASJ9RJVgWw4NUC40UOsKnj3 h65R1TUR5KvFxi_bHD8CFxMsJzB4eUG8_PyrXE4jodX2PsdNg
X-Yahoo-SMTP: ygPrP9CswBCWPbPtKJlJyLY0KMlg
X-Rocket-Received: from [192.168.0.103] (davarish@98.248.42.143 with ) by smtp208.mail.bf1.yahoo.com with SMTP; 06 Aug 2013 06:35:22 -0700 PDT
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com> <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com>
Mime-Version: 1.0 (1.0)
In-Reply-To: <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: quoted-printable
Message-Id: <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com>
X-Mailer: iPhone Mail (10A403)
From: "S. Davari" <davarish@yahoo.com>
Date: Tue, 6 Aug 2013 06:35:15 -0700
To: Alexander Vainshtein <Alexander.Vainshtein@ecitele.com>
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 06 Aug 2013 13:35:33 -0000

Sasha,

If a router is not 1588 aware we want it to switch the packet with minimum j=
itter and delay via high priority queue.=20

Sending it to CPU creates large jitter and delay since the CPUs are generall=
y busy doing other things. Time stamping or updating CF in the CPU won't hel=
p since it does not cover the variable delay that is incurred from the time t=
he packet enters the router to the time the packet gets processed by CPU.

I wish it was that easy!

Regards,
Shahram


On Aug 6, 2013, at 12:38 AM, Alexander Vainshtein <Alexander.Vainshtein@ecit=
ele.com> wrote:

> Shahram,
> I agree with you that the routers that do not recognize a Timing Alert lab=
el would send the packet to the CPU. (This would be easy to arrange since we=
 are speaking about a reserved label.)
>=20
> I do not, however, see this as a serious issue, because the SW running on t=
his CPU would (hopefully) be upgradable to handle the packet correctly i.e.,=
 to forward it to where it should be forwarded and, if it is forwarded as a l=
abeled packet, to prepend the Timing Alert Label on top of its label stack. I=
t could even record the residence time (as observed by the CPU in the proper=
 place in the packet. This would mean that introducing additional error to w=
hatever timing information is associated with this packet - but this is what=
 you should anyway expect if there are non-compliant routers on your path, r=
ight?
>=20
> Regards,
>     Sasha
>=20
>> -----Original Message-----
>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf O=
f
>> Shahram Davari
>> Sent: Monday, August 05, 2013 11:29 PM
>> To: stbryant@cisco.com; S. Davari
>> Cc: mpls@ietf.org; tictoc@ietf.org
>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>=20
>> Hi Stewart,
>>=20
>> Your suggestion of " I suggested an LSP type that has the properties "tim=
estamp
>> and pass to application", is a subset of the existing draft and should wo=
rk. It
>> basically limits the time stamping and correction field update to LERs (w=
hile the
>> draft supports time stamping at LER and LSR).
>>=20
>> However Sasha's suggestion of using a reserved label (RAL, GAL or any oth=
er
>> reserved label) does not satisfy one of the major requirements, which is
>> backward compatibility. Routers that don't understand this reserved label=
 will
>> drop or copy to CPU such packets.
>>=20
>> Regards,
>> Shahram
>>=20
>> -----Original Message-----
>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf O=
f
>> Stewart Bryant
>> Sent: Monday, August 05, 2013 2:29 AM
>> To: S. Davari
>> Cc: mpls@ietf.org; tictoc@ietf.org
>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>=20
>> My concern is that we are creating a new LSP type in MPLS
>> (which is a architectural change and thus needs to be very carefully
>> considered) without asking the question "how can I do this
>> in the most general way, to maximize flexibility/reuse"
>>=20
>> I suggested an LSP type that has the properties "timestamp
>> and pass to application", but I think that Sasha raises a good
>> point about using router alert. However RA has no implicit
>> timestamp, so maybe we need a new type of RA that has the
>> properties "timestamp, and pass top application indicated by
>> the next label". That would be quite useful in a number
>> of OAM applications. There is possibly some GAL variant
>> of that design that should also be considered.
>>=20
>> The application could worry about the path and where it
>> was necessary to skip some hops a hierarchical LSP would
>> accomplish that, with the specific benefit that the application
>> would consciously do this, and may be able to apply some
>> form of compensation within the network.
>>=20
>> - Stewart
>>=20
>> On 04/08/2013 10:40, S. Davari wrote:
>>> Hi Sasha
>>>=20
>>> Perhaps you have not understood the draft well. The main reason that the=

>> draft chose to use a dedicated LSP that is signaled via RSVPTE indicating=
 it is a
>> timing LSP is to be backward compatible with non-1588 nodes. Such nodes
>> simply just switch the packet.
>>>=20
>>> Your proposal had been considered and was rejected since it was not
>> backward compatible. Nodes receiving alert label or TTL =3D 1 send the pa=
cket to
>> CPU. You can refer to the meeting notes.
>>>=20
>>> Regards,
>>> Shahram
>>>=20
>>>=20
>>> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein
>> <Alexander.Vainshtein@ecitele.com> wrote:
>>>=20
>>>> Stewart and all,
>>>> I concur with Stewart's statement that the draft does not define "full
>> interaction with MPLS architecture".
>>>>=20
>>>> E.g., one of the objectives of the draft is to provide a technique that=
 would
>> be backward-compatible with old LSRs that cannot provide on-path support f=
or
>> timing distribution, while the other objective is to make every LSR on th=
e path
>> aware that some MPLS packets are carrying timing-related messages and hen=
ce
>> require on-path support.
>>>>=20
>>>> IMHO and FWIW, the MPLS data plane architecture allows just two methods=

>> for making a transit LSR to provide special processing to a labeled packe=
t:
>>>> - It would carry some kind of an "alert label" on top of the label stac=
k and,
>> specifically, on top of any labels used for actual forwarding
>>>> OR,
>>>> - The TTL in the top label stack entry has been set to 1.
>>>>=20
>>>>=20
>>>> The draft does not follow any of these approaches.
>>>>=20
>>>> I must also admit that Section 12 "OAM, Control and Management" of the
>> draft looks somewhat in
>>>> E.g., I could not understand from the text whether BFD, LSP-Ping etc. O=
AM
>> messages would be subjected to on-path support procedures for timing
>> messages or not.
>>>>=20
>>>> My 2c,
>>>>     Sasha
>>>>=20
>>>>> -----Original Message-----
>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Beha=
lf
>> Of
>>>>> Stewart Bryant
>>>>> Sent: Friday, August 02, 2013 12:01 PM
>>>>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
>>>>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.org=
;
>> draft-
>>>>> ietf-tictoc-1588overmpls@tools.ietf.org
>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
>>>>>=20
>>>>> SB> This draft does not seem to provide a precise definition
>>>>> SB> the properties of the new LSP type that it wishes to
>>>>> SB> define, in particular it does it define the PHB of those
>>>>> SB> LSPs, nor the full interaction with the MPLS
>>>>> SB> architecture.
>>>>> SB>
>>>>> SB> I have not tracked TICTOC for a while but I thought that
>>>>> SB> the original plan was to define the concept of an offset
>>>>> SB> into a packet to do the correction.
>>>>> SB>
>>>>> SB> It is disappointing that the opportunity was not taken
>>>>> SB> to define a timing shim inside the timing LSP so that
>>>>> SB> a time correction could be added to any packet such that
>>>>> SB> the MPLS system was isolated from the details of the
>>>>> SB> complexity of the particular time transfer type.
>>>>>=20
>>>>> SB> I think that much more clarify is needed in terms of
>>>>> SB> definition of the new LSP type, since it is unclear
>>>>> SB> from this text how to implement one.
>>>>> SB>
>>>>> SB> There are a lot of other MPLS services such as
>>>>> SB> LSP ping that need to be considered.
>>>>> SB>
>>>>> SB> Please see inline for more comments. However these
>>>>> SB> comments are made in the context of the text as written
>>>>> SB> whilst I have a fundamental concern that this approach
>>>>> SB> lacks an MPLS architectural soundness that need
>>>>> SB> greater thought with significant impact on the
>>>>> SB> draft.
>>>>>=20
>>>>> - Stewart
>>>>>=20
>>>>>=20
>>>>> TICTOC Working Group                                           S. Dava=
ri
>>>>> Internet-Draft                                                   A. Or=
en
>>>>> Intended status: Standards Track                          Broadcom Cor=
p.
>>>>> Expires: December 17, 2013                                     M. Bhat=
ia
>>>>>                                                               P. Rober=
ts
>>>>>                                                           Alcatel-Luce=
nt
>>>>>                                                               L. Monti=
ni
>>>>>                                                               L. Marti=
ni
>>>>>                                                            Cisco Syste=
ms
>>>>>                                                            June 15, 20=
13
>>>>>=20
>>>>>=20
>>>>>             Transporting Timing messages over MPLS Networks
>>>>>                    draft-ietf-tictoc-1588overmpls-05
>>>>>=20
>>>>> Abstract
>>>>>=20
>>>>>    This document defines the method for transporting Timing messages
>>>>>    such as PTP and NTP over an MPLS network.  The method allows for th=
e
>>>>>    easy identification of these PDUs at the port level to allow for po=
rt
>>>>>=20
>>>>> SB> What is a port
>>>>>=20
>>>>>    level processing of these PDUs in both LERs and LSRs.
>>>>>=20
>>>>>    The basic idea is to transport Timing messages inside dedicated MPL=
S
>>>>>    LSPs.  These LSPs only carry Timing messages and possibly Control a=
nd
>>>>>    Management packets, but they do not carry customer traffic.
>>>>>=20
>>>>> SB> More specifically they only carry traffic associated with the
>>>>> SB> timing service and its support.
>>>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
>>>>> SB> also it gets carried in a structure that causes it to get
>>>>> SB> timestamped.
>>>>>=20
>>>>>    Two methods for transporting Timing messages over MPLS are defined.=

>>>>>=20
>>>>> SB> Perhaps the right approach is to define the new LSP type and then
>>>>> SB> seperately to define  the mapping of the various timing services
>>>>> SB> over that LSP type.
>>>>>=20
>>>>>    The first method is to transport Timing messages directly over the
>>>>>    dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
>>>>>    MPLS networks.  The second method is to transport Timing messages
>>>>>    inside a PW via Ethernet encapsulation.
>>>>>=20
>>>>> SB> I think that we should note that there are some
>>>>> SB> h/w reasons for this preference. A clean sheet approach
>>>>> SB> would have been to use PTP over MPLS with no intermediate
>>>>> SB> layers.
>>>>>=20
>>>>> Status of this Memo
>>>>>=20
>>>>>    This Internet-Draft is submitted in full conformance with the
>>>>>    provisions of BCP 78 and BCP 79.
>>>>>=20
>>>>>    Internet-Drafts are working documents of the Internet Engineering
>>>>>    Task Force (IETF).  Note that other groups may also distribute
>>>>>    working documents as Internet-Drafts.  The list of current Internet=
-
>>>>>    Drafts is at http://datatracker.ietf.org/drafts/current/.
>>>>>=20
>>>>>    Internet-Drafts are draft documents valid for a maximum of six mont=
hs
>>>>>    and may be updated, replaced, or obsoleted by other documents at an=
y
>>>>>    time.  It is inappropriate to use Internet-Drafts as reference
>>>>>    material or to cite them other than as "work in progress."
>>>>>=20
>>>>>    This Internet-Draft will expire on December 17, 2013.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013               [Page 1=
]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> Copyright Notice
>>>>>=20
>>>>>    Copyright (c) 2013 IETF Trust and the persons identified as the
>>>>>    document authors.  All rights reserved.
>>>>>=20
>>>>>    This document is subject to BCP 78 and the IETF Trust's Legal
>>>>>    Provisions Relating to IETF Documents
>>>>>    (http://trustee.ietf.org/license-info) in effect on the date of
>>>>>    publication of this document.  Please review these documents
>>>>>    carefully, as they describe your rights and restrictions with respe=
ct
>>>>>    to this document.  Code Components extracted from this document mus=
t
>>>>>    include Simplified BSD License text as described in Section 4.e of
>>>>>    the Trust Legal Provisions and are provided without warranty as
>>>>>    described in the Simplified BSD License.
>>>>>=20
>>>>>=20
>>>>> Table of Contents
>>>>>=20
>>>>>    1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . . =
 5
>>>>>=20
>>>>>    2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . . =
 7
>>>>>=20
>>>>>    3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . . =
 8
>>>>>=20
>>>>>    4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . . =
 9
>>>>>=20
>>>>>    5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . . 1=
2
>>>>>=20
>>>>>    6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . . 1=
3
>>>>>      6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . . 1=
3
>>>>>      6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . . 1=
3
>>>>>      6.3.  Other Timing Encapsulation methods . . . . . . . . . . . . 1=
4
>>>>>=20
>>>>>    7.  Timing message Processing  . . . . . . . . . . . . . . . . . . 1=
5
>>>>>=20
>>>>>    8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . . 1=
6
>>>>>=20
>>>>>    9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1=
7
>>>>>=20
>>>>>    10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1=
8
>>>>>=20
>>>>>    11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . . 1=
9
>>>>>=20
>>>>>    12. OAM, Control and Management  . . . . . . . . . . . . . . . . . 2=
0
>>>>>=20
>>>>>    13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . . 2=
1
>>>>>=20
>>>>>    14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . . 2=
2
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013               [Page 2=
]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>>    15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . . 2=
3
>>>>>      15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . . 2=
3
>>>>>      15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . . 2=
3
>>>>>      15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . . 2=
4
>>>>>=20
>>>>>    16. Other considerations . . . . . . . . . . . . . . . . . . . . . 2=
5
>>>>>=20
>>>>>    17. Security Considerations  . . . . . . . . . . . . . . . . . . . 2=
6
>>>>>=20
>>>>>    18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 2=
7
>>>>>=20
>>>>>    19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 2=
8
>>>>>=20
>>>>>    20. References . . . . . . . . . . . . . . . . . . . . . . . . . . 2=
9
>>>>>      20.1. Normative References . . . . . . . . . . . . . . . . . . . 2=
9
>>>>>      20.2. Informative References . . . . . . . . . . . . . . . . . . 2=
9
>>>>>=20
>>>>>    Appendix 1.  Routing extensions for Timing-aware Routers . . . . . 3=
2
>>>>>=20
>>>>>    Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . . 3=
3
>>>>>=20
>>>>>    Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 3=
4
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013               [Page 3=
]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>>    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
>> NOT",
>>>>>    "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL"
>> in
>>>>> this
>>>>>    document are to be interpreted as described in RFC2119 [RFC2119].
>>>>>=20
>>>>>    When used in lower case, these words convey their typical use in
>>>>>    common language, and are not to be interpreted as described in
>>>>>    RFC2119 [RFC2119].
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013               [Page 4=
]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 1.  Introduction
>>>>>=20
>>>>>    The objective of Precision Time Protocol (PTP) and Network Timing
>>>>>    Protocol (NTP) are to synchronize independent clocks running on
>>>>>    separate nodes of a distributed system.
>>>>>=20
>>>>>    [IEEE-1588] defines PTP messages for frequency, phase and time
>>>>>    synchronization.  The PTP messages include PTP PDUs over UDP/IP
>>>>>    (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F o=
f
>>>>>    [IEEE-1588]).
>>>>>=20
>>>>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
>>>>> SB> of other PTP mappings if they provide better optimisation.
>>>>>=20
>>>>>    This document defines mapping and transport of the PTP
>>>>>    messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>>>>>    defines several clock types: ordinary clocks, boundary clocks, end-=

>>>>>    to-end transparent clocks, and peer-to-peer transparent clocks.
>>>>>    Transparent clocks require intermediate nodes to update correction
>>>>>    field inside PTP message that reflects the transit time in the node=
.
>>>>>=20
>>>>>    [RFC5905] defines NTP messages for clock and time synchronization.
>>>>>    The PTP messages (PDUs) are transported over UDP/IP.  This document=

>>>>> SB> Should that be NTP messages?
>>>>> SB> It needs to be made clear as soon as you introduce NTP that
>>>>> SB> they use different time representations.
>>>>>=20
>>>>>    defines mapping and transport of the NTP messages defined in
>>>>>    [RFC5905] over MPLS networks.
>>>>>=20
>>>>>    One key attribute of all of these Timing messages is that the Time
>>>>>    stamp processing should occur as close as possible to the actual
>>>>>    transmission and reception at the physical port interface.  This
>>>>>    targets optimal time and/or frequency recovery by avoiding variable=

>>>>>    delay introduced by queues internal to the clocks.
>>>>>=20
>>>>> SB> As I recall NTP has no epoch point defined, and I am not sure
>>>>> SB> where that point is in the case of PTP in this mapping
>>>>> SB> Hopefully this will get defined in due course.
>>>>>=20
>>>>>    To facilitate the fast and efficient recognition of Timing messages=

>>>>>    at the port level when the Timing messages are carried over MPLS
>>>>>    LSPs,
>>>>>=20
>>>>> SB> Over a new LSP type with time optimied characteristics
>>>>>=20
>>>>>    this document defines the specific encapsulations that should
>>>>>    be used.
>>>>> SB> Hopefully it will also define the PHP
>>>>>=20
>>>>>    In addition, it can be expected that there will exist LSR/
>>>>>    LERs where only a subset of the physical ports will have the port-
>>>>>    based Timing message processing capabilities.
>>>>> SB> Do you need to clarify that this only works at base and not in
>>>>> SB> a label heirarchy.
>>>>>=20
>>>>>=20
>>>>>    In order to ensure
>>>>>    that the LSPs carrying Timing packets always enter and exit ports
>>>>>    with this capability, routing extensions are defined to advertise
>>>>>    this capability on a port basis and to allow for the establishment o=
f
>>>>>    LSPs that only transit such ports.  While this path establishment
>>>>>    restriction may be applied only at the LER Ingress and/or egress
>>>>>    ports, it becomes more important when using transparent clock capab=
le
>>>>>    LSRs in the path.
>>>>> SB> I do not understand the implications of the last
>>>>> SB> sentences - starting ", it becomes"
>>>>>=20
>>>>>=20
>>>>>    Port based Timing message processing involves Timing message
>>>>>    recognition.  Once the Timing messages are recognized they can be
>>>>>    modified based on the reception or transmission Time-stamp.
>>>>>=20
>>>>>    This document provides two methods for transporting Timing messages=

>>>>>    over MPLS.  One is applicable to MPLS environment and the other one=

>>>>>    is applicable to MPLS/MPLS-TP environment
>>>>>=20
>>>>> SB> I think the sentence is incomplete.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013               [Page 5=
]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>>    The solution involves transporting Timing messages over dedicated
>>>>>    LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
>>>>>    carry Management and control messages, but not data plane client
>>>>>    traffic.
>>>>>=20
>>>>> SB> It is not clear why this restriction applies.
>>>>>=20
>>>>>    Timing LSPs can be established statically or via signaling.
>>>>> SB> s/statically/by provisioning/network management/
>>>>>=20
>>>>>    Extensions to control plane (OSPF, ISIS, etc.) is required to enabl=
e
>>>>>    routers to distribute their Timing processing capabilities over MPL=
S
>>>>>    to other routers.  However such extensions are outside the scope of=

>>>>>    this document.
>>>>>=20
>>>>>    When signaling is used to setup the PTP LSP, Extensions to signalin=
g
>>>>> SB> is it a PTP LSP or a Timing LSP?
>>>>>=20
>>>>>    protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
>>>>>    However such extensions are outside the scope of this document.
>>>>>=20
>>>>> SB> for mpls-tp GMPLS is the signalling protocol
>>>>>=20
>>>>>    While the techniques included herein allow for the establishment of=

>>>>>    paths optimized to include Time-stamping capable links, the
>>>>>    performance of the Slave clocks is outside the scope of this
>>>>>    document.
>>>>>=20
>>>>>    At the time of publishing this specification, Transparent Clocking
>>>>>    (TC) is only defined for PTP.  Therefore at this time any part of
>>>>>    this specification that talks about Transparent Clocking applies on=
ly
>>>>>    to PTP.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013               [Page 6=
]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 2.  Terminology
>>>>>=20
>>>>>    1588: The timing and synchronization as defined by IEEE 1588.
>>>>>=20
>>>>> SB> I think that there is a more formal name for the 1588 group
>>>>> SB> that needs to be used here.
>>>>> SB> Also do we need to talk about 1588-200? as there is
>>>>> SB> an update in progress
>>>>>=20
>>>>>    NTP: The timing and synchronization protocol defined by IETF RFC-13=
05
>>>>>    and RFC-5905.
>>>>>=20
>>>>>    PTP: The timing and synchronization protocol used by 1588.
>>>>> SB> need the proper name for 1588
>>>>>=20
>>>>>    Master Clock: The source of 1588 timing to a set of slave clocks.
>>>>>=20
>>>>>    Master Port: A port on a ordinary or boundary clock that is in Mast=
er
>>>>>    state.  This is the source of timing toward slave ports.
>>>>>=20
>>>>> SB> I am not sure the reader knows what a port is
>>>>>=20
>>>>>    Slave Clock: A receiver of 1588 timing from a master clock.
>>>>>=20
>>>>>    Slave Port: A port on a boundary clock or ordinary clock that is
>>>>>    receiving timing from a master clock.
>>>>>=20
>>>>>    Ordinary Clock: A device with a single PTP port.
>>>>>=20
>>>>>    Transparent Clock.  A device that measures the time taken for a PTP=

>>>>>    event message to transit the device and then updates the
>>>>>    correctionField of the message with this transit time.
>>>>>=20
>>>>>    Boundary Clock: A device with more than one PTP port.  Generally
>>>>>    boundary clocks will have one port in slave state to receive timing=

>>>>>    and then other ports in master state to re-distribute the timing.
>>>>>=20
>>>>>    PTP LSP: An LSP dedicated to carry PTP messages
>>>>>=20
>>>>> SB> PTP or timing?
>>>>>=20
>>>>>=20
>>>>>    PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>>>>    messages.
>>>>>=20
>>>>> SB> Ah I don't think that PWE3 know what one of these is
>>>>>=20
>>>>>    CW: Pseudowire Control Word
>>>>>=20
>>>>>    LAG: Link Aggregation
>>>>>=20
>>>>>    ECMP: Equal Cost Multipath
>>>>>=20
>>>>>    CF: Correction Field, a field inside certain PTP messages (message
>>>>>    type 0-3)that holds the accumulative transit time inside intermedia=
te
>>>>>    switches
>>>>>=20
>>>>>    Timing messages: Timing Protocol messages that are exchanged
>> between
>>>>>    routers in order to establish a synchronized clock.
>>>>>=20
>>>>> SB> A number of these definitions look like copies of IEEE1588
>>>>> SB> definitions. We need to provide references and note the
>>>>> SB> priority of the IEEE base reference.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013               [Page 7=
]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 3.  Problem Statement
>>>>>=20
>>>>>    [IEEE-1588] has defined methods for transporting PTP messages over
>>>>>    Ethernet and IP networks.  [RFC5905] has defined the method of
>>>>>    transporting NTP messages over IP networks.  There is a need to
>>>>>    transport Timing messages over MPLS networks while supporting the
>>>>>    Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)=

>>>>>    functionality in the LER and LSRs in the MPLS network.
>>>>>=20
>>>>>    There are multiple ways of transporting Timing over MPLS.  However,=

>>>>>    there is a requirement to limit the possible encapsulation options t=
o
>>>>>    simplify the Timing message identification and processing required a=
t
>>>>>    the port level.
>>>>>=20
>>>>>    When Timing-awareness is needed, Timing messages should not be
>>>>>    transported over LSPs or PWs that are carrying customer traffic
>>>>>    because LSRs perform Label switching based on the top label in the
>>>>>    stack.
>>>>>=20
>>>>> SB> Have you explained why?
>>>>>=20
>>>>>    To detect Timing messages inside such LSPs require special
>>>>>    hardware to do deep packet inspection at line rate.  Even if such
>>>>>    hardware exists, the payload can't be deterministically identified b=
y
>>>>>    LSRs because the payload type is a context of the PW label, and the=

>>>>>    PW label and its context are only known to the Edge routers (PEs/
>>>>>    LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CES=
,
>>>>>    etc).  Even if one restricts an LSP to only carry Ethernet PWs, the=

>>>>>    LSRs dont have the knowledge of whether PW Control Word (CW) is
>>>>>    present or not and therefore can not deterministically identify the=

>>>>>    payload.
>>>>>=20
>>>>>    A generic method is defined in this document that does not require
>>>>>    deep packet inspection at line rate, and can deterministically
>>>>>    identify Timing messages.  This method can be used to detect Timing=

>>>>>    Messages in both one-step and two-step clock implementations of
>>>>>    ordinary, boundary and transparent clocks.
>>>>>=20
>>>>> SB> Needs a ref and I am sure many MPLS specialists will not understan=
d
>>>>> SB> the msg types.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013               [Page 8=
]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 4.  Timing over MPLS Architecture
>>>>>=20
>>>>>    Timing messages are exchange between Timing ports on ordinary and
>>>>>=20
>>>>> SB> Have you defined a timing port?
>>>>>=20
>>>>>    boundary clocks.  Boundary clocks terminate the Timing messages and=

>>>>>    act as master for other boundary clocks or for slave clocks.  End-t=
o-
>>>>>    End Transparent clocks do not terminate the Timing messages but the=
y
>>>>>    do modify the contents of the Timing messages as they transit acros=
s
>>>>>    the transparent clock.
>>>>>=20
>>>>>    Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Clo=
ck
>>>>>=20
>>>>>    (TC) could be implemented in either LERs or LSRs.
>>>>>=20
>>>>> SB> LER and LSR need to be expanded
>>>>>=20
>>>>>    An example is shown in Figure 1, where the LERs act as Ordinary Clo=
ck
>>>>>    (OC) and are the initiating/terminating point for Timing messages.
>>>>>    The ingress LER encapsulates the Timing messages in Timing LSP and
>>>>>    the Egress LER terminates the Timing LSP.  The LSRs act as
>>>>>    Transparent Clock (TC) and just update the Timing field in the Timi=
ng
>>>>>    messages.
>>>>>=20
>>>>>=20
>>>>>       +--------+     +-------+     +-------+     +-------+     +------=
--+
>>>>>       |Switch, |     |       |     |       |     |       |     |Switch=
, |
>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Route=
r |
>>>>>       |        |     |  OC   |     |  TC   |     |  OC   |     |      =
  |
>>>>>       +--------+     +-------+     +-------+     +-------+     +------=
--+
>>>>>                      /                                 \
>>>>>       +-------+     /                                   \     +-------=
+
>>>>>       |  LER  |    /                                     \    |  LER  |=

>>>>>       | Master|---/                                       \---| Slave |=

>>>>>       | Clock |                                               | Clock |=

>>>>>       +-------+                                               +-------=
+
>>>>>=20
>>>>>      Figure (1) - Deployment example 1 of timing over MPLS network
>>>>>=20
>>>>>    Another example is shown in Figure2, where LERs terminate the Timin=
g
>>>>>    messages received from switch/routers that are outside of the MPLS
>>>>>    network acting as OC or BC.  In this example LERs regenerate the
>>>>>    clock and initiate timing messages encapsulated in Timing LSP towar=
d
>>>>>    the MPLS network, while the LSRs act as Transparent Clock (TC) and
>>>>>    just update the Timing field in the Timing messages, which are
>>>>>    already encapsulated in Timing LSPs.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013               [Page 9=
]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>>      +--------+     +-------+     +-------+     +-------+     +-------=
-+
>>>>>      |Switch, |     |       |     |       |     |       |     |Switch,=
 |
>>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router=
 |
>>>>>      | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC =
 |
>>>>>      +--------+     +-------+     +-------+     +-------+     +-------=
-+
>>>>>=20
>>>>>      Figure (2) - Deployment example 2 of timing over MPLS network
>>>>>=20
>>>>>=20
>>>>>    Another example is shown in Figure 3, where LERs do not terminate t=
he
>>>>>    Timing messages received from switch/routers that are outside of th=
e
>>>>>    MPLS network acting as OC, TC or BC.  The LERs act as TC and update=

>>>>>    the Timing field in the Timing messages as they transit the LER,
>>>>>    while encapsulating them in timing LSP.  The LSRs also act as
>>>>>    Transparent Clock (TC) and just update the Timing field in the Timi=
ng
>>>>>    messages which are already encapsulated in Timing LSPs.
>>>>>=20
>>>>>       +--------+     +-------+     +-------+     +-------+     +------=
--+
>>>>>       |Switch, |     |       |     |       |     |       |     |Switch=
, |
>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Route=
r |
>>>>>       |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC/=
BC|
>>>>>       +--------+     +-------+     +-------+     +-------+     +------=
--+
>>>>>=20
>>>>>     Figure (3) - Deployment example 3 of timing over MPLS network
>>>>>=20
>>>>>    Another example is shown in Figure 4, where LERs and LSRs support
>>>>>    Boundary Clocks.  A single-hop LSP is created between two adjacent
>>>>>    LSRs engaged in BC operation.  Other methods such as PTP transport
>>>>>    over Ethernet MAY be used for transporting timing messages if the
>>>>>    link between the two routers is Ethernet.
>>>>>=20
>>>>>      +--------+     +-------+     +-------+     +-------+     +-------=
-+
>>>>>      |Switch, |     |       |     |       |     |       |     |Switch,=
 |
>>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router=
 |
>>>>>      | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC =
 |
>>>>>      +--------+     +-------+     +-------+     +-------+     +-------=
-+
>>>>>=20
>>>>>    Figure (4) - Deployment example 3 of timing over MPLS network
>>>>>=20
>>>>>    An MPLS domain MAY serve multiple customers.  In these cases the
>> MPLS
>>>>>    domain (maintained by a service provider) may provide timing servic=
es
>>>>>    to multiple customers, each having their own Timing domain.
>>>>>=20
>>>>>    The Timing over MPLS architecture assumes full mesh of Timing LSPs
>>>>>    between all LERs supporting this specification.
>>>>>=20
>>>>> SB> Note sure this is right - the salves surely do not need to
>>>>> SB> exchange timing amongst themselves
>>>>>=20
>>>>>    It supports
>>>>>    Point-to- point (VPWS) and Multipoint (VPLS) services.
>>>>>=20
>>>>> SB> What does that mean? You do not carry user data traffic?
>>>>> SB> Maybe it's the ordering of the statemnets that is causing
>>>>> SB> confusion.
>>>>>=20
>>>>>    This means
>>>>>    that a customer may purchase a Point-to-point Timing service betwee=
n
>>>>>    two customer sites or a Multipoint Timing service between more than=

>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 1=
0]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>>    two customer sites.
>>>>>=20
>>>>>    The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.=

>>>>>    This means that the Timing Multicast messages such as PTP Multicast=

>>>>>    event messages can be transported over P2MP Timing LSP or be
>>>>>    replicated and transported over many P2P Timing LSPs.
>>>>>=20
>>>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>>>>> SB> nor a P2MP PW, although we are close.
>>>>>=20
>>>>>    Timing messages, that do not require Time stamping or Correction
>>>>>    Field update MAY be transported over Timing LSPs to simplify hardwa=
re
>>>>>    and software.
>>>>>=20
>>>>>    PTP Announce messages that determine the Timing LSP terminating
>> point
>>>>>    behavior such as BC/OC/TC SHOULD be transported over the Timing LSP=

>>>>>    to simplify hardware and software.
>>>>>=20
>>>>> SB> have you defined and referenced PTP announce msgs?
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 1=
1]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 5.  Dedicated LSPs for Timing messages
>>>>>=20
>>>>>    Many methods have been considered for identifying the Timing
>> messages
>>>>>    when they are encapsulated in MPLS such as using GAL/G-ACH or a new=

>>>>>    reserved label.  These methods were not attractive since they eithe=
r
>>>>>    required deep packet inspection at line rate in the intermediate LS=
Rs
>>>>>    or they required use of a scarce new reserved label.  Also one of t=
he
>>>>>    goals was to reuse existing OAM mechanisms.
>>>>>=20
>>>>> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
>>>>>=20
>>>>>    The method defined in this document can be used by LER and LSRs to
>>>>>    identify Timing messages in MPLS tunnels by just looking at the top=

>>>>>    label in the MPLS label stack, which only carry Timing messages as
>>>>>    well as OAM, but not data plane client traffic.
>>>>>=20
>>>>>    Compliant implementations MUST use dedicated LSPs to carry Timing
>>>>>    messages over MPLS.
>>>>>=20
>>>>> SB> I think that we need a definition of the properies of these LSPs
>>>>>=20
>>>>>    These LSPs are herein referred to as "Timing
>>>>>    LSPs" and the labels associated with these LSPs as "Timing LSP
>>>>>    labels".  The Timing LSPs that runs between Ingress and Egress LERs=

>>>>>    MUST be co-routed.  Alternatively, a single bidirectional co-routed=

>>>>>    LSP can be used.
>>>>>=20
>>>>> SB> I though that you said you could use M2MP LSPs - these are not
>>>>> SB> bidirectional.
>>>>>=20
>>>>>    Co-routing of the two directions is required to limit the differenc=
e
>>>>>    in the delays in the Master clock to Slave clock direction compared=

>>>>>    to the Slave clock to Master clock direction.  The Timing LSP MAY b=
e
>>>>>    MPLS/MPLS-TP LSP.
>>>>>=20
>>>>>    The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
>>>>>    New Extensions to RSVP-TE/GMPLS TLVs are required; however they are=

>>>>>    outside the scope of this document.
>>>>>=20
>>>>>    The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such a=
s
>>>>>    BFD and LSP Ping but the LSP data plane client plane traffic MUST b=
e
>>>>>    Timing packets only.
>>>>>=20
>>>>> SB> Why?
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 1=
2]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 6.  Timing over LSP Encapsulation
>>>>>=20
>>>>> The encapsulations is not LSP is it?
>>>>>=20
>>>>>    This document defines two methods for carrying Timing messages over=

>>>>>    MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>>>>    messages over Timing LSPs, and the second method, is carrying
>>>>>    Ethernet encapsulated Timing messages over Ethernet PWs inside
>> Timing
>>>>>    LSPs.
>>>>>=20
>>>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>>>>=20
>>>>>    The simplest method of transporting Timing messages over MPLS is to=

>>>>>    encapsulate Timing PDUs in UDP/IP and then encapsulate them in
>> Timing
>>>>>    LSP.  This format is shown in Figure 4.
>>>>>=20
>>>>>=20
>>>>>                     +----------------------+
>>>>>                     |   Timing LSP Label   |
>>>>>                     +----------------------+
>>>>>                     |        IPv4/6        |
>>>>>                     +----------------------+
>>>>>                     |         UDP          |
>>>>>                     +----------------------+
>>>>>                     |     Timing PDU       |
>>>>>                     +----------------------+
>>>>>=20
>>>>>       Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>>>>=20
>>>>>=20
>>>>>    This encapsulation is very simple and is useful when the network
>>>>>    between Timing Master Clock and Slave Clock is MPLS network.
>>>>>=20
>>>>> SB> Simple is a judgement call
>>>>>=20
>>>>>    In order for an LER/LSR to process Timing messages, the Timing LSP
>>>>>    Label must be at the top label of the label stack.  The LER/LSR MUS=
T
>>>>>    know that the Timing LSP Label is used for carrying Timing messages=
.
>>>>>    This can be accomplished via static configuration or via RSVP-TE
>>>>>    signaling.
>>>>>=20
>>>>>    The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>>>>    [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>>>>>    [RFC5905].
>>>>>=20
>>>>> 6.2.  Timing over PW Encapsulation
>>>>>=20
>>>>>    Another method of transporting Timing over MPLS networks is by
>>>>>    encapsulating Timing PDUs in PW which in turn is transported over
>>>>>    Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
>>>>>    shown in Fig 5(A) MUST be used and the Ethernet encapsulation of PT=
P
>>>>>    MUST follow Annex F of [IEEE-1588].
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 1=
3]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>>    The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
>> the
>>>>>    Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>>>>>    Timing over PW encapsulation MUST use the Control Word (CW) as
>>>>>    specified in [RFC4448] to ensure proper detection of PTP messages
>>>>>    inside the MPLS packets for Timing over LSP and Timing over PW
>>>>>    encapsulation.
>>>>>=20
>>>>> SB> That needs explanation
>>>>>=20
>>>>>    The use of Sequence Number in the CW is optional.
>>>>>=20
>>>>> SB> Given that s/n are never in practice deployed, you could probably
>>>>> SB> simplify things by sayig that they are not used.
>>>>>=20
>>>>>    Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over
>> PW
>>>>>    (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>>>>=20
>>>>>                     +----------------+  +----------------+
>>>>>                      |Timing LSP Label|  |Timing LSP Label|
>>>>>                      +----------------+  +----------------+
>>>>>                      |    PW Label    |  |    PW Label    |
>>>>>                      +----------------+  +----------------+
>>>>>                      |  Control Word  |  |      IP        |
>>>>>                      +----------------+  +----------------+
>>>>>                      |    Ethernet    |  |      UDP       |
>>>>>                      |     Header     |  +----------------+
>>>>>                      +----------------+  |   Timing PDU   |
>>>>>                      |S-VLAN(Optional)|  |                |
>>>>>                      +----------------+  +----------------+
>>>>>                      |C-VLAN(Optional)|        (B)
>>>>>                      +----------------+
>>>>>                      |   Timing PDU   |
>>>>>                      |                |
>>>>>                      +----------------+
>>>>>                             (A)
>>>>>=20
>>>>>               Figure (5) - Timing over PW Encapsulations
>>>>>=20
>>>>>    In order for an LSR to process PTP messages, the top label of the
>>>>>    label stack (the Tunnel Label) MUST be a Timing label.
>>>>>=20
>>>>> S> You said that before.
>>>>>=20
>>>>> 6.3.  Other Timing Encapsulation methods
>>>>>=20
>>>>>    In future other timing encapsulation methods may be introduced, suc=
h
>>>>>    as a new shim header after the Bottom of Stack to carry the Timing
>>>>>    information.  Such new encapsulations are outside the scope of this=

>>>>>    document.
>>>>>=20
>>>>>=20
>>>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>>>>> SB> out of the definition of the LSP
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 1=
4]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> SB> I think we need a section on LSP processing
>>>>>=20
>>>>> 7.  Timing message Processing
>>>>>=20
>>>>>    Each Timing protocol such as PTP and NTP, define their set of Timin=
g
>>>>>    messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>>>>>    FOLLOW_UP, etc messages.
>>>>>=20
>>>>>    Some of the Timing messages require time stamping or correction fie=
ld
>>>>>    update at port level and some dont.  It is the job of the LER/LSR t=
o
>>>>>    parse the timing message and find out the type of the Timing messag=
e
>>>>>    and decide whether and how to Time- stamp it (e.g., BC) or update
>>>>>    correction field(e.g., TC).
>>>>>=20
>>>>>=20
>>>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
>>>>> SB> function rather than the LER?
>>>>>=20
>>>>>    For example the following PTP messages (called Event messages)
>>>>>    require time-stamping or correction field update:
>>>>>=20
>>>>>    o  SYNC
>>>>>=20
>>>>>    o  DELAY_REQ (Delay Request)
>>>>>=20
>>>>>    o  PDELAY_REQ (Peer Delay Request)
>>>>>=20
>>>>>    o  PDELAY_RESP (Peer Delay Response)
>>>>>=20
>>>>>    SYNC and DELAY_REQ are exchanged between Master Clock and Slave
>> Clock
>>>>>    and MUST be transported over PTP LSPs.  PDELAY_REQ and
>> PDELAY_RESP
>>>>>    are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>>>>>    Boundary, or Transparent) and SHOULD be transported over single hop=

>>>>>    PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
>>>>>    and PDELAY_RESP_FOLLOW_UP messages MUST also be transported
>> over
>>>>> the
>>>>>    PTP LSPs.
>>>>>=20
>>>>>    For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
>>>>>    transported over two PTP LSPs that are in opposite directions.  The=
se
>>>>>    PTP LSPs, which are in opposite directions MUST be congruent and co=
-
>>>>>    routed.  Alternatively, a single bidirectional co-routed LSP can be=

>>>>>    used.
>>>>>=20
>>>>>    Except as indicated above for the two-step PTP clocks, Non-Event PT=
P
>>>>>    message types do not need to be processed by intermediate routers.
>>>>>    These message types MAY be carried in PTP Tunnel LSPs.
>>>>>=20
>>>>> SB> Are you saying that a timing P router has to be msg type sensitive=
?
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 1=
5]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 8.  Protection and Redundancy
>>>>>=20
>>>>>=20
>>>>> SB> This is a bit of a jump - I don't know how the LSP itself works ye=
t!
>>>>>=20
>>>>>    In order to ensure continuous uninterrupted operation of slave
>>>>>    clocks, usually as a general practice, slave clocks (or ports) trac=
k
>>>>>    redundant master clocks.
>>>>>=20
>>>>>    It is the responsibility of the network operator to ensure that
>>>>>    physically disjoint Timing LSPs are established between a slave clo=
ck
>>>>>    (or port) and redundant master clocks (or ports).
>>>>>=20
>>>>>    When a slave clock (or port) listens to redundant master clocks or
>>>>>    ports, any prolonged Timing LSP outage will trigger the slave clock=

>>>>>    or port to switch to a redundant master clock or port.
>>>>>=20
>>>>>    LSP/PW protection such as Linear protection Switching (1:1, 1+1),
>>>>>    Ring protection switching or MPLS Fast Reroute (FRR) generally swit=
ch
>>>>>    alternative path that usually cause a change in delay, which if
>>>>>    undetected by slave clock can reduce accuracy of the slave clock.
>>>>>=20
>>>>>    Therefore protection switching MAY be used, as long as phase jumps
>>>>>    upon switchover due to differences in path latency are detected and=

>>>>>    compensated for (such compensation not being required if BCs or pee=
r-
>>>>>    peer TCs are used throughout).
>>>>>=20
>>>>>    Note that any protection or reroute mechanism that adds additional
>>>>>    MPLS label to the label stack, such as Facility Backup Fast Reroute=
,
>>>>>    MUST ensure that the pushed label is also a Timing Label to ensure
>>>>>    recognition of the MPLS frame as containing Timing messages, as it
>>>>>    transits the backup path.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 1=
6]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 9.  ECMP
>>>>>=20
>>>>>    To ensure the optimal operation of slave clocks and avoid error
>>>>>    introduced by forward and reverse path delay asymmetry, the physica=
l
>>>>>    path for Timing messages from master clock to slave Clock and vice
>>>>>    versa must be the same for all Event Timing messages listed in
>>>>>    section 7.
>>>>>=20
>>>>>    Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
>>>>>    Multipath).
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 1=
7]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 10.  PHP
>>>>>=20
>>>>>    To ensure that the label on the top of the label stack is the Timin=
g
>>>>>    LSP Label, PHP MUST not be used.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 1=
8]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 11.  Entropy
>>>>>=20
>>>>>    To ensure all Timing messages in a Timing LSP take the same path,
>>>>>    Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>>>>    Entropy Label MUST NOT be used for the PWs that are carried inside
>>>>>    Timing LSP [RFC6391].
>>>>>=20
>>>>> SB> This is incorrect - you mean that all msgs of the same timing
>>>>> SB> flow need to have the same EL value.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 1=
9]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 12.  OAM, Control and Management
>>>>>=20
>>>>>    In order to monitor Timing LSPs and their encapsulated PWs, they MU=
ST
>>>>>    be able to carry OAM and management messages.  These management
>>>>>    messages MUST be differentiated from Timing messages via already
>>>>>    defined IETF methods.
>>>>>=20
>>>>>    For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run=

>>>>>    over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>>>>>    Management protocols can easily be identified by the UDP Destinatio=
n
>>>>>    Port number or by GAL/G-ACH respectively.
>>>>>=20
>>>>>    Also BFD, LSP-Ping and other management messages MAY run over the
>> PWs
>>>>>    encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3 o=
r
>>>>>    4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is goi=
ng
>>>>>    to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1) o=
r
>>>>>    GAL-ACH are used to identify such management messages.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 2=
0]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 13.  QoS Considerations
>>>>>=20
>>>>>    In network deployments where not every LSR/LER is Timing-aware, it i=
s
>>>>>    important to reduce the impact of the non-Timing-aware LSR/LERs on
>>>>>    the timing recovery in the slave clock.  The Timing messages are ti=
me
>>>>>    critical and must be treated with the highest priority.  Therefore
>>>>>    Timing over MPLS messages must be treated with the highest priority=

>>>>>    in the routers.  This can be achieved by proper setup of Timing LSP=
s.
>>>>>=20
>>>>>    It is recommended that the Timing LSPs are setup or configured
>>>>>    properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC2697=
]
>>>>>    for drop eligibility.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 2=
1]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 14.  FCS and Checksum Recalculation
>>>>>=20
>>>>>    When time-stamp generation and timing packet adjustment is performe=
d
>>>>>    near the physical port hardware, the process MUST include
>>>>>    recalculation of the Ethernet FCS.
>>>>>=20
>>>>> SB> The above is confusing - an LSR always recomputes the link layer
>>>>> SB> CRC which may or may not be Ethernet.
>>>>>=20
>>>>>    Also FCS retention for the
>>>>>    payload Ethernet described in [RFC4720] MUST NOT be used.
>>>>>=20
>>>>>    For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum=

>>>>>    may be required as per UDP transport standards.
>>>>>=20
>>>>> SB> You really need to be working on getting the IPv6 C?S computation
>>>>> SB> removed from PTP msgs.
>>>>>=20
>>>>>    When UDP checksum is used, each Timing-aware LER/LSR must either
>>>>>    incrementally update the UDP checksum after Time stamping or
>>>>>    Correction Field update or verify the UDP checksum on reception fro=
m
>>>>>    upstream and recalculate the checksum completely on transmission to=

>>>>>    downstream node after Time stamping or Correction Field update.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 2=
2]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 15.  Behavior of LER/LSR
>>>>>=20
>>>>>    Timing-capable/aware LERs and LSRs are routers that have one or mor=
e
>>>>>=20
>>>>> SB> You mean physical interfaces?
>>>>>=20
>>>>>    interfaces that can perform Timing operations (OC/BC/TC) on Timing
>>>>>    packets and are configured to do so.  Timing-capable/aware LERs and=

>>>>>    LSRs can advertise their Timing-capability per-interface via contro=
l
>>>>>    plane such as OSPF or IS-IS.
>>>>> SB> ISIS and OSPF are routing protocols.
>>>>>=20
>>>>>   The Timing-capable/aware LERs can then
>>>>>    signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timin=
g
>>>>>    capability of LER and LSRs may be configured in a centralized
>>>>>    controller and the Timing LSP may be setup using manual configurati=
on
>>>>>    or other methods such as SDN.
>>>>>=20
>>>>> SB> it can also be configured individually rather then through
>>>>> SB> a cebtral controllwe
>>>>>=20
>>>>> 15.1.  Behavior of Timing-capable/aware LER
>>>>>=20
>>>>>    When a Timing-capable/aware LER behaves as a Transparent clock and
>>>>>    receives a Timing message from a Timing-capable/aware non-MPLS
>>>>>    interface, the LER updates the Correction Field (CF) and encapsulat=
es
>>>>>    and forwards the timing message over previously established Timing
>>>>>    LSP.
>>>>>=20
>>>>> SB> You need to call out the details so that people properly
>>>>> SB> understand the definition of the new LSP.
>>>>>=20
>>>>>    Also when a Timing message is received from a Timing-capable/
>>>>>    aware MPLS interface, LER updates the Correction Filed (CF) and
>>>>>    decapsulates the MPLS encapsulation and forwards the timing message=

>>>>>    to a non-MPLS interface.
>>>>>=20
>>>>>    When a Timing-capable/aware LER behaves as a Boundary clock and
>>>>>    receives a Timing message from a Timing-capable/aware non MPLS
>>>>>    interface, the LER Timestamps the Timing packet and sends it to the=

>>>>>    LERs Boundary clock processing module.  Also when a Timing message i=
s
>>>>>    received from a Timing- capable/aware MPLS interface, the LER
>>>>>    Timestamps the Timing packet and sends it to the LERs Boundary cloc=
k
>>>>>    processing module.
>>>>>=20
>>>>>    When a Timing-capable/aware LER behaves as an Ordinary Clock toward=

>>>>>    the MPLS network, and receives a Timing message from a Timing-
>>>>>    capable/aware MPLS interface, the LER Timestamps the Timing packet
>>>>>    and sends it to the LERs Ordinary clock processing module.
>>>>>=20
>>>>> 15.2.  Behavior of Timing-capable/aware LSR
>>>>>=20
>>>>>    When a Timing-capable/aware LSR behaves as a Transparent clock and
>>>>>    receives a Timing message from a Timing-capable/aware MPLS
>> interface,
>>>>>    The LSR updates the Correction Filed (CF) and forwards the timing
>>>>>    message over another MPLS interface.
>>>>>=20
>>>>>    When a Timing-capable/aware LSR behaves as a Boundary clock and
>>>>>    receives a Timing message from a Timing-capable/aware MPLS
>> interface.
>>>>>    The LSR performs the functions of a Boundary Clock in terminating t=
he
>>>>>    received Timing message and re-generating a new timing message over=

>>>>>    another (or the same) MPLS interface.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 2=
3]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>>>>=20
>>>>>    It is most beneficial when all LSRs in the path of a Timing LSP be
>>>>>    timing-Capable/aware LSRs.  This would ensure the highest quality
>>>>>    time and clock synchronization by Timing Slave Clocks.  However, th=
is
>>>>>    specification does not mandate that all LSRs in path of a Timing LS=
P
>>>>>    be Timing- capable/aware.
>>>>>=20
>>>>>    Non-Timing-capable/aware LSRs just switch the packets encapsulated i=
n
>>>>>    Timing LSPs and dont perform any Timing operation (TC or BC).
>>>>>    However as explained in QoS section the Timing over MPLS packets
>> MUST
>>>>>    be still be treated with the highest priority based on their Traffi=
c
>>>>>    Class (TC) marking.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 2=
4]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 16.  Other considerations
>>>>>=20
>>>>>    [IEEE-1588] defines an optional peer-to-peer Transparent clocking
>>>>>    that requires peer delay measurement between two adjacent Timing-
>>>>>    capable/ aware routers/switches.  Peer delay measurement messages
>>>>>    need to be time stamped and terminated by the Timing-capable/aware
>>>>>    routers/ switches.  This means that two adjacent LSRs may be engage=
d
>>>>>    in a peer delay measurement.
>>>>>=20
>>>>>    For transporting such peer delay measurement messages a single-hop
>>>>>    LSP SHOULD to be created between the two adjacent LSRs engaged in
>>>>>    peer delay measurement to carry peer delay measurement messages.
>>>>>    Other methods such as PTP transport over Ethernet MAY be used for
>>>>>    transporting peer delay measurement messages if the link between th=
e
>>>>>    two routers is Ethernet.
>>>>>=20
>>>>>    In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ wa=
re
>>>>>    routers/switches MUST maintain a list of all the neighbors it needs=

>>>>>    to send a PDelay_Req to, where each neighbor corresponds to a timin=
g
>>>>>    LSP.
>>>>>=20
>>>>>    The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as l=
ong
>>>>>    as either the Explicit Null label is the bottom of stack label
>>>>>    (applicable only to UDP/IP encapsulation) or the label below the
>>>>>    Explicit Null label is a PTP label.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 2=
5]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 17.  Security Considerations
>>>>>=20
>>>>>    MPLS PW security considerations in general are discussed in [RFC398=
5]
>>>>>    and [RFC4447],and those considerations also apply to this document.=

>>>>>=20
>>>>>    An experimental security protocol is defined in [IEEE-1588].The PTP=

>>>>>    security extension and protocol provides group source authenticatio=
n,
>>>>>    message integrity, and replay attack protection for PTP messages.
>>>>>=20
>>>>>    When the MPLS network (provider network) serves multiple customers,=

>>>>>    it is important to maintain and process each customers clock and
>>>>>    Timing messages separately from other customers to ensure there is n=
o
>>>>>    cross- customer effect.  For example if an LER BC is synchronized t=
o
>>>>>    a specific grandmaster, belonging to customer A, then the LER MUST
>>>>>    use that BC clock only for customer A to ensure that customer A
>>>>>    cannot attack other customers by manipulating its time.
>>>>>=20
>>>>>    Timing messages MAY be encrypted or authenticated, provided that th=
e
>>>>>    LERs/LSRs that are Timing capable/aware can authenticate/ decrypt t=
he
>>>>>    timing messages.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 2=
6]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 18.  Acknowledgements
>>>>>=20
>>>>>    The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizrah=
i,
>>>>>    Stefano Ruffini, Peter Meyer, and other members of IETF for reviewi=
ng
>>>>>    and providing feedback on this draft.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 2=
7]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 19.  IANA Considerations
>>>>>=20
>>>>>    There are no IANA requirements in this specification.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 2=
8]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 20.  References
>>>>>=20
>>>>> 20.1.  Normative References
>>>>>=20
>>>>>    [IEEE-1588]
>>>>>               IEEE 1588-2008, "IEEE Standard for a Precision Clock
>>>>>               Synchronization Protocol for Networked Measurement and
>>>>>               Control Systems".
>>>>>=20
>>>>>    [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>>>>               Requirement Levels", BCP 14, RFC 2119, March 1997.
>>>>>=20
>>>>>    [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
>>>>>               Edge (PWE3) Architecture", RFC 3985, March 2005.
>>>>>=20
>>>>>    [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discover=
y
>>>>>               Proxies (ND Proxy)", RFC 4389, April 2006.
>>>>>=20
>>>>>    [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
>>>>>               Heron, "Pseudowire Setup and Maintenance Using the Label=

>>>>>               Distribution Protocol (LDP)", RFC 4447, April 2006.
>>>>>=20
>>>>>    [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>>>>               "Encapsulation Methods for Transport of Ethernet over MP=
LS
>>>>>               Networks", RFC 4448, April 2006.
>>>>>=20
>>>>>    [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>>>>               Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>>>>               Retention", RFC 4720, November 2006.
>>>>>=20
>>>>>    [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit=

>>>>>               Connectivity Verification (VCCV): A Control Channel for
>>>>>               Pseudowires", RFC 5085, December 2007.
>>>>>=20
>>>>>    [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detectio=
n
>>>>>               (BFD)", RFC 5880, June 2010.
>>>>>=20
>>>>>    [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
>>>>>               "Bidirectional Forwarding Detection (BFD) for MPLS Label=

>>>>>               Switched Paths (LSPs)", RFC 5884, June 2010.
>>>>>=20
>>>>> 20.2.  Informative References
>>>>>=20
>>>>>    [I-D.ietf-pwe3-fat-pw]
>>>>>               Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan=
,
>>>>>               J., and S. Amante, "Flow Aware Transport of Pseudowires
>>>>>               over an MPLS Packet Switched Network",
>>>>>               draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.=

>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 2=
9]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>>    [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate=

>>>>>               system routeing information exchange protocol for use in=

>>>>>               conjunction with the Protocol for providing the
>>>>>               Connectionless-mode Network Service (ISO 8473)".
>>>>>=20
>>>>>    [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
>>>>>               dual environments", RFC 1195, December 1990.
>>>>>=20
>>>>>    [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.=

>>>>>=20
>>>>>    [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
>>>>>               Marker", RFC 2697, September 1999.
>>>>>=20
>>>>>    [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boudec=
,
>>>>>               J., Courtney, W., Davari, S., Firoiu, V., and D.
>>>>>               Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>>>>               Behavior)", RFC 3246, March 2002.
>>>>>=20
>>>>>    [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineeri=
ng
>>>>>               (TE) Extensions to OSPF Version 2", RFC 3630,
>>>>>               September 2003.
>>>>>=20
>>>>>    [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate=

>>>>>               System (IS-IS) Extensions for Traffic Engineering (TE)",=

>>>>>               RFC 3784, June 2004.
>>>>>=20
>>>>>    [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.=

>>>>>               Shaffer, "Extensions to OSPF for Advertising Optional
>>>>>               Router Capabilities", RFC 4970, July 2007.
>>>>>=20
>>>>>    [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
>>>>>               System to Intermediate System (IS-IS) Extensions for
>>>>>               Advertising Router Information", RFC 4971, July 2007.
>>>>>=20
>>>>>    [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
>>>>>               Topology (MT) Routing in Intermediate System to
>>>>>               Intermediate Systems (IS-ISs)", RFC 5120, February 2008.=

>>>>>=20
>>>>>    [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>>>>               Engineering", RFC 5305, October 2008.
>>>>>=20
>>>>>    [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>>>>               "Traffic Engineering Extensions to OSPF Version 3",
>>>>>               RFC 5329, September 2008.
>>>>>=20
>>>>>    [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
>>>>>               for IPv6", RFC 5340, July 2008.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 3=
0]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>>    [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Netwo=
rk
>>>>>               Time Protocol Version 4: Protocol and Algorithms
>>>>>               Specification", RFC 5905, June 2010.
>>>>>=20
>>>>>    [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan=
,
>>>>>               J., and S. Amante, "Flow-Aware Transport of Pseudowires
>>>>>               over an MPLS Packet Switched Network", RFC 6391,
>>>>>               November 2011.
>>>>>=20
>>>>>    [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and=

>>>>>               L. Yong, "The Use of Entropy Labels in MPLS Forwarding",=

>>>>>               RFC 6790, November 2012.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 3=
1]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 1.  Routing extensions for Timing-aware Routers
>>>>>=20
>>>>>    MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] an=
d
>>>>>    IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (TE=
)
>>>>>    link information used for constraint-based routing.
>>>>>=20
>>>>>    Indeed, it is useful to advertise data plane TE router link
>>>>>    capabilities, such as the capability for a router to be Timing-awar=
e.
>>>>>    This capability MUST then be taken into account during path
>>>>>    computation to prefer or even require links that advertise themselv=
es
>>>>>    as Timing-aware.  In this way the path can ensure the entry and exi=
t
>>>>>    points into the LERs and, if desired, the links into the LSRs are
>>>>>    able to perform port based time-stamping thus minimizing their impa=
ct
>>>>>    on the performance of the slave clock.
>>>>>=20
>>>>>    extensions are required to OSPF and IS-IS in order to advertise
>>>>>    Timing-aware capabilities of a link.  Such extensions are outside t=
he
>>>>>    scope of this document; however such extension SHOULD be able to
>>>>>    signal the following information per Router Link:
>>>>>=20
>>>>>    o  Capable of processing PTP, NTP or other Timing flows
>>>>>=20
>>>>>    o  Capable of performing Transparent Clock operation
>>>>>=20
>>>>>    o  Capable of performing Boundary Clock operation
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 3=
2]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> 2.  Signaling Extensions for Creating Timing LSPs
>>>>>=20
>>>>>    RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-=
TE
>>>>>    is used to setup Timing LSPs, some information that indicates that
>>>>>    the LSP is carrying Timing flows MUST be included in the new
>>>>>    Extensions to RSVP-TE:
>>>>>=20
>>>>>    The following information MAY also be included in the new Extension=
s
>>>>>    to RSVP-TE:
>>>>>=20
>>>>>    o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp=

>>>>>       field
>>>>>=20
>>>>>    o  Number of VLANs in case of PW encapsulation
>>>>>=20
>>>>>    o  Timestamp field Type
>>>>>=20
>>>>>       *  Correction Field, Timestamp
>>>>>=20
>>>>>    o  Timestamp Field format
>>>>>=20
>>>>>       *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
>>>>>          NTP, etc.
>>>>>=20
>>>>>    Note that in case the above optional information is signaled with
>>>>>    RSVP-TE for a Timing LSP, all the Timing packets carried in that LS=
P
>>>>>    must have the same signaled characteristics.  For example if
>>>>>    Timestamp format is signaled as 64-bit PTPv1, then all Timing packe=
ts
>>>>>    must use 64-bit PTPv1 time-stamp.
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 3=
3]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>> Authors' Addresses
>>>>>=20
>>>>>    Shahram Davari
>>>>>    Broadcom Corp.
>>>>>    San Jose, CA  95134
>>>>>    USA
>>>>>=20
>>>>>    Email: davari@broadcom.com
>>>>>=20
>>>>>=20
>>>>>    Amit Oren
>>>>>    Broadcom Corp.
>>>>>    San Jose, CA  95134
>>>>>    USA
>>>>>=20
>>>>>    Email: amito@broadcom.com
>>>>>=20
>>>>>=20
>>>>>    Manav Bhatia
>>>>>    Alcatel-Lucent
>>>>>    Bangalore,
>>>>>    India
>>>>>=20
>>>>>    Email: manav.bhatia@alcatel-lucent.com
>>>>>=20
>>>>>=20
>>>>>    Peter Roberts
>>>>>    Alcatel-Lucent
>>>>>    Kanata,
>>>>>    Canada
>>>>>=20
>>>>>    Email: peter.roberts@alcatel-lucent.com
>>>>>=20
>>>>>=20
>>>>>    Laurent Montini
>>>>>    Cisco Systems
>>>>>    San Jose CA
>>>>>    USA
>>>>>=20
>>>>>    Email: lmontini@cisco.com
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 3=
4]
>>>>> Internet-Draft        Transporting Timing over MPLS            June 20=
13
>>>>>=20
>>>>>=20
>>>>>    Luca
>>>>>    Cisco Systems
>>>>>    San Jose CA
>>>>>    USA
>>>>>=20
>>>>>    Email: lmartini@cisco.com
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> Davari, et al.          Expires December 17, 2013              [Page 3=
5]
>>>>> --
>>>>> For corporate legal information go to:
>>>>>=20
>>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>>=20
>>>>> _______________________________________________
>>>>> TICTOC mailing list
>>>>> TICTOC@ietf.org
>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>=20
>>>> This e-mail message is intended for the recipient only and contains
>> information which is CONFIDENTIAL and which may be proprietary to ECI
>> Telecom. If you have received this transmission in error, please inform u=
s by e-
>> mail, phone or fax, and then delete the original and all copies thereof.
>>>>=20
>>>> _______________________________________________
>>>> mpls mailing list
>>>> mpls@ietf.org
>>>> https://www.ietf.org/mailman/listinfo/mpls
>>> .
>>=20
>>=20
>> --
>> For corporate legal information go to:
>>=20
>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>=20
>> _______________________________________________
>> TICTOC mailing list
>> TICTOC@ietf.org
>> https://www.ietf.org/mailman/listinfo/tictoc
>>=20
>>=20
>> _______________________________________________
>> TICTOC mailing list
>> TICTOC@ietf.org
>> https://www.ietf.org/mailman/listinfo/tictoc
>=20
>=20
> This e-mail message is intended for the recipient only and contains inform=
ation which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If y=
ou have received this transmission in error, please inform us by e-mail, pho=
ne or fax, and then delete the original and all copies thereof.
>=20

From Alexander.Vainshtein@ecitele.com  Tue Aug  6 07:49:36 2013
Return-Path: <Alexander.Vainshtein@ecitele.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7BAD021F9FE7 for <tictoc@ietfa.amsl.com>; Tue,  6 Aug 2013 07:49:36 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -4.273
X-Spam-Level: 
X-Spam-Status: No, score=-4.273 tagged_above=-999 required=5 tests=[AWL=0.929,  BAYES_00=-2.599, MIME_QP_LONG_LINE=1.396, RCVD_IN_DNSWL_MED=-4,  UNPARSEABLE_RELAY=0.001]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id UzGXSxfj6T8u for <tictoc@ietfa.amsl.com>; Tue,  6 Aug 2013 07:49:31 -0700 (PDT)
Received: from mail1.bemta4.messagelabs.com (mail1.bemta4.messagelabs.com [85.158.143.250]) by ietfa.amsl.com (Postfix) with ESMTP id 13F5E21F9F74 for <tictoc@ietf.org>; Tue,  6 Aug 2013 07:49:15 -0700 (PDT)
Received: from [85.158.143.99:45136] by server-3.bemta-4.messagelabs.com id A2/26-29480-3EC01025; Tue, 06 Aug 2013 14:49:07 +0000
X-Env-Sender: Alexander.Vainshtein@ecitele.com
X-Msg-Ref: server-6.tower-216.messagelabs.com!1375800544!23150191!1
X-Originating-IP: [147.234.242.234]
X-StarScan-Received: 
X-StarScan-Version: 6.9.11; banners=-,-,-
X-VirusChecked: Checked
Received: (qmail 11666 invoked from network); 6 Aug 2013 14:49:05 -0000
Received: from ilptbmg01-out.ecitele.com (HELO ilptbmg01-out.ecitele.com) (147.234.242.234) by server-6.tower-216.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 6 Aug 2013 14:49:05 -0000
X-AuditID: 93eaf2e7-b7ef66d00000140d-4c-52010cdf4271
Received: from ILPTWPVEXCA02.ecitele.com ( [172.31.244.232]) by ilptbmg01-out.ecitele.com (Symantec Messaging Gateway) with SMTP id B2.25.05133.FDC01025; Tue,  6 Aug 2013 17:49:04 +0300 (IDT)
Received: from ILPTWPVEXMB01.ecitele.com ([fe80::f152:8eaf:8fb0:a5da]) by ILPTWPVEXCA02.ecitele.com ([fe80::c473:490d:3a7e:e34a%12]) with mapi id 14.03.0123.003; Tue, 6 Aug 2013 17:49:03 +0300
From: Alexander Vainshtein <Alexander.Vainshtein@ecitele.com>
To: "S. Davari" <davarish@yahoo.com>
Thread-Topic: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
Thread-Index: AQHOkhqJ+XHALQ/MzkWnZzr9BEl1M5mHyTyggAAz1YCAAEUhkA==
Date: Tue, 6 Aug 2013 14:49:02 +0000
Message-ID: <F9336571731ADE42A5397FC831CEAA0215132B63@ILPTWPVEXMB01.ecitele.com>
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com> <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com> <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com>
In-Reply-To: <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-originating-ip: [10.4.35.10]
Content-Type: text/plain; charset="us-ascii"
content-transfer-encoding: quoted-printable
MIME-Version: 1.0
X-CFilter-Loop: Reflected
X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFprCKsWRmVeSWpSXmKPExsUy+dWnL7oPeBiDDJZ9ULE4+LyJ0eLc0zmM Fn+be9gdmD2m/N7I6rFkyU8mj1mzDjMFMEc1MNok5uXllySWpCqkpBYn2yoFFGWWJSZXKilk ptgqGSopFOQkJqfmpuaV2ColFhSk5qUo2XEpYAAboLLMPIXUvOT8lMy8dFslz2B/XQsLU0td QyU7NWVDY2uukIzMYoVU3dzEzByF3NTi4sT0VAWgSMIW5ozTq+6zF5zoYq2YuHMTawPjqivM XYycHBICJhIvtj5khLDFJC7cW8/WxcjFISRwkFHi/rafYAkhgSOMEnff6YDYbAK2EptW32UD sUUEVCSWbl/KAmIzC8xllOg9YNLFyMEhLOAucfiHH0SJh0Tn1POMELaTxP4zv5lAbBag1nsz l7OD2LwCARJ9jzezQOxtYZaY/WAp2HGcQLtO7moDK2IEOu77qTVMELvEJW49mc8EcbSAxJI9 56GeEZV4+fgfK4QtJ/HkySmo23QkFuz+xAZha0ssW/iaGWKxoMTJmU9YIOolJQ6uuMEygVF8 FpIVs5C0z0LSPgtJ+wJGllWMopk5BSVJuekGhnqpyZklqTmpesn5uZsYIenl+Q7GX/NVDjG6 Aj0+kVmKOzkfmJ7ySuKNDQxwc5TEeZc3hPsLCaQDU052ampBalF8UWlOavEhRiYOTqkGRk6f ExdW7JDt7dq8fIWxfzpb8Qn25dplmvevv1RdI/v4rUOn74M1G9k1BRWu7JQuqVqj8NFDbIqG 7sbLmXfPr/x+rU9z4Tf5LneXdSnnTty06bd4soX5tOVOjSnbXmy+PFHU7lX7tpes6W3zPldF xFcXMcfNUVt1MiXdcu+utq8nqywONX773KPEUpyRaKjFXFScCADq6VZgJwMAAA==
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 06 Aug 2013 14:49:36 -0000

Shahram,
>From my experience (admittedly limited), it could be much easier to detect a=
nd filter out large jitter than very low one.
And, in any case, I would wait for the conclusion of the ITU-T SG15 delibera=
tions on partial on path support for 1588.
 
Regards,
     Sasha


> -----Original Message-----
> From: S. Davari [mailto:davarish@yahoo.com]
> Sent: Tuesday, August 06, 2013 4:35 PM
> To: Alexander Vainshtein
> Cc: Shahram Davari; mpls@ietf.org; tictoc@ietf.org; stbryant@cisco.com
> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
> 
> Sasha,
> 
> If a router is not 1588 aware we want it to switch the packet with minimum
> jitter and delay via high priority queue.
> 
> Sending it to CPU creates large jitter and delay since the CPUs are genera=
lly
> busy doing other things. Time stamping or updating CF in the CPU won't hel=
p
> since it does not cover the variable delay that is incurred from the time=
 the
> packet enters the router to the time the packet gets processed by CPU.
> 
> I wish it was that easy!
> 
> Regards,
> Shahram
> 
> 
> On Aug 6, 2013, at 12:38 AM, Alexander Vainshtein
> <Alexander.Vainshtein@ecitele.com> wrote:
> 
> > Shahram,
> > I agree with you that the routers that do not recognize a Timing Alert l=
abel
> would send the packet to the CPU. (This would be easy to arrange since we=
 are
> speaking about a reserved label.)
> >
> > I do not, however, see this as a serious issue, because the SW running o=
n this
> CPU would (hopefully) be upgradable to handle the packet correctly i.e., t=
o
> forward it to where it should be forwarded and, if it is forwarded as a la=
beled
> packet, to prepend the Timing Alert Label on top of its label stack. It co=
uld even
> record the residence time (as observed by the CPU in the proper place in t=
he
> packet. This would mean that introducing additional error to whatever timi=
ng
> information is associated with this packet - but this is what you should a=
nyway
> expect if there are non-compliant routers on your path, right?
> >
> > Regards,
> >     Sasha
> >
> >> -----Original Message-----
> >> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behal=
f
> Of
> >> Shahram Davari
> >> Sent: Monday, August 05, 2013 11:29 PM
> >> To: stbryant@cisco.com; S. Davari
> >> Cc: mpls@ietf.org; tictoc@ietf.org
> >> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
> >>
> >> Hi Stewart,
> >>
> >> Your suggestion of " I suggested an LSP type that has the properties
> "timestamp
> >> and pass to application", is a subset of the existing draft and should=
 work. It
> >> basically limits the time stamping and correction field update to LERs=
 (while
> the
> >> draft supports time stamping at LER and LSR).
> >>
> >> However Sasha's suggestion of using a reserved label (RAL, GAL or any o=
ther
> >> reserved label) does not satisfy one of the major requirements, which i=
s
> >> backward compatibility. Routers that don't understand this reserved lab=
el
> will
> >> drop or copy to CPU such packets.
> >>
> >> Regards,
> >> Shahram
> >>
> >> -----Original Message-----
> >> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behal=
f
> Of
> >> Stewart Bryant
> >> Sent: Monday, August 05, 2013 2:29 AM
> >> To: S. Davari
> >> Cc: mpls@ietf.org; tictoc@ietf.org
> >> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
> >>
> >> My concern is that we are creating a new LSP type in MPLS
> >> (which is a architectural change and thus needs to be very carefully
> >> considered) without asking the question "how can I do this
> >> in the most general way, to maximize flexibility/reuse"
> >>
> >> I suggested an LSP type that has the properties "timestamp
> >> and pass to application", but I think that Sasha raises a good
> >> point about using router alert. However RA has no implicit
> >> timestamp, so maybe we need a new type of RA that has the
> >> properties "timestamp, and pass top application indicated by
> >> the next label". That would be quite useful in a number
> >> of OAM applications. There is possibly some GAL variant
> >> of that design that should also be considered.
> >>
> >> The application could worry about the path and where it
> >> was necessary to skip some hops a hierarchical LSP would
> >> accomplish that, with the specific benefit that the application
> >> would consciously do this, and may be able to apply some
> >> form of compensation within the network.
> >>
> >> - Stewart
> >>
> >> On 04/08/2013 10:40, S. Davari wrote:
> >>> Hi Sasha
> >>>
> >>> Perhaps you have not understood the draft well. The main reason that t=
he
> >> draft chose to use a dedicated LSP that is signaled via RSVPTE indicati=
ng it is
> a
> >> timing LSP is to be backward compatible with non-1588 nodes. Such nodes
> >> simply just switch the packet.
> >>>
> >>> Your proposal had been considered and was rejected since it was not
> >> backward compatible. Nodes receiving alert label or TTL =3D 1 send the=
 packet
> to
> >> CPU. You can refer to the meeting notes.
> >>>
> >>> Regards,
> >>> Shahram
> >>>
> >>>
> >>> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein
> >> <Alexander.Vainshtein@ecitele.com> wrote:
> >>>
> >>>> Stewart and all,
> >>>> I concur with Stewart's statement that the draft does not define "ful=
l
> >> interaction with MPLS architecture".
> >>>>
> >>>> E.g., one of the objectives of the draft is to provide a technique th=
at would
> >> be backward-compatible with old LSRs that cannot provide on-path suppor=
t
> for
> >> timing distribution, while the other objective is to make every LSR on=
 the
> path
> >> aware that some MPLS packets are carrying timing-related messages and
> hence
> >> require on-path support.
> >>>>
> >>>> IMHO and FWIW, the MPLS data plane architecture allows just two
> methods
> >> for making a transit LSR to provide special processing to a labeled pac=
ket:
> >>>> - It would carry some kind of an "alert label" on top of the label st=
ack and,
> >> specifically, on top of any labels used for actual forwarding
> >>>> OR,
> >>>> - The TTL in the top label stack entry has been set to 1.
> >>>>
> >>>>
> >>>> The draft does not follow any of these approaches.
> >>>>
> >>>> I must also admit that Section 12 "OAM, Control and Management" of th=
e
> >> draft looks somewhat in
> >>>> E.g., I could not understand from the text whether BFD, LSP-Ping etc.=
 OAM
> >> messages would be subjected to on-path support procedures for timing
> >> messages or not.
> >>>>
> >>>> My 2c,
> >>>>     Sasha
> >>>>
> >>>>> -----Original Message-----
> >>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On
> Behalf
> >> Of
> >>>>> Stewart Bryant
> >>>>> Sent: Friday, August 02, 2013 12:01 PM
> >>>>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
> >>>>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.o=
rg;
> >> draft-
> >>>>> ietf-tictoc-1588overmpls@tools.ietf.org
> >>>>> Cc: mpls@ietf.org; tictoc@ietf.org
> >>>>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
> >>>>>
> >>>>> SB> This draft does not seem to provide a precise definition
> >>>>> SB> the properties of the new LSP type that it wishes to
> >>>>> SB> define, in particular it does it define the PHB of those
> >>>>> SB> LSPs, nor the full interaction with the MPLS
> >>>>> SB> architecture.
> >>>>> SB>
> >>>>> SB> I have not tracked TICTOC for a while but I thought that
> >>>>> SB> the original plan was to define the concept of an offset
> >>>>> SB> into a packet to do the correction.
> >>>>> SB>
> >>>>> SB> It is disappointing that the opportunity was not taken
> >>>>> SB> to define a timing shim inside the timing LSP so that
> >>>>> SB> a time correction could be added to any packet such that
> >>>>> SB> the MPLS system was isolated from the details of the
> >>>>> SB> complexity of the particular time transfer type.
> >>>>>
> >>>>> SB> I think that much more clarify is needed in terms of
> >>>>> SB> definition of the new LSP type, since it is unclear
> >>>>> SB> from this text how to implement one.
> >>>>> SB>
> >>>>> SB> There are a lot of other MPLS services such as
> >>>>> SB> LSP ping that need to be considered.
> >>>>> SB>
> >>>>> SB> Please see inline for more comments. However these
> >>>>> SB> comments are made in the context of the text as written
> >>>>> SB> whilst I have a fundamental concern that this approach
> >>>>> SB> lacks an MPLS architectural soundness that need
> >>>>> SB> greater thought with significant impact on the
> >>>>> SB> draft.
> >>>>>
> >>>>> - Stewart
> >>>>>
> >>>>>
> >>>>> TICTOC Working Group                                           S. Da=
vari
> >>>>> Internet-Draft                                                   A.=
 Oren
> >>>>> Intended status: Standards Track                          Broadcom C=
orp.
> >>>>> Expires: December 17, 2013                                     M. Bh=
atia
> >>>>>                                                               P. Rob=
erts
> >>>>>                                                           Alcatel-Lu=
cent
> >>>>>                                                               L. Mon=
tini
> >>>>>                                                               L. Mar=
tini
> >>>>>                                                            Cisco Sys=
tems
> >>>>>                                                            June 15,=
 2013
> >>>>>
> >>>>>
> >>>>>             Transporting Timing messages over MPLS Networks
> >>>>>                    draft-ietf-tictoc-1588overmpls-05
> >>>>>
> >>>>> Abstract
> >>>>>
> >>>>>    This document defines the method for transporting Timing messages
> >>>>>    such as PTP and NTP over an MPLS network.  The method allows for
> the
> >>>>>    easy identification of these PDUs at the port level to allow for=
 port
> >>>>>
> >>>>> SB> What is a port
> >>>>>
> >>>>>    level processing of these PDUs in both LERs and LSRs.
> >>>>>
> >>>>>    The basic idea is to transport Timing messages inside dedicated M=
PLS
> >>>>>    LSPs.  These LSPs only carry Timing messages and possibly Control=
 and
> >>>>>    Management packets, but they do not carry customer traffic.
> >>>>>
> >>>>> SB> More specifically they only carry traffic associated with the
> >>>>> SB> timing service and its support.
> >>>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
> >>>>> SB> also it gets carried in a structure that causes it to get
> >>>>> SB> timestamped.
> >>>>>
> >>>>>    Two methods for transporting Timing messages over MPLS are define=
d.
> >>>>>
> >>>>> SB> Perhaps the right approach is to define the new LSP type and the=
n
> >>>>> SB> seperately to define  the mapping of the various timing services
> >>>>> SB> over that LSP type.
> >>>>>
> >>>>>    The first method is to transport Timing messages directly over th=
e
> >>>>>    dedicated MPLS LSP via UDP/IP encapsulation, which is suitable fo=
r
> >>>>>    MPLS networks.  The second method is to transport Timing messages
> >>>>>    inside a PW via Ethernet encapsulation.
> >>>>>
> >>>>> SB> I think that we should note that there are some
> >>>>> SB> h/w reasons for this preference. A clean sheet approach
> >>>>> SB> would have been to use PTP over MPLS with no intermediate
> >>>>> SB> layers.
> >>>>>
> >>>>> Status of this Memo
> >>>>>
> >>>>>    This Internet-Draft is submitted in full conformance with the
> >>>>>    provisions of BCP 78 and BCP 79.
> >>>>>
> >>>>>    Internet-Drafts are working documents of the Internet Engineering
> >>>>>    Task Force (IETF).  Note that other groups may also distribute
> >>>>>    working documents as Internet-Drafts.  The list of current Intern=
et-
> >>>>>    Drafts is at http://datatracker.ietf.org/drafts/current/.
> >>>>>
> >>>>>    Internet-Drafts are draft documents valid for a maximum of six mo=
nths
> >>>>>    and may be updated, replaced, or obsoleted by other documents at=
 any
> >>>>>    time.  It is inappropriate to use Internet-Drafts as reference
> >>>>>    material or to cite them other than as "work in progress."
> >>>>>
> >>>>>    This Internet-Draft will expire on December 17, 2013.
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 1]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> Copyright Notice
> >>>>>
> >>>>>    Copyright (c) 2013 IETF Trust and the persons identified as the
> >>>>>    document authors.  All rights reserved.
> >>>>>
> >>>>>    This document is subject to BCP 78 and the IETF Trust's Legal
> >>>>>    Provisions Relating to IETF Documents
> >>>>>    (http://trustee.ietf.org/license-info) in effect on the date of
> >>>>>    publication of this document.  Please review these documents
> >>>>>    carefully, as they describe your rights and restrictions with res=
pect
> >>>>>    to this document.  Code Components extracted from this document
> must
> >>>>>    include Simplified BSD License text as described in Section 4.e o=
f
> >>>>>    the Trust Legal Provisions and are provided without warranty as
> >>>>>    described in the Simplified BSD License.
> >>>>>
> >>>>>
> >>>>> Table of Contents
> >>>>>
> >>>>>    1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . .=
 .  5
> >>>>>
> >>>>>    2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . .=
 .  7
> >>>>>
> >>>>>    3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . .=
 .  8
> >>>>>
> >>>>>    4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . .=
 .  9
> >>>>>
> >>>>>    5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . .=
 . 12
> >>>>>
> >>>>>    6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . .=
 . 13
> >>>>>      6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . .=
 . 13
> >>>>>      6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . .=
 . 13
> >>>>>      6.3.  Other Timing Encapsulation methods . . . . . . . . . . .=
 . 14
> >>>>>
> >>>>>    7.  Timing message Processing  . . . . . . . . . . . . . . . . .=
 . 15
> >>>>>
> >>>>>    8.  Protection and Redundancy  . . . . . . . . . . . . . . . . .=
 . 16
> >>>>>
> >>>>>    9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 17
> >>>>>
> >>>>>    10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 18
> >>>>>
> >>>>>    11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 19
> >>>>>
> >>>>>    12. OAM, Control and Management  . . . . . . . . . . . . . . . .=
 . 20
> >>>>>
> >>>>>    13. QoS Considerations . . . . . . . . . . . . . . . . . . . . .=
 . 21
> >>>>>
> >>>>>    14. FCS and Checksum Recalculation . . . . . . . . . . . . . . .=
 . 22
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 2]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>>    15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . .=
 . 23
> >>>>>      15.1. Behavior of Timing-capable/aware LER . . . . . . . . . .=
 . 23
> >>>>>      15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . .=
 . 23
> >>>>>      15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . .=
 . 24
> >>>>>
> >>>>>    16. Other considerations . . . . . . . . . . . . . . . . . . . .=
 . 25
> >>>>>
> >>>>>    17. Security Considerations  . . . . . . . . . . . . . . . . . .=
 . 26
> >>>>>
> >>>>>    18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . .=
 . 27
> >>>>>
> >>>>>    19. IANA Considerations  . . . . . . . . . . . . . . . . . . . .=
 . 28
> >>>>>
> >>>>>    20. References . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 29
> >>>>>      20.1. Normative References . . . . . . . . . . . . . . . . . .=
 . 29
> >>>>>      20.2. Informative References . . . . . . . . . . . . . . . . .=
 . 29
> >>>>>
> >>>>>    Appendix 1.  Routing extensions for Timing-aware Routers . . . .=
 . 32
> >>>>>
> >>>>>    Appendix 2.  Signaling Extensions for Creating Timing LSPs . . .=
 . 33
> >>>>>
> >>>>>    Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . .=
 . 34
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 3]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>>    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
> >> NOT",
> >>>>>    "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and
> "OPTIONAL"
> >> in
> >>>>> this
> >>>>>    document are to be interpreted as described in RFC2119 [RFC2119].
> >>>>>
> >>>>>    When used in lower case, these words convey their typical use in
> >>>>>    common language, and are not to be interpreted as described in
> >>>>>    RFC2119 [RFC2119].
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 4]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 1.  Introduction
> >>>>>
> >>>>>    The objective of Precision Time Protocol (PTP) and Network Timing
> >>>>>    Protocol (NTP) are to synchronize independent clocks running on
> >>>>>    separate nodes of a distributed system.
> >>>>>
> >>>>>    [IEEE-1588] defines PTP messages for frequency, phase and time
> >>>>>    synchronization.  The PTP messages include PTP PDUs over UDP/IP
> >>>>>    (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex=
 F of
> >>>>>    [IEEE-1588]).
> >>>>>
> >>>>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
> >>>>> SB> of other PTP mappings if they provide better optimisation.
> >>>>>
> >>>>>    This document defines mapping and transport of the PTP
> >>>>>    messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
> >>>>>    defines several clock types: ordinary clocks, boundary clocks, en=
d-
> >>>>>    to-end transparent clocks, and peer-to-peer transparent clocks.
> >>>>>    Transparent clocks require intermediate nodes to update correctio=
n
> >>>>>    field inside PTP message that reflects the transit time in the no=
de.
> >>>>>
> >>>>>    [RFC5905] defines NTP messages for clock and time synchronization=
.
> >>>>>    The PTP messages (PDUs) are transported over UDP/IP.  This docume=
nt
> >>>>> SB> Should that be NTP messages?
> >>>>> SB> It needs to be made clear as soon as you introduce NTP that
> >>>>> SB> they use different time representations.
> >>>>>
> >>>>>    defines mapping and transport of the NTP messages defined in
> >>>>>    [RFC5905] over MPLS networks.
> >>>>>
> >>>>>    One key attribute of all of these Timing messages is that the Tim=
e
> >>>>>    stamp processing should occur as close as possible to the actual
> >>>>>    transmission and reception at the physical port interface.  This
> >>>>>    targets optimal time and/or frequency recovery by avoiding variab=
le
> >>>>>    delay introduced by queues internal to the clocks.
> >>>>>
> >>>>> SB> As I recall NTP has no epoch point defined, and I am not sure
> >>>>> SB> where that point is in the case of PTP in this mapping
> >>>>> SB> Hopefully this will get defined in due course.
> >>>>>
> >>>>>    To facilitate the fast and efficient recognition of Timing messag=
es
> >>>>>    at the port level when the Timing messages are carried over MPLS
> >>>>>    LSPs,
> >>>>>
> >>>>> SB> Over a new LSP type with time optimied characteristics
> >>>>>
> >>>>>    this document defines the specific encapsulations that should
> >>>>>    be used.
> >>>>> SB> Hopefully it will also define the PHP
> >>>>>
> >>>>>    In addition, it can be expected that there will exist LSR/
> >>>>>    LERs where only a subset of the physical ports will have the port=
-
> >>>>>    based Timing message processing capabilities.
> >>>>> SB> Do you need to clarify that this only works at base and not in
> >>>>> SB> a label heirarchy.
> >>>>>
> >>>>>
> >>>>>    In order to ensure
> >>>>>    that the LSPs carrying Timing packets always enter and exit ports
> >>>>>    with this capability, routing extensions are defined to advertise
> >>>>>    this capability on a port basis and to allow for the establishmen=
t of
> >>>>>    LSPs that only transit such ports.  While this path establishment
> >>>>>    restriction may be applied only at the LER Ingress and/or egress
> >>>>>    ports, it becomes more important when using transparent clock
> capable
> >>>>>    LSRs in the path.
> >>>>> SB> I do not understand the implications of the last
> >>>>> SB> sentences - starting ", it becomes"
> >>>>>
> >>>>>
> >>>>>    Port based Timing message processing involves Timing message
> >>>>>    recognition.  Once the Timing messages are recognized they can be
> >>>>>    modified based on the reception or transmission Time-stamp.
> >>>>>
> >>>>>    This document provides two methods for transporting Timing
> messages
> >>>>>    over MPLS.  One is applicable to MPLS environment and the other o=
ne
> >>>>>    is applicable to MPLS/MPLS-TP environment
> >>>>>
> >>>>> SB> I think the sentence is incomplete.
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 5]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>>    The solution involves transporting Timing messages over dedicated
> >>>>>    LSPs called Timing LSPs.  These LSPs carry Timing messages and MA=
Y
> >>>>>    carry Management and control messages, but not data plane client
> >>>>>    traffic.
> >>>>>
> >>>>> SB> It is not clear why this restriction applies.
> >>>>>
> >>>>>    Timing LSPs can be established statically or via signaling.
> >>>>> SB> s/statically/by provisioning/network management/
> >>>>>
> >>>>>    Extensions to control plane (OSPF, ISIS, etc.) is required to ena=
ble
> >>>>>    routers to distribute their Timing processing capabilities over M=
PLS
> >>>>>    to other routers.  However such extensions are outside the scope=
 of
> >>>>>    this document.
> >>>>>
> >>>>>    When signaling is used to setup the PTP LSP, Extensions to signal=
ing
> >>>>> SB> is it a PTP LSP or a Timing LSP?
> >>>>>
> >>>>>    protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
> >>>>>    However such extensions are outside the scope of this document.
> >>>>>
> >>>>> SB> for mpls-tp GMPLS is the signalling protocol
> >>>>>
> >>>>>    While the techniques included herein allow for the establishment=
 of
> >>>>>    paths optimized to include Time-stamping capable links, the
> >>>>>    performance of the Slave clocks is outside the scope of this
> >>>>>    document.
> >>>>>
> >>>>>    At the time of publishing this specification, Transparent Clockin=
g
> >>>>>    (TC) is only defined for PTP.  Therefore at this time any part of
> >>>>>    this specification that talks about Transparent Clocking applies=
 only
> >>>>>    to PTP.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 6]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 2.  Terminology
> >>>>>
> >>>>>    1588: The timing and synchronization as defined by IEEE 1588.
> >>>>>
> >>>>> SB> I think that there is a more formal name for the 1588 group
> >>>>> SB> that needs to be used here.
> >>>>> SB> Also do we need to talk about 1588-200? as there is
> >>>>> SB> an update in progress
> >>>>>
> >>>>>    NTP: The timing and synchronization protocol defined by IETF RFC-=
1305
> >>>>>    and RFC-5905.
> >>>>>
> >>>>>    PTP: The timing and synchronization protocol used by 1588.
> >>>>> SB> need the proper name for 1588
> >>>>>
> >>>>>    Master Clock: The source of 1588 timing to a set of slave clocks.
> >>>>>
> >>>>>    Master Port: A port on a ordinary or boundary clock that is in Ma=
ster
> >>>>>    state.  This is the source of timing toward slave ports.
> >>>>>
> >>>>> SB> I am not sure the reader knows what a port is
> >>>>>
> >>>>>    Slave Clock: A receiver of 1588 timing from a master clock.
> >>>>>
> >>>>>    Slave Port: A port on a boundary clock or ordinary clock that is
> >>>>>    receiving timing from a master clock.
> >>>>>
> >>>>>    Ordinary Clock: A device with a single PTP port.
> >>>>>
> >>>>>    Transparent Clock.  A device that measures the time taken for a P=
TP
> >>>>>    event message to transit the device and then updates the
> >>>>>    correctionField of the message with this transit time.
> >>>>>
> >>>>>    Boundary Clock: A device with more than one PTP port.  Generally
> >>>>>    boundary clocks will have one port in slave state to receive timi=
ng
> >>>>>    and then other ports in master state to re-distribute the timing.
> >>>>>
> >>>>>    PTP LSP: An LSP dedicated to carry PTP messages
> >>>>>
> >>>>> SB> PTP or timing?
> >>>>>
> >>>>>
> >>>>>    PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
> >>>>>    messages.
> >>>>>
> >>>>> SB> Ah I don't think that PWE3 know what one of these is
> >>>>>
> >>>>>    CW: Pseudowire Control Word
> >>>>>
> >>>>>    LAG: Link Aggregation
> >>>>>
> >>>>>    ECMP: Equal Cost Multipath
> >>>>>
> >>>>>    CF: Correction Field, a field inside certain PTP messages (messag=
e
> >>>>>    type 0-3)that holds the accumulative transit time inside intermed=
iate
> >>>>>    switches
> >>>>>
> >>>>>    Timing messages: Timing Protocol messages that are exchanged
> >> between
> >>>>>    routers in order to establish a synchronized clock.
> >>>>>
> >>>>> SB> A number of these definitions look like copies of IEEE1588
> >>>>> SB> definitions. We need to provide references and note the
> >>>>> SB> priority of the IEEE base reference.
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 7]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 3.  Problem Statement
> >>>>>
> >>>>>    [IEEE-1588] has defined methods for transporting PTP messages ove=
r
> >>>>>    Ethernet and IP networks.  [RFC5905] has defined the method of
> >>>>>    transporting NTP messages over IP networks.  There is a need to
> >>>>>    transport Timing messages over MPLS networks while supporting the
> >>>>>    Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (O=
C)
> >>>>>    functionality in the LER and LSRs in the MPLS network.
> >>>>>
> >>>>>    There are multiple ways of transporting Timing over MPLS.  Howeve=
r,
> >>>>>    there is a requirement to limit the possible encapsulation option=
s to
> >>>>>    simplify the Timing message identification and processing require=
d at
> >>>>>    the port level.
> >>>>>
> >>>>>    When Timing-awareness is needed, Timing messages should not be
> >>>>>    transported over LSPs or PWs that are carrying customer traffic
> >>>>>    because LSRs perform Label switching based on the top label in th=
e
> >>>>>    stack.
> >>>>>
> >>>>> SB> Have you explained why?
> >>>>>
> >>>>>    To detect Timing messages inside such LSPs require special
> >>>>>    hardware to do deep packet inspection at line rate.  Even if such
> >>>>>    hardware exists, the payload can't be deterministically identifie=
d by
> >>>>>    LSRs because the payload type is a context of the PW label, and t=
he
> >>>>>    PW label and its context are only known to the Edge routers (PEs/
> >>>>>    LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, C=
ES,
> >>>>>    etc).  Even if one restricts an LSP to only carry Ethernet PWs, t=
he
> >>>>>    LSRs dont have the knowledge of whether PW Control Word (CW) is
> >>>>>    present or not and therefore can not deterministically identify t=
he
> >>>>>    payload.
> >>>>>
> >>>>>    A generic method is defined in this document that does not requir=
e
> >>>>>    deep packet inspection at line rate, and can deterministically
> >>>>>    identify Timing messages.  This method can be used to detect Timi=
ng
> >>>>>    Messages in both one-step and two-step clock implementations of
> >>>>>    ordinary, boundary and transparent clocks.
> >>>>>
> >>>>> SB> Needs a ref and I am sure many MPLS specialists will not underst=
and
> >>>>> SB> the msg types.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 8]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 4.  Timing over MPLS Architecture
> >>>>>
> >>>>>    Timing messages are exchange between Timing ports on ordinary and
> >>>>>
> >>>>> SB> Have you defined a timing port?
> >>>>>
> >>>>>    boundary clocks.  Boundary clocks terminate the Timing messages a=
nd
> >>>>>    act as master for other boundary clocks or for slave clocks.  End=
-to-
> >>>>>    End Transparent clocks do not terminate the Timing messages but t=
hey
> >>>>>    do modify the contents of the Timing messages as they transit acr=
oss
> >>>>>    the transparent clock.
> >>>>>
> >>>>>    Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent
> Clock
> >>>>>
> >>>>>    (TC) could be implemented in either LERs or LSRs.
> >>>>>
> >>>>> SB> LER and LSR need to be expanded
> >>>>>
> >>>>>    An example is shown in Figure 1, where the LERs act as Ordinary C=
lock
> >>>>>    (OC) and are the initiating/terminating point for Timing messages=
.
> >>>>>    The ingress LER encapsulates the Timing messages in Timing LSP an=
d
> >>>>>    the Egress LER terminates the Timing LSP.  The LSRs act as
> >>>>>    Transparent Clock (TC) and just update the Timing field in the Ti=
ming
> >>>>>    messages.
> >>>>>
> >>>>>
> >>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
> >>>>>       |Switch, |     |       |     |       |     |       |     |Swit=
ch, |
> >>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rou=
ter |
> >>>>>       |        |     |  OC   |     |  TC   |     |  OC   |     |   =
     |
> >>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
> >>>>>                      /                                 \
> >>>>>       +-------+     /                                   \     +-----=
--+
> >>>>>       |  LER  |    /                                     \    |  LER=
  |
> >>>>>       | Master|---/                                       \---| Slav=
e |
> >>>>>       | Clock |                                               | Cloc=
k |
> >>>>>       +-------+                                               +-----=
--+
> >>>>>
> >>>>>      Figure (1) - Deployment example 1 of timing over MPLS network
> >>>>>
> >>>>>    Another example is shown in Figure2, where LERs terminate the Tim=
ing
> >>>>>    messages received from switch/routers that are outside of the MPL=
S
> >>>>>    network acting as OC or BC.  In this example LERs regenerate the
> >>>>>    clock and initiate timing messages encapsulated in Timing LSP tow=
ard
> >>>>>    the MPLS network, while the LSRs act as Transparent Clock (TC) an=
d
> >>>>>    just update the Timing field in the Timing messages, which are
> >>>>>    already encapsulated in Timing LSPs.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 9]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
> >>>>>      |Switch, |     |       |     |       |     |       |     |Switc=
h, |
> >>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
> >>>>>      | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/B=
C  |
> >>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
> >>>>>
> >>>>>      Figure (2) - Deployment example 2 of timing over MPLS network
> >>>>>
> >>>>>
> >>>>>    Another example is shown in Figure 3, where LERs do not terminate=
 the
> >>>>>    Timing messages received from switch/routers that are outside of=
 the
> >>>>>    MPLS network acting as OC, TC or BC.  The LERs act as TC and upda=
te
> >>>>>    the Timing field in the Timing messages as they transit the LER,
> >>>>>    while encapsulating them in timing LSP.  The LSRs also act as
> >>>>>    Transparent Clock (TC) and just update the Timing field in the Ti=
ming
> >>>>>    messages which are already encapsulated in Timing LSPs.
> >>>>>
> >>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
> >>>>>       |Switch, |     |       |     |       |     |       |     |Swit=
ch, |
> >>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rou=
ter |
> >>>>>       |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/T=
C/BC|
> >>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
> >>>>>
> >>>>>     Figure (3) - Deployment example 3 of timing over MPLS network
> >>>>>
> >>>>>    Another example is shown in Figure 4, where LERs and LSRs support
> >>>>>    Boundary Clocks.  A single-hop LSP is created between two adjacen=
t
> >>>>>    LSRs engaged in BC operation.  Other methods such as PTP transpor=
t
> >>>>>    over Ethernet MAY be used for transporting timing messages if the
> >>>>>    link between the two routers is Ethernet.
> >>>>>
> >>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
> >>>>>      |Switch, |     |       |     |       |     |       |     |Switc=
h, |
> >>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
> >>>>>      | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/B=
C  |
> >>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
> >>>>>
> >>>>>    Figure (4) - Deployment example 3 of timing over MPLS network
> >>>>>
> >>>>>    An MPLS domain MAY serve multiple customers.  In these cases the
> >> MPLS
> >>>>>    domain (maintained by a service provider) may provide timing serv=
ices
> >>>>>    to multiple customers, each having their own Timing domain.
> >>>>>
> >>>>>    The Timing over MPLS architecture assumes full mesh of Timing LSP=
s
> >>>>>    between all LERs supporting this specification.
> >>>>>
> >>>>> SB> Note sure this is right - the salves surely do not need to
> >>>>> SB> exchange timing amongst themselves
> >>>>>
> >>>>>    It supports
> >>>>>    Point-to- point (VPWS) and Multipoint (VPLS) services.
> >>>>>
> >>>>> SB> What does that mean? You do not carry user data traffic?
> >>>>> SB> Maybe it's the ordering of the statemnets that is causing
> >>>>> SB> confusion.
> >>>>>
> >>>>>    This means
> >>>>>    that a customer may purchase a Point-to-point Timing service betw=
een
> >>>>>    two customer sites or a Multipoint Timing service between more th=
an
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 10]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>>    two customer sites.
> >>>>>
> >>>>>    The Timing over MPLS architecture supports P2P or P2MP Timing LSP=
s.
> >>>>>    This means that the Timing Multicast messages such as PTP Multica=
st
> >>>>>    event messages can be transported over P2MP Timing LSP or be
> >>>>>    replicated and transported over many P2P Timing LSPs.
> >>>>>
> >>>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
> >>>>> SB> nor a P2MP PW, although we are close.
> >>>>>
> >>>>>    Timing messages, that do not require Time stamping or Correction
> >>>>>    Field update MAY be transported over Timing LSPs to simplify hard=
ware
> >>>>>    and software.
> >>>>>
> >>>>>    PTP Announce messages that determine the Timing LSP terminating
> >> point
> >>>>>    behavior such as BC/OC/TC SHOULD be transported over the Timing
> LSP
> >>>>>    to simplify hardware and software.
> >>>>>
> >>>>> SB> have you defined and referenced PTP announce msgs?
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 11]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 5.  Dedicated LSPs for Timing messages
> >>>>>
> >>>>>    Many methods have been considered for identifying the Timing
> >> messages
> >>>>>    when they are encapsulated in MPLS such as using GAL/G-ACH or a
> new
> >>>>>    reserved label.  These methods were not attractive since they eit=
her
> >>>>>    required deep packet inspection at line rate in the intermediate=
 LSRs
> >>>>>    or they required use of a scarce new reserved label.  Also one of=
 the
> >>>>>    goals was to reuse existing OAM mechanisms.
> >>>>>
> >>>>> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
> >>>>>
> >>>>>    The method defined in this document can be used by LER and LSRs t=
o
> >>>>>    identify Timing messages in MPLS tunnels by just looking at the t=
op
> >>>>>    label in the MPLS label stack, which only carry Timing messages a=
s
> >>>>>    well as OAM, but not data plane client traffic.
> >>>>>
> >>>>>    Compliant implementations MUST use dedicated LSPs to carry Timing
> >>>>>    messages over MPLS.
> >>>>>
> >>>>> SB> I think that we need a definition of the properies of these LSPs
> >>>>>
> >>>>>    These LSPs are herein referred to as "Timing
> >>>>>    LSPs" and the labels associated with these LSPs as "Timing LSP
> >>>>>    labels".  The Timing LSPs that runs between Ingress and Egress LE=
Rs
> >>>>>    MUST be co-routed.  Alternatively, a single bidirectional co-rout=
ed
> >>>>>    LSP can be used.
> >>>>>
> >>>>> SB> I though that you said you could use M2MP LSPs - these are not
> >>>>> SB> bidirectional.
> >>>>>
> >>>>>    Co-routing of the two directions is required to limit the differe=
nce
> >>>>>    in the delays in the Master clock to Slave clock direction compar=
ed
> >>>>>    to the Slave clock to Master clock direction.  The Timing LSP MAY=
 be
> >>>>>    MPLS/MPLS-TP LSP.
> >>>>>
> >>>>>    The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS=
.
> >>>>>    New Extensions to RSVP-TE/GMPLS TLVs are required; however they
> are
> >>>>>    outside the scope of this document.
> >>>>>
> >>>>>    The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such
> as
> >>>>>    BFD and LSP Ping but the LSP data plane client plane traffic MUST=
 be
> >>>>>    Timing packets only.
> >>>>>
> >>>>> SB> Why?
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 12]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 6.  Timing over LSP Encapsulation
> >>>>>
> >>>>> The encapsulations is not LSP is it?
> >>>>>
> >>>>>    This document defines two methods for carrying Timing messages ov=
er
> >>>>>    MPLS.  The first method is carrying UDP/IP encapsulated Timing
> >>>>>    messages over Timing LSPs, and the second method, is carrying
> >>>>>    Ethernet encapsulated Timing messages over Ethernet PWs inside
> >> Timing
> >>>>>    LSPs.
> >>>>>
> >>>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
> >>>>>
> >>>>>    The simplest method of transporting Timing messages over MPLS is=
 to
> >>>>>    encapsulate Timing PDUs in UDP/IP and then encapsulate them in
> >> Timing
> >>>>>    LSP.  This format is shown in Figure 4.
> >>>>>
> >>>>>
> >>>>>                     +----------------------+
> >>>>>                     |   Timing LSP Label   |
> >>>>>                     +----------------------+
> >>>>>                     |        IPv4/6        |
> >>>>>                     +----------------------+
> >>>>>                     |         UDP          |
> >>>>>                     +----------------------+
> >>>>>                     |     Timing PDU       |
> >>>>>                     +----------------------+
> >>>>>
> >>>>>       Figure (4) - Timing over UDP/IP over MPLS Encapsulation
> >>>>>
> >>>>>
> >>>>>    This encapsulation is very simple and is useful when the network
> >>>>>    between Timing Master Clock and Slave Clock is MPLS network.
> >>>>>
> >>>>> SB> Simple is a judgement call
> >>>>>
> >>>>>    In order for an LER/LSR to process Timing messages, the Timing LS=
P
> >>>>>    Label must be at the top label of the label stack.  The LER/LSR M=
UST
> >>>>>    know that the Timing LSP Label is used for carrying Timing messag=
es.
> >>>>>    This can be accomplished via static configuration or via RSVP-TE
> >>>>>    signaling.
> >>>>>
> >>>>>    The UDP/IP encapsulation of PTP MUST follow Annex D and E of
> >>>>>    [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
> >>>>>    [RFC5905].
> >>>>>
> >>>>> 6.2.  Timing over PW Encapsulation
> >>>>>
> >>>>>    Another method of transporting Timing over MPLS networks is by
> >>>>>    encapsulating Timing PDUs in PW which in turn is transported over
> >>>>>    Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448]=
,
> >>>>>    shown in Fig 5(A) MUST be used and the Ethernet encapsulation of=
 PTP
> >>>>>    MUST follow Annex F of [IEEE-1588].
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 13]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>>    The RAW mode or Tagged mode defined in [RFC4448] MAY be used
> and
> >> the
> >>>>>    Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
> >>>>>    Timing over PW encapsulation MUST use the Control Word (CW) as
> >>>>>    specified in [RFC4448] to ensure proper detection of PTP messages
> >>>>>    inside the MPLS packets for Timing over LSP and Timing over PW
> >>>>>    encapsulation.
> >>>>>
> >>>>> SB> That needs explanation
> >>>>>
> >>>>>    The use of Sequence Number in the CW is optional.
> >>>>>
> >>>>> SB> Given that s/n are never in practice deployed, you could probabl=
y
> >>>>> SB> simplify things by sayig that they are not used.
> >>>>>
> >>>>>    Timing over PW encapsulation for NTP MUST use NTP over UDP/IP ove=
r
> >> PW
> >>>>>    (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
> >>>>>
> >>>>>                     +----------------+  +----------------+
> >>>>>                      |Timing LSP Label|  |Timing LSP Label|
> >>>>>                      +----------------+  +----------------+
> >>>>>                      |    PW Label    |  |    PW Label    |
> >>>>>                      +----------------+  +----------------+
> >>>>>                      |  Control Word  |  |      IP        |
> >>>>>                      +----------------+  +----------------+
> >>>>>                      |    Ethernet    |  |      UDP       |
> >>>>>                      |     Header     |  +----------------+
> >>>>>                      +----------------+  |   Timing PDU   |
> >>>>>                      |S-VLAN(Optional)|  |                |
> >>>>>                      +----------------+  +----------------+
> >>>>>                      |C-VLAN(Optional)|        (B)
> >>>>>                      +----------------+
> >>>>>                      |   Timing PDU   |
> >>>>>                      |                |
> >>>>>                      +----------------+
> >>>>>                             (A)
> >>>>>
> >>>>>               Figure (5) - Timing over PW Encapsulations
> >>>>>
> >>>>>    In order for an LSR to process PTP messages, the top label of the
> >>>>>    label stack (the Tunnel Label) MUST be a Timing label.
> >>>>>
> >>>>> S> You said that before.
> >>>>>
> >>>>> 6.3.  Other Timing Encapsulation methods
> >>>>>
> >>>>>    In future other timing encapsulation methods may be introduced, s=
uch
> >>>>>    as a new shim header after the Bottom of Stack to carry the Timin=
g
> >>>>>    information.  Such new encapsulations are outside the scope of th=
is
> >>>>>    document.
> >>>>>
> >>>>>
> >>>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
> >>>>> SB> out of the definition of the LSP
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 14]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> SB> I think we need a section on LSP processing
> >>>>>
> >>>>> 7.  Timing message Processing
> >>>>>
> >>>>>    Each Timing protocol such as PTP and NTP, define their set of Tim=
ing
> >>>>>    messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
> >>>>>    FOLLOW_UP, etc messages.
> >>>>>
> >>>>>    Some of the Timing messages require time stamping or correction f=
ield
> >>>>>    update at port level and some dont.  It is the job of the LER/LSR=
 to
> >>>>>    parse the timing message and find out the type of the Timing mess=
age
> >>>>>    and decide whether and how to Time- stamp it (e.g., BC) or update
> >>>>>    correction field(e.g., TC).
> >>>>>
> >>>>>
> >>>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
> >>>>> SB> function rather than the LER?
> >>>>>
> >>>>>    For example the following PTP messages (called Event messages)
> >>>>>    require time-stamping or correction field update:
> >>>>>
> >>>>>    o  SYNC
> >>>>>
> >>>>>    o  DELAY_REQ (Delay Request)
> >>>>>
> >>>>>    o  PDELAY_REQ (Peer Delay Request)
> >>>>>
> >>>>>    o  PDELAY_RESP (Peer Delay Response)
> >>>>>
> >>>>>    SYNC and DELAY_REQ are exchanged between Master Clock and Slave
> >> Clock
> >>>>>    and MUST be transported over PTP LSPs.  PDELAY_REQ and
> >> PDELAY_RESP
> >>>>>    are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
> >>>>>    Boundary, or Transparent) and SHOULD be transported over single h=
op
> >>>>>    PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP=
,
> >>>>>    and PDELAY_RESP_FOLLOW_UP messages MUST also be transported
> >> over
> >>>>> the
> >>>>>    PTP LSPs.
> >>>>>
> >>>>>    For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
> >>>>>    transported over two PTP LSPs that are in opposite directions.  T=
hese
> >>>>>    PTP LSPs, which are in opposite directions MUST be congruent and=
 co-
> >>>>>    routed.  Alternatively, a single bidirectional co-routed LSP can=
 be
> >>>>>    used.
> >>>>>
> >>>>>    Except as indicated above for the two-step PTP clocks, Non-Event=
 PTP
> >>>>>    message types do not need to be processed by intermediate routers=
.
> >>>>>    These message types MAY be carried in PTP Tunnel LSPs.
> >>>>>
> >>>>> SB> Are you saying that a timing P router has to be msg type sensiti=
ve?
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 15]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 8.  Protection and Redundancy
> >>>>>
> >>>>>
> >>>>> SB> This is a bit of a jump - I don't know how the LSP itself works=
 yet!
> >>>>>
> >>>>>    In order to ensure continuous uninterrupted operation of slave
> >>>>>    clocks, usually as a general practice, slave clocks (or ports) tr=
ack
> >>>>>    redundant master clocks.
> >>>>>
> >>>>>    It is the responsibility of the network operator to ensure that
> >>>>>    physically disjoint Timing LSPs are established between a slave c=
lock
> >>>>>    (or port) and redundant master clocks (or ports).
> >>>>>
> >>>>>    When a slave clock (or port) listens to redundant master clocks o=
r
> >>>>>    ports, any prolonged Timing LSP outage will trigger the slave clo=
ck
> >>>>>    or port to switch to a redundant master clock or port.
> >>>>>
> >>>>>    LSP/PW protection such as Linear protection Switching (1:1, 1+1),
> >>>>>    Ring protection switching or MPLS Fast Reroute (FRR) generally sw=
itch
> >>>>>    alternative path that usually cause a change in delay, which if
> >>>>>    undetected by slave clock can reduce accuracy of the slave clock.
> >>>>>
> >>>>>    Therefore protection switching MAY be used, as long as phase jump=
s
> >>>>>    upon switchover due to differences in path latency are detected a=
nd
> >>>>>    compensated for (such compensation not being required if BCs or p=
eer-
> >>>>>    peer TCs are used throughout).
> >>>>>
> >>>>>    Note that any protection or reroute mechanism that adds additiona=
l
> >>>>>    MPLS label to the label stack, such as Facility Backup Fast Rerou=
te,
> >>>>>    MUST ensure that the pushed label is also a Timing Label to ensur=
e
> >>>>>    recognition of the MPLS frame as containing Timing messages, as i=
t
> >>>>>    transits the backup path.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 16]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 9.  ECMP
> >>>>>
> >>>>>    To ensure the optimal operation of slave clocks and avoid error
> >>>>>    introduced by forward and reverse path delay asymmetry, the physi=
cal
> >>>>>    path for Timing messages from master clock to slave Clock and vic=
e
> >>>>>    versa must be the same for all Event Timing messages listed in
> >>>>>    section 7.
> >>>>>
> >>>>>    Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
> >>>>>    Multipath).
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 17]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 10.  PHP
> >>>>>
> >>>>>    To ensure that the label on the top of the label stack is the Tim=
ing
> >>>>>    LSP Label, PHP MUST not be used.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 18]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 11.  Entropy
> >>>>>
> >>>>>    To ensure all Timing messages in a Timing LSP take the same path,
> >>>>>    Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
> >>>>>    Entropy Label MUST NOT be used for the PWs that are carried insid=
e
> >>>>>    Timing LSP [RFC6391].
> >>>>>
> >>>>> SB> This is incorrect - you mean that all msgs of the same timing
> >>>>> SB> flow need to have the same EL value.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 19]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 12.  OAM, Control and Management
> >>>>>
> >>>>>    In order to monitor Timing LSPs and their encapsulated PWs, they
> MUST
> >>>>>    be able to carry OAM and management messages.  These
> management
> >>>>>    messages MUST be differentiated from Timing messages via already
> >>>>>    defined IETF methods.
> >>>>>
> >>>>>    For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY
> run
> >>>>>    over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
> >>>>>    Management protocols can easily be identified by the UDP Destinat=
ion
> >>>>>    Port number or by GAL/G-ACH respectively.
> >>>>>
> >>>>>    Also BFD, LSP-Ping and other management messages MAY run over the
> >> PWs
> >>>>>    encapsulated in Timing LSP via one of the defined VCCVs (Type 1,=
 3 or
> >>>>>    4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is g=
oing
> >>>>>    to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1=
) or
> >>>>>    GAL-ACH are used to identify such management messages.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 20]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 13.  QoS Considerations
> >>>>>
> >>>>>    In network deployments where not every LSR/LER is Timing-aware, i=
t is
> >>>>>    important to reduce the impact of the non-Timing-aware LSR/LERs o=
n
> >>>>>    the timing recovery in the slave clock.  The Timing messages are=
 time
> >>>>>    critical and must be treated with the highest priority.  Therefor=
e
> >>>>>    Timing over MPLS messages must be treated with the highest priori=
ty
> >>>>>    in the routers.  This can be achieved by proper setup of Timing L=
SPs.
> >>>>>
> >>>>>    It is recommended that the Timing LSPs are setup or configured
> >>>>>    properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC26=
97]
> >>>>>    for drop eligibility.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 21]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 14.  FCS and Checksum Recalculation
> >>>>>
> >>>>>    When time-stamp generation and timing packet adjustment is
> performed
> >>>>>    near the physical port hardware, the process MUST include
> >>>>>    recalculation of the Ethernet FCS.
> >>>>>
> >>>>> SB> The above is confusing - an LSR always recomputes the link layer
> >>>>> SB> CRC which may or may not be Ethernet.
> >>>>>
> >>>>>    Also FCS retention for the
> >>>>>    payload Ethernet described in [RFC4720] MUST NOT be used.
> >>>>>
> >>>>>    For UDP/IP encapsulation mode of Timing over MPLS, the UDP
> checksum
> >>>>>    may be required as per UDP transport standards.
> >>>>>
> >>>>> SB> You really need to be working on getting the IPv6 C?S computatio=
n
> >>>>> SB> removed from PTP msgs.
> >>>>>
> >>>>>    When UDP checksum is used, each Timing-aware LER/LSR must either
> >>>>>    incrementally update the UDP checksum after Time stamping or
> >>>>>    Correction Field update or verify the UDP checksum on reception f=
rom
> >>>>>    upstream and recalculate the checksum completely on transmission=
 to
> >>>>>    downstream node after Time stamping or Correction Field update.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 22]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 15.  Behavior of LER/LSR
> >>>>>
> >>>>>    Timing-capable/aware LERs and LSRs are routers that have one or
> more
> >>>>>
> >>>>> SB> You mean physical interfaces?
> >>>>>
> >>>>>    interfaces that can perform Timing operations (OC/BC/TC) on Timin=
g
> >>>>>    packets and are configured to do so.  Timing-capable/aware LERs a=
nd
> >>>>>    LSRs can advertise their Timing-capability per-interface via cont=
rol
> >>>>>    plane such as OSPF or IS-IS.
> >>>>> SB> ISIS and OSPF are routing protocols.
> >>>>>
> >>>>>   The Timing-capable/aware LERs can then
> >>>>>    signals Timing LSPs via RSVP-TE signaling.  Alternatively the Tim=
ing
> >>>>>    capability of LER and LSRs may be configured in a centralized
> >>>>>    controller and the Timing LSP may be setup using manual configura=
tion
> >>>>>    or other methods such as SDN.
> >>>>>
> >>>>> SB> it can also be configured individually rather then through
> >>>>> SB> a cebtral controllwe
> >>>>>
> >>>>> 15.1.  Behavior of Timing-capable/aware LER
> >>>>>
> >>>>>    When a Timing-capable/aware LER behaves as a Transparent clock
> and
> >>>>>    receives a Timing message from a Timing-capable/aware non-MPLS
> >>>>>    interface, the LER updates the Correction Field (CF) and encapsul=
ates
> >>>>>    and forwards the timing message over previously established Timin=
g
> >>>>>    LSP.
> >>>>>
> >>>>> SB> You need to call out the details so that people properly
> >>>>> SB> understand the definition of the new LSP.
> >>>>>
> >>>>>    Also when a Timing message is received from a Timing-capable/
> >>>>>    aware MPLS interface, LER updates the Correction Filed (CF) and
> >>>>>    decapsulates the MPLS encapsulation and forwards the timing
> message
> >>>>>    to a non-MPLS interface.
> >>>>>
> >>>>>    When a Timing-capable/aware LER behaves as a Boundary clock and
> >>>>>    receives a Timing message from a Timing-capable/aware non MPLS
> >>>>>    interface, the LER Timestamps the Timing packet and sends it to t=
he
> >>>>>    LERs Boundary clock processing module.  Also when a Timing messag=
e
> is
> >>>>>    received from a Timing- capable/aware MPLS interface, the LER
> >>>>>    Timestamps the Timing packet and sends it to the LERs Boundary cl=
ock
> >>>>>    processing module.
> >>>>>
> >>>>>    When a Timing-capable/aware LER behaves as an Ordinary Clock
> toward
> >>>>>    the MPLS network, and receives a Timing message from a Timing-
> >>>>>    capable/aware MPLS interface, the LER Timestamps the Timing packe=
t
> >>>>>    and sends it to the LERs Ordinary clock processing module.
> >>>>>
> >>>>> 15.2.  Behavior of Timing-capable/aware LSR
> >>>>>
> >>>>>    When a Timing-capable/aware LSR behaves as a Transparent clock an=
d
> >>>>>    receives a Timing message from a Timing-capable/aware MPLS
> >> interface,
> >>>>>    The LSR updates the Correction Filed (CF) and forwards the timing
> >>>>>    message over another MPLS interface.
> >>>>>
> >>>>>    When a Timing-capable/aware LSR behaves as a Boundary clock and
> >>>>>    receives a Timing message from a Timing-capable/aware MPLS
> >> interface.
> >>>>>    The LSR performs the functions of a Boundary Clock in terminating=
 the
> >>>>>    received Timing message and re-generating a new timing message
> over
> >>>>>    another (or the same) MPLS interface.
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 23]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 15.3.  Behavior of non-Timing-capable/aware LSR
> >>>>>
> >>>>>    It is most beneficial when all LSRs in the path of a Timing LSP b=
e
> >>>>>    timing-Capable/aware LSRs.  This would ensure the highest quality
> >>>>>    time and clock synchronization by Timing Slave Clocks.  However,=
 this
> >>>>>    specification does not mandate that all LSRs in path of a Timing=
 LSP
> >>>>>    be Timing- capable/aware.
> >>>>>
> >>>>>    Non-Timing-capable/aware LSRs just switch the packets encapsulate=
d
> in
> >>>>>    Timing LSPs and dont perform any Timing operation (TC or BC).
> >>>>>    However as explained in QoS section the Timing over MPLS packets
> >> MUST
> >>>>>    be still be treated with the highest priority based on their Traf=
fic
> >>>>>    Class (TC) marking.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 24]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 16.  Other considerations
> >>>>>
> >>>>>    [IEEE-1588] defines an optional peer-to-peer Transparent clocking
> >>>>>    that requires peer delay measurement between two adjacent Timing-
> >>>>>    capable/ aware routers/switches.  Peer delay measurement messages
> >>>>>    need to be time stamped and terminated by the Timing-capable/awar=
e
> >>>>>    routers/ switches.  This means that two adjacent LSRs may be enga=
ged
> >>>>>    in a peer delay measurement.
> >>>>>
> >>>>>    For transporting such peer delay measurement messages a single-ho=
p
> >>>>>    LSP SHOULD to be created between the two adjacent LSRs engaged in
> >>>>>    peer delay measurement to carry peer delay measurement messages.
> >>>>>    Other methods such as PTP transport over Ethernet MAY be used for
> >>>>>    transporting peer delay measurement messages if the link between=
 the
> >>>>>    two routers is Ethernet.
> >>>>>
> >>>>>    In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/=
 ware
> >>>>>    routers/switches MUST maintain a list of all the neighbors it nee=
ds
> >>>>>    to send a PDelay_Req to, where each neighbor corresponds to a tim=
ing
> >>>>>    LSP.
> >>>>>
> >>>>>    The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as=
 long
> >>>>>    as either the Explicit Null label is the bottom of stack label
> >>>>>    (applicable only to UDP/IP encapsulation) or the label below the
> >>>>>    Explicit Null label is a PTP label.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 25]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 17.  Security Considerations
> >>>>>
> >>>>>    MPLS PW security considerations in general are discussed in
> [RFC3985]
> >>>>>    and [RFC4447],and those considerations also apply to this documen=
t.
> >>>>>
> >>>>>    An experimental security protocol is defined in [IEEE-1588].The P=
TP
> >>>>>    security extension and protocol provides group source authenticat=
ion,
> >>>>>    message integrity, and replay attack protection for PTP messages.
> >>>>>
> >>>>>    When the MPLS network (provider network) serves multiple customer=
s,
> >>>>>    it is important to maintain and process each customers clock and
> >>>>>    Timing messages separately from other customers to ensure there i=
s
> no
> >>>>>    cross- customer effect.  For example if an LER BC is synchronized=
 to
> >>>>>    a specific grandmaster, belonging to customer A, then the LER MUS=
T
> >>>>>    use that BC clock only for customer A to ensure that customer A
> >>>>>    cannot attack other customers by manipulating its time.
> >>>>>
> >>>>>    Timing messages MAY be encrypted or authenticated, provided that
> the
> >>>>>    LERs/LSRs that are Timing capable/aware can authenticate/ decrypt
> the
> >>>>>    timing messages.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 26]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 18.  Acknowledgements
> >>>>>
> >>>>>    The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizr=
ahi,
> >>>>>    Stefano Ruffini, Peter Meyer, and other members of IETF for revie=
wing
> >>>>>    and providing feedback on this draft.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 27]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 19.  IANA Considerations
> >>>>>
> >>>>>    There are no IANA requirements in this specification.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 28]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 20.  References
> >>>>>
> >>>>> 20.1.  Normative References
> >>>>>
> >>>>>    [IEEE-1588]
> >>>>>               IEEE 1588-2008, "IEEE Standard for a Precision Clock
> >>>>>               Synchronization Protocol for Networked Measurement and
> >>>>>               Control Systems".
> >>>>>
> >>>>>    [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
> >>>>>               Requirement Levels", BCP 14, RFC 2119, March 1997.
> >>>>>
> >>>>>    [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to=
-
> >>>>>               Edge (PWE3) Architecture", RFC 3985, March 2005.
> >>>>>
> >>>>>    [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discov=
ery
> >>>>>               Proxies (ND Proxy)", RFC 4389, April 2006.
> >>>>>
> >>>>>    [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G=
.
> >>>>>               Heron, "Pseudowire Setup and Maintenance Using the Lab=
el
> >>>>>               Distribution Protocol (LDP)", RFC 4447, April 2006.
> >>>>>
> >>>>>    [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
> >>>>>               "Encapsulation Methods for Transport of Ethernet over=
 MPLS
> >>>>>               Networks", RFC 4448, April 2006.
> >>>>>
> >>>>>    [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
> >>>>>               Emulation Edge-to-Edge (PWE3) Frame Check Sequence
> >>>>>               Retention", RFC 4720, November 2006.
> >>>>>
> >>>>>    [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circu=
it
> >>>>>               Connectivity Verification (VCCV): A Control Channel fo=
r
> >>>>>               Pseudowires", RFC 5085, December 2007.
> >>>>>
> >>>>>    [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detect=
ion
> >>>>>               (BFD)", RFC 5880, June 2010.
> >>>>>
> >>>>>    [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow=
,
> >>>>>               "Bidirectional Forwarding Detection (BFD) for MPLS Lab=
el
> >>>>>               Switched Paths (LSPs)", RFC 5884, June 2010.
> >>>>>
> >>>>> 20.2.  Informative References
> >>>>>
> >>>>>    [I-D.ietf-pwe3-fat-pw]
> >>>>>               Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Reg=
an,
> >>>>>               J., and S. Amante, "Flow Aware Transport of Pseudowire=
s
> >>>>>               over an MPLS Packet Switched Network",
> >>>>>               draft-ietf-pwe3-fat-pw-07 (work in progress), July 201=
1.
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 29]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>>    [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermedia=
te
> >>>>>               system routeing information exchange protocol for use=
 in
> >>>>>               conjunction with the Protocol for providing the
> >>>>>               Connectionless-mode Network Service (ISO 8473)".
> >>>>>
> >>>>>    [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP an=
d
> >>>>>               dual environments", RFC 1195, December 1990.
> >>>>>
> >>>>>    [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 199=
8.
> >>>>>
> >>>>>    [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
> >>>>>               Marker", RFC 2697, September 1999.
> >>>>>
> >>>>>    [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boud=
ec,
> >>>>>               J., Courtney, W., Davari, S., Firoiu, V., and D.
> >>>>>               Stiliadis, "An Expedited Forwarding PHB (Per-Hop
> >>>>>               Behavior)", RFC 3246, March 2002.
> >>>>>
> >>>>>    [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Enginee=
ring
> >>>>>               (TE) Extensions to OSPF Version 2", RFC 3630,
> >>>>>               September 2003.
> >>>>>
> >>>>>    [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermedia=
te
> >>>>>               System (IS-IS) Extensions for Traffic Engineering (TE)=
",
> >>>>>               RFC 3784, June 2004.
> >>>>>
> >>>>>    [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and=
 S.
> >>>>>               Shaffer, "Extensions to OSPF for Advertising Optional
> >>>>>               Router Capabilities", RFC 4970, July 2007.
> >>>>>
> >>>>>    [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
> >>>>>               System to Intermediate System (IS-IS) Extensions for
> >>>>>               Advertising Router Information", RFC 4971, July 2007.
> >>>>>
> >>>>>    [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
> >>>>>               Topology (MT) Routing in Intermediate System to
> >>>>>               Intermediate Systems (IS-ISs)", RFC 5120, February 200=
8.
> >>>>>
> >>>>>    [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
> >>>>>               Engineering", RFC 5305, October 2008.
> >>>>>
> >>>>>    [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
> >>>>>               "Traffic Engineering Extensions to OSPF Version 3",
> >>>>>               RFC 5329, September 2008.
> >>>>>
> >>>>>    [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSP=
F
> >>>>>               for IPv6", RFC 5340, July 2008.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 30]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>>    [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Net=
work
> >>>>>               Time Protocol Version 4: Protocol and Algorithms
> >>>>>               Specification", RFC 5905, June 2010.
> >>>>>
> >>>>>    [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Reg=
an,
> >>>>>               J., and S. Amante, "Flow-Aware Transport of Pseudowire=
s
> >>>>>               over an MPLS Packet Switched Network", RFC 6391,
> >>>>>               November 2011.
> >>>>>
> >>>>>    [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., a=
nd
> >>>>>               L. Yong, "The Use of Entropy Labels in MPLS Forwarding=
",
> >>>>>               RFC 6790, November 2012.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 31]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 1.  Routing extensions for Timing-aware Routers
> >>>>>
> >>>>>    MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340]
> and
> >>>>>    IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (=
TE)
> >>>>>    link information used for constraint-based routing.
> >>>>>
> >>>>>    Indeed, it is useful to advertise data plane TE router link
> >>>>>    capabilities, such as the capability for a router to be Timing-aw=
are.
> >>>>>    This capability MUST then be taken into account during path
> >>>>>    computation to prefer or even require links that advertise themse=
lves
> >>>>>    as Timing-aware.  In this way the path can ensure the entry and e=
xit
> >>>>>    points into the LERs and, if desired, the links into the LSRs are
> >>>>>    able to perform port based time-stamping thus minimizing their im=
pact
> >>>>>    on the performance of the slave clock.
> >>>>>
> >>>>>    extensions are required to OSPF and IS-IS in order to advertise
> >>>>>    Timing-aware capabilities of a link.  Such extensions are outside=
 the
> >>>>>    scope of this document; however such extension SHOULD be able to
> >>>>>    signal the following information per Router Link:
> >>>>>
> >>>>>    o  Capable of processing PTP, NTP or other Timing flows
> >>>>>
> >>>>>    o  Capable of performing Transparent Clock operation
> >>>>>
> >>>>>    o  Capable of performing Boundary Clock operation
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 32]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> 2.  Signaling Extensions for Creating Timing LSPs
> >>>>>
> >>>>>    RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSV=
P-
> TE
> >>>>>    is used to setup Timing LSPs, some information that indicates tha=
t
> >>>>>    the LSP is carrying Timing flows MUST be included in the new
> >>>>>    Extensions to RSVP-TE:
> >>>>>
> >>>>>    The following information MAY also be included in the new Extensi=
ons
> >>>>>    to RSVP-TE:
> >>>>>
> >>>>>    o  Offset from Bottom of Stack (BoS) to the start of the Time-sta=
mp
> >>>>>       field
> >>>>>
> >>>>>    o  Number of VLANs in case of PW encapsulation
> >>>>>
> >>>>>    o  Timestamp field Type
> >>>>>
> >>>>>       *  Correction Field, Timestamp
> >>>>>
> >>>>>    o  Timestamp Field format
> >>>>>
> >>>>>       *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
> >>>>>          NTP, etc.
> >>>>>
> >>>>>    Note that in case the above optional information is signaled with
> >>>>>    RSVP-TE for a Timing LSP, all the Timing packets carried in that=
 LSP
> >>>>>    must have the same signaled characteristics.  For example if
> >>>>>    Timestamp format is signaled as 64-bit PTPv1, then all Timing pac=
kets
> >>>>>    must use 64-bit PTPv1 time-stamp.
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 33]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>> Authors' Addresses
> >>>>>
> >>>>>    Shahram Davari
> >>>>>    Broadcom Corp.
> >>>>>    San Jose, CA  95134
> >>>>>    USA
> >>>>>
> >>>>>    Email: davari@broadcom.com
> >>>>>
> >>>>>
> >>>>>    Amit Oren
> >>>>>    Broadcom Corp.
> >>>>>    San Jose, CA  95134
> >>>>>    USA
> >>>>>
> >>>>>    Email: amito@broadcom.com
> >>>>>
> >>>>>
> >>>>>    Manav Bhatia
> >>>>>    Alcatel-Lucent
> >>>>>    Bangalore,
> >>>>>    India
> >>>>>
> >>>>>    Email: manav.bhatia@alcatel-lucent.com
> >>>>>
> >>>>>
> >>>>>    Peter Roberts
> >>>>>    Alcatel-Lucent
> >>>>>    Kanata,
> >>>>>    Canada
> >>>>>
> >>>>>    Email: peter.roberts@alcatel-lucent.com
> >>>>>
> >>>>>
> >>>>>    Laurent Montini
> >>>>>    Cisco Systems
> >>>>>    San Jose CA
> >>>>>    USA
> >>>>>
> >>>>>    Email: lmontini@cisco.com
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 34]
> >>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
> >>>>>
> >>>>>
> >>>>>    Luca
> >>>>>    Cisco Systems
> >>>>>    San Jose CA
> >>>>>    USA
> >>>>>
> >>>>>    Email: lmartini@cisco.com
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> Davari, et al.          Expires December 17, 2013              [Page=
 35]
> >>>>> --
> >>>>> For corporate legal information go to:
> >>>>>
> >>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
> >>>>>
> >>>>> _______________________________________________
> >>>>> TICTOC mailing list
> >>>>> TICTOC@ietf.org
> >>>>> https://www.ietf.org/mailman/listinfo/tictoc
> >>>>
> >>>> This e-mail message is intended for the recipient only and contains
> >> information which is CONFIDENTIAL and which may be proprietary to ECI
> >> Telecom. If you have received this transmission in error, please inform=
 us by
> e-
> >> mail, phone or fax, and then delete the original and all copies thereof=
.
> >>>>
> >>>> _______________________________________________
> >>>> mpls mailing list
> >>>> mpls@ietf.org
> >>>> https://www.ietf.org/mailman/listinfo/mpls
> >>> .
> >>
> >>
> >> --
> >> For corporate legal information go to:
> >>
> >> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
> >>
> >> _______________________________________________
> >> TICTOC mailing list
> >> TICTOC@ietf.org
> >> https://www.ietf.org/mailman/listinfo/tictoc
> >>
> >>
> >> _______________________________________________
> >> TICTOC mailing list
> >> TICTOC@ietf.org
> >> https://www.ietf.org/mailman/listinfo/tictoc
> >
> >
> > This e-mail message is intended for the recipient only and contains
> information which is CONFIDENTIAL and which may be proprietary to ECI
> Telecom. If you have received this transmission in error, please inform us=
 by e-
> mail, phone or fax, and then delete the original and all copies thereof.
> >


This e-mail message is intended for the recipient only and contains informat=
ion which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If yo=
u have received this transmission in error, please inform us by e-mail, phon=
e or fax, and then delete the original and all copies thereof.


From mills@udel.edu  Tue Aug  6 18:02:17 2013
Return-Path: <mills@udel.edu>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 4644621F9D4A for <tictoc@ietfa.amsl.com>; Tue,  6 Aug 2013 18:02:17 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.598
X-Spam-Level: 
X-Spam-Status: No, score=-6.598 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id dxaB5fORV3G1 for <tictoc@ietfa.amsl.com>; Tue,  6 Aug 2013 18:02:11 -0700 (PDT)
Received: from mail.eecis.udel.edu (louie.udel.edu [128.4.40.12]) by ietfa.amsl.com (Postfix) with ESMTP id 4A39D21F9D38 for <tictoc@ietf.org>; Tue,  6 Aug 2013 18:02:10 -0700 (PDT)
Received: from [192.168.1.2] (pool-74-109-13-78.phlapa.fios.verizon.net [74.109.13.78]) (Authenticated sender: mills@mail.eecis.udel.edu) by mail.eecis.udel.edu (Postfix) with ESMTP id 51D9A1071; Tue,  6 Aug 2013 21:01:50 -0400 (EDT)
Message-ID: <52019C7B.9070602@udel.edu>
Date: Wed, 07 Aug 2013 01:01:47 +0000
From: "David L. Mills" <mills@udel.edu>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax)
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Kurt Roeckx <kurt@roeckx.be>
References: <51FAC820.3090401@udel.edu> <20130803174008.GA17578@roeckx.be>
In-Reply-To: <20130803174008.GA17578@roeckx.be>
Content-Type: multipart/alternative; boundary="------------040201000803070007040309"
Cc: NTP Working Group <ntpwg@lists.ntp.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [ntpwg] security ID submitted for review
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 07 Aug 2013 01:02:17 -0000

This is a multi-part message in MIME format.
--------------040201000803070007040309
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Kurt,

Thanks for your reply.  As a window of history, the Autokey protocol was 
devised during several walks in Regents Park, London, in 1996.  I 
thought the most significant hazzard was the replay of client messages 
requiring extensive server cryptographic computations.  At that time, I 
did not think that a middle-man attack would be significant.  Both of 
those assumptions are probably inadvisable today. 

One problem with Autokey is the assumption that secret keys for each 
client can be manufactured using packet header fields and a hash of a 
server private value.  A middle-man can manipulate the header fields 
compromising the client private key.  There does not seem to be a 
solution for this problem other than requiring server state.

There are two problems with requiring a server state variable associated 
with each client.  The first is how to reliably match the server state 
with the client packet.  For public time servers with thousands of 
clients, the storage of something like a client certificate for each 
client could be prohibitive.  The storage could be dynamically allocated 
as a field in the MRU list maintained in the current NTP implimentation; 
however, the MRU list for several hundred entries has a lifetime of only 
15 seconds, at NIST, for example. 

In the current Autokey design, the MAC computation covers the header and 
all extension fields.  However, the key used for the MAC computation in 
packets with extension fields is not private.  Therefore, in a 
cut-and-paste attack, an intruder could mix and match extension fields 
and simply recalculate the MAC. 

RFC 5906 defines the extension field including the filestamp and 
timestamp.  The filestamp was designed to reliably associate the 
cryptographic media with the current version, as it can be frequently 
changed.  This might not be useful in current designs.  The timestamp 
field was designed to detect replays requiring expensive cryptographic 
computations which could take up to a second.  However, the "call gap" 
(minimum headway) provisions nicely deflect such attacks, so the 
timestamp field might not be useful in current designs.   Lastly, in the 
current design, each extension field is separately signed  at the time 
the data were constructed.  In current designs the signature might be 
computed at the time of transmission.

The more I think of it, the better I like a design where the existing 
MAC is replaced by a single extension field computed as the signature of 
the header.  I previously suggested this for broadcast mode, but I think 
this could be useful in other modes as well.  It would be hightly 
interesting to determine the overhead for such a signature using modern 
processors.  An experiment could be designed for the current NTP 
implementation.  It would consist of an adaptor routine that matches an 
OpenSSL signature routine to the authencrypt/authdecrypt interface 
routine.  I would do this myself, but with my current vision condition, 
it would not be practical.  For someone with knowledge of the NTP 
programming style, this would be an interesting weekend project.

In theis proposed design a busy server could be presented with thousands 
of signed client requests, each requiring  a signed server response.  
However, the server might ignore the client request signature on the 
expectation that the  server response could be reliably determined.  
Actually, the NTP on-wire protocol is very good for this.  As an aside, 
the protocol analysis and simulation described in the white paper at the 
NTP project page could be done for PTP as well.

Dave

Kurt Roeckx wrote:

>On Thu, Aug 01, 2013 at 08:42:08PM +0000, David L. Mills wrote:
>  
>
>>Guys,
>>
>>Here are some general comments on the ID.  I apologize if they are a
>>little rough as I have to dictate this message instead of my usual
>>polished wordsmithing.
>>
>>format.
>>The ID does not specify an extension field format.
>>    
>>
>
>The draft only seems to specify the the general idea of how it
>should work and doesn't seem to go into detail how this should
>work at the protocol level.  I assume he first wants to know if
>the solution is something that can work and that others see
>as a good solution before he works out the details.
>
>  
>
>>In particular, the private key for the message
>>authentication code (MAC) , if used, must be known beforehand.  This
>>is where the "cookie snatcher" issue comes up as described in
>>previous messages.  In particular, an important constraint for
>>normal operation is that extension fields are not used unless
>>required for cookie determination.
>>    
>>
>
>I'm not sure I understand your remark.  As far as I know the
>"cookie snatcher" you talk about is the normal "man in the middle"
>attack.  As I understand we would change the verification by
>replacing it by the normal X.509 verification as done for instance
>SSL/TLS.
>
>I assume you meant public key and note private key.
>
>  
>
>>Glaring vulnerability exists when using extension fields, as the key
>>value is public and must be known in advance.  This invites a
>>cut-and-paste attack, as well as header changes, even if
>>subsequently detected by the on-wire protocol.
>>    
>>
>
>I've seen this cut-and-paste attack mentioned before, but I've
>never really understood it.  As I understand things, the MAC
>is generated over the NTP + extensions.  I don't see how you
>can take the extension from 1 packet into an other and still
>have a valid MAC.  Or are you saying that you do something
>with a packet even when it doesn't have a valid MAC?
>
>  
>
>>certificate.
>>Both RFC 5906 and the ID require the server certificate.  The server
>>certificate can be conveyed, either off-line, or directly or
>>indirectly via a certificate trail as described in RFC 5906 or the
>>ID.  If obtained via the trail, there is a hazzard due to a
>>middle-man attack and masquerade.
>>    
>>
>
>I really don't understand how such a man in the middle attack
>can work.  Either the certificate + the chain validate, or they
>don't.  It is of course important to you check that the CommonName
>in the certificate matches the server you're trying to reach,
>and that the root CA is in your list of trusted CAs.  But there
>really isn't anything new or hard about this.
>
>  
>
>>In broadcast mode, the protocols described in RFC 5906 and the ID
>>are messy and with several possible vulnerabilities.  Here is a
>>suggestion that might be useful.  Define a new broadcast packet
>>format as the NTP header followed by one extension field
>>representing the MAC.  The extension field represents the signature
>>of the header computed using the private key of the server and can
>>be verified by the client using the public key in the server
>>certificate.  Originally, this was considered an unacceptable
>>dilution of accuracy due to the variability of the signature
>>computing time and the length of the signature.  However, at least
>>in modern NTP, the interleaved broadcast mode can be used to correct
>>for this overhead.  This assumes that the overhead for either the
>>server or client is reduced using fast modern processors.
>>    
>>
>
>I think that in all cases where it's no problem to keep track of
>all clients you should use the interleaved mode, and we can use
>something like the Diffie-Hellman key agreement for those.
>
>But there clearly is also a need for servers that can't keep
>the state of each client, and we need something else for that.
>So maybe we need to define 2 ways of working.
>
>
>
>Kurt
>
>  
>


--------------040201000803070007040309
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
  <title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
Kurt,<br>
<br>
Thanks for your reply.&nbsp; As a window of history, the Autokey protocol
was devised during several walks in Regents Park, London, in 1996.&nbsp; I
thought the most significant hazzard was the replay of client messages
requiring extensive server cryptographic computations.&nbsp; At that time, I
did not think that a middle-man attack would be significant.&nbsp; Both of
those assumptions are probably inadvisable today.&nbsp; <br>
<br>
One problem with Autokey is the assumption that secret keys for each
client can be manufactured using packet header fields and a hash of a
server private value.&nbsp; A middle-man can manipulate the header fields
compromising the client private key.&nbsp; There does not seem to be a
solution for this problem other than requiring server state. <br>
<br>
There are two problems with requiring a server state variable
associated with each client.&nbsp; The first is how to reliably match the
server state with the client packet.&nbsp; For public time servers with
thousands of clients, the storage of something like a client
certificate for each client could be prohibitive.&nbsp; The storage could be
dynamically allocated as a field in the MRU list maintained in the
current NTP implimentation; however, the MRU list for several hundred
entries has a lifetime of only 15 seconds, at NIST, for example.&nbsp; <br>
<br>
In the current Autokey design, the MAC computation covers the header
and all extension fields.&nbsp; However, the key used for the MAC
computation in packets with extension fields is not private.&nbsp;
Therefore, in a cut-and-paste attack, an intruder could mix and match
extension fields and simply recalculate the MAC.&nbsp; <br>
<br>
RFC 5906 defines the extension field including the filestamp and
timestamp.&nbsp; The filestamp was designed to reliably associate the
cryptographic media with the current version, as it can be frequently
changed.&nbsp; This might not be useful in current designs.&nbsp; The timestamp
field was designed to detect replays requiring expensive cryptographic
computations which could take up to a second.&nbsp; However, the "call gap"
(minimum headway) provisions nicely deflect such attacks, so the
timestamp field might not be useful in current designs.&nbsp;&nbsp; Lastly, in
the current design, each extension field is separately signed&nbsp; at the
time the data were constructed.&nbsp; In current designs the signature might
be computed at the time of transmission.<br>
<br>
The more I think of it, the better I like a design where the existing
MAC is replaced by a single extension field computed as the signature
of the header.&nbsp; I previously suggested this for broadcast mode, but I
think this could be useful in other modes as well.&nbsp; It would be hightly
interesting to determine the overhead for such a signature using modern
processors.&nbsp; An experiment could be designed for the current NTP
implementation.&nbsp; It would consist of an adaptor routine that matches an
OpenSSL signature routine to the authencrypt/authdecrypt interface
routine.&nbsp; I would do this myself, but with my current vision condition,
it would not be practical.&nbsp; For someone with knowledge of the NTP
programming style, this would be an interesting weekend project.<br>
<br>
In theis proposed design a busy server could be presented with
thousands of signed client requests, each requiring&nbsp; a signed server
response.&nbsp; However, the server might ignore the client request
signature on the expectation that the&nbsp; server response could be
reliably determined.&nbsp; Actually, the NTP on-wire protocol is very good
for this.&nbsp; As an aside, the protocol analysis and simulation described
in the white paper at the NTP project page could be done for PTP as
well.<br>
<br>
Dave<br>
<br>
Kurt Roeckx wrote:
<blockquote cite="mid20130803174008.GA17578@roeckx.be" type="cite">
  <pre wrap="">On Thu, Aug 01, 2013 at 08:42:08PM +0000, David L. Mills wrote:
  </pre>
  <blockquote type="cite">
    <pre wrap="">Guys,

Here are some general comments on the ID.  I apologize if they are a
little rough as I have to dictate this message instead of my usual
polished wordsmithing.

format.
The ID does not specify an extension field format.
    </pre>
  </blockquote>
  <pre wrap=""><!---->
The draft only seems to specify the the general idea of how it
should work and doesn't seem to go into detail how this should
work at the protocol level.  I assume he first wants to know if
the solution is something that can work and that others see
as a good solution before he works out the details.

  </pre>
  <blockquote type="cite">
    <pre wrap="">In particular, the private key for the message
authentication code (MAC) , if used, must be known beforehand.  This
is where the "cookie snatcher" issue comes up as described in
previous messages.  In particular, an important constraint for
normal operation is that extension fields are not used unless
required for cookie determination.
    </pre>
  </blockquote>
  <pre wrap=""><!---->
I'm not sure I understand your remark.  As far as I know the
"cookie snatcher" you talk about is the normal "man in the middle"
attack.  As I understand we would change the verification by
replacing it by the normal X.509 verification as done for instance
SSL/TLS.

I assume you meant public key and note private key.

  </pre>
  <blockquote type="cite">
    <pre wrap="">Glaring vulnerability exists when using extension fields, as the key
value is public and must be known in advance.  This invites a
cut-and-paste attack, as well as header changes, even if
subsequently detected by the on-wire protocol.
    </pre>
  </blockquote>
  <pre wrap=""><!---->
I've seen this cut-and-paste attack mentioned before, but I've
never really understood it.  As I understand things, the MAC
is generated over the NTP + extensions.  I don't see how you
can take the extension from 1 packet into an other and still
have a valid MAC.  Or are you saying that you do something
with a packet even when it doesn't have a valid MAC?

  </pre>
  <blockquote type="cite">
    <pre wrap="">certificate.
Both RFC 5906 and the ID require the server certificate.  The server
certificate can be conveyed, either off-line, or directly or
indirectly via a certificate trail as described in RFC 5906 or the
ID.  If obtained via the trail, there is a hazzard due to a
middle-man attack and masquerade.
    </pre>
  </blockquote>
  <pre wrap=""><!---->
I really don't understand how such a man in the middle attack
can work.  Either the certificate + the chain validate, or they
don't.  It is of course important to you check that the CommonName
in the certificate matches the server you're trying to reach,
and that the root CA is in your list of trusted CAs.  But there
really isn't anything new or hard about this.

  </pre>
  <blockquote type="cite">
    <pre wrap="">In broadcast mode, the protocols described in RFC 5906 and the ID
are messy and with several possible vulnerabilities.  Here is a
suggestion that might be useful.  Define a new broadcast packet
format as the NTP header followed by one extension field
representing the MAC.  The extension field represents the signature
of the header computed using the private key of the server and can
be verified by the client using the public key in the server
certificate.  Originally, this was considered an unacceptable
dilution of accuracy due to the variability of the signature
computing time and the length of the signature.  However, at least
in modern NTP, the interleaved broadcast mode can be used to correct
for this overhead.  This assumes that the overhead for either the
server or client is reduced using fast modern processors.
    </pre>
  </blockquote>
  <pre wrap=""><!---->
I think that in all cases where it's no problem to keep track of
all clients you should use the interleaved mode, and we can use
something like the Diffie-Hellman key agreement for those.

But there clearly is also a need for servers that can't keep
the state of each client, and we need something else for that.
So maybe we need to define 2 ways of working.



Kurt

  </pre>
</blockquote>
<br>
</body>
</html>

--------------040201000803070007040309--

From kurt@roeckx.be  Wed Aug  7 03:23:36 2013
Return-Path: <kurt@roeckx.be>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7747921E80BE for <tictoc@ietfa.amsl.com>; Wed,  7 Aug 2013 03:23:35 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.998
X-Spam-Level: 
X-Spam-Status: No, score=-2.998 tagged_above=-999 required=5 tests=[AWL=-0.399, BAYES_00=-2.599]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id wMwP+L0XPdGp for <tictoc@ietfa.amsl.com>; Wed,  7 Aug 2013 03:23:29 -0700 (PDT)
Received: from gerard.telenet-ops.be (gerard.telenet-ops.be [195.130.132.48]) by ietfa.amsl.com (Postfix) with ESMTP id C8C4A21E80D7 for <tictoc@ietf.org>; Wed,  7 Aug 2013 03:23:27 -0700 (PDT)
Received: from intrepid.roeckx.be ([94.226.199.45]) by gerard.telenet-ops.be with bizsmtp id 9mPQ1m00G0zFtyu0HmPQkB; Wed, 07 Aug 2013 12:23:26 +0200
Received: by intrepid.roeckx.be (Postfix, from userid 1000) id 19564EB1C9; Wed,  7 Aug 2013 12:23:24 +0200 (CEST)
Date: Wed, 7 Aug 2013 12:23:24 +0200
From: Kurt Roeckx <kurt@roeckx.be>
To: "David L. Mills" <mills@udel.edu>
Message-ID: <20130807102324.GA17618@roeckx.be>
References: <51FAC820.3090401@udel.edu> <20130803174008.GA17578@roeckx.be> <52019C7B.9070602@udel.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <52019C7B.9070602@udel.edu>
User-Agent: Mutt/1.5.21 (2010-09-15)
Cc: NTP Working Group <ntpwg@lists.ntp.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [ntpwg] security ID submitted for review
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 07 Aug 2013 10:23:36 -0000

Hi Dave,

On Wed, Aug 07, 2013 at 01:01:47AM +0000, David L. Mills wrote:
> Kurt,
> 
> Thanks for your reply.  As a window of history, the Autokey protocol
> was devised during several walks in Regents Park, London, in 1996.
> I thought the most significant hazzard was the replay of client
> messages requiring extensive server cryptographic computations.  At
> that time, I did not think that a middle-man attack would be
> significant.  Both of those assumptions are probably inadvisable
> today.
> 
> One problem with Autokey is the assumption that secret keys for each
> client can be manufactured using packet header fields and a hash of
> a server private value.  A middle-man can manipulate the header
> fields compromising the client private key.  There does not seem to
> be a solution for this problem other than requiring server state.

This proposal also generates a cookie based on the something the
client sends and a secret in the server, which is then encrypted
using the client's public key.

The ID does not mention that this message is signed by the server,
but Stephen's mail said it was.  This part is important to avoid
the man in the middle attack.

So as result the client got a unique secret for him, and only
the client and the server known it.  It can also be sure that
it really comes from the server, so that there can be no man
in the middle attack.  And the server doesn't need to keep
state since it can calculate it again based on what the client
sends him.

> In the current Autokey design, the MAC computation covers the header
> and all extension fields.  However, the key used for the MAC
> computation in packets with extension fields is not private.
> Therefore, in a cut-and-paste attack, an intruder could mix and
> match extension fields and simply recalculate the MAC.

If the key is not private, by adding extentions to a packet an
attacker can just generate anything he wants?

> RFC 5906 defines the extension field including the filestamp and
> timestamp.  The filestamp was designed to reliably associate the
> cryptographic media with the current version, as it can be
> frequently changed.  This might not be useful in current designs.
> The timestamp field was designed to detect replays requiring
> expensive cryptographic computations which could take up to a
> second.  However, the "call gap" (minimum headway) provisions nicely
> deflect such attacks, so the timestamp field might not be useful in
> current designs.   Lastly, in the current design, each extension
> field is separately signed  at the time the data were constructed.
> In current designs the signature might be computed at the time of

I'm not sure I understand what you're trying to say.

I'm not sure what the filestamps are used for.  But I assume
that for instance for the leap seconds you send when it was
last updates.  I guess you can sign this once and always send
this signed in an extension, even when you're otherwise not
using any encryption.  Since you already signed it, the overhead
of adding that extension is small.

The problem I see with only sending the MAC over the ntp headers
and not over the extensions is that the extensions can be removed
or added and the client can't tell.  So it could for instance
be replaced with an old version of the extension. I think it's
important that the MAC covers everything.


Kurt


From kurt@roeckx.be  Wed Aug  7 04:45:45 2013
Return-Path: <kurt@roeckx.be>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 154AA21F9BD8 for <tictoc@ietfa.amsl.com>; Wed,  7 Aug 2013 04:45:45 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.17
X-Spam-Level: 
X-Spam-Status: No, score=-3.17 tagged_above=-999 required=5 tests=[AWL=-0.571,  BAYES_00=-2.599]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id HYVvQq8nwqCz for <tictoc@ietfa.amsl.com>; Wed,  7 Aug 2013 04:45:39 -0700 (PDT)
Received: from jacques.telenet-ops.be (jacques.telenet-ops.be [195.130.132.50]) by ietfa.amsl.com (Postfix) with ESMTP id CA37621F9A40 for <tictoc@ietf.org>; Wed,  7 Aug 2013 04:45:38 -0700 (PDT)
Received: from intrepid.roeckx.be ([94.226.199.45]) by jacques.telenet-ops.be with bizsmtp id 9nlX1m00R0zFtyu0JnlXda; Wed, 07 Aug 2013 13:45:32 +0200
Received: by intrepid.roeckx.be (Postfix, from userid 1000) id 9F804EB1C9; Wed,  7 Aug 2013 13:45:30 +0200 (CEST)
Date: Wed, 7 Aug 2013 13:45:30 +0200
From: Kurt Roeckx <kurt@roeckx.be>
To: "David L. Mills" <mills@udel.edu>
Message-ID: <20130807114530.GA20391@roeckx.be>
References: <51FAC820.3090401@udel.edu> <20130803174008.GA17578@roeckx.be> <52019C7B.9070602@udel.edu> <20130807102324.GA17618@roeckx.be>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20130807102324.GA17618@roeckx.be>
User-Agent: Mutt/1.5.21 (2010-09-15)
Cc: NTP Working Group <ntpwg@lists.ntp.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [ntpwg] security ID submitted for review
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 07 Aug 2013 11:45:45 -0000

On Wed, Aug 07, 2013 at 12:23:24PM +0200, Kurt Roeckx wrote:
> The problem I see with only sending the MAC over the ntp headers
> and not over the extensions is that the extensions can be removed
> or added and the client can't tell.  So it could for instance
> be replaced with an old version of the extension. I think it's
> important that the MAC covers everything.

You could of course also add a timestamp indicating when the
signature of the extension expires.  I think we should do
that in any case.

Then I'm not sure if there is still a need to to sign the
whole packet.  Can all extensions be signed? Is there a
need to sign them all?


Kurt


From dieter.sibold@ptb.de  Wed Aug  7 05:19:44 2013
Return-Path: <dieter.sibold@ptb.de>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B049D11E8129 for <tictoc@ietfa.amsl.com>; Wed,  7 Aug 2013 05:19:44 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.248
X-Spam-Level: 
X-Spam-Status: No, score=-2.248 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HELO_EQ_DE=0.35, HTML_MESSAGE=0.001]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ezuAaqQhKjip for <tictoc@ietfa.amsl.com>; Wed,  7 Aug 2013 05:19:39 -0700 (PDT)
Received: from mx1.bs.ptb.de (mx1.bs.ptb.de [192.53.103.106]) by ietfa.amsl.com (Postfix) with ESMTP id 4F7FF21E80D7 for <tictoc@ietf.org>; Wed,  7 Aug 2013 05:19:39 -0700 (PDT)
Received: from mx1.bs.ptb.de (localhost [127.0.0.1]) by localhost (Postfix) with SMTP id EF793CE67F1; Wed,  7 Aug 2013 14:19:37 +0200 (CEST)
Received: from rose.bs.ptb.de (rose.bs.ptb.de [141.25.85.201]) by mx1.bs.ptb.de (Postfix) with ESMTP id D3BF2CE2DEB; Wed,  7 Aug 2013 14:19:37 +0200 (CEST)
In-Reply-To: <20130803174008.GA17578@roeckx.be>
References: <51FAC820.3090401@udel.edu> <20130803174008.GA17578@roeckx.be>
To: Kurt Roeckx <kurt@roeckx.be>
MIME-Version: 1.0
X-KeepSent: CF046B68:1B709B01-C1257BC0:002ECD08; type=4; name=$KeepSent
X-Mailer: IBM Notes Release 9.0 March 08, 2013
Message-ID: <OFCF046B68.1B709B01-ONC1257BC0.002ECD08-C1257BC0.0043B5D5@ptb.de>
From: dieter.sibold@ptb.de
Date: Wed, 7 Aug 2013 14:19:33 +0200
X-MIMETrack: S/MIME Sign by Notes Client on Dieter Sibold/PTB(Release 9.0|March 08, 2013) at 07.08.2013 14:19:34, Serialize by Notes Client on Dieter Sibold/PTB(Release 9.0|March 08, 2013) at 07.08.2013 14:19:34, Serialize complete at 07.08.2013 14:19:34, Itemize by Notes Client on Dieter Sibold/PTB(Release 9.0|March 08, 2013) at 07.08.2013 14:19:36, S/MIME Sign complete at 07.08.2013 14:19:36, Serialize by Router on ROSE/PTB(Release 8.5.3FP3HF370 | April 24, 2013) at 08/07/2013 14:19:36, Serialize complete at 08/07/2013 14:19:36
Content-Type: multipart/signed; protocol="application/x-pkcs7-signature"; micalg=sha1; boundary=-------z50770_boundary_sign
Cc: NTP Working Group <ntpwg@lists.ntp.org>, "tictoc@ietf.org" <tictoc@ietf.org>, "David L. Mills" <mills@udel.edu>
Subject: Re: [TICTOC] [ntpwg] security ID submitted for review
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 07 Aug 2013 12:19:44 -0000

This is an S/MIME signed message.

---------z50770_boundary_sign
Content-Type: multipart/alternative; boundary="=_alternative 0043B5D4C1257BC0_="

Dies ist eine mehrteilige Nachricht im MIME-Format.
--=_alternative 0043B5D4C1257BC0_=
Content-Type: text/plain; charset="US-ASCII"

Hello Dave and Kurt,

thanks for your comments:

As Kurt already pointed out. Currently the ID specifies only the design of 
the security protocol. We believe that we need to be clear about the 
design questions before specifying the technical details. 

The private key of the client is not needed for the MAC. Instead the 
calculation of the MAC includes the cookie and the whole NTP packet (i.e. 
the extension fields also). The cookie is calculated from the server seed 
and the client's public key, which the client attaches to each time 
request (which requires a new extension field as pointed out in the 
draft). During cookie exchange the server encrypts the cookie with the 
client's public key before transmission. An adversary is therefore not 
able to abuse a cookie which he requests in behalf of the client. 

As Kurt pointed out the certificate exchange works as in a standard PKI. I 
emphasize that this is different to autokey. In autokey the chain of trust 
and time are equal whereas in this ID chain of trust and time might be 
disjunct.

Regarding the broadcast mode. I agree with Dave that the use of TESLA for 
the broadcast mode adds additional complexity.  It also has the 
disadvantage that the authenticity of a packet can only be verified in 
retrospect. But I don't see security vulnerabilities. Dave's suggestion to 
use a combination of interleave mode with asymmetric cryptography is 
interesting. We should consider this. The argument that asymmetric 
cryptography might be feasible because of modern power of CPUs is okay for 
traditional computer systems but I'm not sure whether this assumption also 
applies for small devices with constrained resources. 

Regards
Dieter


----------------------------------
Physikalisch-Technische Bundesanstalt
Dr. Dieter Sibold
Q.42 - Serversysteme und Datenhaltung
QM-Verantwortlicher der Stelle IT-Infrastruktur
Bundesallee 100 
D-38116 Braunschweig
Tel: +49-531-592-84 20
E-Mail: dieter.sibold@ptb.de



Von:    Kurt Roeckx <kurt@roeckx.be>
An:     "David L. Mills" <mills@udel.edu>
Kopie:  NTP Working Group <ntpwg@lists.ntp.org>, "tictoc@ietf.org" 
<tictoc@ietf.org>
Datum:  03.08.2013 20:04
Betreff:        Re: [ntpwg] security ID submitted for review
Gesendet von:   ntpwg-bounces+dieter.sibold=ptb.de@lists.ntp.org



On Thu, Aug 01, 2013 at 08:42:08PM +0000, David L. Mills wrote:
> Guys,
> 
> Here are some general comments on the ID.  I apologize if they are a
> little rough as I have to dictate this message instead of my usual
> polished wordsmithing.
> 
> format.
> The ID does not specify an extension field format.

The draft only seems to specify the the general idea of how it
should work and doesn't seem to go into detail how this should
work at the protocol level.  I assume he first wants to know if
the solution is something that can work and that others see
as a good solution before he works out the details.

> In particular, the private key for the message
> authentication code (MAC) , if used, must be known beforehand.  This
> is where the "cookie snatcher" issue comes up as described in
> previous messages.  In particular, an important constraint for
> normal operation is that extension fields are not used unless
> required for cookie determination.

I'm not sure I understand your remark.  As far as I know the
"cookie snatcher" you talk about is the normal "man in the middle"
attack.  As I understand we would change the verification by
replacing it by the normal X.509 verification as done for instance
SSL/TLS.

I assume you meant public key and note private key.

> Glaring vulnerability exists when using extension fields, as the key
> value is public and must be known in advance.  This invites a
> cut-and-paste attack, as well as header changes, even if
> subsequently detected by the on-wire protocol.

I've seen this cut-and-paste attack mentioned before, but I've
never really understood it.  As I understand things, the MAC
is generated over the NTP + extensions.  I don't see how you
can take the extension from 1 packet into an other and still
have a valid MAC.  Or are you saying that you do something
with a packet even when it doesn't have a valid MAC?

> certificate.
> Both RFC 5906 and the ID require the server certificate.  The server
> certificate can be conveyed, either off-line, or directly or
> indirectly via a certificate trail as described in RFC 5906 or the
> ID.  If obtained via the trail, there is a hazzard due to a
> middle-man attack and masquerade.

I really don't understand how such a man in the middle attack
can work.  Either the certificate + the chain validate, or they
don't.  It is of course important to you check that the CommonName
in the certificate matches the server you're trying to reach,
and that the root CA is in your list of trusted CAs.  But there
really isn't anything new or hard about this.

> In broadcast mode, the protocols described in RFC 5906 and the ID
> are messy and with several possible vulnerabilities.  Here is a
> suggestion that might be useful.  Define a new broadcast packet
> format as the NTP header followed by one extension field
> representing the MAC.  The extension field represents the signature
> of the header computed using the private key of the server and can
> be verified by the client using the public key in the server
> certificate.  Originally, this was considered an unacceptable
> dilution of accuracy due to the variability of the signature
> computing time and the length of the signature.  However, at least
> in modern NTP, the interleaved broadcast mode can be used to correct
> for this overhead.  This assumes that the overhead for either the
> server or client is reduced using fast modern processors.

I think that in all cases where it's no problem to keep track of
all clients you should use the interleaved mode, and we can use
something like the Diffie-Hellman key agreement for those.

But there clearly is also a need for servers that can't keep
the state of each client, and we need something else for that.
So maybe we need to define 2 ways of working.



Kurt

_______________________________________________
ntpwg mailing list
ntpwg@lists.ntp.org
http://lists.ntp.org/listinfo/ntpwg


--=_alternative 0043B5D4C1257BC0_=
Content-Type: text/html; charset="US-ASCII"

<font size=2 face="sans-serif">Hello Dave and Kurt,</font>
<br>
<br><font size=2 face="sans-serif">thanks for your comments:</font>
<br>
<br><font size=2 face="sans-serif">As Kurt already pointed out. Currently
the ID specifies only the design of the security protocol. We believe that
we need to be clear about the design questions before specifying the technical
details. </font>
<br>
<br><font size=2 face="sans-serif">The private key of the client is not
needed for the MAC. Instead the calculation of the MAC includes the cookie
and the whole NTP packet (i.e. the extension fields also). The cookie is
calculated from the server seed and the client's public key, which the
client attaches to each time request (which requires a new extension field
as pointed out in the draft). During cookie exchange the server encrypts
the cookie with the client's public key before transmission. An adversary
is therefore not able to abuse a cookie which he requests in behalf of
the client. </font>
<br>
<br><font size=2 face="sans-serif">As Kurt pointed out the certificate
exchange works as in a standard PKI. I emphasize that this is different
to autokey. In autokey the chain of trust and time are equal whereas in
this ID chain of trust and time might be disjunct.</font>
<br>
<br><font size=2 face="sans-serif">Regarding the broadcast mode. I agree
with Dave that the use of TESLA for the broadcast mode adds additional
complexity. &nbsp;It also has the disadvantage that the authenticity of
a packet can only be verified in retrospect. But I don't see security vulnerabilities.
Dave's suggestion to use a combination of interleave mode with asymmetric
cryptography is interesting. We should consider this. The argument that
asymmetric cryptography might be feasible because of modern power of CPUs
is okay for traditional computer systems but I'm not sure whether this
assumption also applies for small devices with constrained resources. </font>
<br>
<br><font size=2 face="sans-serif">Regards</font>
<br><font size=2 face="sans-serif">Dieter</font>
<br><font size=2 face="sans-serif"><br>
<br>
----------------------------------<br>
Physikalisch-Technische Bundesanstalt<br>
Dr. Dieter Sibold<br>
Q.42 - Serversysteme und Datenhaltung<br>
QM-Verantwortlicher der Stelle IT-Infrastruktur<br>
Bundesallee 100 <br>
D-38116 Braunschweig<br>
Tel: +49-531-592-84 20<br>
E-Mail: dieter.sibold@ptb.de</font>
<br>
<br>
<br>
<br><font size=1 color=#5f5f5f face="sans-serif">Von: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font size=1 face="sans-serif">Kurt Roeckx &lt;kurt@roeckx.be&gt;</font>
<br><font size=1 color=#5f5f5f face="sans-serif">An: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font size=1 face="sans-serif">&quot;David L. Mills&quot;
&lt;mills@udel.edu&gt;</font>
<br><font size=1 color=#5f5f5f face="sans-serif">Kopie: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font size=1 face="sans-serif">NTP Working Group &lt;ntpwg@lists.ntp.org&gt;,
&quot;tictoc@ietf.org&quot; &lt;tictoc@ietf.org&gt;</font>
<br><font size=1 color=#5f5f5f face="sans-serif">Datum: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font size=1 face="sans-serif">03.08.2013 20:04</font>
<br><font size=1 color=#5f5f5f face="sans-serif">Betreff: &nbsp; &nbsp;
&nbsp; &nbsp;</font><font size=1 face="sans-serif">Re: [ntpwg]
security ID submitted for review</font>
<br><font size=1 color=#5f5f5f face="sans-serif">Gesendet von: &nbsp; &nbsp;
&nbsp; &nbsp;</font><font size=1 face="sans-serif">ntpwg-bounces+dieter.sibold=ptb.de@lists.ntp.org</font>
<br>
<hr noshade>
<br>
<br>
<br><tt><font size=2>On Thu, Aug 01, 2013 at 08:42:08PM +0000, David L.
Mills wrote:<br>
&gt; Guys,<br>
&gt; <br>
&gt; Here are some general comments on the ID. &nbsp;I apologize if they
are a<br>
&gt; little rough as I have to dictate this message instead of my usual<br>
&gt; polished wordsmithing.<br>
&gt; <br>
&gt; format.<br>
&gt; The ID does not specify an extension field format.<br>
<br>
The draft only seems to specify the the general idea of how it<br>
should work and doesn't seem to go into detail how this should<br>
work at the protocol level. &nbsp;I assume he first wants to know if<br>
the solution is something that can work and that others see<br>
as a good solution before he works out the details.<br>
<br>
&gt; In particular, the private key for the message<br>
&gt; authentication code (MAC) , if used, must be known beforehand. &nbsp;This<br>
&gt; is where the &quot;cookie snatcher&quot; issue comes up as described
in<br>
&gt; previous messages. &nbsp;In particular, an important constraint for<br>
&gt; normal operation is that extension fields are not used unless<br>
&gt; required for cookie determination.<br>
<br>
I'm not sure I understand your remark. &nbsp;As far as I know the<br>
&quot;cookie snatcher&quot; you talk about is the normal &quot;man in the
middle&quot;<br>
attack. &nbsp;As I understand we would change the verification by<br>
replacing it by the normal X.509 verification as done for instance<br>
SSL/TLS.<br>
<br>
I assume you meant public key and note private key.<br>
<br>
&gt; Glaring vulnerability exists when using extension fields, as the key<br>
&gt; value is public and must be known in advance. &nbsp;This invites a<br>
&gt; cut-and-paste attack, as well as header changes, even if<br>
&gt; subsequently detected by the on-wire protocol.<br>
<br>
I've seen this cut-and-paste attack mentioned before, but I've<br>
never really understood it. &nbsp;As I understand things, the MAC<br>
is generated over the NTP + extensions. &nbsp;I don't see how you<br>
can take the extension from 1 packet into an other and still<br>
have a valid MAC. &nbsp;Or are you saying that you do something<br>
with a packet even when it doesn't have a valid MAC?<br>
<br>
&gt; certificate.<br>
&gt; Both RFC 5906 and the ID require the server certificate. &nbsp;The
server<br>
&gt; certificate can be conveyed, either off-line, or directly or<br>
&gt; indirectly via a certificate trail as described in RFC 5906 or the<br>
&gt; ID. &nbsp;If obtained via the trail, there is a hazzard due to a<br>
&gt; middle-man attack and masquerade.<br>
<br>
I really don't understand how such a man in the middle attack<br>
can work. &nbsp;Either the certificate + the chain validate, or they<br>
don't. &nbsp;It is of course important to you check that the CommonName<br>
in the certificate matches the server you're trying to reach,<br>
and that the root CA is in your list of trusted CAs. &nbsp;But there<br>
really isn't anything new or hard about this.<br>
<br>
&gt; In broadcast mode, the protocols described in RFC 5906 and the ID<br>
&gt; are messy and with several possible vulnerabilities. &nbsp;Here is
a<br>
&gt; suggestion that might be useful. &nbsp;Define a new broadcast packet<br>
&gt; format as the NTP header followed by one extension field<br>
&gt; representing the MAC. &nbsp;The extension field represents the signature<br>
&gt; of the header computed using the private key of the server and can<br>
&gt; be verified by the client using the public key in the server<br>
&gt; certificate. &nbsp;Originally, this was considered an unacceptable<br>
&gt; dilution of accuracy due to the variability of the signature<br>
&gt; computing time and the length of the signature. &nbsp;However, at
least<br>
&gt; in modern NTP, the interleaved broadcast mode can be used to correct<br>
&gt; for this overhead. &nbsp;This assumes that the overhead for either
the<br>
&gt; server or client is reduced using fast modern processors.<br>
<br>
I think that in all cases where it's no problem to keep track of<br>
all clients you should use the interleaved mode, and we can use<br>
something like the Diffie-Hellman key agreement for those.<br>
<br>
But there clearly is also a need for servers that can't keep<br>
the state of each client, and we need something else for that.<br>
So maybe we need to define 2 ways of working.<br>
<br>
<br>
<br>
Kurt<br>
<br>
_______________________________________________<br>
ntpwg mailing list<br>
ntpwg@lists.ntp.org<br>
</font></tt><a href=http://lists.ntp.org/listinfo/ntpwg><tt><font size=2>http://lists.ntp.org/listinfo/ntpwg</font></tt></a><tt><font size=2><br>
</font></tt>
<br>
--=_alternative 0043B5D4C1257BC0_=--

---------z50770_boundary_sign
Content-Type: application/x-pkcs7-signature; name="smime.p7s"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="smime.p7s"
Content-Description: S/MIME Cryptographic Signature
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---------z50770_boundary_sign--

From dieter.sibold@ptb.de  Wed Aug  7 05:45:29 2013
Return-Path: <dieter.sibold@ptb.de>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2363D21F9808 for <tictoc@ietfa.amsl.com>; Wed,  7 Aug 2013 05:45:29 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.336
X-Spam-Level: 
X-Spam-Status: No, score=-3.336 tagged_above=-999 required=5 tests=[AWL=-1.087, BAYES_00=-2.599, HELO_EQ_DE=0.35, HTML_MESSAGE=0.001]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id JC1mg2xzDb4b for <tictoc@ietfa.amsl.com>; Wed,  7 Aug 2013 05:45:24 -0700 (PDT)
Received: from mx1.bs.ptb.de (mx1.bs.ptb.de [192.53.103.106]) by ietfa.amsl.com (Postfix) with ESMTP id D7CF011E811E for <tictoc@ietf.org>; Wed,  7 Aug 2013 05:45:23 -0700 (PDT)
Received: from mx1.bs.ptb.de (localhost [127.0.0.1]) by localhost (Postfix) with SMTP id A71EECE709A; Wed,  7 Aug 2013 14:45:22 +0200 (CEST)
Received: from rose.bs.ptb.de (rose.bs.ptb.de [141.25.85.201]) by mx1.bs.ptb.de (Postfix) with ESMTP id 916E9CE7082; Wed,  7 Aug 2013 14:45:22 +0200 (CEST)
In-Reply-To: <20130807102324.GA17618@roeckx.be>
References: <51FAC820.3090401@udel.edu> <20130803174008.GA17578@roeckx.be>	<52019C7B.9070602@udel.edu> <20130807102324.GA17618@roeckx.be>
To: Kurt Roeckx <kurt@roeckx.be>
MIME-Version: 1.0
X-KeepSent: 0D16CB83:CE739C1F-C1257BC0:0045BC00; type=4; name=$KeepSent
X-Mailer: IBM Notes Release 9.0 March 08, 2013
Message-ID: <OF0D16CB83.CE739C1F-ONC1257BC0.0045BC00-C1257BC0.004610E0@ptb.de>
From: dieter.sibold@ptb.de
Date: Wed, 7 Aug 2013 14:45:18 +0200
X-MIMETrack: S/MIME Sign by Notes Client on Dieter Sibold/PTB(Release 9.0|March 08, 2013) at 07.08.2013 14:45:18, Serialize by Notes Client on Dieter Sibold/PTB(Release 9.0|March 08, 2013) at 07.08.2013 14:45:18, Serialize complete at 07.08.2013 14:45:18, Itemize by Notes Client on Dieter Sibold/PTB(Release 9.0|March 08, 2013) at 07.08.2013 14:45:20, S/MIME Sign complete at 07.08.2013 14:45:20, Serialize by Router on ROSE/PTB(Release 8.5.3FP3HF370 | April 24, 2013) at 08/07/2013 14:45:22, Serialize complete at 08/07/2013 14:45:22
Content-Type: multipart/signed; protocol="application/x-pkcs7-signature"; micalg=sha1; boundary=-------z21885_boundary_sign
Cc: NTP Working Group <ntpwg@lists.ntp.org>, "tictoc@ietf.org" <tictoc@ietf.org>, ntpwg-bounces+dieter.sibold=ptb.de@lists.ntp.org, "David L. Mills" <mills@udel.edu>
Subject: Re: [TICTOC] [ntpwg] security ID submitted for review
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 07 Aug 2013 12:45:29 -0000

This is an S/MIME signed message.

---------z21885_boundary_sign
Content-Type: multipart/alternative; boundary="=_alternative 004610DEC1257BC0_="

Dies ist eine mehrteilige Nachricht im MIME-Format.
--=_alternative 004610DEC1257BC0_=
Content-Type: text/plain; charset="US-ASCII"

Hello Kurt,

thanks for your comments. Your are right that the draft currently does not 
specifies that the server has to sign the message during cookie message 
exchange. We shall add this to the next version of the draft.

Dieter


----------------------------------
Physikalisch-Technische Bundesanstalt
Dr. Dieter Sibold
Q.42 - Serversysteme und Datenhaltung
QM-Verantwortlicher der Stelle IT-Infrastruktur
Bundesallee 100 
D-38116 Braunschweig
Tel: +49-531-592-84 20
E-Mail: dieter.sibold@ptb.de



Von:    Kurt Roeckx <kurt@roeckx.be>
An:     "David L. Mills" <mills@udel.edu>
Kopie:  NTP Working Group <ntpwg@lists.ntp.org>, "tictoc@ietf.org" 
<tictoc@ietf.org>
Datum:  07.08.2013 12:23
Betreff:        Re: [ntpwg] security ID submitted for review
Gesendet von:   ntpwg-bounces+dieter.sibold=ptb.de@lists.ntp.org



Hi Dave,

On Wed, Aug 07, 2013 at 01:01:47AM +0000, David L. Mills wrote:
> Kurt,
> 
> Thanks for your reply.  As a window of history, the Autokey protocol
> was devised during several walks in Regents Park, London, in 1996.
> I thought the most significant hazzard was the replay of client
> messages requiring extensive server cryptographic computations.  At
> that time, I did not think that a middle-man attack would be
> significant.  Both of those assumptions are probably inadvisable
> today.
> 
> One problem with Autokey is the assumption that secret keys for each
> client can be manufactured using packet header fields and a hash of
> a server private value.  A middle-man can manipulate the header
> fields compromising the client private key.  There does not seem to
> be a solution for this problem other than requiring server state.

This proposal also generates a cookie based on the something the
client sends and a secret in the server, which is then encrypted
using the client's public key.

The ID does not mention that this message is signed by the server,
but Stephen's mail said it was.  This part is important to avoid
the man in the middle attack.

So as result the client got a unique secret for him, and only
the client and the server known it.  It can also be sure that
it really comes from the server, so that there can be no man
in the middle attack.  And the server doesn't need to keep
state since it can calculate it again based on what the client
sends him.

> In the current Autokey design, the MAC computation covers the header
> and all extension fields.  However, the key used for the MAC
> computation in packets with extension fields is not private.
> Therefore, in a cut-and-paste attack, an intruder could mix and
> match extension fields and simply recalculate the MAC.

If the key is not private, by adding extentions to a packet an
attacker can just generate anything he wants?

> RFC 5906 defines the extension field including the filestamp and
> timestamp.  The filestamp was designed to reliably associate the
> cryptographic media with the current version, as it can be
> frequently changed.  This might not be useful in current designs.
> The timestamp field was designed to detect replays requiring
> expensive cryptographic computations which could take up to a
> second.  However, the "call gap" (minimum headway) provisions nicely
> deflect such attacks, so the timestamp field might not be useful in
> current designs.   Lastly, in the current design, each extension
> field is separately signed  at the time the data were constructed.
> In current designs the signature might be computed at the time of

I'm not sure I understand what you're trying to say.

I'm not sure what the filestamps are used for.  But I assume
that for instance for the leap seconds you send when it was
last updates.  I guess you can sign this once and always send
this signed in an extension, even when you're otherwise not
using any encryption.  Since you already signed it, the overhead
of adding that extension is small.

The problem I see with only sending the MAC over the ntp headers
and not over the extensions is that the extensions can be removed
or added and the client can't tell.  So it could for instance
be replaced with an old version of the extension. I think it's
important that the MAC covers everything.


Kurt

_______________________________________________
ntpwg mailing list
ntpwg@lists.ntp.org
http://lists.ntp.org/listinfo/ntpwg


--=_alternative 004610DEC1257BC0_=
Content-Type: text/html; charset="US-ASCII"

<font size=2 face="sans-serif">Hello Kurt,</font>
<br>
<br><font size=2 face="sans-serif">thanks for your comments. Your are right
that the draft currently does not specifies that the server has to sign
the message during cookie message exchange. We shall add this to the next
version of the draft.</font>
<br>
<br><font size=2 face="sans-serif">Dieter</font>
<br><font size=2 face="sans-serif"><br>
<br>
----------------------------------<br>
Physikalisch-Technische Bundesanstalt<br>
Dr. Dieter Sibold<br>
Q.42 - Serversysteme und Datenhaltung<br>
QM-Verantwortlicher der Stelle IT-Infrastruktur<br>
Bundesallee 100 <br>
D-38116 Braunschweig<br>
Tel: +49-531-592-84 20<br>
E-Mail: dieter.sibold@ptb.de</font>
<br>
<br>
<br>
<br><font size=1 color=#5f5f5f face="sans-serif">Von: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font size=1 face="sans-serif">Kurt Roeckx &lt;kurt@roeckx.be&gt;</font>
<br><font size=1 color=#5f5f5f face="sans-serif">An: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font size=1 face="sans-serif">&quot;David L. Mills&quot;
&lt;mills@udel.edu&gt;</font>
<br><font size=1 color=#5f5f5f face="sans-serif">Kopie: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font size=1 face="sans-serif">NTP Working Group &lt;ntpwg@lists.ntp.org&gt;,
&quot;tictoc@ietf.org&quot; &lt;tictoc@ietf.org&gt;</font>
<br><font size=1 color=#5f5f5f face="sans-serif">Datum: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font size=1 face="sans-serif">07.08.2013 12:23</font>
<br><font size=1 color=#5f5f5f face="sans-serif">Betreff: &nbsp; &nbsp;
&nbsp; &nbsp;</font><font size=1 face="sans-serif">Re: [ntpwg]
security ID submitted for review</font>
<br><font size=1 color=#5f5f5f face="sans-serif">Gesendet von: &nbsp; &nbsp;
&nbsp; &nbsp;</font><font size=1 face="sans-serif">ntpwg-bounces+dieter.sibold=ptb.de@lists.ntp.org</font>
<br>
<hr noshade>
<br>
<br>
<br><tt><font size=2>Hi Dave,<br>
<br>
On Wed, Aug 07, 2013 at 01:01:47AM +0000, David L. Mills wrote:<br>
&gt; Kurt,<br>
&gt; <br>
&gt; Thanks for your reply. &nbsp;As a window of history, the Autokey protocol<br>
&gt; was devised during several walks in Regents Park, London, in 1996.<br>
&gt; I thought the most significant hazzard was the replay of client<br>
&gt; messages requiring extensive server cryptographic computations. &nbsp;At<br>
&gt; that time, I did not think that a middle-man attack would be<br>
&gt; significant. &nbsp;Both of those assumptions are probably inadvisable<br>
&gt; today.<br>
&gt; <br>
&gt; One problem with Autokey is the assumption that secret keys for each<br>
&gt; client can be manufactured using packet header fields and a hash of<br>
&gt; a server private value. &nbsp;A middle-man can manipulate the header<br>
&gt; fields compromising the client private key. &nbsp;There does not seem
to<br>
&gt; be a solution for this problem other than requiring server state.<br>
<br>
This proposal also generates a cookie based on the something the<br>
client sends and a secret in the server, which is then encrypted<br>
using the client's public key.<br>
<br>
The ID does not mention that this message is signed by the server,<br>
but Stephen's mail said it was. &nbsp;This part is important to avoid<br>
the man in the middle attack.<br>
<br>
So as result the client got a unique secret for him, and only<br>
the client and the server known it. &nbsp;It can also be sure that<br>
it really comes from the server, so that there can be no man<br>
in the middle attack. &nbsp;And the server doesn't need to keep<br>
state since it can calculate it again based on what the client<br>
sends him.<br>
<br>
&gt; In the current Autokey design, the MAC computation covers the header<br>
&gt; and all extension fields. &nbsp;However, the key used for the MAC<br>
&gt; computation in packets with extension fields is not private.<br>
&gt; Therefore, in a cut-and-paste attack, an intruder could mix and<br>
&gt; match extension fields and simply recalculate the MAC.<br>
<br>
If the key is not private, by adding extentions to a packet an<br>
attacker can just generate anything he wants?<br>
<br>
&gt; RFC 5906 defines the extension field including the filestamp and<br>
&gt; timestamp. &nbsp;The filestamp was designed to reliably associate
the<br>
&gt; cryptographic media with the current version, as it can be<br>
&gt; frequently changed. &nbsp;This might not be useful in current designs.<br>
&gt; The timestamp field was designed to detect replays requiring<br>
&gt; expensive cryptographic computations which could take up to a<br>
&gt; second. &nbsp;However, the &quot;call gap&quot; (minimum headway)
provisions nicely<br>
&gt; deflect such attacks, so the timestamp field might not be useful in<br>
&gt; current designs. &nbsp; Lastly, in the current design, each extension<br>
&gt; field is separately signed &nbsp;at the time the data were constructed.<br>
&gt; In current designs the signature might be computed at the time of<br>
<br>
I'm not sure I understand what you're trying to say.<br>
<br>
I'm not sure what the filestamps are used for. &nbsp;But I assume<br>
that for instance for the leap seconds you send when it was<br>
last updates. &nbsp;I guess you can sign this once and always send<br>
this signed in an extension, even when you're otherwise not<br>
using any encryption. &nbsp;Since you already signed it, the overhead<br>
of adding that extension is small.<br>
<br>
The problem I see with only sending the MAC over the ntp headers<br>
and not over the extensions is that the extensions can be removed<br>
or added and the client can't tell. &nbsp;So it could for instance<br>
be replaced with an old version of the extension. I think it's<br>
important that the MAC covers everything.<br>
<br>
<br>
Kurt<br>
<br>
_______________________________________________<br>
ntpwg mailing list<br>
ntpwg@lists.ntp.org<br>
</font></tt><a href=http://lists.ntp.org/listinfo/ntpwg><tt><font size=2>http://lists.ntp.org/listinfo/ntpwg</font></tt></a><tt><font size=2><br>
</font></tt>
<br>
--=_alternative 004610DEC1257BC0_=--

---------z21885_boundary_sign
Content-Type: application/x-pkcs7-signature; name="smime.p7s"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="smime.p7s"
Content-Description: S/MIME Cryptographic Signature
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---------z21885_boundary_sign--

From kurt@roeckx.be  Wed Aug  7 08:16:45 2013
Return-Path: <kurt@roeckx.be>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 46E1921E8137 for <tictoc@ietfa.amsl.com>; Wed,  7 Aug 2013 08:16:45 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.265
X-Spam-Level: 
X-Spam-Status: No, score=-3.265 tagged_above=-999 required=5 tests=[AWL=-0.666, BAYES_00=-2.599]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 6Ip7X2iEIZpO for <tictoc@ietfa.amsl.com>; Wed,  7 Aug 2013 08:16:39 -0700 (PDT)
Received: from jacques.telenet-ops.be (jacques.telenet-ops.be [195.130.132.50]) by ietfa.amsl.com (Postfix) with ESMTP id 0BD8911E8143 for <tictoc@ietf.org>; Wed,  7 Aug 2013 08:16:38 -0700 (PDT)
Received: from intrepid.roeckx.be ([94.226.199.45]) by jacques.telenet-ops.be with bizsmtp id 9rGd1m00o0zFtyu0JrGdxs; Wed, 07 Aug 2013 17:16:37 +0200
Received: by intrepid.roeckx.be (Postfix, from userid 1000) id 2C9B0EB20B; Wed,  7 Aug 2013 17:16:37 +0200 (CEST)
Date: Wed, 7 Aug 2013 17:16:37 +0200
From: Kurt Roeckx <kurt@roeckx.be>
To: NTP Working Group <ntpwg@lists.ntp.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Message-ID: <20130807151637.GA31292@roeckx.be>
References: <51FAC820.3090401@udel.edu> <20130803174008.GA17578@roeckx.be>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20130803174008.GA17578@roeckx.be>
User-Agent: Mutt/1.5.21 (2010-09-15)
Subject: Re: [TICTOC] [ntpwg] security ID submitted for review
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 07 Aug 2013 15:16:45 -0000

On Sat, Aug 03, 2013 at 07:40:08PM +0200, Kurt Roeckx wrote:
> 
> I really don't understand how such a man in the middle attack
> can work.  Either the certificate + the chain validate, or they
> don't.  It is of course important to you check that the CommonName
> in the certificate matches the server you're trying to reach,
> and that the root CA is in your list of trusted CAs.  But there
> really isn't anything new or hard about this.

So thinking about this so more, I do see 1 problem with this.  I
would like to be able to use this for the pool for those that wish
to use it.  But if you're verify the certificate you're not going
to know which server you're talking too, so there is no way to
check the CommonName.  Does it make sense to just skip this
check in case of the pool?


Kurt


From mills@udel.edu  Tue Aug 13 08:10:14 2013
Return-Path: <mills@udel.edu>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 861B021E808C for <tictoc@ietfa.amsl.com>; Tue, 13 Aug 2013 08:10:14 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.998
X-Spam-Level: 
X-Spam-Status: No, score=-3.998 tagged_above=-999 required=5 tests=[BAYES_50=0.001, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id dbRvvAs6Reo3 for <tictoc@ietfa.amsl.com>; Tue, 13 Aug 2013 08:10:09 -0700 (PDT)
Received: from mail.eecis.udel.edu (louie.udel.edu [128.4.40.12]) by ietfa.amsl.com (Postfix) with ESMTP id E7FC521F8DDD for <tictoc@ietf.org>; Tue, 13 Aug 2013 08:10:07 -0700 (PDT)
Received: from [192.168.1.2] (pool-74-109-13-78.phlapa.fios.verizon.net [74.109.13.78]) (Authenticated sender: mills@mail.eecis.udel.edu) by mail.eecis.udel.edu (Postfix) with ESMTP id 5186B1BDA; Tue, 13 Aug 2013 11:10:06 -0400 (EDT)
Message-ID: <520A4C4F.2050400@udel.edu>
Date: Tue, 13 Aug 2013 15:10:07 +0000
From: "David L. Mills" <mills@udel.edu>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax)
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: dieter.sibold@ptb.de
References: <51FAC820.3090401@udel.edu> <20130803174008.GA17578@roeckx.be> <OFCF046B68.1B709B01-ONC1257BC0.002ECD08-C1257BC0.0043B5D5@ptb.de>
In-Reply-To: <OFCF046B68.1B709B01-ONC1257BC0.002ECD08-C1257BC0.0043B5D5@ptb.de>
Content-Type: multipart/alternative; boundary="------------080405070303000203070401"
Cc: NTP Working Group <ntpwg@lists.ntp.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [ntpwg] security ID submitted for review
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 13 Aug 2013 15:10:14 -0000

This is a multi-part message in MIME format.
--------------080405070303000203070401
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

dieter.sibold@ptb.de wrote:

> Dieter,

There are several minor points I might mention.  First the easy ones.  
You mentioned TESLA , which sound very much like the existing Autokey in 
broadcast mode.  Actually, the Autokey scheme was stolen from S-KEY of 
many years ago.   I still think this should all be replaced by a 
signature using the server public key. This is complicated in the case 
of multiple broadcast servers implemented in the same system.  In these 
cases,  the network delay calibration might be different, as is the case 
in the current NTP implementation. In order to support this requirement, 
the extension field for the broadcast message needs to include the 
process ID for the broadcast association.  The process ID is used for 
the particular security protocol in the delay calculation.  Other means 
might be devised in the new protocol to perform this function.

For client/server and symmetric modes, I suggest the following design 
candidate.  This would be appropriate for national time servers operated 
by NIST , USNO and NRC.  A certain number of these servers currently 
support symmetric key verification, where each key is chosen for each 
subscriber upon payment of a fee.  For instance, NIST has over 25 
million clients of their public servers, and where probably several 
million use the bolder complex with no cryptographic support.  One 
machine includes support for symmetric key cryptography, so probably 
several hundred different keys could be supported. 

Symmetric key cryptography becomes unwieldy when the number of customer 
clients much exceeds this number.  An alternative design using public 
key cryptography might be more appropriate.  This design would use 
Diffy-Hellman key agreement.  The client would send the public value to 
the server signed by the client public key.  The server would respond 
with its public value signed by the server public key.  The agreed key 
for both the client and server would be split in two portions, the 
32-bit key identifier and a 160-bit symmetric key.  Both the client and 
the server would save the values in the same data structure used for 
static symmetric keys.  Subsequently, operation would continue as in 
ordinary static symmetric keys. 

The devil is in the details.  The server needs a table with entries 
corresponding to the client IP address and certificate. This is what the 
subscriber pays for.  Additional provisions are necessary to time out 
old keys and regenerate accordingly. 

This scheme avoids the use of extension fields in all cases except when 
a new agreement is negotiated.  A question remains on whether the 
sgreement messages are to be authenticated by other than the agreed key 
itself.  My suggestion is simply to avoid this step on condition that 
the NTP header time values are ignored. 

My experience over 17 years operating experiment configurations on up to 
17 different machines running various combinations of unsecured 
symmetric and public key combinations suggests that the supporting 
infrastructure is of serious concern.  For one thing, certificate 
lifetimes are meaningless until the correct time has been verified.  
There is no a-priori assumption about the certificate lifetimes other 
than a continuous trail, possibly spanning several intermediate 
servers,   Consider the effect of replacing a public/private key when a 
server fails and is replaced.  Consider what happens when an 
intermediate server rolls a new certificate which interupts  the trail 
continuity.  Remember, a typical NTP client runs multiple servers, each 
of which might have a different trail to the same server or servers.  A 
large crop of obscure interworking issues have been found and corrented 
over the years.  I suspect similar issues and corrections will be found 
in PTP applications.

The particular set of rules and constraints described in previous 
reports and RFC 5906 were constructed with great care.  They established 
the security model and partial-order precidents relations considered at 
a meeting of computer science theorists and myself at the Dagstuhl 
(Germany) meeting in 1996.   The rules require that all security changes 
conform to Lamport's "happens before" relation.  The filestamps and 
timestamps described in RFC 5906 conform to these rules.  Unfortunately, 
to my knowledge, current applications do not necessarily conform to 
these rules.  It may be a lost cause to incorporate these rules in new 
NTP or PTP protocols and applications.

I hope these comments are useful in the evolution of NTP and PTP 
security protocols.

Dave


> Hello Dave and Kurt,
>
> thanks for your comments:
>
> As Kurt already pointed out. Currently the ID specifies only the 
> design of the security protocol. We believe that we need to be clear 
> about the design questions before specifying the technical details.
>
> The private key of the client is not needed for the MAC. Instead the 
> calculation of the MAC includes the cookie and the whole NTP packet 
> (i.e. the extension fields also). The cookie is calculated from the 
> server seed and the client's public key, which the client attaches to 
> each time request (which requires a new extension field as pointed out 
> in the draft). During cookie exchange the server encrypts the cookie 
> with the client's public key before transmission. An adversary is 
> therefore not able to abuse a cookie which he requests in behalf of 
> the client.
>
> As Kurt pointed out the certificate exchange works as in a standard 
> PKI. I emphasize that this is different to autokey. In autokey the 
> chain of trust and time are equal whereas in this ID chain of trust 
> and time might be disjunct.
>
> Regarding the broadcast mode. I agree with Dave that the use of TESLA 
> for the broadcast mode adds additional complexity.  It also has the 
> disadvantage that the authenticity of a packet can only be verified in 
> retrospect. But I don't see security vulnerabilities. Dave's 
> suggestion to use a combination of interleave mode with asymmetric 
> cryptography is interesting. We should consider this. The argument 
> that asymmetric cryptography might be feasible because of modern power 
> of CPUs is okay for traditional computer systems but I'm not sure 
> whether this assumption also applies for small devices with 
> constrained resources.
>
> Regards
> Dieter
>
>
> ----------------------------------
> Physikalisch-Technische Bundesanstalt
> Dr. Dieter Sibold
> Q.42 - Serversysteme und Datenhaltung
> QM-Verantwortlicher der Stelle IT-Infrastruktur
> Bundesallee 100
> D-38116 Braunschweig
> Tel: +49-531-592-84 20
> E-Mail: dieter.sibold@ptb.de
>
>
>
> Von:        Kurt Roeckx <kurt@roeckx.be>
> An:        "David L. Mills" <mills@udel.edu>
> Kopie:        NTP Working Group <ntpwg@lists.ntp.org>, 
> "tictoc@ietf.org" <tictoc@ietf.org>
> Datum:        03.08.2013 20:04
> Betreff:        Re: [ntpwg] security ID submitted for review
> Gesendet von:        ntpwg-bounces+dieter.sibold=ptb.de@lists.ntp.org
> ------------------------------------------------------------------------
>
>
>
> On Thu, Aug 01, 2013 at 08:42:08PM +0000, David L. Mills wrote:
> > Guys,
> >
> > Here are some general comments on the ID.  I apologize if they are a
> > little rough as I have to dictate this message instead of my usual
> > polished wordsmithing.
> >
> > format.
> > The ID does not specify an extension field format.
>
> The draft only seems to specify the the general idea of how it
> should work and doesn't seem to go into detail how this should
> work at the protocol level.  I assume he first wants to know if
> the solution is something that can work and that others see
> as a good solution before he works out the details.
>
> > In particular, the private key for the message
> > authentication code (MAC) , if used, must be known beforehand.  This
> > is where the "cookie snatcher" issue comes up as described in
> > previous messages.  In particular, an important constraint for
> > normal operation is that extension fields are not used unless
> > required for cookie determination.
>
> I'm not sure I understand your remark.  As far as I know the
> "cookie snatcher" you talk about is the normal "man in the middle"
> attack.  As I understand we would change the verification by
> replacing it by the normal X.509 verification as done for instance
> SSL/TLS.
>
> I assume you meant public key and note private key.
>
> > Glaring vulnerability exists when using extension fields, as the key
> > value is public and must be known in advance.  This invites a
> > cut-and-paste attack, as well as header changes, even if
> > subsequently detected by the on-wire protocol.
>
> I've seen this cut-and-paste attack mentioned before, but I've
> never really understood it.  As I understand things, the MAC
> is generated over the NTP + extensions.  I don't see how you
> can take the extension from 1 packet into an other and still
> have a valid MAC.  Or are you saying that you do something
> with a packet even when it doesn't have a valid MAC?
>
> > certificate.
> > Both RFC 5906 and the ID require the server certificate.  The server
> > certificate can be conveyed, either off-line, or directly or
> > indirectly via a certificate trail as described in RFC 5906 or the
> > ID.  If obtained via the trail, there is a hazzard due to a
> > middle-man attack and masquerade.
>
> I really don't understand how such a man in the middle attack
> can work.  Either the certificate + the chain validate, or they
> don't.  It is of course important to you check that the CommonName
> in the certificate matches the server you're trying to reach,
> and that the root CA is in your list of trusted CAs.  But there
> really isn't anything new or hard about this.
>
> > In broadcast mode, the protocols described in RFC 5906 and the ID
> > are messy and with several possible vulnerabilities.  Here is a
> > suggestion that might be useful.  Define a new broadcast packet
> > format as the NTP header followed by one extension field
> > representing the MAC.  The extension field represents the signature
> > of the header computed using the private key of the server and can
> > be verified by the client using the public key in the server
> > certificate.  Originally, this was considered an unacceptable
> > dilution of accuracy due to the variability of the signature
> > computing time and the length of the signature.  However, at least
> > in modern NTP, the interleaved broadcast mode can be used to correct
> > for this overhead.  This assumes that the overhead for either the
> > server or client is reduced using fast modern processors.
>
> I think that in all cases where it's no problem to keep track of
> all clients you should use the interleaved mode, and we can use
> something like the Diffie-Hellman key agreement for those.
>
> But there clearly is also a need for servers that can't keep
> the state of each client, and we need something else for that.
> So maybe we need to define 2 ways of working.
>
>
>
> Kurt
>
> _______________________________________________
> ntpwg mailing list
> ntpwg@lists.ntp.org
> http://lists.ntp.org/listinfo/ntpwg
>


--------------080405070303000203070401
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
  <title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
<a class="moz-txt-link-abbreviated" href="mailto:dieter.sibold@ptb.de">dieter.sibold@ptb.de</a> wrote:
<blockquote
 cite="midOFCF046B68.1B709B01-ONC1257BC0.002ECD08-C1257BC0.0043B5D5@ptb.de"
 type="cite"><font face="sans-serif" size="2">Dieter,</font><br>
</blockquote>
There are several minor points I might mention.&nbsp; First the easy ones.&nbsp;
You mentioned TESLA , which sound very much like the existing Autokey
in broadcast mode.&nbsp; Actually, the Autokey scheme was stolen from S-KEY
of many years ago.&nbsp;&nbsp; I still think this should all be replaced by a
signature using the server public key. This is complicated in the case
of multiple broadcast servers implemented in the same system.&nbsp; In these
cases,&nbsp; the network delay calibration might be different, as is the
case in the current NTP implementation. In order to support this
requirement, the extension field for the broadcast message needs to
include the process ID for the broadcast association.&nbsp; The process ID
is used for the particular security protocol in the delay calculation.&nbsp;
Other means might be devised in the new protocol to perform this
function.<br>
<br>
For client/server and symmetric modes, I suggest the following design
candidate.&nbsp; This would be appropriate for national time servers
operated by NIST , USNO and NRC.&nbsp; A certain number of these servers
currently support symmetric key verification, where each key is chosen
for each subscriber upon payment of a fee.&nbsp; For instance, NIST has over
25 million clients of their public servers, and where probably several
million use the bolder complex with no cryptographic support.&nbsp; One
machine includes support for symmetric key cryptography, so probably
several hundred different keys could be supported.&nbsp; <br>
<br>
Symmetric key cryptography becomes unwieldy when the number of customer
clients much exceeds this number.&nbsp; An alternative design using public
key cryptography might be more appropriate.&nbsp; This design would use
Diffy-Hellman key agreement.&nbsp; The client would send the public value to
the server signed by the client public key.&nbsp; The server would respond
with its public value signed by the server public key.&nbsp; The agreed key
for both the client and server would be split in two portions, the
32-bit key identifier and a 160-bit symmetric key.&nbsp; Both the client and
the server would save the values in the same data structure used for
static symmetric keys.&nbsp; Subsequently, operation would continue as in
ordinary static symmetric keys.&nbsp; <br>
<br>
The devil is in the details.&nbsp; The server needs a table with entries
corresponding to the client IP address and certificate. This is what
the subscriber pays for.&nbsp; Additional provisions are necessary to time
out old keys and regenerate accordingly.&nbsp; <br>
<br>
This scheme avoids the use of extension fields in all cases except when
a new agreement is negotiated.&nbsp; A question remains on whether the
sgreement messages are to be authenticated by other than the agreed key
itself.&nbsp; My suggestion is simply to avoid this step on condition that
the NTP header time values are ignored.&nbsp; <br>
<br>
My experience over 17 years operating experiment configurations on up
to 17 different machines running various combinations of unsecured
symmetric and public key combinations suggests that the supporting
infrastructure is of serious concern.&nbsp; For one thing, certificate
lifetimes are meaningless until the correct time has been verified.&nbsp;
There is no a-priori assumption about the certificate lifetimes other
than a continuous trail, possibly spanning several intermediate
servers,&nbsp;&nbsp; Consider the effect of replacing a public/private key when a
server fails and is replaced.&nbsp; Consider what happens when an
intermediate server rolls a new certificate which interupts&nbsp; the trail
continuity.&nbsp; Remember, a typical NTP client runs multiple servers, each
of which might have a different trail to the same server or servers.&nbsp; A
large crop of obscure interworking issues have been found and corrented
over the years.&nbsp; I suspect similar issues and corrections will be found
in PTP applications. <br>
<br>
The particular set of rules and constraints described in previous
reports and RFC 5906 were constructed with great care.&nbsp; They
established the security model and partial-order precidents relations
considered at a meeting of computer science theorists and myself at the
Dagstuhl (Germany) meeting in 1996.&nbsp;&nbsp; The rules require that all
security changes conform to Lamport's "happens before" relation.&nbsp; The
filestamps and timestamps described in RFC 5906 conform to these
rules.&nbsp; Unfortunately, to my knowledge, current applications do not
necessarily conform to these rules.&nbsp; It may be a lost cause to
incorporate these rules in new NTP or PTP protocols and applications.<br>
<br>
I hope these comments are useful in the evolution of NTP and PTP
security protocols.<br>
<br>
Dave<br>
<br>
<br>
<blockquote
 cite="midOFCF046B68.1B709B01-ONC1257BC0.002ECD08-C1257BC0.0043B5D5@ptb.de"
 type="cite"><font face="sans-serif" size="2">Hello Dave and Kurt,</font>
  <br>
  <br>
  <font face="sans-serif" size="2">thanks for your comments:</font>
  <br>
  <br>
  <font face="sans-serif" size="2">As Kurt already pointed out.
Currently
the ID specifies only the design of the security protocol. We believe
that
we need to be clear about the design questions before specifying the
technical
details. </font>
  <br>
  <br>
  <font face="sans-serif" size="2">The private key of the client is not
needed for the MAC. Instead the calculation of the MAC includes the
cookie
and the whole NTP packet (i.e. the extension fields also). The cookie
is
calculated from the server seed and the client's public key, which the
client attaches to each time request (which requires a new extension
field
as pointed out in the draft). During cookie exchange the server
encrypts
the cookie with the client's public key before transmission. An
adversary
is therefore not able to abuse a cookie which he requests in behalf of
the client. </font>
  <br>
  <br>
  <font face="sans-serif" size="2">As Kurt pointed out the certificate
exchange works as in a standard PKI. I emphasize that this is different
to autokey. In autokey the chain of trust and time are equal whereas in
this ID chain of trust and time might be disjunct.</font>
  <br>
  <br>
  <font face="sans-serif" size="2">Regarding the broadcast mode. I
agree
with Dave that the use of TESLA for the broadcast mode adds additional
complexity. &nbsp;It also has the disadvantage that the authenticity of
a packet can only be verified in retrospect. But I don't see security
vulnerabilities.
Dave's suggestion to use a combination of interleave mode with
asymmetric
cryptography is interesting. We should consider this. The argument that
asymmetric cryptography might be feasible because of modern power of
CPUs
is okay for traditional computer systems but I'm not sure whether this
assumption also applies for small devices with constrained resources. </font>
  <br>
  <br>
  <font face="sans-serif" size="2">Regards</font>
  <br>
  <font face="sans-serif" size="2">Dieter</font>
  <br>
  <font face="sans-serif" size="2"><br>
  <br>
----------------------------------<br>
Physikalisch-Technische Bundesanstalt<br>
Dr. Dieter Sibold<br>
Q.42 - Serversysteme und Datenhaltung<br>
QM-Verantwortlicher der Stelle IT-Infrastruktur<br>
Bundesallee 100 <br>
D-38116 Braunschweig<br>
Tel: +49-531-592-84 20<br>
E-Mail: <a class="moz-txt-link-abbreviated" href="mailto:dieter.sibold@ptb.de">dieter.sibold@ptb.de</a></font>
  <br>
  <br>
  <br>
  <br>
  <font color="#5f5f5f" face="sans-serif" size="1">Von: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font face="sans-serif" size="1">Kurt Roeckx
<a class="moz-txt-link-rfc2396E" href="mailto:kurt@roeckx.be">&lt;kurt@roeckx.be&gt;</a></font>
  <br>
  <font color="#5f5f5f" face="sans-serif" size="1">An: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font face="sans-serif" size="1">"David L. Mills"
<a class="moz-txt-link-rfc2396E" href="mailto:mills@udel.edu">&lt;mills@udel.edu&gt;</a></font>
  <br>
  <font color="#5f5f5f" face="sans-serif" size="1">Kopie: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font face="sans-serif" size="1">NTP Working Group
<a class="moz-txt-link-rfc2396E" href="mailto:ntpwg@lists.ntp.org">&lt;ntpwg@lists.ntp.org&gt;</a>,
<a class="moz-txt-link-rfc2396E" href="mailto:tictoc@ietf.org">"tictoc@ietf.org"</a> <a class="moz-txt-link-rfc2396E" href="mailto:tictoc@ietf.org">&lt;tictoc@ietf.org&gt;</a></font>
  <br>
  <font color="#5f5f5f" face="sans-serif" size="1">Datum: &nbsp; &nbsp; &nbsp;
&nbsp;</font><font face="sans-serif" size="1">03.08.2013 20:04</font>
  <br>
  <font color="#5f5f5f" face="sans-serif" size="1">Betreff: &nbsp; &nbsp;
&nbsp; &nbsp;</font><font face="sans-serif" size="1">Re: [ntpwg]
security ID submitted for review</font>
  <br>
  <font color="#5f5f5f" face="sans-serif" size="1">Gesendet von: &nbsp; &nbsp;
&nbsp; &nbsp;</font><font face="sans-serif" size="1"><a class="moz-txt-link-abbreviated" href="mailto:ntpwg-bounces+dieter.sibold=ptb.de@lists.ntp.org">ntpwg-bounces+dieter.sibold=ptb.de@lists.ntp.org</a></font>
  <br>
  <hr noshade="noshade">
  <br>
  <br>
  <br>
  <tt><font size="2">On Thu, Aug 01, 2013 at 08:42:08PM +0000, David L.
Mills wrote:<br>
&gt; Guys,<br>
&gt; <br>
&gt; Here are some general comments on the ID. &nbsp;I apologize if they
are a<br>
&gt; little rough as I have to dictate this message instead of my usual<br>
&gt; polished wordsmithing.<br>
&gt; <br>
&gt; format.<br>
&gt; The ID does not specify an extension field format.<br>
  <br>
The draft only seems to specify the the general idea of how it<br>
should work and doesn't seem to go into detail how this should<br>
work at the protocol level. &nbsp;I assume he first wants to know if<br>
the solution is something that can work and that others see<br>
as a good solution before he works out the details.<br>
  <br>
&gt; In particular, the private key for the message<br>
&gt; authentication code (MAC) , if used, must be known beforehand.
&nbsp;This<br>
&gt; is where the "cookie snatcher" issue comes up as described
in<br>
&gt; previous messages. &nbsp;In particular, an important constraint for<br>
&gt; normal operation is that extension fields are not used unless<br>
&gt; required for cookie determination.<br>
  <br>
I'm not sure I understand your remark. &nbsp;As far as I know the<br>
"cookie snatcher" you talk about is the normal "man in the
middle"<br>
attack. &nbsp;As I understand we would change the verification by<br>
replacing it by the normal X.509 verification as done for instance<br>
SSL/TLS.<br>
  <br>
I assume you meant public key and note private key.<br>
  <br>
&gt; Glaring vulnerability exists when using extension fields, as the
key<br>
&gt; value is public and must be known in advance. &nbsp;This invites a<br>
&gt; cut-and-paste attack, as well as header changes, even if<br>
&gt; subsequently detected by the on-wire protocol.<br>
  <br>
I've seen this cut-and-paste attack mentioned before, but I've<br>
never really understood it. &nbsp;As I understand things, the MAC<br>
is generated over the NTP + extensions. &nbsp;I don't see how you<br>
can take the extension from 1 packet into an other and still<br>
have a valid MAC. &nbsp;Or are you saying that you do something<br>
with a packet even when it doesn't have a valid MAC?<br>
  <br>
&gt; certificate.<br>
&gt; Both RFC 5906 and the ID require the server certificate. &nbsp;The
server<br>
&gt; certificate can be conveyed, either off-line, or directly or<br>
&gt; indirectly via a certificate trail as described in RFC 5906 or the<br>
&gt; ID. &nbsp;If obtained via the trail, there is a hazzard due to a<br>
&gt; middle-man attack and masquerade.<br>
  <br>
I really don't understand how such a man in the middle attack<br>
can work. &nbsp;Either the certificate + the chain validate, or they<br>
don't. &nbsp;It is of course important to you check that the CommonName<br>
in the certificate matches the server you're trying to reach,<br>
and that the root CA is in your list of trusted CAs. &nbsp;But there<br>
really isn't anything new or hard about this.<br>
  <br>
&gt; In broadcast mode, the protocols described in RFC 5906 and the ID<br>
&gt; are messy and with several possible vulnerabilities. &nbsp;Here is
a<br>
&gt; suggestion that might be useful. &nbsp;Define a new broadcast packet<br>
&gt; format as the NTP header followed by one extension field<br>
&gt; representing the MAC. &nbsp;The extension field represents the signature<br>
&gt; of the header computed using the private key of the server and can<br>
&gt; be verified by the client using the public key in the server<br>
&gt; certificate. &nbsp;Originally, this was considered an unacceptable<br>
&gt; dilution of accuracy due to the variability of the signature<br>
&gt; computing time and the length of the signature. &nbsp;However, at
least<br>
&gt; in modern NTP, the interleaved broadcast mode can be used to
correct<br>
&gt; for this overhead. &nbsp;This assumes that the overhead for either
the<br>
&gt; server or client is reduced using fast modern processors.<br>
  <br>
I think that in all cases where it's no problem to keep track of<br>
all clients you should use the interleaved mode, and we can use<br>
something like the Diffie-Hellman key agreement for those.<br>
  <br>
But there clearly is also a need for servers that can't keep<br>
the state of each client, and we need something else for that.<br>
So maybe we need to define 2 ways of working.<br>
  <br>
  <br>
  <br>
Kurt<br>
  <br>
_______________________________________________<br>
ntpwg mailing list<br>
<a class="moz-txt-link-abbreviated" href="mailto:ntpwg@lists.ntp.org">ntpwg@lists.ntp.org</a><br>
  </font></tt><a href="http://lists.ntp.org/listinfo/ntpwg"><tt><font
 size="2">http://lists.ntp.org/listinfo/ntpwg</font></tt></a><tt><font
 size="2"><br>
  </font></tt>
  <br>
</blockquote>
<br>
</body>
</html>

--------------080405070303000203070401--

From stbryant@cisco.com  Tue Aug 20 09:28:05 2013
Return-Path: <stbryant@cisco.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 514F511E812C; Tue, 20 Aug 2013 09:28:05 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -110.599
X-Spam-Level: 
X-Spam-Status: No, score=-110.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 27c4TH9UkWxI; Tue, 20 Aug 2013 09:28:00 -0700 (PDT)
Received: from ams-iport-3.cisco.com (ams-iport-3.cisco.com [144.254.224.146]) by ietfa.amsl.com (Postfix) with ESMTP id B487B11E80E4; Tue, 20 Aug 2013 09:27:58 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=79744; q=dns/txt; s=iport; t=1377016079; x=1378225679; h=message-id:date:from:reply-to:mime-version:to:cc:subject: references:in-reply-to:content-transfer-encoding; bh=pBPUw7dQhvNfFR+bza+wfJUV3NQHWdfTkqAGRvyFqpw=; b=hiH+IIeFUN27rrKLhJ7MueYijEXZ9rS0dIBzcvls1yKo5AzdvmlcWEcw tmO8/b08jELPJfGCz9Xe/oxtA4iGoZBzSurF/sgc/hH9hxLjiHsMbn4bI d314wy20J3igH8yA/NadEWHLD4e2kZIZLkFJ0gMaw/j4vC6uzxUdjNcPf A=;
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AhIFADOYE1KQ/khM/2dsb2JhbABQCoMFNcAqgSQWdIIkAQEBAwEBAQEXAR0uAQcKAQwECxEBAwEBAQkMCggHCQMCAQIBFR8DBggGCgMBBQIBAReHbwYMrS6PAQIJBgUFgRciBwYEhAoDjESHCEGDWIEthXGKOIMdgWgIFw
X-IronPort-AV: E=Sophos;i="4.89,921,1367971200"; d="scan'208";a="16852392"
Received: from ams-core-3.cisco.com ([144.254.72.76]) by ams-iport-3.cisco.com with ESMTP; 20 Aug 2013 16:27:54 +0000
Received: from cisco.com (mrwint.cisco.com [64.103.70.36]) by ams-core-3.cisco.com (8.14.5/8.14.5) with ESMTP id r7KGRqbB016277 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 20 Aug 2013 16:27:52 GMT
Received: from [IPv6:::1] (localhost [127.0.0.1]) by cisco.com (8.14.4+Sun/8.8.8) with ESMTP id r7KGRlAL026894; Tue, 20 Aug 2013 17:27:47 +0100 (BST)
Message-ID: <52139903.8040801@cisco.com>
Date: Tue, 20 Aug 2013 17:27:47 +0100
From: Stewart Bryant <stbryant@cisco.com>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:17.0) Gecko/20130620 Thunderbird/17.0.7
MIME-Version: 1.0
To: "S. Davari" <davarish@yahoo.com>
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com> <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com> <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com>
In-Reply-To: <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
Reply-To: stbryant@cisco.com
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 20 Aug 2013 16:28:05 -0000

On 06/08/2013 14:35, S. Davari wrote:
> Sasha,
>
> If a router is not 1588 aware we want it to switch the packet with minimum jitter and delay via high priority queue.
Yes, but you can explicitly tunnel across such island of incompatibility

Stewart

>   
>
> Sending it to CPU creates large jitter and delay since the CPUs are generally busy doing other things. Time stamping or updating CF in the CPU won't help since it does not cover the variable delay that is incurred from the time the packet enters the router to the time the packet gets processed by CPU.
>
> I wish it was that easy!
>
> Regards,
> Shahram
>
>
> On Aug 6, 2013, at 12:38 AM, Alexander Vainshtein <Alexander.Vainshtein@ecitele.com> wrote:
>
>> Shahram,
>> I agree with you that the routers that do not recognize a Timing Alert label would send the packet to the CPU. (This would be easy to arrange since we are speaking about a reserved label.)
>>
>> I do not, however, see this as a serious issue, because the SW running on this CPU would (hopefully) be upgradable to handle the packet correctly i.e., to forward it to where it should be forwarded and, if it is forwarded as a labeled packet, to prepend the Timing Alert Label on top of its label stack. It could even record the residence time (as observed by the CPU in the proper place in the packet. This would mean that introducing additional error to whatever timing information is associated with this packet - but this is what you should anyway expect if there are non-compliant routers on your path, right?
>>
>> Regards,
>>      Sasha
>>
>>> -----Original Message-----
>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf Of
>>> Shahram Davari
>>> Sent: Monday, August 05, 2013 11:29 PM
>>> To: stbryant@cisco.com; S. Davari
>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>>
>>> Hi Stewart,
>>>
>>> Your suggestion of " I suggested an LSP type that has the properties "timestamp
>>> and pass to application", is a subset of the existing draft and should work. It
>>> basically limits the time stamping and correction field update to LERs (while the
>>> draft supports time stamping at LER and LSR).
>>>
>>> However Sasha's suggestion of using a reserved label (RAL, GAL or any other
>>> reserved label) does not satisfy one of the major requirements, which is
>>> backward compatibility. Routers that don't understand this reserved label will
>>> drop or copy to CPU such packets.
>>>
>>> Regards,
>>> Shahram
>>>
>>> -----Original Message-----
>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf Of
>>> Stewart Bryant
>>> Sent: Monday, August 05, 2013 2:29 AM
>>> To: S. Davari
>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>>
>>> My concern is that we are creating a new LSP type in MPLS
>>> (which is a architectural change and thus needs to be very carefully
>>> considered) without asking the question "how can I do this
>>> in the most general way, to maximize flexibility/reuse"
>>>
>>> I suggested an LSP type that has the properties "timestamp
>>> and pass to application", but I think that Sasha raises a good
>>> point about using router alert. However RA has no implicit
>>> timestamp, so maybe we need a new type of RA that has the
>>> properties "timestamp, and pass top application indicated by
>>> the next label". That would be quite useful in a number
>>> of OAM applications. There is possibly some GAL variant
>>> of that design that should also be considered.
>>>
>>> The application could worry about the path and where it
>>> was necessary to skip some hops a hierarchical LSP would
>>> accomplish that, with the specific benefit that the application
>>> would consciously do this, and may be able to apply some
>>> form of compensation within the network.
>>>
>>> - Stewart
>>>
>>> On 04/08/2013 10:40, S. Davari wrote:
>>>> Hi Sasha
>>>>
>>>> Perhaps you have not understood the draft well. The main reason that the
>>> draft chose to use a dedicated LSP that is signaled via RSVPTE indicating it is a
>>> timing LSP is to be backward compatible with non-1588 nodes. Such nodes
>>> simply just switch the packet.
>>>> Your proposal had been considered and was rejected since it was not
>>> backward compatible. Nodes receiving alert label or TTL = 1 send the packet to
>>> CPU. You can refer to the meeting notes.
>>>> Regards,
>>>> Shahram
>>>>
>>>>
>>>> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein
>>> <Alexander.Vainshtein@ecitele.com> wrote:
>>>>> Stewart and all,
>>>>> I concur with Stewart's statement that the draft does not define "full
>>> interaction with MPLS architecture".
>>>>> E.g., one of the objectives of the draft is to provide a technique that would
>>> be backward-compatible with old LSRs that cannot provide on-path support for
>>> timing distribution, while the other objective is to make every LSR on the path
>>> aware that some MPLS packets are carrying timing-related messages and hence
>>> require on-path support.
>>>>> IMHO and FWIW, the MPLS data plane architecture allows just two methods
>>> for making a transit LSR to provide special processing to a labeled packet:
>>>>> - It would carry some kind of an "alert label" on top of the label stack and,
>>> specifically, on top of any labels used for actual forwarding
>>>>> OR,
>>>>> - The TTL in the top label stack entry has been set to 1.
>>>>>
>>>>>
>>>>> The draft does not follow any of these approaches.
>>>>>
>>>>> I must also admit that Section 12 "OAM, Control and Management" of the
>>> draft looks somewhat in
>>>>> E.g., I could not understand from the text whether BFD, LSP-Ping etc. OAM
>>> messages would be subjected to on-path support procedures for timing
>>> messages or not.
>>>>> My 2c,
>>>>>      Sasha
>>>>>
>>>>>> -----Original Message-----
>>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf
>>> Of
>>>>>> Stewart Bryant
>>>>>> Sent: Friday, August 02, 2013 12:01 PM
>>>>>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
>>>>>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.org;
>>> draft-
>>>>>> ietf-tictoc-1588overmpls@tools.ietf.org
>>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
>>>>>>
>>>>>> SB> This draft does not seem to provide a precise definition
>>>>>> SB> the properties of the new LSP type that it wishes to
>>>>>> SB> define, in particular it does it define the PHB of those
>>>>>> SB> LSPs, nor the full interaction with the MPLS
>>>>>> SB> architecture.
>>>>>> SB>
>>>>>> SB> I have not tracked TICTOC for a while but I thought that
>>>>>> SB> the original plan was to define the concept of an offset
>>>>>> SB> into a packet to do the correction.
>>>>>> SB>
>>>>>> SB> It is disappointing that the opportunity was not taken
>>>>>> SB> to define a timing shim inside the timing LSP so that
>>>>>> SB> a time correction could be added to any packet such that
>>>>>> SB> the MPLS system was isolated from the details of the
>>>>>> SB> complexity of the particular time transfer type.
>>>>>>
>>>>>> SB> I think that much more clarify is needed in terms of
>>>>>> SB> definition of the new LSP type, since it is unclear
>>>>>> SB> from this text how to implement one.
>>>>>> SB>
>>>>>> SB> There are a lot of other MPLS services such as
>>>>>> SB> LSP ping that need to be considered.
>>>>>> SB>
>>>>>> SB> Please see inline for more comments. However these
>>>>>> SB> comments are made in the context of the text as written
>>>>>> SB> whilst I have a fundamental concern that this approach
>>>>>> SB> lacks an MPLS architectural soundness that need
>>>>>> SB> greater thought with significant impact on the
>>>>>> SB> draft.
>>>>>>
>>>>>> - Stewart
>>>>>>
>>>>>>
>>>>>> TICTOC Working Group                                           S. Davari
>>>>>> Internet-Draft                                                   A. Oren
>>>>>> Intended status: Standards Track                          Broadcom Corp.
>>>>>> Expires: December 17, 2013                                     M. Bhatia
>>>>>>                                                                P. Roberts
>>>>>>                                                            Alcatel-Lucent
>>>>>>                                                                L. Montini
>>>>>>                                                                L. Martini
>>>>>>                                                             Cisco Systems
>>>>>>                                                             June 15, 2013
>>>>>>
>>>>>>
>>>>>>              Transporting Timing messages over MPLS Networks
>>>>>>                     draft-ietf-tictoc-1588overmpls-05
>>>>>>
>>>>>> Abstract
>>>>>>
>>>>>>     This document defines the method for transporting Timing messages
>>>>>>     such as PTP and NTP over an MPLS network.  The method allows for the
>>>>>>     easy identification of these PDUs at the port level to allow for port
>>>>>>
>>>>>> SB> What is a port
>>>>>>
>>>>>>     level processing of these PDUs in both LERs and LSRs.
>>>>>>
>>>>>>     The basic idea is to transport Timing messages inside dedicated MPLS
>>>>>>     LSPs.  These LSPs only carry Timing messages and possibly Control and
>>>>>>     Management packets, but they do not carry customer traffic.
>>>>>>
>>>>>> SB> More specifically they only carry traffic associated with the
>>>>>> SB> timing service and its support.
>>>>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
>>>>>> SB> also it gets carried in a structure that causes it to get
>>>>>> SB> timestamped.
>>>>>>
>>>>>>     Two methods for transporting Timing messages over MPLS are defined.
>>>>>>
>>>>>> SB> Perhaps the right approach is to define the new LSP type and then
>>>>>> SB> seperately to define  the mapping of the various timing services
>>>>>> SB> over that LSP type.
>>>>>>
>>>>>>     The first method is to transport Timing messages directly over the
>>>>>>     dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
>>>>>>     MPLS networks.  The second method is to transport Timing messages
>>>>>>     inside a PW via Ethernet encapsulation.
>>>>>>
>>>>>> SB> I think that we should note that there are some
>>>>>> SB> h/w reasons for this preference. A clean sheet approach
>>>>>> SB> would have been to use PTP over MPLS with no intermediate
>>>>>> SB> layers.
>>>>>>
>>>>>> Status of this Memo
>>>>>>
>>>>>>     This Internet-Draft is submitted in full conformance with the
>>>>>>     provisions of BCP 78 and BCP 79.
>>>>>>
>>>>>>     Internet-Drafts are working documents of the Internet Engineering
>>>>>>     Task Force (IETF).  Note that other groups may also distribute
>>>>>>     working documents as Internet-Drafts.  The list of current Internet-
>>>>>>     Drafts is at http://datatracker.ietf.org/drafts/current/.
>>>>>>
>>>>>>     Internet-Drafts are draft documents valid for a maximum of six months
>>>>>>     and may be updated, replaced, or obsoleted by other documents at any
>>>>>>     time.  It is inappropriate to use Internet-Drafts as reference
>>>>>>     material or to cite them other than as "work in progress."
>>>>>>
>>>>>>     This Internet-Draft will expire on December 17, 2013.
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page 1]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> Copyright Notice
>>>>>>
>>>>>>     Copyright (c) 2013 IETF Trust and the persons identified as the
>>>>>>     document authors.  All rights reserved.
>>>>>>
>>>>>>     This document is subject to BCP 78 and the IETF Trust's Legal
>>>>>>     Provisions Relating to IETF Documents
>>>>>>     (http://trustee.ietf.org/license-info) in effect on the date of
>>>>>>     publication of this document.  Please review these documents
>>>>>>     carefully, as they describe your rights and restrictions with respect
>>>>>>     to this document.  Code Components extracted from this document must
>>>>>>     include Simplified BSD License text as described in Section 4.e of
>>>>>>     the Trust Legal Provisions and are provided without warranty as
>>>>>>     described in the Simplified BSD License.
>>>>>>
>>>>>>
>>>>>> Table of Contents
>>>>>>
>>>>>>     1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  5
>>>>>>
>>>>>>     2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .  7
>>>>>>
>>>>>>     3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .  8
>>>>>>
>>>>>>     4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .  9
>>>>>>
>>>>>>     5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . . 12
>>>>>>
>>>>>>     6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . . 13
>>>>>>       6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . . 13
>>>>>>       6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . . 13
>>>>>>       6.3.  Other Timing Encapsulation methods . . . . . . . . . . . . 14
>>>>>>
>>>>>>     7.  Timing message Processing  . . . . . . . . . . . . . . . . . . 15
>>>>>>
>>>>>>     8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . . 16
>>>>>>
>>>>>>     9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
>>>>>>
>>>>>>     10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
>>>>>>
>>>>>>     11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
>>>>>>
>>>>>>     12. OAM, Control and Management  . . . . . . . . . . . . . . . . . 20
>>>>>>
>>>>>>     13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . . 21
>>>>>>
>>>>>>     14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . . 22
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page 2]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>>     15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . . 23
>>>>>>       15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . . 23
>>>>>>       15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . . 23
>>>>>>       15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . . 24
>>>>>>
>>>>>>     16. Other considerations . . . . . . . . . . . . . . . . . . . . . 25
>>>>>>
>>>>>>     17. Security Considerations  . . . . . . . . . . . . . . . . . . . 26
>>>>>>
>>>>>>     18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 27
>>>>>>
>>>>>>     19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 28
>>>>>>
>>>>>>     20. References . . . . . . . . . . . . . . . . . . . . . . . . . . 29
>>>>>>       20.1. Normative References . . . . . . . . . . . . . . . . . . . 29
>>>>>>       20.2. Informative References . . . . . . . . . . . . . . . . . . 29
>>>>>>
>>>>>>     Appendix 1.  Routing extensions for Timing-aware Routers . . . . . 32
>>>>>>
>>>>>>     Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . . 33
>>>>>>
>>>>>>     Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 34
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page 3]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>>     The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
>>> NOT",
>>>>>>     "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL"
>>> in
>>>>>> this
>>>>>>     document are to be interpreted as described in RFC2119 [RFC2119].
>>>>>>
>>>>>>     When used in lower case, these words convey their typical use in
>>>>>>     common language, and are not to be interpreted as described in
>>>>>>     RFC2119 [RFC2119].
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page 4]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 1.  Introduction
>>>>>>
>>>>>>     The objective of Precision Time Protocol (PTP) and Network Timing
>>>>>>     Protocol (NTP) are to synchronize independent clocks running on
>>>>>>     separate nodes of a distributed system.
>>>>>>
>>>>>>     [IEEE-1588] defines PTP messages for frequency, phase and time
>>>>>>     synchronization.  The PTP messages include PTP PDUs over UDP/IP
>>>>>>     (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F of
>>>>>>     [IEEE-1588]).
>>>>>>
>>>>>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
>>>>>> SB> of other PTP mappings if they provide better optimisation.
>>>>>>
>>>>>>     This document defines mapping and transport of the PTP
>>>>>>     messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>>>>>>     defines several clock types: ordinary clocks, boundary clocks, end-
>>>>>>     to-end transparent clocks, and peer-to-peer transparent clocks.
>>>>>>     Transparent clocks require intermediate nodes to update correction
>>>>>>     field inside PTP message that reflects the transit time in the node.
>>>>>>
>>>>>>     [RFC5905] defines NTP messages for clock and time synchronization.
>>>>>>     The PTP messages (PDUs) are transported over UDP/IP.  This document
>>>>>> SB> Should that be NTP messages?
>>>>>> SB> It needs to be made clear as soon as you introduce NTP that
>>>>>> SB> they use different time representations.
>>>>>>
>>>>>>     defines mapping and transport of the NTP messages defined in
>>>>>>     [RFC5905] over MPLS networks.
>>>>>>
>>>>>>     One key attribute of all of these Timing messages is that the Time
>>>>>>     stamp processing should occur as close as possible to the actual
>>>>>>     transmission and reception at the physical port interface.  This
>>>>>>     targets optimal time and/or frequency recovery by avoiding variable
>>>>>>     delay introduced by queues internal to the clocks.
>>>>>>
>>>>>> SB> As I recall NTP has no epoch point defined, and I am not sure
>>>>>> SB> where that point is in the case of PTP in this mapping
>>>>>> SB> Hopefully this will get defined in due course.
>>>>>>
>>>>>>     To facilitate the fast and efficient recognition of Timing messages
>>>>>>     at the port level when the Timing messages are carried over MPLS
>>>>>>     LSPs,
>>>>>>
>>>>>> SB> Over a new LSP type with time optimied characteristics
>>>>>>
>>>>>>     this document defines the specific encapsulations that should
>>>>>>     be used.
>>>>>> SB> Hopefully it will also define the PHP
>>>>>>
>>>>>>     In addition, it can be expected that there will exist LSR/
>>>>>>     LERs where only a subset of the physical ports will have the port-
>>>>>>     based Timing message processing capabilities.
>>>>>> SB> Do you need to clarify that this only works at base and not in
>>>>>> SB> a label heirarchy.
>>>>>>
>>>>>>
>>>>>>     In order to ensure
>>>>>>     that the LSPs carrying Timing packets always enter and exit ports
>>>>>>     with this capability, routing extensions are defined to advertise
>>>>>>     this capability on a port basis and to allow for the establishment of
>>>>>>     LSPs that only transit such ports.  While this path establishment
>>>>>>     restriction may be applied only at the LER Ingress and/or egress
>>>>>>     ports, it becomes more important when using transparent clock capable
>>>>>>     LSRs in the path.
>>>>>> SB> I do not understand the implications of the last
>>>>>> SB> sentences - starting ", it becomes"
>>>>>>
>>>>>>
>>>>>>     Port based Timing message processing involves Timing message
>>>>>>     recognition.  Once the Timing messages are recognized they can be
>>>>>>     modified based on the reception or transmission Time-stamp.
>>>>>>
>>>>>>     This document provides two methods for transporting Timing messages
>>>>>>     over MPLS.  One is applicable to MPLS environment and the other one
>>>>>>     is applicable to MPLS/MPLS-TP environment
>>>>>>
>>>>>> SB> I think the sentence is incomplete.
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page 5]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>>     The solution involves transporting Timing messages over dedicated
>>>>>>     LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
>>>>>>     carry Management and control messages, but not data plane client
>>>>>>     traffic.
>>>>>>
>>>>>> SB> It is not clear why this restriction applies.
>>>>>>
>>>>>>     Timing LSPs can be established statically or via signaling.
>>>>>> SB> s/statically/by provisioning/network management/
>>>>>>
>>>>>>     Extensions to control plane (OSPF, ISIS, etc.) is required to enable
>>>>>>     routers to distribute their Timing processing capabilities over MPLS
>>>>>>     to other routers.  However such extensions are outside the scope of
>>>>>>     this document.
>>>>>>
>>>>>>     When signaling is used to setup the PTP LSP, Extensions to signaling
>>>>>> SB> is it a PTP LSP or a Timing LSP?
>>>>>>
>>>>>>     protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
>>>>>>     However such extensions are outside the scope of this document.
>>>>>>
>>>>>> SB> for mpls-tp GMPLS is the signalling protocol
>>>>>>
>>>>>>     While the techniques included herein allow for the establishment of
>>>>>>     paths optimized to include Time-stamping capable links, the
>>>>>>     performance of the Slave clocks is outside the scope of this
>>>>>>     document.
>>>>>>
>>>>>>     At the time of publishing this specification, Transparent Clocking
>>>>>>     (TC) is only defined for PTP.  Therefore at this time any part of
>>>>>>     this specification that talks about Transparent Clocking applies only
>>>>>>     to PTP.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page 6]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 2.  Terminology
>>>>>>
>>>>>>     1588: The timing and synchronization as defined by IEEE 1588.
>>>>>>
>>>>>> SB> I think that there is a more formal name for the 1588 group
>>>>>> SB> that needs to be used here.
>>>>>> SB> Also do we need to talk about 1588-200? as there is
>>>>>> SB> an update in progress
>>>>>>
>>>>>>     NTP: The timing and synchronization protocol defined by IETF RFC-1305
>>>>>>     and RFC-5905.
>>>>>>
>>>>>>     PTP: The timing and synchronization protocol used by 1588.
>>>>>> SB> need the proper name for 1588
>>>>>>
>>>>>>     Master Clock: The source of 1588 timing to a set of slave clocks.
>>>>>>
>>>>>>     Master Port: A port on a ordinary or boundary clock that is in Master
>>>>>>     state.  This is the source of timing toward slave ports.
>>>>>>
>>>>>> SB> I am not sure the reader knows what a port is
>>>>>>
>>>>>>     Slave Clock: A receiver of 1588 timing from a master clock.
>>>>>>
>>>>>>     Slave Port: A port on a boundary clock or ordinary clock that is
>>>>>>     receiving timing from a master clock.
>>>>>>
>>>>>>     Ordinary Clock: A device with a single PTP port.
>>>>>>
>>>>>>     Transparent Clock.  A device that measures the time taken for a PTP
>>>>>>     event message to transit the device and then updates the
>>>>>>     correctionField of the message with this transit time.
>>>>>>
>>>>>>     Boundary Clock: A device with more than one PTP port.  Generally
>>>>>>     boundary clocks will have one port in slave state to receive timing
>>>>>>     and then other ports in master state to re-distribute the timing.
>>>>>>
>>>>>>     PTP LSP: An LSP dedicated to carry PTP messages
>>>>>>
>>>>>> SB> PTP or timing?
>>>>>>
>>>>>>
>>>>>>     PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>>>>>     messages.
>>>>>>
>>>>>> SB> Ah I don't think that PWE3 know what one of these is
>>>>>>
>>>>>>     CW: Pseudowire Control Word
>>>>>>
>>>>>>     LAG: Link Aggregation
>>>>>>
>>>>>>     ECMP: Equal Cost Multipath
>>>>>>
>>>>>>     CF: Correction Field, a field inside certain PTP messages (message
>>>>>>     type 0-3)that holds the accumulative transit time inside intermediate
>>>>>>     switches
>>>>>>
>>>>>>     Timing messages: Timing Protocol messages that are exchanged
>>> between
>>>>>>     routers in order to establish a synchronized clock.
>>>>>>
>>>>>> SB> A number of these definitions look like copies of IEEE1588
>>>>>> SB> definitions. We need to provide references and note the
>>>>>> SB> priority of the IEEE base reference.
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page 7]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 3.  Problem Statement
>>>>>>
>>>>>>     [IEEE-1588] has defined methods for transporting PTP messages over
>>>>>>     Ethernet and IP networks.  [RFC5905] has defined the method of
>>>>>>     transporting NTP messages over IP networks.  There is a need to
>>>>>>     transport Timing messages over MPLS networks while supporting the
>>>>>>     Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
>>>>>>     functionality in the LER and LSRs in the MPLS network.
>>>>>>
>>>>>>     There are multiple ways of transporting Timing over MPLS.  However,
>>>>>>     there is a requirement to limit the possible encapsulation options to
>>>>>>     simplify the Timing message identification and processing required at
>>>>>>     the port level.
>>>>>>
>>>>>>     When Timing-awareness is needed, Timing messages should not be
>>>>>>     transported over LSPs or PWs that are carrying customer traffic
>>>>>>     because LSRs perform Label switching based on the top label in the
>>>>>>     stack.
>>>>>>
>>>>>> SB> Have you explained why?
>>>>>>
>>>>>>     To detect Timing messages inside such LSPs require special
>>>>>>     hardware to do deep packet inspection at line rate.  Even if such
>>>>>>     hardware exists, the payload can't be deterministically identified by
>>>>>>     LSRs because the payload type is a context of the PW label, and the
>>>>>>     PW label and its context are only known to the Edge routers (PEs/
>>>>>>     LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CES,
>>>>>>     etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
>>>>>>     LSRs dont have the knowledge of whether PW Control Word (CW) is
>>>>>>     present or not and therefore can not deterministically identify the
>>>>>>     payload.
>>>>>>
>>>>>>     A generic method is defined in this document that does not require
>>>>>>     deep packet inspection at line rate, and can deterministically
>>>>>>     identify Timing messages.  This method can be used to detect Timing
>>>>>>     Messages in both one-step and two-step clock implementations of
>>>>>>     ordinary, boundary and transparent clocks.
>>>>>>
>>>>>> SB> Needs a ref and I am sure many MPLS specialists will not understand
>>>>>> SB> the msg types.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page 8]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 4.  Timing over MPLS Architecture
>>>>>>
>>>>>>     Timing messages are exchange between Timing ports on ordinary and
>>>>>>
>>>>>> SB> Have you defined a timing port?
>>>>>>
>>>>>>     boundary clocks.  Boundary clocks terminate the Timing messages and
>>>>>>     act as master for other boundary clocks or for slave clocks.  End-to-
>>>>>>     End Transparent clocks do not terminate the Timing messages but they
>>>>>>     do modify the contents of the Timing messages as they transit across
>>>>>>     the transparent clock.
>>>>>>
>>>>>>     Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Clock
>>>>>>
>>>>>>     (TC) could be implemented in either LERs or LSRs.
>>>>>>
>>>>>> SB> LER and LSR need to be expanded
>>>>>>
>>>>>>     An example is shown in Figure 1, where the LERs act as Ordinary Clock
>>>>>>     (OC) and are the initiating/terminating point for Timing messages.
>>>>>>     The ingress LER encapsulates the Timing messages in Timing LSP and
>>>>>>     the Egress LER terminates the Timing LSP.  The LSRs act as
>>>>>>     Transparent Clock (TC) and just update the Timing field in the Timing
>>>>>>     messages.
>>>>>>
>>>>>>
>>>>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>        |Switch, |     |       |     |       |     |       |     |Switch, |
>>>>>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>>>>        |        |     |  OC   |     |  TC   |     |  OC   |     |        |
>>>>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>                       /                                 \
>>>>>>        +-------+     /                                   \     +-------+
>>>>>>        |  LER  |    /                                     \    |  LER  |
>>>>>>        | Master|---/                                       \---| Slave |
>>>>>>        | Clock |                                               | Clock |
>>>>>>        +-------+                                               +-------+
>>>>>>
>>>>>>       Figure (1) - Deployment example 1 of timing over MPLS network
>>>>>>
>>>>>>     Another example is shown in Figure2, where LERs terminate the Timing
>>>>>>     messages received from switch/routers that are outside of the MPLS
>>>>>>     network acting as OC or BC.  In this example LERs regenerate the
>>>>>>     clock and initiate timing messages encapsulated in Timing LSP toward
>>>>>>     the MPLS network, while the LSRs act as Transparent Clock (TC) and
>>>>>>     just update the Timing field in the Timing messages, which are
>>>>>>     already encapsulated in Timing LSPs.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page 9]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>       |Switch, |     |       |     |       |     |       |     |Switch, |
>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>>>>       | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC  |
>>>>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>
>>>>>>       Figure (2) - Deployment example 2 of timing over MPLS network
>>>>>>
>>>>>>
>>>>>>     Another example is shown in Figure 3, where LERs do not terminate the
>>>>>>     Timing messages received from switch/routers that are outside of the
>>>>>>     MPLS network acting as OC, TC or BC.  The LERs act as TC and update
>>>>>>     the Timing field in the Timing messages as they transit the LER,
>>>>>>     while encapsulating them in timing LSP.  The LSRs also act as
>>>>>>     Transparent Clock (TC) and just update the Timing field in the Timing
>>>>>>     messages which are already encapsulated in Timing LSPs.
>>>>>>
>>>>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>        |Switch, |     |       |     |       |     |       |     |Switch, |
>>>>>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>>>>        |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC/BC|
>>>>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>
>>>>>>      Figure (3) - Deployment example 3 of timing over MPLS network
>>>>>>
>>>>>>     Another example is shown in Figure 4, where LERs and LSRs support
>>>>>>     Boundary Clocks.  A single-hop LSP is created between two adjacent
>>>>>>     LSRs engaged in BC operation.  Other methods such as PTP transport
>>>>>>     over Ethernet MAY be used for transporting timing messages if the
>>>>>>     link between the two routers is Ethernet.
>>>>>>
>>>>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>       |Switch, |     |       |     |       |     |       |     |Switch, |
>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>>>>       | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC  |
>>>>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>
>>>>>>     Figure (4) - Deployment example 3 of timing over MPLS network
>>>>>>
>>>>>>     An MPLS domain MAY serve multiple customers.  In these cases the
>>> MPLS
>>>>>>     domain (maintained by a service provider) may provide timing services
>>>>>>     to multiple customers, each having their own Timing domain.
>>>>>>
>>>>>>     The Timing over MPLS architecture assumes full mesh of Timing LSPs
>>>>>>     between all LERs supporting this specification.
>>>>>>
>>>>>> SB> Note sure this is right - the salves surely do not need to
>>>>>> SB> exchange timing amongst themselves
>>>>>>
>>>>>>     It supports
>>>>>>     Point-to- point (VPWS) and Multipoint (VPLS) services.
>>>>>>
>>>>>> SB> What does that mean? You do not carry user data traffic?
>>>>>> SB> Maybe it's the ordering of the statemnets that is causing
>>>>>> SB> confusion.
>>>>>>
>>>>>>     This means
>>>>>>     that a customer may purchase a Point-to-point Timing service between
>>>>>>     two customer sites or a Multipoint Timing service between more than
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 10]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>>     two customer sites.
>>>>>>
>>>>>>     The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
>>>>>>     This means that the Timing Multicast messages such as PTP Multicast
>>>>>>     event messages can be transported over P2MP Timing LSP or be
>>>>>>     replicated and transported over many P2P Timing LSPs.
>>>>>>
>>>>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>>>>>> SB> nor a P2MP PW, although we are close.
>>>>>>
>>>>>>     Timing messages, that do not require Time stamping or Correction
>>>>>>     Field update MAY be transported over Timing LSPs to simplify hardware
>>>>>>     and software.
>>>>>>
>>>>>>     PTP Announce messages that determine the Timing LSP terminating
>>> point
>>>>>>     behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
>>>>>>     to simplify hardware and software.
>>>>>>
>>>>>> SB> have you defined and referenced PTP announce msgs?
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 11]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 5.  Dedicated LSPs for Timing messages
>>>>>>
>>>>>>     Many methods have been considered for identifying the Timing
>>> messages
>>>>>>     when they are encapsulated in MPLS such as using GAL/G-ACH or a new
>>>>>>     reserved label.  These methods were not attractive since they either
>>>>>>     required deep packet inspection at line rate in the intermediate LSRs
>>>>>>     or they required use of a scarce new reserved label.  Also one of the
>>>>>>     goals was to reuse existing OAM mechanisms.
>>>>>>
>>>>>> SB> RLs = SPLs are not so rare now. In any case needs a ref.
>>>>>>
>>>>>>     The method defined in this document can be used by LER and LSRs to
>>>>>>     identify Timing messages in MPLS tunnels by just looking at the top
>>>>>>     label in the MPLS label stack, which only carry Timing messages as
>>>>>>     well as OAM, but not data plane client traffic.
>>>>>>
>>>>>>     Compliant implementations MUST use dedicated LSPs to carry Timing
>>>>>>     messages over MPLS.
>>>>>>
>>>>>> SB> I think that we need a definition of the properies of these LSPs
>>>>>>
>>>>>>     These LSPs are herein referred to as "Timing
>>>>>>     LSPs" and the labels associated with these LSPs as "Timing LSP
>>>>>>     labels".  The Timing LSPs that runs between Ingress and Egress LERs
>>>>>>     MUST be co-routed.  Alternatively, a single bidirectional co-routed
>>>>>>     LSP can be used.
>>>>>>
>>>>>> SB> I though that you said you could use M2MP LSPs - these are not
>>>>>> SB> bidirectional.
>>>>>>
>>>>>>     Co-routing of the two directions is required to limit the difference
>>>>>>     in the delays in the Master clock to Slave clock direction compared
>>>>>>     to the Slave clock to Master clock direction.  The Timing LSP MAY be
>>>>>>     MPLS/MPLS-TP LSP.
>>>>>>
>>>>>>     The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
>>>>>>     New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
>>>>>>     outside the scope of this document.
>>>>>>
>>>>>>     The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such as
>>>>>>     BFD and LSP Ping but the LSP data plane client plane traffic MUST be
>>>>>>     Timing packets only.
>>>>>>
>>>>>> SB> Why?
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 12]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 6.  Timing over LSP Encapsulation
>>>>>>
>>>>>> The encapsulations is not LSP is it?
>>>>>>
>>>>>>     This document defines two methods for carrying Timing messages over
>>>>>>     MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>>>>>     messages over Timing LSPs, and the second method, is carrying
>>>>>>     Ethernet encapsulated Timing messages over Ethernet PWs inside
>>> Timing
>>>>>>     LSPs.
>>>>>>
>>>>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>>>>>
>>>>>>     The simplest method of transporting Timing messages over MPLS is to
>>>>>>     encapsulate Timing PDUs in UDP/IP and then encapsulate them in
>>> Timing
>>>>>>     LSP.  This format is shown in Figure 4.
>>>>>>
>>>>>>
>>>>>>                      +----------------------+
>>>>>>                      |   Timing LSP Label   |
>>>>>>                      +----------------------+
>>>>>>                      |        IPv4/6        |
>>>>>>                      +----------------------+
>>>>>>                      |         UDP          |
>>>>>>                      +----------------------+
>>>>>>                      |     Timing PDU       |
>>>>>>                      +----------------------+
>>>>>>
>>>>>>        Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>>>>>
>>>>>>
>>>>>>     This encapsulation is very simple and is useful when the network
>>>>>>     between Timing Master Clock and Slave Clock is MPLS network.
>>>>>>
>>>>>> SB> Simple is a judgement call
>>>>>>
>>>>>>     In order for an LER/LSR to process Timing messages, the Timing LSP
>>>>>>     Label must be at the top label of the label stack.  The LER/LSR MUST
>>>>>>     know that the Timing LSP Label is used for carrying Timing messages.
>>>>>>     This can be accomplished via static configuration or via RSVP-TE
>>>>>>     signaling.
>>>>>>
>>>>>>     The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>>>>>     [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>>>>>>     [RFC5905].
>>>>>>
>>>>>> 6.2.  Timing over PW Encapsulation
>>>>>>
>>>>>>     Another method of transporting Timing over MPLS networks is by
>>>>>>     encapsulating Timing PDUs in PW which in turn is transported over
>>>>>>     Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
>>>>>>     shown in Fig 5(A) MUST be used and the Ethernet encapsulation of PTP
>>>>>>     MUST follow Annex F of [IEEE-1588].
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 13]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>>     The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
>>> the
>>>>>>     Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>>>>>>     Timing over PW encapsulation MUST use the Control Word (CW) as
>>>>>>     specified in [RFC4448] to ensure proper detection of PTP messages
>>>>>>     inside the MPLS packets for Timing over LSP and Timing over PW
>>>>>>     encapsulation.
>>>>>>
>>>>>> SB> That needs explanation
>>>>>>
>>>>>>     The use of Sequence Number in the CW is optional.
>>>>>>
>>>>>> SB> Given that s/n are never in practice deployed, you could probably
>>>>>> SB> simplify things by sayig that they are not used.
>>>>>>
>>>>>>     Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over
>>> PW
>>>>>>     (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>>>>>
>>>>>>                      +----------------+  +----------------+
>>>>>>                       |Timing LSP Label|  |Timing LSP Label|
>>>>>>                       +----------------+  +----------------+
>>>>>>                       |    PW Label    |  |    PW Label    |
>>>>>>                       +----------------+  +----------------+
>>>>>>                       |  Control Word  |  |      IP        |
>>>>>>                       +----------------+  +----------------+
>>>>>>                       |    Ethernet    |  |      UDP       |
>>>>>>                       |     Header     |  +----------------+
>>>>>>                       +----------------+  |   Timing PDU   |
>>>>>>                       |S-VLAN(Optional)|  |                |
>>>>>>                       +----------------+  +----------------+
>>>>>>                       |C-VLAN(Optional)|        (B)
>>>>>>                       +----------------+
>>>>>>                       |   Timing PDU   |
>>>>>>                       |                |
>>>>>>                       +----------------+
>>>>>>                              (A)
>>>>>>
>>>>>>                Figure (5) - Timing over PW Encapsulations
>>>>>>
>>>>>>     In order for an LSR to process PTP messages, the top label of the
>>>>>>     label stack (the Tunnel Label) MUST be a Timing label.
>>>>>>
>>>>>> S> You said that before.
>>>>>>
>>>>>> 6.3.  Other Timing Encapsulation methods
>>>>>>
>>>>>>     In future other timing encapsulation methods may be introduced, such
>>>>>>     as a new shim header after the Bottom of Stack to carry the Timing
>>>>>>     information.  Such new encapsulations are outside the scope of this
>>>>>>     document.
>>>>>>
>>>>>>
>>>>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>>>>>> SB> out of the definition of the LSP
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 14]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> SB> I think we need a section on LSP processing
>>>>>>
>>>>>> 7.  Timing message Processing
>>>>>>
>>>>>>     Each Timing protocol such as PTP and NTP, define their set of Timing
>>>>>>     messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>>>>>>     FOLLOW_UP, etc messages.
>>>>>>
>>>>>>     Some of the Timing messages require time stamping or correction field
>>>>>>     update at port level and some dont.  It is the job of the LER/LSR to
>>>>>>     parse the timing message and find out the type of the Timing message
>>>>>>     and decide whether and how to Time- stamp it (e.g., BC) or update
>>>>>>     correction field(e.g., TC).
>>>>>>
>>>>>>
>>>>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
>>>>>> SB> function rather than the LER?
>>>>>>
>>>>>>     For example the following PTP messages (called Event messages)
>>>>>>     require time-stamping or correction field update:
>>>>>>
>>>>>>     o  SYNC
>>>>>>
>>>>>>     o  DELAY_REQ (Delay Request)
>>>>>>
>>>>>>     o  PDELAY_REQ (Peer Delay Request)
>>>>>>
>>>>>>     o  PDELAY_RESP (Peer Delay Response)
>>>>>>
>>>>>>     SYNC and DELAY_REQ are exchanged between Master Clock and Slave
>>> Clock
>>>>>>     and MUST be transported over PTP LSPs.  PDELAY_REQ and
>>> PDELAY_RESP
>>>>>>     are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>>>>>>     Boundary, or Transparent) and SHOULD be transported over single hop
>>>>>>     PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
>>>>>>     and PDELAY_RESP_FOLLOW_UP messages MUST also be transported
>>> over
>>>>>> the
>>>>>>     PTP LSPs.
>>>>>>
>>>>>>     For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
>>>>>>     transported over two PTP LSPs that are in opposite directions.  These
>>>>>>     PTP LSPs, which are in opposite directions MUST be congruent and co-
>>>>>>     routed.  Alternatively, a single bidirectional co-routed LSP can be
>>>>>>     used.
>>>>>>
>>>>>>     Except as indicated above for the two-step PTP clocks, Non-Event PTP
>>>>>>     message types do not need to be processed by intermediate routers.
>>>>>>     These message types MAY be carried in PTP Tunnel LSPs.
>>>>>>
>>>>>> SB> Are you saying that a timing P router has to be msg type sensitive?
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 15]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 8.  Protection and Redundancy
>>>>>>
>>>>>>
>>>>>> SB> This is a bit of a jump - I don't know how the LSP itself works yet!
>>>>>>
>>>>>>     In order to ensure continuous uninterrupted operation of slave
>>>>>>     clocks, usually as a general practice, slave clocks (or ports) track
>>>>>>     redundant master clocks.
>>>>>>
>>>>>>     It is the responsibility of the network operator to ensure that
>>>>>>     physically disjoint Timing LSPs are established between a slave clock
>>>>>>     (or port) and redundant master clocks (or ports).
>>>>>>
>>>>>>     When a slave clock (or port) listens to redundant master clocks or
>>>>>>     ports, any prolonged Timing LSP outage will trigger the slave clock
>>>>>>     or port to switch to a redundant master clock or port.
>>>>>>
>>>>>>     LSP/PW protection such as Linear protection Switching (1:1, 1+1),
>>>>>>     Ring protection switching or MPLS Fast Reroute (FRR) generally switch
>>>>>>     alternative path that usually cause a change in delay, which if
>>>>>>     undetected by slave clock can reduce accuracy of the slave clock.
>>>>>>
>>>>>>     Therefore protection switching MAY be used, as long as phase jumps
>>>>>>     upon switchover due to differences in path latency are detected and
>>>>>>     compensated for (such compensation not being required if BCs or peer-
>>>>>>     peer TCs are used throughout).
>>>>>>
>>>>>>     Note that any protection or reroute mechanism that adds additional
>>>>>>     MPLS label to the label stack, such as Facility Backup Fast Reroute,
>>>>>>     MUST ensure that the pushed label is also a Timing Label to ensure
>>>>>>     recognition of the MPLS frame as containing Timing messages, as it
>>>>>>     transits the backup path.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 16]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 9.  ECMP
>>>>>>
>>>>>>     To ensure the optimal operation of slave clocks and avoid error
>>>>>>     introduced by forward and reverse path delay asymmetry, the physical
>>>>>>     path for Timing messages from master clock to slave Clock and vice
>>>>>>     versa must be the same for all Event Timing messages listed in
>>>>>>     section 7.
>>>>>>
>>>>>>     Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
>>>>>>     Multipath).
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 17]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 10.  PHP
>>>>>>
>>>>>>     To ensure that the label on the top of the label stack is the Timing
>>>>>>     LSP Label, PHP MUST not be used.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 18]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 11.  Entropy
>>>>>>
>>>>>>     To ensure all Timing messages in a Timing LSP take the same path,
>>>>>>     Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>>>>>     Entropy Label MUST NOT be used for the PWs that are carried inside
>>>>>>     Timing LSP [RFC6391].
>>>>>>
>>>>>> SB> This is incorrect - you mean that all msgs of the same timing
>>>>>> SB> flow need to have the same EL value.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 19]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 12.  OAM, Control and Management
>>>>>>
>>>>>>     In order to monitor Timing LSPs and their encapsulated PWs, they MUST
>>>>>>     be able to carry OAM and management messages.  These management
>>>>>>     messages MUST be differentiated from Timing messages via already
>>>>>>     defined IETF methods.
>>>>>>
>>>>>>     For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
>>>>>>     over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>>>>>>     Management protocols can easily be identified by the UDP Destination
>>>>>>     Port number or by GAL/G-ACH respectively.
>>>>>>
>>>>>>     Also BFD, LSP-Ping and other management messages MAY run over the
>>> PWs
>>>>>>     encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3 or
>>>>>>     4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is going
>>>>>>     to be deprecated by IETF).  In this case G-ACH, PW label (TTL=1) or
>>>>>>     GAL-ACH are used to identify such management messages.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 20]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 13.  QoS Considerations
>>>>>>
>>>>>>     In network deployments where not every LSR/LER is Timing-aware, it is
>>>>>>     important to reduce the impact of the non-Timing-aware LSR/LERs on
>>>>>>     the timing recovery in the slave clock.  The Timing messages are time
>>>>>>     critical and must be treated with the highest priority.  Therefore
>>>>>>     Timing over MPLS messages must be treated with the highest priority
>>>>>>     in the routers.  This can be achieved by proper setup of Timing LSPs.
>>>>>>
>>>>>>     It is recommended that the Timing LSPs are setup or configured
>>>>>>     properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC2697]
>>>>>>     for drop eligibility.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 21]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 14.  FCS and Checksum Recalculation
>>>>>>
>>>>>>     When time-stamp generation and timing packet adjustment is performed
>>>>>>     near the physical port hardware, the process MUST include
>>>>>>     recalculation of the Ethernet FCS.
>>>>>>
>>>>>> SB> The above is confusing - an LSR always recomputes the link layer
>>>>>> SB> CRC which may or may not be Ethernet.
>>>>>>
>>>>>>     Also FCS retention for the
>>>>>>     payload Ethernet described in [RFC4720] MUST NOT be used.
>>>>>>
>>>>>>     For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
>>>>>>     may be required as per UDP transport standards.
>>>>>>
>>>>>> SB> You really need to be working on getting the IPv6 C?S computation
>>>>>> SB> removed from PTP msgs.
>>>>>>
>>>>>>     When UDP checksum is used, each Timing-aware LER/LSR must either
>>>>>>     incrementally update the UDP checksum after Time stamping or
>>>>>>     Correction Field update or verify the UDP checksum on reception from
>>>>>>     upstream and recalculate the checksum completely on transmission to
>>>>>>     downstream node after Time stamping or Correction Field update.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 22]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 15.  Behavior of LER/LSR
>>>>>>
>>>>>>     Timing-capable/aware LERs and LSRs are routers that have one or more
>>>>>>
>>>>>> SB> You mean physical interfaces?
>>>>>>
>>>>>>     interfaces that can perform Timing operations (OC/BC/TC) on Timing
>>>>>>     packets and are configured to do so.  Timing-capable/aware LERs and
>>>>>>     LSRs can advertise their Timing-capability per-interface via control
>>>>>>     plane such as OSPF or IS-IS.
>>>>>> SB> ISIS and OSPF are routing protocols.
>>>>>>
>>>>>>    The Timing-capable/aware LERs can then
>>>>>>     signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timing
>>>>>>     capability of LER and LSRs may be configured in a centralized
>>>>>>     controller and the Timing LSP may be setup using manual configuration
>>>>>>     or other methods such as SDN.
>>>>>>
>>>>>> SB> it can also be configured individually rather then through
>>>>>> SB> a cebtral controllwe
>>>>>>
>>>>>> 15.1.  Behavior of Timing-capable/aware LER
>>>>>>
>>>>>>     When a Timing-capable/aware LER behaves as a Transparent clock and
>>>>>>     receives a Timing message from a Timing-capable/aware non-MPLS
>>>>>>     interface, the LER updates the Correction Field (CF) and encapsulates
>>>>>>     and forwards the timing message over previously established Timing
>>>>>>     LSP.
>>>>>>
>>>>>> SB> You need to call out the details so that people properly
>>>>>> SB> understand the definition of the new LSP.
>>>>>>
>>>>>>     Also when a Timing message is received from a Timing-capable/
>>>>>>     aware MPLS interface, LER updates the Correction Filed (CF) and
>>>>>>     decapsulates the MPLS encapsulation and forwards the timing message
>>>>>>     to a non-MPLS interface.
>>>>>>
>>>>>>     When a Timing-capable/aware LER behaves as a Boundary clock and
>>>>>>     receives a Timing message from a Timing-capable/aware non MPLS
>>>>>>     interface, the LER Timestamps the Timing packet and sends it to the
>>>>>>     LERs Boundary clock processing module.  Also when a Timing message is
>>>>>>     received from a Timing- capable/aware MPLS interface, the LER
>>>>>>     Timestamps the Timing packet and sends it to the LERs Boundary clock
>>>>>>     processing module.
>>>>>>
>>>>>>     When a Timing-capable/aware LER behaves as an Ordinary Clock toward
>>>>>>     the MPLS network, and receives a Timing message from a Timing-
>>>>>>     capable/aware MPLS interface, the LER Timestamps the Timing packet
>>>>>>     and sends it to the LERs Ordinary clock processing module.
>>>>>>
>>>>>> 15.2.  Behavior of Timing-capable/aware LSR
>>>>>>
>>>>>>     When a Timing-capable/aware LSR behaves as a Transparent clock and
>>>>>>     receives a Timing message from a Timing-capable/aware MPLS
>>> interface,
>>>>>>     The LSR updates the Correction Filed (CF) and forwards the timing
>>>>>>     message over another MPLS interface.
>>>>>>
>>>>>>     When a Timing-capable/aware LSR behaves as a Boundary clock and
>>>>>>     receives a Timing message from a Timing-capable/aware MPLS
>>> interface.
>>>>>>     The LSR performs the functions of a Boundary Clock in terminating the
>>>>>>     received Timing message and re-generating a new timing message over
>>>>>>     another (or the same) MPLS interface.
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 23]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>>>>>
>>>>>>     It is most beneficial when all LSRs in the path of a Timing LSP be
>>>>>>     timing-Capable/aware LSRs.  This would ensure the highest quality
>>>>>>     time and clock synchronization by Timing Slave Clocks.  However, this
>>>>>>     specification does not mandate that all LSRs in path of a Timing LSP
>>>>>>     be Timing- capable/aware.
>>>>>>
>>>>>>     Non-Timing-capable/aware LSRs just switch the packets encapsulated in
>>>>>>     Timing LSPs and dont perform any Timing operation (TC or BC).
>>>>>>     However as explained in QoS section the Timing over MPLS packets
>>> MUST
>>>>>>     be still be treated with the highest priority based on their Traffic
>>>>>>     Class (TC) marking.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 24]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 16.  Other considerations
>>>>>>
>>>>>>     [IEEE-1588] defines an optional peer-to-peer Transparent clocking
>>>>>>     that requires peer delay measurement between two adjacent Timing-
>>>>>>     capable/ aware routers/switches.  Peer delay measurement messages
>>>>>>     need to be time stamped and terminated by the Timing-capable/aware
>>>>>>     routers/ switches.  This means that two adjacent LSRs may be engaged
>>>>>>     in a peer delay measurement.
>>>>>>
>>>>>>     For transporting such peer delay measurement messages a single-hop
>>>>>>     LSP SHOULD to be created between the two adjacent LSRs engaged in
>>>>>>     peer delay measurement to carry peer delay measurement messages.
>>>>>>     Other methods such as PTP transport over Ethernet MAY be used for
>>>>>>     transporting peer delay measurement messages if the link between the
>>>>>>     two routers is Ethernet.
>>>>>>
>>>>>>     In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ ware
>>>>>>     routers/switches MUST maintain a list of all the neighbors it needs
>>>>>>     to send a PDelay_Req to, where each neighbor corresponds to a timing
>>>>>>     LSP.
>>>>>>
>>>>>>     The use of Explicit Null Label (Label= 0 or 2) is acceptable as long
>>>>>>     as either the Explicit Null label is the bottom of stack label
>>>>>>     (applicable only to UDP/IP encapsulation) or the label below the
>>>>>>     Explicit Null label is a PTP label.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 25]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 17.  Security Considerations
>>>>>>
>>>>>>     MPLS PW security considerations in general are discussed in [RFC3985]
>>>>>>     and [RFC4447],and those considerations also apply to this document.
>>>>>>
>>>>>>     An experimental security protocol is defined in [IEEE-1588].The PTP
>>>>>>     security extension and protocol provides group source authentication,
>>>>>>     message integrity, and replay attack protection for PTP messages.
>>>>>>
>>>>>>     When the MPLS network (provider network) serves multiple customers,
>>>>>>     it is important to maintain and process each customers clock and
>>>>>>     Timing messages separately from other customers to ensure there is no
>>>>>>     cross- customer effect.  For example if an LER BC is synchronized to
>>>>>>     a specific grandmaster, belonging to customer A, then the LER MUST
>>>>>>     use that BC clock only for customer A to ensure that customer A
>>>>>>     cannot attack other customers by manipulating its time.
>>>>>>
>>>>>>     Timing messages MAY be encrypted or authenticated, provided that the
>>>>>>     LERs/LSRs that are Timing capable/aware can authenticate/ decrypt the
>>>>>>     timing messages.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 26]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 18.  Acknowledgements
>>>>>>
>>>>>>     The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizrahi,
>>>>>>     Stefano Ruffini, Peter Meyer, and other members of IETF for reviewing
>>>>>>     and providing feedback on this draft.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 27]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 19.  IANA Considerations
>>>>>>
>>>>>>     There are no IANA requirements in this specification.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 28]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 20.  References
>>>>>>
>>>>>> 20.1.  Normative References
>>>>>>
>>>>>>     [IEEE-1588]
>>>>>>                IEEE 1588-2008, "IEEE Standard for a Precision Clock
>>>>>>                Synchronization Protocol for Networked Measurement and
>>>>>>                Control Systems".
>>>>>>
>>>>>>     [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>>>>>                Requirement Levels", BCP 14, RFC 2119, March 1997.
>>>>>>
>>>>>>     [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
>>>>>>                Edge (PWE3) Architecture", RFC 3985, March 2005.
>>>>>>
>>>>>>     [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
>>>>>>                Proxies (ND Proxy)", RFC 4389, April 2006.
>>>>>>
>>>>>>     [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
>>>>>>                Heron, "Pseudowire Setup and Maintenance Using the Label
>>>>>>                Distribution Protocol (LDP)", RFC 4447, April 2006.
>>>>>>
>>>>>>     [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>>>>>                "Encapsulation Methods for Transport of Ethernet over MPLS
>>>>>>                Networks", RFC 4448, April 2006.
>>>>>>
>>>>>>     [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>>>>>                Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>>>>>                Retention", RFC 4720, November 2006.
>>>>>>
>>>>>>     [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
>>>>>>                Connectivity Verification (VCCV): A Control Channel for
>>>>>>                Pseudowires", RFC 5085, December 2007.
>>>>>>
>>>>>>     [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detection
>>>>>>                (BFD)", RFC 5880, June 2010.
>>>>>>
>>>>>>     [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
>>>>>>                "Bidirectional Forwarding Detection (BFD) for MPLS Label
>>>>>>                Switched Paths (LSPs)", RFC 5884, June 2010.
>>>>>>
>>>>>> 20.2.  Informative References
>>>>>>
>>>>>>     [I-D.ietf-pwe3-fat-pw]
>>>>>>                Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>>>>>>                J., and S. Amante, "Flow Aware Transport of Pseudowires
>>>>>>                over an MPLS Packet Switched Network",
>>>>>>                draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 29]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>>     [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
>>>>>>                system routeing information exchange protocol for use in
>>>>>>                conjunction with the Protocol for providing the
>>>>>>                Connectionless-mode Network Service (ISO 8473)".
>>>>>>
>>>>>>     [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
>>>>>>                dual environments", RFC 1195, December 1990.
>>>>>>
>>>>>>     [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
>>>>>>
>>>>>>     [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
>>>>>>                Marker", RFC 2697, September 1999.
>>>>>>
>>>>>>     [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boudec,
>>>>>>                J., Courtney, W., Davari, S., Firoiu, V., and D.
>>>>>>                Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>>>>>                Behavior)", RFC 3246, March 2002.
>>>>>>
>>>>>>     [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineering
>>>>>>                (TE) Extensions to OSPF Version 2", RFC 3630,
>>>>>>                September 2003.
>>>>>>
>>>>>>     [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
>>>>>>                System (IS-IS) Extensions for Traffic Engineering (TE)",
>>>>>>                RFC 3784, June 2004.
>>>>>>
>>>>>>     [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
>>>>>>                Shaffer, "Extensions to OSPF for Advertising Optional
>>>>>>                Router Capabilities", RFC 4970, July 2007.
>>>>>>
>>>>>>     [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
>>>>>>                System to Intermediate System (IS-IS) Extensions for
>>>>>>                Advertising Router Information", RFC 4971, July 2007.
>>>>>>
>>>>>>     [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
>>>>>>                Topology (MT) Routing in Intermediate System to
>>>>>>                Intermediate Systems (IS-ISs)", RFC 5120, February 2008.
>>>>>>
>>>>>>     [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>>>>>                Engineering", RFC 5305, October 2008.
>>>>>>
>>>>>>     [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>>>>>                "Traffic Engineering Extensions to OSPF Version 3",
>>>>>>                RFC 5329, September 2008.
>>>>>>
>>>>>>     [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
>>>>>>                for IPv6", RFC 5340, July 2008.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 30]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>>     [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network
>>>>>>                Time Protocol Version 4: Protocol and Algorithms
>>>>>>                Specification", RFC 5905, June 2010.
>>>>>>
>>>>>>     [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>>>>>>                J., and S. Amante, "Flow-Aware Transport of Pseudowires
>>>>>>                over an MPLS Packet Switched Network", RFC 6391,
>>>>>>                November 2011.
>>>>>>
>>>>>>     [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
>>>>>>                L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
>>>>>>                RFC 6790, November 2012.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 31]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 1.  Routing extensions for Timing-aware Routers
>>>>>>
>>>>>>     MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] and
>>>>>>     IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (TE)
>>>>>>     link information used for constraint-based routing.
>>>>>>
>>>>>>     Indeed, it is useful to advertise data plane TE router link
>>>>>>     capabilities, such as the capability for a router to be Timing-aware.
>>>>>>     This capability MUST then be taken into account during path
>>>>>>     computation to prefer or even require links that advertise themselves
>>>>>>     as Timing-aware.  In this way the path can ensure the entry and exit
>>>>>>     points into the LERs and, if desired, the links into the LSRs are
>>>>>>     able to perform port based time-stamping thus minimizing their impact
>>>>>>     on the performance of the slave clock.
>>>>>>
>>>>>>     extensions are required to OSPF and IS-IS in order to advertise
>>>>>>     Timing-aware capabilities of a link.  Such extensions are outside the
>>>>>>     scope of this document; however such extension SHOULD be able to
>>>>>>     signal the following information per Router Link:
>>>>>>
>>>>>>     o  Capable of processing PTP, NTP or other Timing flows
>>>>>>
>>>>>>     o  Capable of performing Transparent Clock operation
>>>>>>
>>>>>>     o  Capable of performing Boundary Clock operation
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 32]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> 2.  Signaling Extensions for Creating Timing LSPs
>>>>>>
>>>>>>     RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-TE
>>>>>>     is used to setup Timing LSPs, some information that indicates that
>>>>>>     the LSP is carrying Timing flows MUST be included in the new
>>>>>>     Extensions to RSVP-TE:
>>>>>>
>>>>>>     The following information MAY also be included in the new Extensions
>>>>>>     to RSVP-TE:
>>>>>>
>>>>>>     o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
>>>>>>        field
>>>>>>
>>>>>>     o  Number of VLANs in case of PW encapsulation
>>>>>>
>>>>>>     o  Timestamp field Type
>>>>>>
>>>>>>        *  Correction Field, Timestamp
>>>>>>
>>>>>>     o  Timestamp Field format
>>>>>>
>>>>>>        *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
>>>>>>           NTP, etc.
>>>>>>
>>>>>>     Note that in case the above optional information is signaled with
>>>>>>     RSVP-TE for a Timing LSP, all the Timing packets carried in that LSP
>>>>>>     must have the same signaled characteristics.  For example if
>>>>>>     Timestamp format is signaled as 64-bit PTPv1, then all Timing packets
>>>>>>     must use 64-bit PTPv1 time-stamp.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 33]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>> Authors' Addresses
>>>>>>
>>>>>>     Shahram Davari
>>>>>>     Broadcom Corp.
>>>>>>     San Jose, CA  95134
>>>>>>     USA
>>>>>>
>>>>>>     Email: davari@broadcom.com
>>>>>>
>>>>>>
>>>>>>     Amit Oren
>>>>>>     Broadcom Corp.
>>>>>>     San Jose, CA  95134
>>>>>>     USA
>>>>>>
>>>>>>     Email: amito@broadcom.com
>>>>>>
>>>>>>
>>>>>>     Manav Bhatia
>>>>>>     Alcatel-Lucent
>>>>>>     Bangalore,
>>>>>>     India
>>>>>>
>>>>>>     Email: manav.bhatia@alcatel-lucent.com
>>>>>>
>>>>>>
>>>>>>     Peter Roberts
>>>>>>     Alcatel-Lucent
>>>>>>     Kanata,
>>>>>>     Canada
>>>>>>
>>>>>>     Email: peter.roberts@alcatel-lucent.com
>>>>>>
>>>>>>
>>>>>>     Laurent Montini
>>>>>>     Cisco Systems
>>>>>>     San Jose CA
>>>>>>     USA
>>>>>>
>>>>>>     Email: lmontini@cisco.com
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 34]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>
>>>>>>
>>>>>>     Luca
>>>>>>     Cisco Systems
>>>>>>     San Jose CA
>>>>>>     USA
>>>>>>
>>>>>>     Email: lmartini@cisco.com
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page 35]
>>>>>> --
>>>>>> For corporate legal information go to:
>>>>>>
>>>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>>>
>>>>>> _______________________________________________
>>>>>> TICTOC mailing list
>>>>>> TICTOC@ietf.org
>>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>> This e-mail message is intended for the recipient only and contains
>>> information which is CONFIDENTIAL and which may be proprietary to ECI
>>> Telecom. If you have received this transmission in error, please inform us by e-
>>> mail, phone or fax, and then delete the original and all copies thereof.
>>>>> _______________________________________________
>>>>> mpls mailing list
>>>>> mpls@ietf.org
>>>>> https://www.ietf.org/mailman/listinfo/mpls
>>>> .
>>>
>>> --
>>> For corporate legal information go to:
>>>
>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>
>>> _______________________________________________
>>> TICTOC mailing list
>>> TICTOC@ietf.org
>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>
>>>
>>> _______________________________________________
>>> TICTOC mailing list
>>> TICTOC@ietf.org
>>> https://www.ietf.org/mailman/listinfo/tictoc
>>
>> This e-mail message is intended for the recipient only and contains information which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If you have received this transmission in error, please inform us by e-mail, phone or fax, and then delete the original and all copies thereof.
>>
> .
>


-- 
For corporate legal information go to:

http://www.cisco.com/web/about/doing_business/legal/cri/index.html


From davarish@yahoo.com  Tue Aug 20 09:50:07 2013
Return-Path: <davarish@yahoo.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id AC91511E8251 for <tictoc@ietfa.amsl.com>; Tue, 20 Aug 2013 09:50:07 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.203
X-Spam-Level: 
X-Spam-Status: No, score=-1.203 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, MIME_QP_LONG_LINE=1.396]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id g38Jm6zc2mVr for <tictoc@ietfa.amsl.com>; Tue, 20 Aug 2013 09:50:02 -0700 (PDT)
Received: from nm43-vm5.bullet.mail.bf1.yahoo.com (nm43-vm5.bullet.mail.bf1.yahoo.com [216.109.114.236]) by ietfa.amsl.com (Postfix) with ESMTP id 33EBA11E8260 for <tictoc@ietf.org>; Tue, 20 Aug 2013 09:49:46 -0700 (PDT)
Received: from [98.139.212.150] by nm43.bullet.mail.bf1.yahoo.com with NNFMP; 20 Aug 2013 16:49:40 -0000
Received: from [98.139.211.192] by tm7.bullet.mail.bf1.yahoo.com with NNFMP; 20 Aug 2013 16:49:40 -0000
Received: from [127.0.0.1] by smtp201.mail.bf1.yahoo.com with NNFMP; 20 Aug 2013 16:49:40 -0000
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yahoo.com; s=s1024; t=1377017380; bh=gQcetVzAzrB5HjM0s50W+cCFVDk/x189lM6m53hK5kU=; h=X-Yahoo-Newman-Id:X-Yahoo-Newman-Property:X-YMail-OSG:X-Yahoo-SMTP:X-Rocket-Received:References:Mime-Version:In-Reply-To:Content-Type:Content-Transfer-Encoding:Message-Id:Cc:X-Mailer:From:Subject:Date:To; b=oXoNeVagRu3M3VoLkW9a62esCE8FbTLc/xkw7Xabx6UksLgHCdIMg8y1c4dw76HrnW31j1SKwXKz02p3J41OnliYYbvoIv4P/rDhGO+TrM5J8lg8r4cQT8eEOC0qDJcMOCfcF5YG7ONjoP5rygvdEE2EmaZWJvJaf0cuRwLI/To=
X-Yahoo-Newman-Id: 597071.38204.bm@smtp201.mail.bf1.yahoo.com
X-Yahoo-Newman-Property: ymail-3
X-YMail-OSG: nQ.vlFMVM1nCrFqlgKNSmFwGxmFIdLIw73xQzeJnseZnYmx vrBcxsLEHr.BoDL._Ppr577pNxwEi705HXSU6UQuHUhGc.yQxNxMpxvKN2dM BVXTalyCnrtY.xjMaj.PRFjCuOtZk.G1ldDqCKvip6I0uHXyvJ68pKp6z2SP szoLs63Lfq_hxQW0wLqi69dzWVha0ZDCvpCOGHsmlDM94YQL0Oy3zmE7GJx1 0g8beTCc.C8yTO_wa_BQHstsIZhN7BdHB0nwg6m.9Pl8bYVQeXhHBqj.a3X9 wv6Y1hRNYNcTwXQT5yZ_UBOcXUYwdqEQKxQoPICwkzKHbcIwhmPTTBj6mmmU E2sZJ6e.rKJ2FZ1Aoy8WAjmmWYNY6uLZuBxm9Ux0zFHL82rsh9qapUb7mazL I877NWsAA18CmecvYKoifsUC1hjJYWKpg87EFB3yXoMSfz02QSIlt6S5q7_v hODSP25cM7jrVHukCGCIkQjMLno1BkY95pdqVIGB5tgLd1T7ih_GWnZhY1b2 8HkK79TSCsO9Ip3q0fOdNt_znVsqtWNdvncVWlnNAdA71Y_SF
X-Yahoo-SMTP: ygPrP9CswBCWPbPtKJlJyLY0KMlg
X-Rocket-Received: from [192.168.0.103] (davarish@98.248.42.143 with ) by smtp201.mail.bf1.yahoo.com with SMTP; 20 Aug 2013 09:49:40 -0700 PDT
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com> <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com> <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com> <52139903.8040801@cisco.com>
Mime-Version: 1.0 (1.0)
In-Reply-To: <52139903.8040801@cisco.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: quoted-printable
Message-Id: <9A8CF696-5D44-4C6C-925B-2DC670374671@yahoo.com>
X-Mailer: iPhone Mail (10A403)
From: "S. Davari" <davarish@yahoo.com>
Date: Tue, 20 Aug 2013 09:49:31 -0700
To: "stbryant@cisco.com" <stbryant@cisco.com>
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 20 Aug 2013 16:50:07 -0000

Hi Stewart,

Agree that can be done, but not with a reserved label on top if the stack.

Regards,
Shahram


On Aug 20, 2013, at 9:27 AM, Stewart Bryant <stbryant@cisco.com> wrote:

> On 06/08/2013 14:35, S. Davari wrote:
>> Sasha,
>>=20
>> If a router is not 1588 aware we want it to switch the packet with minimu=
m jitter and delay via high priority queue.
> Yes, but you can explicitly tunnel across such island of incompatibility
>=20
> Stewart
>=20
>> =20
>> Sending it to CPU creates large jitter and delay since the CPUs are gener=
ally busy doing other things. Time stamping or updating CF in the CPU won't h=
elp since it does not cover the variable delay that is incurred from the tim=
e the packet enters the router to the time the packet gets processed by CPU.=

>>=20
>> I wish it was that easy!
>>=20
>> Regards,
>> Shahram
>>=20
>>=20
>> On Aug 6, 2013, at 12:38 AM, Alexander Vainshtein <Alexander.Vainshtein@e=
citele.com> wrote:
>>=20
>>> Shahram,
>>> I agree with you that the routers that do not recognize a Timing Alert l=
abel would send the packet to the CPU. (This would be easy to arrange since w=
e are speaking about a reserved label.)
>>>=20
>>> I do not, however, see this as a serious issue, because the SW running o=
n this CPU would (hopefully) be upgradable to handle the packet correctly i.=
e., to forward it to where it should be forwarded and, if it is forwarded as=
 a labeled packet, to prepend the Timing Alert Label on top of its label sta=
ck. It could even record the residence time (as observed by the CPU in the p=
roper place in the packet. This would mean that introducing additional error=
 to whatever timing information is associated with this packet - but this is=
 what you should anyway expect if there are non-compliant routers on your pa=
th, right?
>>>=20
>>> Regards,
>>>     Sasha
>>>=20
>>>> -----Original Message-----
>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behal=
f Of
>>>> Shahram Davari
>>>> Sent: Monday, August 05, 2013 11:29 PM
>>>> To: stbryant@cisco.com; S. Davari
>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls=

>>>>=20
>>>> Hi Stewart,
>>>>=20
>>>> Your suggestion of " I suggested an LSP type that has the properties "t=
imestamp
>>>> and pass to application", is a subset of the existing draft and should w=
ork. It
>>>> basically limits the time stamping and correction field update to LERs (=
while the
>>>> draft supports time stamping at LER and LSR).
>>>>=20
>>>> However Sasha's suggestion of using a reserved label (RAL, GAL or any o=
ther
>>>> reserved label) does not satisfy one of the major requirements, which i=
s
>>>> backward compatibility. Routers that don't understand this reserved lab=
el will
>>>> drop or copy to CPU such packets.
>>>>=20
>>>> Regards,
>>>> Shahram
>>>>=20
>>>> -----Original Message-----
>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behal=
f Of
>>>> Stewart Bryant
>>>> Sent: Monday, August 05, 2013 2:29 AM
>>>> To: S. Davari
>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls=

>>>>=20
>>>> My concern is that we are creating a new LSP type in MPLS
>>>> (which is a architectural change and thus needs to be very carefully
>>>> considered) without asking the question "how can I do this
>>>> in the most general way, to maximize flexibility/reuse"
>>>>=20
>>>> I suggested an LSP type that has the properties "timestamp
>>>> and pass to application", but I think that Sasha raises a good
>>>> point about using router alert. However RA has no implicit
>>>> timestamp, so maybe we need a new type of RA that has the
>>>> properties "timestamp, and pass top application indicated by
>>>> the next label". That would be quite useful in a number
>>>> of OAM applications. There is possibly some GAL variant
>>>> of that design that should also be considered.
>>>>=20
>>>> The application could worry about the path and where it
>>>> was necessary to skip some hops a hierarchical LSP would
>>>> accomplish that, with the specific benefit that the application
>>>> would consciously do this, and may be able to apply some
>>>> form of compensation within the network.
>>>>=20
>>>> - Stewart
>>>>=20
>>>> On 04/08/2013 10:40, S. Davari wrote:
>>>>> Hi Sasha
>>>>>=20
>>>>> Perhaps you have not understood the draft well. The main reason that t=
he
>>>> draft chose to use a dedicated LSP that is signaled via RSVPTE indicati=
ng it is a
>>>> timing LSP is to be backward compatible with non-1588 nodes. Such nodes=

>>>> simply just switch the packet.
>>>>> Your proposal had been considered and was rejected since it was not
>>>> backward compatible. Nodes receiving alert label or TTL =3D 1 send the p=
acket to
>>>> CPU. You can refer to the meeting notes.
>>>>> Regards,
>>>>> Shahram
>>>>>=20
>>>>>=20
>>>>> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein
>>>> <Alexander.Vainshtein@ecitele.com> wrote:
>>>>>> Stewart and all,
>>>>>> I concur with Stewart's statement that the draft does not define "ful=
l
>>>> interaction with MPLS architecture".
>>>>>> E.g., one of the objectives of the draft is to provide a technique th=
at would
>>>> be backward-compatible with old LSRs that cannot provide on-path suppor=
t for
>>>> timing distribution, while the other objective is to make every LSR on t=
he path
>>>> aware that some MPLS packets are carrying timing-related messages and h=
ence
>>>> require on-path support.
>>>>>> IMHO and FWIW, the MPLS data plane architecture allows just two metho=
ds
>>>> for making a transit LSR to provide special processing to a labeled pac=
ket:
>>>>>> - It would carry some kind of an "alert label" on top of the label st=
ack and,
>>>> specifically, on top of any labels used for actual forwarding
>>>>>> OR,
>>>>>> - The TTL in the top label stack entry has been set to 1.
>>>>>>=20
>>>>>>=20
>>>>>> The draft does not follow any of these approaches.
>>>>>>=20
>>>>>> I must also admit that Section 12 "OAM, Control and Management" of th=
e
>>>> draft looks somewhat in
>>>>>> E.g., I could not understand from the text whether BFD, LSP-Ping etc.=
 OAM
>>>> messages would be subjected to on-path support procedures for timing
>>>> messages or not.
>>>>>> My 2c,
>>>>>>     Sasha
>>>>>>=20
>>>>>>> -----Original Message-----
>>>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Be=
half
>>>> Of
>>>>>>> Stewart Bryant
>>>>>>> Sent: Friday, August 02, 2013 12:01 PM
>>>>>>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
>>>>>>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.o=
rg;
>>>> draft-
>>>>>>> ietf-tictoc-1588overmpls@tools.ietf.org
>>>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>>>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
>>>>>>>=20
>>>>>>> SB> This draft does not seem to provide a precise definition
>>>>>>> SB> the properties of the new LSP type that it wishes to
>>>>>>> SB> define, in particular it does it define the PHB of those
>>>>>>> SB> LSPs, nor the full interaction with the MPLS
>>>>>>> SB> architecture.
>>>>>>> SB>
>>>>>>> SB> I have not tracked TICTOC for a while but I thought that
>>>>>>> SB> the original plan was to define the concept of an offset
>>>>>>> SB> into a packet to do the correction.
>>>>>>> SB>
>>>>>>> SB> It is disappointing that the opportunity was not taken
>>>>>>> SB> to define a timing shim inside the timing LSP so that
>>>>>>> SB> a time correction could be added to any packet such that
>>>>>>> SB> the MPLS system was isolated from the details of the
>>>>>>> SB> complexity of the particular time transfer type.
>>>>>>>=20
>>>>>>> SB> I think that much more clarify is needed in terms of
>>>>>>> SB> definition of the new LSP type, since it is unclear
>>>>>>> SB> from this text how to implement one.
>>>>>>> SB>
>>>>>>> SB> There are a lot of other MPLS services such as
>>>>>>> SB> LSP ping that need to be considered.
>>>>>>> SB>
>>>>>>> SB> Please see inline for more comments. However these
>>>>>>> SB> comments are made in the context of the text as written
>>>>>>> SB> whilst I have a fundamental concern that this approach
>>>>>>> SB> lacks an MPLS architectural soundness that need
>>>>>>> SB> greater thought with significant impact on the
>>>>>>> SB> draft.
>>>>>>>=20
>>>>>>> - Stewart
>>>>>>>=20
>>>>>>>=20
>>>>>>> TICTOC Working Group                                           S. Da=
vari
>>>>>>> Internet-Draft                                                   A. O=
ren
>>>>>>> Intended status: Standards Track                          Broadcom C=
orp.
>>>>>>> Expires: December 17, 2013                                     M. Bh=
atia
>>>>>>>                                                               P. Rob=
erts
>>>>>>>                                                           Alcatel-Lu=
cent
>>>>>>>                                                               L. Mon=
tini
>>>>>>>                                                               L. Mar=
tini
>>>>>>>                                                            Cisco Sys=
tems
>>>>>>>                                                            June 15, 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>             Transporting Timing messages over MPLS Networks
>>>>>>>                    draft-ietf-tictoc-1588overmpls-05
>>>>>>>=20
>>>>>>> Abstract
>>>>>>>=20
>>>>>>>    This document defines the method for transporting Timing messages=

>>>>>>>    such as PTP and NTP over an MPLS network.  The method allows for t=
he
>>>>>>>    easy identification of these PDUs at the port level to allow for p=
ort
>>>>>>>=20
>>>>>>> SB> What is a port
>>>>>>>=20
>>>>>>>    level processing of these PDUs in both LERs and LSRs.
>>>>>>>=20
>>>>>>>    The basic idea is to transport Timing messages inside dedicated M=
PLS
>>>>>>>    LSPs.  These LSPs only carry Timing messages and possibly Control=
 and
>>>>>>>    Management packets, but they do not carry customer traffic.
>>>>>>>=20
>>>>>>> SB> More specifically they only carry traffic associated with the
>>>>>>> SB> timing service and its support.
>>>>>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
>>>>>>> SB> also it gets carried in a structure that causes it to get
>>>>>>> SB> timestamped.
>>>>>>>=20
>>>>>>>    Two methods for transporting Timing messages over MPLS are define=
d.
>>>>>>>=20
>>>>>>> SB> Perhaps the right approach is to define the new LSP type and the=
n
>>>>>>> SB> seperately to define  the mapping of the various timing services=

>>>>>>> SB> over that LSP type.
>>>>>>>=20
>>>>>>>    The first method is to transport Timing messages directly over th=
e
>>>>>>>    dedicated MPLS LSP via UDP/IP encapsulation, which is suitable fo=
r
>>>>>>>    MPLS networks.  The second method is to transport Timing messages=

>>>>>>>    inside a PW via Ethernet encapsulation.
>>>>>>>=20
>>>>>>> SB> I think that we should note that there are some
>>>>>>> SB> h/w reasons for this preference. A clean sheet approach
>>>>>>> SB> would have been to use PTP over MPLS with no intermediate
>>>>>>> SB> layers.
>>>>>>>=20
>>>>>>> Status of this Memo
>>>>>>>=20
>>>>>>>    This Internet-Draft is submitted in full conformance with the
>>>>>>>    provisions of BCP 78 and BCP 79.
>>>>>>>=20
>>>>>>>    Internet-Drafts are working documents of the Internet Engineering=

>>>>>>>    Task Force (IETF).  Note that other groups may also distribute
>>>>>>>    working documents as Internet-Drafts.  The list of current Intern=
et-
>>>>>>>    Drafts is at http://datatracker.ietf.org/drafts/current/.
>>>>>>>=20
>>>>>>>    Internet-Drafts are draft documents valid for a maximum of six mo=
nths
>>>>>>>    and may be updated, replaced, or obsoleted by other documents at a=
ny
>>>>>>>    time.  It is inappropriate to use Internet-Drafts as reference
>>>>>>>    material or to cite them other than as "work in progress."
>>>>>>>=20
>>>>>>>    This Internet-Draft will expire on December 17, 2013.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 1]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> Copyright Notice
>>>>>>>=20
>>>>>>>    Copyright (c) 2013 IETF Trust and the persons identified as the
>>>>>>>    document authors.  All rights reserved.
>>>>>>>=20
>>>>>>>    This document is subject to BCP 78 and the IETF Trust's Legal
>>>>>>>    Provisions Relating to IETF Documents
>>>>>>>    (http://trustee.ietf.org/license-info) in effect on the date of
>>>>>>>    publication of this document.  Please review these documents
>>>>>>>    carefully, as they describe your rights and restrictions with res=
pect
>>>>>>>    to this document.  Code Components extracted from this document m=
ust
>>>>>>>    include Simplified BSD License text as described in Section 4.e o=
f
>>>>>>>    the Trust Legal Provisions and are provided without warranty as
>>>>>>>    described in the Simplified BSD License.
>>>>>>>=20
>>>>>>>=20
>>>>>>> Table of Contents
>>>>>>>=20
>>>>>>>    1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .=
  5
>>>>>>>=20
>>>>>>>    2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .=
  7
>>>>>>>=20
>>>>>>>    3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .=
  8
>>>>>>>=20
>>>>>>>    4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .=
  9
>>>>>>>=20
>>>>>>>    5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . .=
 12
>>>>>>>=20
>>>>>>>    6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . .=
 13
>>>>>>>      6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . .=
 13
>>>>>>>      6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . .=
 13
>>>>>>>      6.3.  Other Timing Encapsulation methods . . . . . . . . . . . .=
 14
>>>>>>>=20
>>>>>>>    7.  Timing message Processing  . . . . . . . . . . . . . . . . . .=
 15
>>>>>>>=20
>>>>>>>    8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . .=
 16
>>>>>>>=20
>>>>>>>    9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 17
>>>>>>>=20
>>>>>>>    10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 18
>>>>>>>=20
>>>>>>>    11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 19
>>>>>>>=20
>>>>>>>    12. OAM, Control and Management  . . . . . . . . . . . . . . . . .=
 20
>>>>>>>=20
>>>>>>>    13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . .=
 21
>>>>>>>=20
>>>>>>>    14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . .=
 22
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 2]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . .=
 23
>>>>>>>      15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . .=
 23
>>>>>>>      15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . .=
 23
>>>>>>>      15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . .=
 24
>>>>>>>=20
>>>>>>>    16. Other considerations . . . . . . . . . . . . . . . . . . . . .=
 25
>>>>>>>=20
>>>>>>>    17. Security Considerations  . . . . . . . . . . . . . . . . . . .=
 26
>>>>>>>=20
>>>>>>>    18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . .=
 27
>>>>>>>=20
>>>>>>>    19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . .=
 28
>>>>>>>=20
>>>>>>>    20. References . . . . . . . . . . . . . . . . . . . . . . . . . .=
 29
>>>>>>>      20.1. Normative References . . . . . . . . . . . . . . . . . . .=
 29
>>>>>>>      20.2. Informative References . . . . . . . . . . . . . . . . . .=
 29
>>>>>>>=20
>>>>>>>    Appendix 1.  Routing extensions for Timing-aware Routers . . . . .=
 32
>>>>>>>=20
>>>>>>>    Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . .=
 33
>>>>>>>=20
>>>>>>>    Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . .=
 34
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 3]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
>>>> NOT",
>>>>>>>    "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL"
>>>> in
>>>>>>> this
>>>>>>>    document are to be interpreted as described in RFC2119 [RFC2119].=

>>>>>>>=20
>>>>>>>    When used in lower case, these words convey their typical use in
>>>>>>>    common language, and are not to be interpreted as described in
>>>>>>>    RFC2119 [RFC2119].
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 4]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 1.  Introduction
>>>>>>>=20
>>>>>>>    The objective of Precision Time Protocol (PTP) and Network Timing=

>>>>>>>    Protocol (NTP) are to synchronize independent clocks running on
>>>>>>>    separate nodes of a distributed system.
>>>>>>>=20
>>>>>>>    [IEEE-1588] defines PTP messages for frequency, phase and time
>>>>>>>    synchronization.  The PTP messages include PTP PDUs over UDP/IP
>>>>>>>    (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex =
F of
>>>>>>>    [IEEE-1588]).
>>>>>>>=20
>>>>>>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
>>>>>>> SB> of other PTP mappings if they provide better optimisation.
>>>>>>>=20
>>>>>>>    This document defines mapping and transport of the PTP
>>>>>>>    messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>>>>>>>    defines several clock types: ordinary clocks, boundary clocks, en=
d-
>>>>>>>    to-end transparent clocks, and peer-to-peer transparent clocks.
>>>>>>>    Transparent clocks require intermediate nodes to update correctio=
n
>>>>>>>    field inside PTP message that reflects the transit time in the no=
de.
>>>>>>>=20
>>>>>>>    [RFC5905] defines NTP messages for clock and time synchronization=
.
>>>>>>>    The PTP messages (PDUs) are transported over UDP/IP.  This docume=
nt
>>>>>>> SB> Should that be NTP messages?
>>>>>>> SB> It needs to be made clear as soon as you introduce NTP that
>>>>>>> SB> they use different time representations.
>>>>>>>=20
>>>>>>>    defines mapping and transport of the NTP messages defined in
>>>>>>>    [RFC5905] over MPLS networks.
>>>>>>>=20
>>>>>>>    One key attribute of all of these Timing messages is that the Tim=
e
>>>>>>>    stamp processing should occur as close as possible to the actual
>>>>>>>    transmission and reception at the physical port interface.  This
>>>>>>>    targets optimal time and/or frequency recovery by avoiding variab=
le
>>>>>>>    delay introduced by queues internal to the clocks.
>>>>>>>=20
>>>>>>> SB> As I recall NTP has no epoch point defined, and I am not sure
>>>>>>> SB> where that point is in the case of PTP in this mapping
>>>>>>> SB> Hopefully this will get defined in due course.
>>>>>>>=20
>>>>>>>    To facilitate the fast and efficient recognition of Timing messag=
es
>>>>>>>    at the port level when the Timing messages are carried over MPLS
>>>>>>>    LSPs,
>>>>>>>=20
>>>>>>> SB> Over a new LSP type with time optimied characteristics
>>>>>>>=20
>>>>>>>    this document defines the specific encapsulations that should
>>>>>>>    be used.
>>>>>>> SB> Hopefully it will also define the PHP
>>>>>>>=20
>>>>>>>    In addition, it can be expected that there will exist LSR/
>>>>>>>    LERs where only a subset of the physical ports will have the port=
-
>>>>>>>    based Timing message processing capabilities.
>>>>>>> SB> Do you need to clarify that this only works at base and not in
>>>>>>> SB> a label heirarchy.
>>>>>>>=20
>>>>>>>=20
>>>>>>>    In order to ensure
>>>>>>>    that the LSPs carrying Timing packets always enter and exit ports=

>>>>>>>    with this capability, routing extensions are defined to advertise=

>>>>>>>    this capability on a port basis and to allow for the establishmen=
t of
>>>>>>>    LSPs that only transit such ports.  While this path establishment=

>>>>>>>    restriction may be applied only at the LER Ingress and/or egress
>>>>>>>    ports, it becomes more important when using transparent clock cap=
able
>>>>>>>    LSRs in the path.
>>>>>>> SB> I do not understand the implications of the last
>>>>>>> SB> sentences - starting ", it becomes"
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Port based Timing message processing involves Timing message
>>>>>>>    recognition.  Once the Timing messages are recognized they can be=

>>>>>>>    modified based on the reception or transmission Time-stamp.
>>>>>>>=20
>>>>>>>    This document provides two methods for transporting Timing messag=
es
>>>>>>>    over MPLS.  One is applicable to MPLS environment and the other o=
ne
>>>>>>>    is applicable to MPLS/MPLS-TP environment
>>>>>>>=20
>>>>>>> SB> I think the sentence is incomplete.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 5]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    The solution involves transporting Timing messages over dedicated=

>>>>>>>    LSPs called Timing LSPs.  These LSPs carry Timing messages and MA=
Y
>>>>>>>    carry Management and control messages, but not data plane client
>>>>>>>    traffic.
>>>>>>>=20
>>>>>>> SB> It is not clear why this restriction applies.
>>>>>>>=20
>>>>>>>    Timing LSPs can be established statically or via signaling.
>>>>>>> SB> s/statically/by provisioning/network management/
>>>>>>>=20
>>>>>>>    Extensions to control plane (OSPF, ISIS, etc.) is required to ena=
ble
>>>>>>>    routers to distribute their Timing processing capabilities over M=
PLS
>>>>>>>    to other routers.  However such extensions are outside the scope o=
f
>>>>>>>    this document.
>>>>>>>=20
>>>>>>>    When signaling is used to setup the PTP LSP, Extensions to signal=
ing
>>>>>>> SB> is it a PTP LSP or a Timing LSP?
>>>>>>>=20
>>>>>>>    protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.=

>>>>>>>    However such extensions are outside the scope of this document.
>>>>>>>=20
>>>>>>> SB> for mpls-tp GMPLS is the signalling protocol
>>>>>>>=20
>>>>>>>    While the techniques included herein allow for the establishment o=
f
>>>>>>>    paths optimized to include Time-stamping capable links, the
>>>>>>>    performance of the Slave clocks is outside the scope of this
>>>>>>>    document.
>>>>>>>=20
>>>>>>>    At the time of publishing this specification, Transparent Clockin=
g
>>>>>>>    (TC) is only defined for PTP.  Therefore at this time any part of=

>>>>>>>    this specification that talks about Transparent Clocking applies o=
nly
>>>>>>>    to PTP.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 6]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 2.  Terminology
>>>>>>>=20
>>>>>>>    1588: The timing and synchronization as defined by IEEE 1588.
>>>>>>>=20
>>>>>>> SB> I think that there is a more formal name for the 1588 group
>>>>>>> SB> that needs to be used here.
>>>>>>> SB> Also do we need to talk about 1588-200? as there is
>>>>>>> SB> an update in progress
>>>>>>>=20
>>>>>>>    NTP: The timing and synchronization protocol defined by IETF RFC-=
1305
>>>>>>>    and RFC-5905.
>>>>>>>=20
>>>>>>>    PTP: The timing and synchronization protocol used by 1588.
>>>>>>> SB> need the proper name for 1588
>>>>>>>=20
>>>>>>>    Master Clock: The source of 1588 timing to a set of slave clocks.=

>>>>>>>=20
>>>>>>>    Master Port: A port on a ordinary or boundary clock that is in Ma=
ster
>>>>>>>    state.  This is the source of timing toward slave ports.
>>>>>>>=20
>>>>>>> SB> I am not sure the reader knows what a port is
>>>>>>>=20
>>>>>>>    Slave Clock: A receiver of 1588 timing from a master clock.
>>>>>>>=20
>>>>>>>    Slave Port: A port on a boundary clock or ordinary clock that is
>>>>>>>    receiving timing from a master clock.
>>>>>>>=20
>>>>>>>    Ordinary Clock: A device with a single PTP port.
>>>>>>>=20
>>>>>>>    Transparent Clock.  A device that measures the time taken for a P=
TP
>>>>>>>    event message to transit the device and then updates the
>>>>>>>    correctionField of the message with this transit time.
>>>>>>>=20
>>>>>>>    Boundary Clock: A device with more than one PTP port.  Generally
>>>>>>>    boundary clocks will have one port in slave state to receive timi=
ng
>>>>>>>    and then other ports in master state to re-distribute the timing.=

>>>>>>>=20
>>>>>>>    PTP LSP: An LSP dedicated to carry PTP messages
>>>>>>>=20
>>>>>>> SB> PTP or timing?
>>>>>>>=20
>>>>>>>=20
>>>>>>>    PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>>>>>>    messages.
>>>>>>>=20
>>>>>>> SB> Ah I don't think that PWE3 know what one of these is
>>>>>>>=20
>>>>>>>    CW: Pseudowire Control Word
>>>>>>>=20
>>>>>>>    LAG: Link Aggregation
>>>>>>>=20
>>>>>>>    ECMP: Equal Cost Multipath
>>>>>>>=20
>>>>>>>    CF: Correction Field, a field inside certain PTP messages (messag=
e
>>>>>>>    type 0-3)that holds the accumulative transit time inside intermed=
iate
>>>>>>>    switches
>>>>>>>=20
>>>>>>>    Timing messages: Timing Protocol messages that are exchanged
>>>> between
>>>>>>>    routers in order to establish a synchronized clock.
>>>>>>>=20
>>>>>>> SB> A number of these definitions look like copies of IEEE1588
>>>>>>> SB> definitions. We need to provide references and note the
>>>>>>> SB> priority of the IEEE base reference.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 7]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 3.  Problem Statement
>>>>>>>=20
>>>>>>>    [IEEE-1588] has defined methods for transporting PTP messages ove=
r
>>>>>>>    Ethernet and IP networks.  [RFC5905] has defined the method of
>>>>>>>    transporting NTP messages over IP networks.  There is a need to
>>>>>>>    transport Timing messages over MPLS networks while supporting the=

>>>>>>>    Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (O=
C)
>>>>>>>    functionality in the LER and LSRs in the MPLS network.
>>>>>>>=20
>>>>>>>    There are multiple ways of transporting Timing over MPLS.  Howeve=
r,
>>>>>>>    there is a requirement to limit the possible encapsulation option=
s to
>>>>>>>    simplify the Timing message identification and processing require=
d at
>>>>>>>    the port level.
>>>>>>>=20
>>>>>>>    When Timing-awareness is needed, Timing messages should not be
>>>>>>>    transported over LSPs or PWs that are carrying customer traffic
>>>>>>>    because LSRs perform Label switching based on the top label in th=
e
>>>>>>>    stack.
>>>>>>>=20
>>>>>>> SB> Have you explained why?
>>>>>>>=20
>>>>>>>    To detect Timing messages inside such LSPs require special
>>>>>>>    hardware to do deep packet inspection at line rate.  Even if such=

>>>>>>>    hardware exists, the payload can't be deterministically identifie=
d by
>>>>>>>    LSRs because the payload type is a context of the PW label, and t=
he
>>>>>>>    PW label and its context are only known to the Edge routers (PEs/=

>>>>>>>    LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, C=
ES,
>>>>>>>    etc).  Even if one restricts an LSP to only carry Ethernet PWs, t=
he
>>>>>>>    LSRs dont have the knowledge of whether PW Control Word (CW) is
>>>>>>>    present or not and therefore can not deterministically identify t=
he
>>>>>>>    payload.
>>>>>>>=20
>>>>>>>    A generic method is defined in this document that does not requir=
e
>>>>>>>    deep packet inspection at line rate, and can deterministically
>>>>>>>    identify Timing messages.  This method can be used to detect Timi=
ng
>>>>>>>    Messages in both one-step and two-step clock implementations of
>>>>>>>    ordinary, boundary and transparent clocks.
>>>>>>>=20
>>>>>>> SB> Needs a ref and I am sure many MPLS specialists will not underst=
and
>>>>>>> SB> the msg types.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 8]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 4.  Timing over MPLS Architecture
>>>>>>>=20
>>>>>>>    Timing messages are exchange between Timing ports on ordinary and=

>>>>>>>=20
>>>>>>> SB> Have you defined a timing port?
>>>>>>>=20
>>>>>>>    boundary clocks.  Boundary clocks terminate the Timing messages a=
nd
>>>>>>>    act as master for other boundary clocks or for slave clocks.  End=
-to-
>>>>>>>    End Transparent clocks do not terminate the Timing messages but t=
hey
>>>>>>>    do modify the contents of the Timing messages as they transit acr=
oss
>>>>>>>    the transparent clock.
>>>>>>>=20
>>>>>>>    Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent C=
lock
>>>>>>>=20
>>>>>>>    (TC) could be implemented in either LERs or LSRs.
>>>>>>>=20
>>>>>>> SB> LER and LSR need to be expanded
>>>>>>>=20
>>>>>>>    An example is shown in Figure 1, where the LERs act as Ordinary C=
lock
>>>>>>>    (OC) and are the initiating/terminating point for Timing messages=
.
>>>>>>>    The ingress LER encapsulates the Timing messages in Timing LSP an=
d
>>>>>>>    the Egress LER terminates the Timing LSP.  The LSRs act as
>>>>>>>    Transparent Clock (TC) and just update the Timing field in the Ti=
ming
>>>>>>>    messages.
>>>>>>>=20
>>>>>>>=20
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>       |Switch, |     |       |     |       |     |       |     |Swit=
ch, |
>>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rou=
ter |
>>>>>>>       |        |     |  OC   |     |  TC   |     |  OC   |     |    =
    |
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>                      /                                 \
>>>>>>>       +-------+     /                                   \     +-----=
--+
>>>>>>>       |  LER  |    /                                     \    |  LER=
  |
>>>>>>>       | Master|---/                                       \---| Slav=
e |
>>>>>>>       | Clock |                                               | Cloc=
k |
>>>>>>>       +-------+                                               +-----=
--+
>>>>>>>=20
>>>>>>>      Figure (1) - Deployment example 1 of timing over MPLS network
>>>>>>>=20
>>>>>>>    Another example is shown in Figure2, where LERs terminate the Tim=
ing
>>>>>>>    messages received from switch/routers that are outside of the MPL=
S
>>>>>>>    network acting as OC or BC.  In this example LERs regenerate the
>>>>>>>    clock and initiate timing messages encapsulated in Timing LSP tow=
ard
>>>>>>>    the MPLS network, while the LSRs act as Transparent Clock (TC) an=
d
>>>>>>>    just update the Timing field in the Timing messages, which are
>>>>>>>    already encapsulated in Timing LSPs.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 9]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>      |Switch, |     |       |     |       |     |       |     |Switc=
h, |
>>>>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
>>>>>>>      | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/B=
C  |
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>=20
>>>>>>>      Figure (2) - Deployment example 2 of timing over MPLS network
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Another example is shown in Figure 3, where LERs do not terminate=
 the
>>>>>>>    Timing messages received from switch/routers that are outside of t=
he
>>>>>>>    MPLS network acting as OC, TC or BC.  The LERs act as TC and upda=
te
>>>>>>>    the Timing field in the Timing messages as they transit the LER,
>>>>>>>    while encapsulating them in timing LSP.  The LSRs also act as
>>>>>>>    Transparent Clock (TC) and just update the Timing field in the Ti=
ming
>>>>>>>    messages which are already encapsulated in Timing LSPs.
>>>>>>>=20
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>       |Switch, |     |       |     |       |     |       |     |Swit=
ch, |
>>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rou=
ter |
>>>>>>>       |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/T=
C/BC|
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>=20
>>>>>>>     Figure (3) - Deployment example 3 of timing over MPLS network
>>>>>>>=20
>>>>>>>    Another example is shown in Figure 4, where LERs and LSRs support=

>>>>>>>    Boundary Clocks.  A single-hop LSP is created between two adjacen=
t
>>>>>>>    LSRs engaged in BC operation.  Other methods such as PTP transpor=
t
>>>>>>>    over Ethernet MAY be used for transporting timing messages if the=

>>>>>>>    link between the two routers is Ethernet.
>>>>>>>=20
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>      |Switch, |     |       |     |       |     |       |     |Switc=
h, |
>>>>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
>>>>>>>      | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/B=
C  |
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>=20
>>>>>>>    Figure (4) - Deployment example 3 of timing over MPLS network
>>>>>>>=20
>>>>>>>    An MPLS domain MAY serve multiple customers.  In these cases the
>>>> MPLS
>>>>>>>    domain (maintained by a service provider) may provide timing serv=
ices
>>>>>>>    to multiple customers, each having their own Timing domain.
>>>>>>>=20
>>>>>>>    The Timing over MPLS architecture assumes full mesh of Timing LSP=
s
>>>>>>>    between all LERs supporting this specification.
>>>>>>>=20
>>>>>>> SB> Note sure this is right - the salves surely do not need to
>>>>>>> SB> exchange timing amongst themselves
>>>>>>>=20
>>>>>>>    It supports
>>>>>>>    Point-to- point (VPWS) and Multipoint (VPLS) services.
>>>>>>>=20
>>>>>>> SB> What does that mean? You do not carry user data traffic?
>>>>>>> SB> Maybe it's the ordering of the statemnets that is causing
>>>>>>> SB> confusion.
>>>>>>>=20
>>>>>>>    This means
>>>>>>>    that a customer may purchase a Point-to-point Timing service betw=
een
>>>>>>>    two customer sites or a Multipoint Timing service between more th=
an
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 10]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    two customer sites.
>>>>>>>=20
>>>>>>>    The Timing over MPLS architecture supports P2P or P2MP Timing LSP=
s.
>>>>>>>    This means that the Timing Multicast messages such as PTP Multica=
st
>>>>>>>    event messages can be transported over P2MP Timing LSP or be
>>>>>>>    replicated and transported over many P2P Timing LSPs.
>>>>>>>=20
>>>>>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>>>>>>> SB> nor a P2MP PW, although we are close.
>>>>>>>=20
>>>>>>>    Timing messages, that do not require Time stamping or Correction
>>>>>>>    Field update MAY be transported over Timing LSPs to simplify hard=
ware
>>>>>>>    and software.
>>>>>>>=20
>>>>>>>    PTP Announce messages that determine the Timing LSP terminating
>>>> point
>>>>>>>    behavior such as BC/OC/TC SHOULD be transported over the Timing L=
SP
>>>>>>>    to simplify hardware and software.
>>>>>>>=20
>>>>>>> SB> have you defined and referenced PTP announce msgs?
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 11]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 5.  Dedicated LSPs for Timing messages
>>>>>>>=20
>>>>>>>    Many methods have been considered for identifying the Timing
>>>> messages
>>>>>>>    when they are encapsulated in MPLS such as using GAL/G-ACH or a n=
ew
>>>>>>>    reserved label.  These methods were not attractive since they eit=
her
>>>>>>>    required deep packet inspection at line rate in the intermediate L=
SRs
>>>>>>>    or they required use of a scarce new reserved label.  Also one of=
 the
>>>>>>>    goals was to reuse existing OAM mechanisms.
>>>>>>>=20
>>>>>>> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
>>>>>>>=20
>>>>>>>    The method defined in this document can be used by LER and LSRs t=
o
>>>>>>>    identify Timing messages in MPLS tunnels by just looking at the t=
op
>>>>>>>    label in the MPLS label stack, which only carry Timing messages a=
s
>>>>>>>    well as OAM, but not data plane client traffic.
>>>>>>>=20
>>>>>>>    Compliant implementations MUST use dedicated LSPs to carry Timing=

>>>>>>>    messages over MPLS.
>>>>>>>=20
>>>>>>> SB> I think that we need a definition of the properies of these LSPs=

>>>>>>>=20
>>>>>>>    These LSPs are herein referred to as "Timing
>>>>>>>    LSPs" and the labels associated with these LSPs as "Timing LSP
>>>>>>>    labels".  The Timing LSPs that runs between Ingress and Egress LE=
Rs
>>>>>>>    MUST be co-routed.  Alternatively, a single bidirectional co-rout=
ed
>>>>>>>    LSP can be used.
>>>>>>>=20
>>>>>>> SB> I though that you said you could use M2MP LSPs - these are not
>>>>>>> SB> bidirectional.
>>>>>>>=20
>>>>>>>    Co-routing of the two directions is required to limit the differe=
nce
>>>>>>>    in the delays in the Master clock to Slave clock direction compar=
ed
>>>>>>>    to the Slave clock to Master clock direction.  The Timing LSP MAY=
 be
>>>>>>>    MPLS/MPLS-TP LSP.
>>>>>>>=20
>>>>>>>    The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS=
.
>>>>>>>    New Extensions to RSVP-TE/GMPLS TLVs are required; however they a=
re
>>>>>>>    outside the scope of this document.
>>>>>>>=20
>>>>>>>    The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such=
 as
>>>>>>>    BFD and LSP Ping but the LSP data plane client plane traffic MUST=
 be
>>>>>>>    Timing packets only.
>>>>>>>=20
>>>>>>> SB> Why?
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 12]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 6.  Timing over LSP Encapsulation
>>>>>>>=20
>>>>>>> The encapsulations is not LSP is it?
>>>>>>>=20
>>>>>>>    This document defines two methods for carrying Timing messages ov=
er
>>>>>>>    MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>>>>>>    messages over Timing LSPs, and the second method, is carrying
>>>>>>>    Ethernet encapsulated Timing messages over Ethernet PWs inside
>>>> Timing
>>>>>>>    LSPs.
>>>>>>>=20
>>>>>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>>>>>>=20
>>>>>>>    The simplest method of transporting Timing messages over MPLS is t=
o
>>>>>>>    encapsulate Timing PDUs in UDP/IP and then encapsulate them in
>>>> Timing
>>>>>>>    LSP.  This format is shown in Figure 4.
>>>>>>>=20
>>>>>>>=20
>>>>>>>                     +----------------------+
>>>>>>>                     |   Timing LSP Label   |
>>>>>>>                     +----------------------+
>>>>>>>                     |        IPv4/6        |
>>>>>>>                     +----------------------+
>>>>>>>                     |         UDP          |
>>>>>>>                     +----------------------+
>>>>>>>                     |     Timing PDU       |
>>>>>>>                     +----------------------+
>>>>>>>=20
>>>>>>>       Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>>>>>>=20
>>>>>>>=20
>>>>>>>    This encapsulation is very simple and is useful when the network
>>>>>>>    between Timing Master Clock and Slave Clock is MPLS network.
>>>>>>>=20
>>>>>>> SB> Simple is a judgement call
>>>>>>>=20
>>>>>>>    In order for an LER/LSR to process Timing messages, the Timing LS=
P
>>>>>>>    Label must be at the top label of the label stack.  The LER/LSR M=
UST
>>>>>>>    know that the Timing LSP Label is used for carrying Timing messag=
es.
>>>>>>>    This can be accomplished via static configuration or via RSVP-TE
>>>>>>>    signaling.
>>>>>>>=20
>>>>>>>    The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>>>>>>    [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>>>>>>>    [RFC5905].
>>>>>>>=20
>>>>>>> 6.2.  Timing over PW Encapsulation
>>>>>>>=20
>>>>>>>    Another method of transporting Timing over MPLS networks is by
>>>>>>>    encapsulating Timing PDUs in PW which in turn is transported over=

>>>>>>>    Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448]=
,
>>>>>>>    shown in Fig 5(A) MUST be used and the Ethernet encapsulation of P=
TP
>>>>>>>    MUST follow Annex F of [IEEE-1588].
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 13]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
>>>> the
>>>>>>>    Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>>>>>>>    Timing over PW encapsulation MUST use the Control Word (CW) as
>>>>>>>    specified in [RFC4448] to ensure proper detection of PTP messages=

>>>>>>>    inside the MPLS packets for Timing over LSP and Timing over PW
>>>>>>>    encapsulation.
>>>>>>>=20
>>>>>>> SB> That needs explanation
>>>>>>>=20
>>>>>>>    The use of Sequence Number in the CW is optional.
>>>>>>>=20
>>>>>>> SB> Given that s/n are never in practice deployed, you could probabl=
y
>>>>>>> SB> simplify things by sayig that they are not used.
>>>>>>>=20
>>>>>>>    Timing over PW encapsulation for NTP MUST use NTP over UDP/IP ove=
r
>>>> PW
>>>>>>>    (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>>>>>>=20
>>>>>>>                     +----------------+  +----------------+
>>>>>>>                      |Timing LSP Label|  |Timing LSP Label|
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |    PW Label    |  |    PW Label    |
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |  Control Word  |  |      IP        |
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |    Ethernet    |  |      UDP       |
>>>>>>>                      |     Header     |  +----------------+
>>>>>>>                      +----------------+  |   Timing PDU   |
>>>>>>>                      |S-VLAN(Optional)|  |                |
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |C-VLAN(Optional)|        (B)
>>>>>>>                      +----------------+
>>>>>>>                      |   Timing PDU   |
>>>>>>>                      |                |
>>>>>>>                      +----------------+
>>>>>>>                             (A)
>>>>>>>=20
>>>>>>>               Figure (5) - Timing over PW Encapsulations
>>>>>>>=20
>>>>>>>    In order for an LSR to process PTP messages, the top label of the=

>>>>>>>    label stack (the Tunnel Label) MUST be a Timing label.
>>>>>>>=20
>>>>>>> S> You said that before.
>>>>>>>=20
>>>>>>> 6.3.  Other Timing Encapsulation methods
>>>>>>>=20
>>>>>>>    In future other timing encapsulation methods may be introduced, s=
uch
>>>>>>>    as a new shim header after the Bottom of Stack to carry the Timin=
g
>>>>>>>    information.  Such new encapsulations are outside the scope of th=
is
>>>>>>>    document.
>>>>>>>=20
>>>>>>>=20
>>>>>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>>>>>>> SB> out of the definition of the LSP
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 14]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> SB> I think we need a section on LSP processing
>>>>>>>=20
>>>>>>> 7.  Timing message Processing
>>>>>>>=20
>>>>>>>    Each Timing protocol such as PTP and NTP, define their set of Tim=
ing
>>>>>>>    messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>>>>>>>    FOLLOW_UP, etc messages.
>>>>>>>=20
>>>>>>>    Some of the Timing messages require time stamping or correction f=
ield
>>>>>>>    update at port level and some dont.  It is the job of the LER/LSR=
 to
>>>>>>>    parse the timing message and find out the type of the Timing mess=
age
>>>>>>>    and decide whether and how to Time- stamp it (e.g., BC) or update=

>>>>>>>    correction field(e.g., TC).
>>>>>>>=20
>>>>>>>=20
>>>>>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
>>>>>>> SB> function rather than the LER?
>>>>>>>=20
>>>>>>>    For example the following PTP messages (called Event messages)
>>>>>>>    require time-stamping or correction field update:
>>>>>>>=20
>>>>>>>    o  SYNC
>>>>>>>=20
>>>>>>>    o  DELAY_REQ (Delay Request)
>>>>>>>=20
>>>>>>>    o  PDELAY_REQ (Peer Delay Request)
>>>>>>>=20
>>>>>>>    o  PDELAY_RESP (Peer Delay Response)
>>>>>>>=20
>>>>>>>    SYNC and DELAY_REQ are exchanged between Master Clock and Slave
>>>> Clock
>>>>>>>    and MUST be transported over PTP LSPs.  PDELAY_REQ and
>>>> PDELAY_RESP
>>>>>>>    are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>>>>>>>    Boundary, or Transparent) and SHOULD be transported over single h=
op
>>>>>>>    PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP=
,
>>>>>>>    and PDELAY_RESP_FOLLOW_UP messages MUST also be transported
>>>> over
>>>>>>> the
>>>>>>>    PTP LSPs.
>>>>>>>=20
>>>>>>>    For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be=

>>>>>>>    transported over two PTP LSPs that are in opposite directions.  T=
hese
>>>>>>>    PTP LSPs, which are in opposite directions MUST be congruent and c=
o-
>>>>>>>    routed.  Alternatively, a single bidirectional co-routed LSP can b=
e
>>>>>>>    used.
>>>>>>>=20
>>>>>>>    Except as indicated above for the two-step PTP clocks, Non-Event P=
TP
>>>>>>>    message types do not need to be processed by intermediate routers=
.
>>>>>>>    These message types MAY be carried in PTP Tunnel LSPs.
>>>>>>>=20
>>>>>>> SB> Are you saying that a timing P router has to be msg type sensiti=
ve?
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 15]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 8.  Protection and Redundancy
>>>>>>>=20
>>>>>>>=20
>>>>>>> SB> This is a bit of a jump - I don't know how the LSP itself works y=
et!
>>>>>>>=20
>>>>>>>    In order to ensure continuous uninterrupted operation of slave
>>>>>>>    clocks, usually as a general practice, slave clocks (or ports) tr=
ack
>>>>>>>    redundant master clocks.
>>>>>>>=20
>>>>>>>    It is the responsibility of the network operator to ensure that
>>>>>>>    physically disjoint Timing LSPs are established between a slave c=
lock
>>>>>>>    (or port) and redundant master clocks (or ports).
>>>>>>>=20
>>>>>>>    When a slave clock (or port) listens to redundant master clocks o=
r
>>>>>>>    ports, any prolonged Timing LSP outage will trigger the slave clo=
ck
>>>>>>>    or port to switch to a redundant master clock or port.
>>>>>>>=20
>>>>>>>    LSP/PW protection such as Linear protection Switching (1:1, 1+1),=

>>>>>>>    Ring protection switching or MPLS Fast Reroute (FRR) generally sw=
itch
>>>>>>>    alternative path that usually cause a change in delay, which if
>>>>>>>    undetected by slave clock can reduce accuracy of the slave clock.=

>>>>>>>=20
>>>>>>>    Therefore protection switching MAY be used, as long as phase jump=
s
>>>>>>>    upon switchover due to differences in path latency are detected a=
nd
>>>>>>>    compensated for (such compensation not being required if BCs or p=
eer-
>>>>>>>    peer TCs are used throughout).
>>>>>>>=20
>>>>>>>    Note that any protection or reroute mechanism that adds additiona=
l
>>>>>>>    MPLS label to the label stack, such as Facility Backup Fast Rerou=
te,
>>>>>>>    MUST ensure that the pushed label is also a Timing Label to ensur=
e
>>>>>>>    recognition of the MPLS frame as containing Timing messages, as i=
t
>>>>>>>    transits the backup path.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 16]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 9.  ECMP
>>>>>>>=20
>>>>>>>    To ensure the optimal operation of slave clocks and avoid error
>>>>>>>    introduced by forward and reverse path delay asymmetry, the physi=
cal
>>>>>>>    path for Timing messages from master clock to slave Clock and vic=
e
>>>>>>>    versa must be the same for all Event Timing messages listed in
>>>>>>>    section 7.
>>>>>>>=20
>>>>>>>    Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost=

>>>>>>>    Multipath).
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 17]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 10.  PHP
>>>>>>>=20
>>>>>>>    To ensure that the label on the top of the label stack is the Tim=
ing
>>>>>>>    LSP Label, PHP MUST not be used.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 18]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 11.  Entropy
>>>>>>>=20
>>>>>>>    To ensure all Timing messages in a Timing LSP take the same path,=

>>>>>>>    Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>>>>>>    Entropy Label MUST NOT be used for the PWs that are carried insid=
e
>>>>>>>    Timing LSP [RFC6391].
>>>>>>>=20
>>>>>>> SB> This is incorrect - you mean that all msgs of the same timing
>>>>>>> SB> flow need to have the same EL value.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 19]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 12.  OAM, Control and Management
>>>>>>>=20
>>>>>>>    In order to monitor Timing LSPs and their encapsulated PWs, they M=
UST
>>>>>>>    be able to carry OAM and management messages.  These management
>>>>>>>    messages MUST be differentiated from Timing messages via already
>>>>>>>    defined IETF methods.
>>>>>>>=20
>>>>>>>    For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY r=
un
>>>>>>>    over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>>>>>>>    Management protocols can easily be identified by the UDP Destinat=
ion
>>>>>>>    Port number or by GAL/G-ACH respectively.
>>>>>>>=20
>>>>>>>    Also BFD, LSP-Ping and other management messages MAY run over the=

>>>> PWs
>>>>>>>    encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3=
 or
>>>>>>>    4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is g=
oing
>>>>>>>    to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1=
) or
>>>>>>>    GAL-ACH are used to identify such management messages.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 20]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 13.  QoS Considerations
>>>>>>>=20
>>>>>>>    In network deployments where not every LSR/LER is Timing-aware, i=
t is
>>>>>>>    important to reduce the impact of the non-Timing-aware LSR/LERs o=
n
>>>>>>>    the timing recovery in the slave clock.  The Timing messages are t=
ime
>>>>>>>    critical and must be treated with the highest priority.  Therefor=
e
>>>>>>>    Timing over MPLS messages must be treated with the highest priori=
ty
>>>>>>>    in the routers.  This can be achieved by proper setup of Timing L=
SPs.
>>>>>>>=20
>>>>>>>    It is recommended that the Timing LSPs are setup or configured
>>>>>>>    properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC26=
97]
>>>>>>>    for drop eligibility.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 21]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 14.  FCS and Checksum Recalculation
>>>>>>>=20
>>>>>>>    When time-stamp generation and timing packet adjustment is perfor=
med
>>>>>>>    near the physical port hardware, the process MUST include
>>>>>>>    recalculation of the Ethernet FCS.
>>>>>>>=20
>>>>>>> SB> The above is confusing - an LSR always recomputes the link layer=

>>>>>>> SB> CRC which may or may not be Ethernet.
>>>>>>>=20
>>>>>>>    Also FCS retention for the
>>>>>>>    payload Ethernet described in [RFC4720] MUST NOT be used.
>>>>>>>=20
>>>>>>>    For UDP/IP encapsulation mode of Timing over MPLS, the UDP checks=
um
>>>>>>>    may be required as per UDP transport standards.
>>>>>>>=20
>>>>>>> SB> You really need to be working on getting the IPv6 C?S computatio=
n
>>>>>>> SB> removed from PTP msgs.
>>>>>>>=20
>>>>>>>    When UDP checksum is used, each Timing-aware LER/LSR must either
>>>>>>>    incrementally update the UDP checksum after Time stamping or
>>>>>>>    Correction Field update or verify the UDP checksum on reception f=
rom
>>>>>>>    upstream and recalculate the checksum completely on transmission t=
o
>>>>>>>    downstream node after Time stamping or Correction Field update.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 22]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 15.  Behavior of LER/LSR
>>>>>>>=20
>>>>>>>    Timing-capable/aware LERs and LSRs are routers that have one or m=
ore
>>>>>>>=20
>>>>>>> SB> You mean physical interfaces?
>>>>>>>=20
>>>>>>>    interfaces that can perform Timing operations (OC/BC/TC) on Timin=
g
>>>>>>>    packets and are configured to do so.  Timing-capable/aware LERs a=
nd
>>>>>>>    LSRs can advertise their Timing-capability per-interface via cont=
rol
>>>>>>>    plane such as OSPF or IS-IS.
>>>>>>> SB> ISIS and OSPF are routing protocols.
>>>>>>>=20
>>>>>>>   The Timing-capable/aware LERs can then
>>>>>>>    signals Timing LSPs via RSVP-TE signaling.  Alternatively the Tim=
ing
>>>>>>>    capability of LER and LSRs may be configured in a centralized
>>>>>>>    controller and the Timing LSP may be setup using manual configura=
tion
>>>>>>>    or other methods such as SDN.
>>>>>>>=20
>>>>>>> SB> it can also be configured individually rather then through
>>>>>>> SB> a cebtral controllwe
>>>>>>>=20
>>>>>>> 15.1.  Behavior of Timing-capable/aware LER
>>>>>>>=20
>>>>>>>    When a Timing-capable/aware LER behaves as a Transparent clock an=
d
>>>>>>>    receives a Timing message from a Timing-capable/aware non-MPLS
>>>>>>>    interface, the LER updates the Correction Field (CF) and encapsul=
ates
>>>>>>>    and forwards the timing message over previously established Timin=
g
>>>>>>>    LSP.
>>>>>>>=20
>>>>>>> SB> You need to call out the details so that people properly
>>>>>>> SB> understand the definition of the new LSP.
>>>>>>>=20
>>>>>>>    Also when a Timing message is received from a Timing-capable/
>>>>>>>    aware MPLS interface, LER updates the Correction Filed (CF) and
>>>>>>>    decapsulates the MPLS encapsulation and forwards the timing messa=
ge
>>>>>>>    to a non-MPLS interface.
>>>>>>>=20
>>>>>>>    When a Timing-capable/aware LER behaves as a Boundary clock and
>>>>>>>    receives a Timing message from a Timing-capable/aware non MPLS
>>>>>>>    interface, the LER Timestamps the Timing packet and sends it to t=
he
>>>>>>>    LERs Boundary clock processing module.  Also when a Timing messag=
e is
>>>>>>>    received from a Timing- capable/aware MPLS interface, the LER
>>>>>>>    Timestamps the Timing packet and sends it to the LERs Boundary cl=
ock
>>>>>>>    processing module.
>>>>>>>=20
>>>>>>>    When a Timing-capable/aware LER behaves as an Ordinary Clock towa=
rd
>>>>>>>    the MPLS network, and receives a Timing message from a Timing-
>>>>>>>    capable/aware MPLS interface, the LER Timestamps the Timing packe=
t
>>>>>>>    and sends it to the LERs Ordinary clock processing module.
>>>>>>>=20
>>>>>>> 15.2.  Behavior of Timing-capable/aware LSR
>>>>>>>=20
>>>>>>>    When a Timing-capable/aware LSR behaves as a Transparent clock an=
d
>>>>>>>    receives a Timing message from a Timing-capable/aware MPLS
>>>> interface,
>>>>>>>    The LSR updates the Correction Filed (CF) and forwards the timing=

>>>>>>>    message over another MPLS interface.
>>>>>>>=20
>>>>>>>    When a Timing-capable/aware LSR behaves as a Boundary clock and
>>>>>>>    receives a Timing message from a Timing-capable/aware MPLS
>>>> interface.
>>>>>>>    The LSR performs the functions of a Boundary Clock in terminating=
 the
>>>>>>>    received Timing message and re-generating a new timing message ov=
er
>>>>>>>    another (or the same) MPLS interface.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 23]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>>>>>>=20
>>>>>>>    It is most beneficial when all LSRs in the path of a Timing LSP b=
e
>>>>>>>    timing-Capable/aware LSRs.  This would ensure the highest quality=

>>>>>>>    time and clock synchronization by Timing Slave Clocks.  However, t=
his
>>>>>>>    specification does not mandate that all LSRs in path of a Timing L=
SP
>>>>>>>    be Timing- capable/aware.
>>>>>>>=20
>>>>>>>    Non-Timing-capable/aware LSRs just switch the packets encapsulate=
d in
>>>>>>>    Timing LSPs and dont perform any Timing operation (TC or BC).
>>>>>>>    However as explained in QoS section the Timing over MPLS packets
>>>> MUST
>>>>>>>    be still be treated with the highest priority based on their Traf=
fic
>>>>>>>    Class (TC) marking.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 24]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 16.  Other considerations
>>>>>>>=20
>>>>>>>    [IEEE-1588] defines an optional peer-to-peer Transparent clocking=

>>>>>>>    that requires peer delay measurement between two adjacent Timing-=

>>>>>>>    capable/ aware routers/switches.  Peer delay measurement messages=

>>>>>>>    need to be time stamped and terminated by the Timing-capable/awar=
e
>>>>>>>    routers/ switches.  This means that two adjacent LSRs may be enga=
ged
>>>>>>>    in a peer delay measurement.
>>>>>>>=20
>>>>>>>    For transporting such peer delay measurement messages a single-ho=
p
>>>>>>>    LSP SHOULD to be created between the two adjacent LSRs engaged in=

>>>>>>>    peer delay measurement to carry peer delay measurement messages.
>>>>>>>    Other methods such as PTP transport over Ethernet MAY be used for=

>>>>>>>    transporting peer delay measurement messages if the link between t=
he
>>>>>>>    two routers is Ethernet.
>>>>>>>=20
>>>>>>>    In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ w=
are
>>>>>>>    routers/switches MUST maintain a list of all the neighbors it nee=
ds
>>>>>>>    to send a PDelay_Req to, where each neighbor corresponds to a tim=
ing
>>>>>>>    LSP.
>>>>>>>=20
>>>>>>>    The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as=
 long
>>>>>>>    as either the Explicit Null label is the bottom of stack label
>>>>>>>    (applicable only to UDP/IP encapsulation) or the label below the
>>>>>>>    Explicit Null label is a PTP label.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 25]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 17.  Security Considerations
>>>>>>>=20
>>>>>>>    MPLS PW security considerations in general are discussed in [RFC3=
985]
>>>>>>>    and [RFC4447],and those considerations also apply to this documen=
t.
>>>>>>>=20
>>>>>>>    An experimental security protocol is defined in [IEEE-1588].The P=
TP
>>>>>>>    security extension and protocol provides group source authenticat=
ion,
>>>>>>>    message integrity, and replay attack protection for PTP messages.=

>>>>>>>=20
>>>>>>>    When the MPLS network (provider network) serves multiple customer=
s,
>>>>>>>    it is important to maintain and process each customers clock and
>>>>>>>    Timing messages separately from other customers to ensure there i=
s no
>>>>>>>    cross- customer effect.  For example if an LER BC is synchronized=
 to
>>>>>>>    a specific grandmaster, belonging to customer A, then the LER MUS=
T
>>>>>>>    use that BC clock only for customer A to ensure that customer A
>>>>>>>    cannot attack other customers by manipulating its time.
>>>>>>>=20
>>>>>>>    Timing messages MAY be encrypted or authenticated, provided that t=
he
>>>>>>>    LERs/LSRs that are Timing capable/aware can authenticate/ decrypt=
 the
>>>>>>>    timing messages.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 26]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 18.  Acknowledgements
>>>>>>>=20
>>>>>>>    The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizr=
ahi,
>>>>>>>    Stefano Ruffini, Peter Meyer, and other members of IETF for revie=
wing
>>>>>>>    and providing feedback on this draft.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 27]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 19.  IANA Considerations
>>>>>>>=20
>>>>>>>    There are no IANA requirements in this specification.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 28]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 20.  References
>>>>>>>=20
>>>>>>> 20.1.  Normative References
>>>>>>>=20
>>>>>>>    [IEEE-1588]
>>>>>>>               IEEE 1588-2008, "IEEE Standard for a Precision Clock
>>>>>>>               Synchronization Protocol for Networked Measurement and=

>>>>>>>               Control Systems".
>>>>>>>=20
>>>>>>>    [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>>>>>>               Requirement Levels", BCP 14, RFC 2119, March 1997.
>>>>>>>=20
>>>>>>>    [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to=
-
>>>>>>>               Edge (PWE3) Architecture", RFC 3985, March 2005.
>>>>>>>=20
>>>>>>>    [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discov=
ery
>>>>>>>               Proxies (ND Proxy)", RFC 4389, April 2006.
>>>>>>>=20
>>>>>>>    [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G=
.
>>>>>>>               Heron, "Pseudowire Setup and Maintenance Using the Lab=
el
>>>>>>>               Distribution Protocol (LDP)", RFC 4447, April 2006.
>>>>>>>=20
>>>>>>>    [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>>>>>>               "Encapsulation Methods for Transport of Ethernet over M=
PLS
>>>>>>>               Networks", RFC 4448, April 2006.
>>>>>>>=20
>>>>>>>    [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>>>>>>               Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>>>>>>               Retention", RFC 4720, November 2006.
>>>>>>>=20
>>>>>>>    [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circu=
it
>>>>>>>               Connectivity Verification (VCCV): A Control Channel fo=
r
>>>>>>>               Pseudowires", RFC 5085, December 2007.
>>>>>>>=20
>>>>>>>    [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detect=
ion
>>>>>>>               (BFD)", RFC 5880, June 2010.
>>>>>>>=20
>>>>>>>    [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow=
,
>>>>>>>               "Bidirectional Forwarding Detection (BFD) for MPLS Lab=
el
>>>>>>>               Switched Paths (LSPs)", RFC 5884, June 2010.
>>>>>>>=20
>>>>>>> 20.2.  Informative References
>>>>>>>=20
>>>>>>>    [I-D.ietf-pwe3-fat-pw]
>>>>>>>               Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Reg=
an,
>>>>>>>               J., and S. Amante, "Flow Aware Transport of Pseudowire=
s
>>>>>>>               over an MPLS Packet Switched Network",
>>>>>>>               draft-ietf-pwe3-fat-pw-07 (work in progress), July 201=
1.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 29]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermedia=
te
>>>>>>>               system routeing information exchange protocol for use i=
n
>>>>>>>               conjunction with the Protocol for providing the
>>>>>>>               Connectionless-mode Network Service (ISO 8473)".
>>>>>>>=20
>>>>>>>    [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP an=
d
>>>>>>>               dual environments", RFC 1195, December 1990.
>>>>>>>=20
>>>>>>>    [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 199=
8.
>>>>>>>=20
>>>>>>>    [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color=

>>>>>>>               Marker", RFC 2697, September 1999.
>>>>>>>=20
>>>>>>>    [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boud=
ec,
>>>>>>>               J., Courtney, W., Davari, S., Firoiu, V., and D.
>>>>>>>               Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>>>>>>               Behavior)", RFC 3246, March 2002.
>>>>>>>=20
>>>>>>>    [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Enginee=
ring
>>>>>>>               (TE) Extensions to OSPF Version 2", RFC 3630,
>>>>>>>               September 2003.
>>>>>>>=20
>>>>>>>    [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermedia=
te
>>>>>>>               System (IS-IS) Extensions for Traffic Engineering (TE)=
",
>>>>>>>               RFC 3784, June 2004.
>>>>>>>=20
>>>>>>>    [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S=
.
>>>>>>>               Shaffer, "Extensions to OSPF for Advertising Optional
>>>>>>>               Router Capabilities", RFC 4970, July 2007.
>>>>>>>=20
>>>>>>>    [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate=

>>>>>>>               System to Intermediate System (IS-IS) Extensions for
>>>>>>>               Advertising Router Information", RFC 4971, July 2007.
>>>>>>>=20
>>>>>>>    [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi=

>>>>>>>               Topology (MT) Routing in Intermediate System to
>>>>>>>               Intermediate Systems (IS-ISs)", RFC 5120, February 200=
8.
>>>>>>>=20
>>>>>>>    [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>>>>>>               Engineering", RFC 5305, October 2008.
>>>>>>>=20
>>>>>>>    [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>>>>>>               "Traffic Engineering Extensions to OSPF Version 3",
>>>>>>>               RFC 5329, September 2008.
>>>>>>>=20
>>>>>>>    [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSP=
F
>>>>>>>               for IPv6", RFC 5340, July 2008.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 30]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Net=
work
>>>>>>>               Time Protocol Version 4: Protocol and Algorithms
>>>>>>>               Specification", RFC 5905, June 2010.
>>>>>>>=20
>>>>>>>    [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Reg=
an,
>>>>>>>               J., and S. Amante, "Flow-Aware Transport of Pseudowire=
s
>>>>>>>               over an MPLS Packet Switched Network", RFC 6391,
>>>>>>>               November 2011.
>>>>>>>=20
>>>>>>>    [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., a=
nd
>>>>>>>               L. Yong, "The Use of Entropy Labels in MPLS Forwarding=
",
>>>>>>>               RFC 6790, November 2012.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 31]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 1.  Routing extensions for Timing-aware Routers
>>>>>>>=20
>>>>>>>    MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] a=
nd
>>>>>>>    IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (=
TE)
>>>>>>>    link information used for constraint-based routing.
>>>>>>>=20
>>>>>>>    Indeed, it is useful to advertise data plane TE router link
>>>>>>>    capabilities, such as the capability for a router to be Timing-aw=
are.
>>>>>>>    This capability MUST then be taken into account during path
>>>>>>>    computation to prefer or even require links that advertise themse=
lves
>>>>>>>    as Timing-aware.  In this way the path can ensure the entry and e=
xit
>>>>>>>    points into the LERs and, if desired, the links into the LSRs are=

>>>>>>>    able to perform port based time-stamping thus minimizing their im=
pact
>>>>>>>    on the performance of the slave clock.
>>>>>>>=20
>>>>>>>    extensions are required to OSPF and IS-IS in order to advertise
>>>>>>>    Timing-aware capabilities of a link.  Such extensions are outside=
 the
>>>>>>>    scope of this document; however such extension SHOULD be able to
>>>>>>>    signal the following information per Router Link:
>>>>>>>=20
>>>>>>>    o  Capable of processing PTP, NTP or other Timing flows
>>>>>>>=20
>>>>>>>    o  Capable of performing Transparent Clock operation
>>>>>>>=20
>>>>>>>    o  Capable of performing Boundary Clock operation
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 32]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 2.  Signaling Extensions for Creating Timing LSPs
>>>>>>>=20
>>>>>>>    RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSV=
P-TE
>>>>>>>    is used to setup Timing LSPs, some information that indicates tha=
t
>>>>>>>    the LSP is carrying Timing flows MUST be included in the new
>>>>>>>    Extensions to RSVP-TE:
>>>>>>>=20
>>>>>>>    The following information MAY also be included in the new Extensi=
ons
>>>>>>>    to RSVP-TE:
>>>>>>>=20
>>>>>>>    o  Offset from Bottom of Stack (BoS) to the start of the Time-sta=
mp
>>>>>>>       field
>>>>>>>=20
>>>>>>>    o  Number of VLANs in case of PW encapsulation
>>>>>>>=20
>>>>>>>    o  Timestamp field Type
>>>>>>>=20
>>>>>>>       *  Correction Field, Timestamp
>>>>>>>=20
>>>>>>>    o  Timestamp Field format
>>>>>>>=20
>>>>>>>       *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit=

>>>>>>>          NTP, etc.
>>>>>>>=20
>>>>>>>    Note that in case the above optional information is signaled with=

>>>>>>>    RSVP-TE for a Timing LSP, all the Timing packets carried in that L=
SP
>>>>>>>    must have the same signaled characteristics.  For example if
>>>>>>>    Timestamp format is signaled as 64-bit PTPv1, then all Timing pac=
kets
>>>>>>>    must use 64-bit PTPv1 time-stamp.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 33]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> Authors' Addresses
>>>>>>>=20
>>>>>>>    Shahram Davari
>>>>>>>    Broadcom Corp.
>>>>>>>    San Jose, CA  95134
>>>>>>>    USA
>>>>>>>=20
>>>>>>>    Email: davari@broadcom.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Amit Oren
>>>>>>>    Broadcom Corp.
>>>>>>>    San Jose, CA  95134
>>>>>>>    USA
>>>>>>>=20
>>>>>>>    Email: amito@broadcom.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Manav Bhatia
>>>>>>>    Alcatel-Lucent
>>>>>>>    Bangalore,
>>>>>>>    India
>>>>>>>=20
>>>>>>>    Email: manav.bhatia@alcatel-lucent.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Peter Roberts
>>>>>>>    Alcatel-Lucent
>>>>>>>    Kanata,
>>>>>>>    Canada
>>>>>>>=20
>>>>>>>    Email: peter.roberts@alcatel-lucent.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Laurent Montini
>>>>>>>    Cisco Systems
>>>>>>>    San Jose CA
>>>>>>>    USA
>>>>>>>=20
>>>>>>>    Email: lmontini@cisco.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 34]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Luca
>>>>>>>    Cisco Systems
>>>>>>>    San Jose CA
>>>>>>>    USA
>>>>>>>=20
>>>>>>>    Email: lmartini@cisco.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 35]
>>>>>>> --
>>>>>>> For corporate legal information go to:
>>>>>>>=20
>>>>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>>>>=20
>>>>>>> _______________________________________________
>>>>>>> TICTOC mailing list
>>>>>>> TICTOC@ietf.org
>>>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>>> This e-mail message is intended for the recipient only and contains
>>>> information which is CONFIDENTIAL and which may be proprietary to ECI
>>>> Telecom. If you have received this transmission in error, please inform=
 us by e-
>>>> mail, phone or fax, and then delete the original and all copies thereof=
.
>>>>>> _______________________________________________
>>>>>> mpls mailing list
>>>>>> mpls@ietf.org
>>>>>> https://www.ietf.org/mailman/listinfo/mpls
>>>>> .
>>>>=20
>>>> --
>>>> For corporate legal information go to:
>>>>=20
>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>=20
>>>> _______________________________________________
>>>> TICTOC mailing list
>>>> TICTOC@ietf.org
>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>=20
>>>>=20
>>>> _______________________________________________
>>>> TICTOC mailing list
>>>> TICTOC@ietf.org
>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>=20
>>> This e-mail message is intended for the recipient only and contains info=
rmation which is CONFIDENTIAL and which may be proprietary to ECI Telecom. I=
f you have received this transmission in error, please inform us by e-mail, p=
hone or fax, and then delete the original and all copies thereof.
>> .
>=20
>=20
> --=20
> For corporate legal information go to:
>=20
> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>=20

From Alexander.Vainshtein@ecitele.com  Tue Aug 20 09:53:27 2013
Return-Path: <Alexander.Vainshtein@ecitele.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 65E1421F8948; Tue, 20 Aug 2013 09:53:27 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.202
X-Spam-Level: 
X-Spam-Status: No, score=-2.202 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, MIME_QP_LONG_LINE=1.396, RCVD_IN_DNSWL_LOW=-1,  UNPARSEABLE_RELAY=0.001]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 049bivCKOzWW; Tue, 20 Aug 2013 09:53:22 -0700 (PDT)
Received: from mail1.bemta3.messagelabs.com (mail1.bemta3.messagelabs.com [195.245.230.169]) by ietfa.amsl.com (Postfix) with ESMTP id 58D2B21F888F; Tue, 20 Aug 2013 09:53:18 -0700 (PDT)
Received: from [85.158.137.68:17382] by server-9.bemta-3.messagelabs.com id 06/E1-15303-CFE93125; Tue, 20 Aug 2013 16:53:16 +0000
X-Env-Sender: Alexander.Vainshtein@ecitele.com
X-Msg-Ref: server-11.tower-31.messagelabs.com!1377017593!2818503!1
X-Originating-IP: [147.234.242.234]
X-StarScan-Received: 
X-StarScan-Version: 6.9.11; banners=-,-,-
X-VirusChecked: Checked
Received: (qmail 23282 invoked from network); 20 Aug 2013 16:53:14 -0000
Received: from ilptbmg01-out.ecitele.com (HELO ilptbmg01-out.ecitele.com) (147.234.242.234) by server-11.tower-31.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 20 Aug 2013 16:53:14 -0000
X-AuditID: 93eaf2e7-b7f0e6d000004f79-b0-52139ef49565
Received: from ILPTWPVEXCA02.ecitele.com ( [172.31.244.232]) by ilptbmg01-out.ecitele.com (Symantec Messaging Gateway) with SMTP id 58.63.20345.4FE93125; Tue, 20 Aug 2013 19:53:08 +0300 (IDT)
Received: from ILPTWPVEXMB01.ecitele.com ([fe80::f152:8eaf:8fb0:a5da]) by ILPTWPVEXCA02.ecitele.com ([fe80::c473:490d:3a7e:e34a%12]) with mapi id 14.03.0123.003; Tue, 20 Aug 2013 19:53:07 +0300
From: Alexander Vainshtein <Alexander.Vainshtein@ecitele.com>
To: "stbryant@cisco.com" <stbryant@cisco.com>, "S. Davari" <davarish@yahoo.com>
Thread-Topic: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
Thread-Index: AQHOkhqJ+XHALQ/MzkWnZzr9BEl1M5mHyTyggBaXBuSAAAM16Q==
Date: Tue, 20 Aug 2013 16:53:07 +0000
Message-ID: <F9336571731ADE42A5397FC831CEAA021513A083@ILPTWPVEXMB01.ecitele.com>
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com> <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com> <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com>, <52139903.8040801@cisco.com>
In-Reply-To: <52139903.8040801@cisco.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-originating-ip: [147.234.1.1]
Content-Type: text/plain; charset="us-ascii"
content-transfer-encoding: quoted-printable
MIME-Version: 1.0
X-CFilter-Loop: Reflected
X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFuplleLIzCtJLcpLzFFi42KZ/OrTF92f84SDDJpXKFocfN7EaHFr6UpW i7/NPewOzB5Llvxk8pg16zBTAFNUA6NNYl5efkliSapCSmpxsq1SQFFmWWJypZJCZoqtkqGS QkFOYnJqbmpeia1SYkFBal6Kkh2XAgawASrLzFNIzUvOT8nMS7dV8gz217WwMLXUNVSyU1M2 NLbmCsnILFZI1c1NzMxRyE0tLk5MT1UAiiRsYc5YfHUNS8GHdtaKqz8/szYw3rjD3MXIySEh YCKxYdomJghbTOLCvfVsXYxcHEICBxklvsz9BuUcZZRYvnkGWBWbgK3EptV32UBsEYFAia7j b1hBbGaBAonnm18D1XBwCAu4Sxz+4QdR4iHROfU8I0hYRMBJ4sY/LZAwi4CqxMJVV1hBwrwC ARI3l+RBbFrILDHh31Y2kDingKbE+R4ZkHJGoNO+n1rDBLFIXOLWk/lQJwtILNlzHuoVUYmX j/+xQtjyEhc/PICq15FYsPsTG4StLbFs4Wuwel4BQYmTM5+wQNRLShxccYNlAqP4LCQrZiFp n4WkfRaS9gWMLKsYRTNzCkqSctMNDPVSkzNLUnNS9ZLzczcxQhLK8x2Mv+arHGJ0BXp7IrMU d3I+MCHllcQbGxjg5iiJ8y5vCPcXEkgHppvs1NSC1KL4otKc1OJDjEwcnFINjNrrwhWDkyLC rjieSv5v6mWUuzAv79xmrkAJ+wO/HazP/zEMmiRXUm4c3Drn17ojBf+5hAtj8hcefCe1Klxf tK/h5Kwzu7bJpHqdtPuwR/nMtM+8VTeqFIqYnu5YePLyxZeqff/CgiScq35kf9BO79Peznld ch3bmjlP25/f/rX3sCzv8o2Lm5VYijMSDbWYi4oTAeplDWMgAwAA
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 20 Aug 2013 16:53:27 -0000

Stewart, Shahram and all,

A couple of (hopefully, last) comments:

1. Supporting a "Timing Alert" label at the top of the label stack introduce=
s a backward compatibility problem. We differ in our estimate of the importa=
nce of this problem, but we all agree that it exists.

2. On the other hand, this problem does not exist if the "Timing Alert" labe=
l is inserted at the bottom of the label stack. At the same time, HW that an=
alyzes the label stack hopefully could analyze the bottom label (almost) as=
 easily as it would analyze the top label. Thus we could both provide backwa=
rd compatibility (incompatible transit LSRs simply would not notice the "Tim=
ing Alert" label") and avoid the need for dedicated "Timing-only" LSPs. In p=
articular, all the MPLS OAM tools (including MPLS-TP) would work without any=
 changes. 

3. A "timing shim header" (between the bottom of the label stack and the beg=
inning of the timing  packet) has been mentioned by Stewart as an already pr=
oposed mechanism. IMHO and FWIW its usage would eliminate the need to unders=
tand format of the specific timing distribution protocol in the transit LSRs=
.  It would further solve the problem of secure transport of timing packets=
 since security mechanisms would be applied to its original encapsulation (o=
ver IP or over Ethernet) without involving the transit LSRs.

Did I miss something substantial?

Regards, and lots of thanks in advance,
     Sasha


________________________________________
From: Stewart Bryant [stbryant@cisco.com]
Sent: Tuesday, August 20, 2013 6:27 PM
To: S. Davari
Cc: Alexander Vainshtein; Shahram Davari; mpls@ietf.org; tictoc@ietf.org
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls

On 06/08/2013 14:35, S. Davari wrote:
> Sasha,
>
> If a router is not 1588 aware we want it to switch the packet with minimum=
 jitter and delay via high priority queue.
Yes, but you can explicitly tunnel across such island of incompatibility

Stewart

>
>
> Sending it to CPU creates large jitter and delay since the CPUs are genera=
lly busy doing other things. Time stamping or updating CF in the CPU won't h=
elp since it does not cover the variable delay that is incurred from the tim=
e the packet enters the router to the time the packet gets processed by CPU.
>
> I wish it was that easy!
>
> Regards,
> Shahram
>
>
> On Aug 6, 2013, at 12:38 AM, Alexander Vainshtein <Alexander.Vainshtein@ec=
itele.com> wrote:
>
>> Shahram,
>> I agree with you that the routers that do not recognize a Timing Alert la=
bel would send the packet to the CPU. (This would be easy to arrange since w=
e are speaking about a reserved label.)
>>
>> I do not, however, see this as a serious issue, because the SW running on=
 this CPU would (hopefully) be upgradable to handle the packet correctly i.e=
., to forward it to where it should be forwarded and, if it is forwarded as=
 a labeled packet, to prepend the Timing Alert Label on top of its label sta=
ck. It could even record the residence time (as observed by the CPU in the p=
roper place in the packet. This would mean that introducing additional error=
 to whatever timing information is associated with this packet - but this is=
 what you should anyway expect if there are non-compliant routers on your pa=
th, right?
>>
>> Regards,
>>      Sasha
>>
>>> -----Original Message-----
>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf=
 Of
>>> Shahram Davari
>>> Sent: Monday, August 05, 2013 11:29 PM
>>> To: stbryant@cisco.com; S. Davari
>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>>
>>> Hi Stewart,
>>>
>>> Your suggestion of " I suggested an LSP type that has the properties "ti=
mestamp
>>> and pass to application", is a subset of the existing draft and should w=
ork. It
>>> basically limits the time stamping and correction field update to LERs (=
while the
>>> draft supports time stamping at LER and LSR).
>>>
>>> However Sasha's suggestion of using a reserved label (RAL, GAL or any ot=
her
>>> reserved label) does not satisfy one of the major requirements, which is
>>> backward compatibility. Routers that don't understand this reserved labe=
l will
>>> drop or copy to CPU such packets.
>>>
>>> Regards,
>>> Shahram
>>>
>>> -----Original Message-----
>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf=
 Of
>>> Stewart Bryant
>>> Sent: Monday, August 05, 2013 2:29 AM
>>> To: S. Davari
>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>>
>>> My concern is that we are creating a new LSP type in MPLS
>>> (which is a architectural change and thus needs to be very carefully
>>> considered) without asking the question "how can I do this
>>> in the most general way, to maximize flexibility/reuse"
>>>
>>> I suggested an LSP type that has the properties "timestamp
>>> and pass to application", but I think that Sasha raises a good
>>> point about using router alert. However RA has no implicit
>>> timestamp, so maybe we need a new type of RA that has the
>>> properties "timestamp, and pass top application indicated by
>>> the next label". That would be quite useful in a number
>>> of OAM applications. There is possibly some GAL variant
>>> of that design that should also be considered.
>>>
>>> The application could worry about the path and where it
>>> was necessary to skip some hops a hierarchical LSP would
>>> accomplish that, with the specific benefit that the application
>>> would consciously do this, and may be able to apply some
>>> form of compensation within the network.
>>>
>>> - Stewart
>>>
>>> On 04/08/2013 10:40, S. Davari wrote:
>>>> Hi Sasha
>>>>
>>>> Perhaps you have not understood the draft well. The main reason that th=
e
>>> draft chose to use a dedicated LSP that is signaled via RSVPTE indicatin=
g it is a
>>> timing LSP is to be backward compatible with non-1588 nodes. Such nodes
>>> simply just switch the packet.
>>>> Your proposal had been considered and was rejected since it was not
>>> backward compatible. Nodes receiving alert label or TTL =3D 1 send the p=
acket to
>>> CPU. You can refer to the meeting notes.
>>>> Regards,
>>>> Shahram
>>>>
>>>>
>>>> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein
>>> <Alexander.Vainshtein@ecitele.com> wrote:
>>>>> Stewart and all,
>>>>> I concur with Stewart's statement that the draft does not define "full
>>> interaction with MPLS architecture".
>>>>> E.g., one of the objectives of the draft is to provide a technique tha=
t would
>>> be backward-compatible with old LSRs that cannot provide on-path support=
 for
>>> timing distribution, while the other objective is to make every LSR on t=
he path
>>> aware that some MPLS packets are carrying timing-related messages and he=
nce
>>> require on-path support.
>>>>> IMHO and FWIW, the MPLS data plane architecture allows just two method=
s
>>> for making a transit LSR to provide special processing to a labeled pack=
et:
>>>>> - It would carry some kind of an "alert label" on top of the label sta=
ck and,
>>> specifically, on top of any labels used for actual forwarding
>>>>> OR,
>>>>> - The TTL in the top label stack entry has been set to 1.
>>>>>
>>>>>
>>>>> The draft does not follow any of these approaches.
>>>>>
>>>>> I must also admit that Section 12 "OAM, Control and Management" of the
>>> draft looks somewhat in
>>>>> E.g., I could not understand from the text whether BFD, LSP-Ping etc.=
 OAM
>>> messages would be subjected to on-path support procedures for timing
>>> messages or not.
>>>>> My 2c,
>>>>>      Sasha
>>>>>
>>>>>> -----Original Message-----
>>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Beh=
alf
>>> Of
>>>>>> Stewart Bryant
>>>>>> Sent: Friday, August 02, 2013 12:01 PM
>>>>>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
>>>>>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.or=
g;
>>> draft-
>>>>>> ietf-tictoc-1588overmpls@tools.ietf.org
>>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
>>>>>>
>>>>>> SB> This draft does not seem to provide a precise definition
>>>>>> SB> the properties of the new LSP type that it wishes to
>>>>>> SB> define, in particular it does it define the PHB of those
>>>>>> SB> LSPs, nor the full interaction with the MPLS
>>>>>> SB> architecture.
>>>>>> SB>
>>>>>> SB> I have not tracked TICTOC for a while but I thought that
>>>>>> SB> the original plan was to define the concept of an offset
>>>>>> SB> into a packet to do the correction.
>>>>>> SB>
>>>>>> SB> It is disappointing that the opportunity was not taken
>>>>>> SB> to define a timing shim inside the timing LSP so that
>>>>>> SB> a time correction could be added to any packet such that
>>>>>> SB> the MPLS system was isolated from the details of the
>>>>>> SB> complexity of the particular time transfer type.
>>>>>>
>>>>>> SB> I think that much more clarify is needed in terms of
>>>>>> SB> definition of the new LSP type, since it is unclear
>>>>>> SB> from this text how to implement one.
>>>>>> SB>
>>>>>> SB> There are a lot of other MPLS services such as
>>>>>> SB> LSP ping that need to be considered.
>>>>>> SB>
>>>>>> SB> Please see inline for more comments. However these
>>>>>> SB> comments are made in the context of the text as written
>>>>>> SB> whilst I have a fundamental concern that this approach
>>>>>> SB> lacks an MPLS architectural soundness that need
>>>>>> SB> greater thought with significant impact on the
>>>>>> SB> draft.
>>>>>>
>>>>>> - Stewart
>>>>>>
>>>>>>
>>>>>> TICTOC Working Group                                           S. Dav=
ari
>>>>>> Internet-Draft                                                   A. O=
ren
>>>>>> Intended status: Standards Track                          Broadcom Co=
rp.
>>>>>> Expires: December 17, 2013                                     M. Bha=
tia
>>>>>>                                                                P. Rob=
erts
>>>>>>                                                            Alcatel-Lu=
cent
>>>>>>                                                                L. Mon=
tini
>>>>>>                                                                L. Mar=
tini
>>>>>>                                                             Cisco Sys=
tems
>>>>>>                                                             June 15,=
 2013
>>>>>>
>>>>>>
>>>>>>              Transporting Timing messages over MPLS Networks
>>>>>>                     draft-ietf-tictoc-1588overmpls-05
>>>>>>
>>>>>> Abstract
>>>>>>
>>>>>>     This document defines the method for transporting Timing messages
>>>>>>     such as PTP and NTP over an MPLS network.  The method allows for=
 the
>>>>>>     easy identification of these PDUs at the port level to allow for=
 port
>>>>>>
>>>>>> SB> What is a port
>>>>>>
>>>>>>     level processing of these PDUs in both LERs and LSRs.
>>>>>>
>>>>>>     The basic idea is to transport Timing messages inside dedicated M=
PLS
>>>>>>     LSPs.  These LSPs only carry Timing messages and possibly Control=
 and
>>>>>>     Management packets, but they do not carry customer traffic.
>>>>>>
>>>>>> SB> More specifically they only carry traffic associated with the
>>>>>> SB> timing service and its support.
>>>>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
>>>>>> SB> also it gets carried in a structure that causes it to get
>>>>>> SB> timestamped.
>>>>>>
>>>>>>     Two methods for transporting Timing messages over MPLS are define=
d.
>>>>>>
>>>>>> SB> Perhaps the right approach is to define the new LSP type and then
>>>>>> SB> seperately to define  the mapping of the various timing services
>>>>>> SB> over that LSP type.
>>>>>>
>>>>>>     The first method is to transport Timing messages directly over th=
e
>>>>>>     dedicated MPLS LSP via UDP/IP encapsulation, which is suitable fo=
r
>>>>>>     MPLS networks.  The second method is to transport Timing messages
>>>>>>     inside a PW via Ethernet encapsulation.
>>>>>>
>>>>>> SB> I think that we should note that there are some
>>>>>> SB> h/w reasons for this preference. A clean sheet approach
>>>>>> SB> would have been to use PTP over MPLS with no intermediate
>>>>>> SB> layers.
>>>>>>
>>>>>> Status of this Memo
>>>>>>
>>>>>>     This Internet-Draft is submitted in full conformance with the
>>>>>>     provisions of BCP 78 and BCP 79.
>>>>>>
>>>>>>     Internet-Drafts are working documents of the Internet Engineering
>>>>>>     Task Force (IETF).  Note that other groups may also distribute
>>>>>>     working documents as Internet-Drafts.  The list of current Intern=
et-
>>>>>>     Drafts is at http://datatracker.ietf.org/drafts/current/.
>>>>>>
>>>>>>     Internet-Drafts are draft documents valid for a maximum of six mo=
nths
>>>>>>     and may be updated, replaced, or obsoleted by other documents at=
 any
>>>>>>     time.  It is inappropriate to use Internet-Drafts as reference
>>>>>>     material or to cite them other than as "work in progress."
>>>>>>
>>>>>>     This Internet-Draft will expire on December 17, 2013.
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page=
 1]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> Copyright Notice
>>>>>>
>>>>>>     Copyright (c) 2013 IETF Trust and the persons identified as the
>>>>>>     document authors.  All rights reserved.
>>>>>>
>>>>>>     This document is subject to BCP 78 and the IETF Trust's Legal
>>>>>>     Provisions Relating to IETF Documents
>>>>>>     (http://trustee.ietf.org/license-info) in effect on the date of
>>>>>>     publication of this document.  Please review these documents
>>>>>>     carefully, as they describe your rights and restrictions with res=
pect
>>>>>>     to this document.  Code Components extracted from this document m=
ust
>>>>>>     include Simplified BSD License text as described in Section 4.e o=
f
>>>>>>     the Trust Legal Provisions and are provided without warranty as
>>>>>>     described in the Simplified BSD License.
>>>>>>
>>>>>>
>>>>>> Table of Contents
>>>>>>
>>>>>>     1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . .=
 .  5
>>>>>>
>>>>>>     2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . .=
 .  7
>>>>>>
>>>>>>     3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . .=
 .  8
>>>>>>
>>>>>>     4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . .=
 .  9
>>>>>>
>>>>>>     5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . .=
 . 12
>>>>>>
>>>>>>     6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . .=
 . 13
>>>>>>       6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . .=
 . 13
>>>>>>       6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . .=
 . 13
>>>>>>       6.3.  Other Timing Encapsulation methods . . . . . . . . . . .=
 . 14
>>>>>>
>>>>>>     7.  Timing message Processing  . . . . . . . . . . . . . . . . .=
 . 15
>>>>>>
>>>>>>     8.  Protection and Redundancy  . . . . . . . . . . . . . . . . .=
 . 16
>>>>>>
>>>>>>     9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 17
>>>>>>
>>>>>>     10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 18
>>>>>>
>>>>>>     11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 19
>>>>>>
>>>>>>     12. OAM, Control and Management  . . . . . . . . . . . . . . . .=
 . 20
>>>>>>
>>>>>>     13. QoS Considerations . . . . . . . . . . . . . . . . . . . . .=
 . 21
>>>>>>
>>>>>>     14. FCS and Checksum Recalculation . . . . . . . . . . . . . . .=
 . 22
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page=
 2]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>>     15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . .=
 . 23
>>>>>>       15.1. Behavior of Timing-capable/aware LER . . . . . . . . . .=
 . 23
>>>>>>       15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . .=
 . 23
>>>>>>       15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . .=
 . 24
>>>>>>
>>>>>>     16. Other considerations . . . . . . . . . . . . . . . . . . . .=
 . 25
>>>>>>
>>>>>>     17. Security Considerations  . . . . . . . . . . . . . . . . . .=
 . 26
>>>>>>
>>>>>>     18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . .=
 . 27
>>>>>>
>>>>>>     19. IANA Considerations  . . . . . . . . . . . . . . . . . . . .=
 . 28
>>>>>>
>>>>>>     20. References . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 29
>>>>>>       20.1. Normative References . . . . . . . . . . . . . . . . . .=
 . 29
>>>>>>       20.2. Informative References . . . . . . . . . . . . . . . . .=
 . 29
>>>>>>
>>>>>>     Appendix 1.  Routing extensions for Timing-aware Routers . . . .=
 . 32
>>>>>>
>>>>>>     Appendix 2.  Signaling Extensions for Creating Timing LSPs . . .=
 . 33
>>>>>>
>>>>>>     Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . .=
 . 34
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page=
 3]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>>     The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
>>> NOT",
>>>>>>     "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL"
>>> in
>>>>>> this
>>>>>>     document are to be interpreted as described in RFC2119 [RFC2119].
>>>>>>
>>>>>>     When used in lower case, these words convey their typical use in
>>>>>>     common language, and are not to be interpreted as described in
>>>>>>     RFC2119 [RFC2119].
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page=
 4]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 1.  Introduction
>>>>>>
>>>>>>     The objective of Precision Time Protocol (PTP) and Network Timing
>>>>>>     Protocol (NTP) are to synchronize independent clocks running on
>>>>>>     separate nodes of a distributed system.
>>>>>>
>>>>>>     [IEEE-1588] defines PTP messages for frequency, phase and time
>>>>>>     synchronization.  The PTP messages include PTP PDUs over UDP/IP
>>>>>>     (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex=
 F of
>>>>>>     [IEEE-1588]).
>>>>>>
>>>>>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
>>>>>> SB> of other PTP mappings if they provide better optimisation.
>>>>>>
>>>>>>     This document defines mapping and transport of the PTP
>>>>>>     messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>>>>>>     defines several clock types: ordinary clocks, boundary clocks, en=
d-
>>>>>>     to-end transparent clocks, and peer-to-peer transparent clocks.
>>>>>>     Transparent clocks require intermediate nodes to update correctio=
n
>>>>>>     field inside PTP message that reflects the transit time in the no=
de.
>>>>>>
>>>>>>     [RFC5905] defines NTP messages for clock and time synchronization=
.
>>>>>>     The PTP messages (PDUs) are transported over UDP/IP.  This docume=
nt
>>>>>> SB> Should that be NTP messages?
>>>>>> SB> It needs to be made clear as soon as you introduce NTP that
>>>>>> SB> they use different time representations.
>>>>>>
>>>>>>     defines mapping and transport of the NTP messages defined in
>>>>>>     [RFC5905] over MPLS networks.
>>>>>>
>>>>>>     One key attribute of all of these Timing messages is that the Tim=
e
>>>>>>     stamp processing should occur as close as possible to the actual
>>>>>>     transmission and reception at the physical port interface.  This
>>>>>>     targets optimal time and/or frequency recovery by avoiding variab=
le
>>>>>>     delay introduced by queues internal to the clocks.
>>>>>>
>>>>>> SB> As I recall NTP has no epoch point defined, and I am not sure
>>>>>> SB> where that point is in the case of PTP in this mapping
>>>>>> SB> Hopefully this will get defined in due course.
>>>>>>
>>>>>>     To facilitate the fast and efficient recognition of Timing messag=
es
>>>>>>     at the port level when the Timing messages are carried over MPLS
>>>>>>     LSPs,
>>>>>>
>>>>>> SB> Over a new LSP type with time optimied characteristics
>>>>>>
>>>>>>     this document defines the specific encapsulations that should
>>>>>>     be used.
>>>>>> SB> Hopefully it will also define the PHP
>>>>>>
>>>>>>     In addition, it can be expected that there will exist LSR/
>>>>>>     LERs where only a subset of the physical ports will have the port=
-
>>>>>>     based Timing message processing capabilities.
>>>>>> SB> Do you need to clarify that this only works at base and not in
>>>>>> SB> a label heirarchy.
>>>>>>
>>>>>>
>>>>>>     In order to ensure
>>>>>>     that the LSPs carrying Timing packets always enter and exit ports
>>>>>>     with this capability, routing extensions are defined to advertise
>>>>>>     this capability on a port basis and to allow for the establishmen=
t of
>>>>>>     LSPs that only transit such ports.  While this path establishment
>>>>>>     restriction may be applied only at the LER Ingress and/or egress
>>>>>>     ports, it becomes more important when using transparent clock cap=
able
>>>>>>     LSRs in the path.
>>>>>> SB> I do not understand the implications of the last
>>>>>> SB> sentences - starting ", it becomes"
>>>>>>
>>>>>>
>>>>>>     Port based Timing message processing involves Timing message
>>>>>>     recognition.  Once the Timing messages are recognized they can be
>>>>>>     modified based on the reception or transmission Time-stamp.
>>>>>>
>>>>>>     This document provides two methods for transporting Timing messag=
es
>>>>>>     over MPLS.  One is applicable to MPLS environment and the other o=
ne
>>>>>>     is applicable to MPLS/MPLS-TP environment
>>>>>>
>>>>>> SB> I think the sentence is incomplete.
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page=
 5]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>>     The solution involves transporting Timing messages over dedicated
>>>>>>     LSPs called Timing LSPs.  These LSPs carry Timing messages and MA=
Y
>>>>>>     carry Management and control messages, but not data plane client
>>>>>>     traffic.
>>>>>>
>>>>>> SB> It is not clear why this restriction applies.
>>>>>>
>>>>>>     Timing LSPs can be established statically or via signaling.
>>>>>> SB> s/statically/by provisioning/network management/
>>>>>>
>>>>>>     Extensions to control plane (OSPF, ISIS, etc.) is required to ena=
ble
>>>>>>     routers to distribute their Timing processing capabilities over M=
PLS
>>>>>>     to other routers.  However such extensions are outside the scope=
 of
>>>>>>     this document.
>>>>>>
>>>>>>     When signaling is used to setup the PTP LSP, Extensions to signal=
ing
>>>>>> SB> is it a PTP LSP or a Timing LSP?
>>>>>>
>>>>>>     protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
>>>>>>     However such extensions are outside the scope of this document.
>>>>>>
>>>>>> SB> for mpls-tp GMPLS is the signalling protocol
>>>>>>
>>>>>>     While the techniques included herein allow for the establishment=
 of
>>>>>>     paths optimized to include Time-stamping capable links, the
>>>>>>     performance of the Slave clocks is outside the scope of this
>>>>>>     document.
>>>>>>
>>>>>>     At the time of publishing this specification, Transparent Clockin=
g
>>>>>>     (TC) is only defined for PTP.  Therefore at this time any part of
>>>>>>     this specification that talks about Transparent Clocking applies=
 only
>>>>>>     to PTP.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page=
 6]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 2.  Terminology
>>>>>>
>>>>>>     1588: The timing and synchronization as defined by IEEE 1588.
>>>>>>
>>>>>> SB> I think that there is a more formal name for the 1588 group
>>>>>> SB> that needs to be used here.
>>>>>> SB> Also do we need to talk about 1588-200? as there is
>>>>>> SB> an update in progress
>>>>>>
>>>>>>     NTP: The timing and synchronization protocol defined by IETF RFC-=
1305
>>>>>>     and RFC-5905.
>>>>>>
>>>>>>     PTP: The timing and synchronization protocol used by 1588.
>>>>>> SB> need the proper name for 1588
>>>>>>
>>>>>>     Master Clock: The source of 1588 timing to a set of slave clocks.
>>>>>>
>>>>>>     Master Port: A port on a ordinary or boundary clock that is in Ma=
ster
>>>>>>     state.  This is the source of timing toward slave ports.
>>>>>>
>>>>>> SB> I am not sure the reader knows what a port is
>>>>>>
>>>>>>     Slave Clock: A receiver of 1588 timing from a master clock.
>>>>>>
>>>>>>     Slave Port: A port on a boundary clock or ordinary clock that is
>>>>>>     receiving timing from a master clock.
>>>>>>
>>>>>>     Ordinary Clock: A device with a single PTP port.
>>>>>>
>>>>>>     Transparent Clock.  A device that measures the time taken for a P=
TP
>>>>>>     event message to transit the device and then updates the
>>>>>>     correctionField of the message with this transit time.
>>>>>>
>>>>>>     Boundary Clock: A device with more than one PTP port.  Generally
>>>>>>     boundary clocks will have one port in slave state to receive timi=
ng
>>>>>>     and then other ports in master state to re-distribute the timing.
>>>>>>
>>>>>>     PTP LSP: An LSP dedicated to carry PTP messages
>>>>>>
>>>>>> SB> PTP or timing?
>>>>>>
>>>>>>
>>>>>>     PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>>>>>     messages.
>>>>>>
>>>>>> SB> Ah I don't think that PWE3 know what one of these is
>>>>>>
>>>>>>     CW: Pseudowire Control Word
>>>>>>
>>>>>>     LAG: Link Aggregation
>>>>>>
>>>>>>     ECMP: Equal Cost Multipath
>>>>>>
>>>>>>     CF: Correction Field, a field inside certain PTP messages (messag=
e
>>>>>>     type 0-3)that holds the accumulative transit time inside intermed=
iate
>>>>>>     switches
>>>>>>
>>>>>>     Timing messages: Timing Protocol messages that are exchanged
>>> between
>>>>>>     routers in order to establish a synchronized clock.
>>>>>>
>>>>>> SB> A number of these definitions look like copies of IEEE1588
>>>>>> SB> definitions. We need to provide references and note the
>>>>>> SB> priority of the IEEE base reference.
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page=
 7]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 3.  Problem Statement
>>>>>>
>>>>>>     [IEEE-1588] has defined methods for transporting PTP messages ove=
r
>>>>>>     Ethernet and IP networks.  [RFC5905] has defined the method of
>>>>>>     transporting NTP messages over IP networks.  There is a need to
>>>>>>     transport Timing messages over MPLS networks while supporting the
>>>>>>     Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (O=
C)
>>>>>>     functionality in the LER and LSRs in the MPLS network.
>>>>>>
>>>>>>     There are multiple ways of transporting Timing over MPLS.  Howeve=
r,
>>>>>>     there is a requirement to limit the possible encapsulation option=
s to
>>>>>>     simplify the Timing message identification and processing require=
d at
>>>>>>     the port level.
>>>>>>
>>>>>>     When Timing-awareness is needed, Timing messages should not be
>>>>>>     transported over LSPs or PWs that are carrying customer traffic
>>>>>>     because LSRs perform Label switching based on the top label in th=
e
>>>>>>     stack.
>>>>>>
>>>>>> SB> Have you explained why?
>>>>>>
>>>>>>     To detect Timing messages inside such LSPs require special
>>>>>>     hardware to do deep packet inspection at line rate.  Even if such
>>>>>>     hardware exists, the payload can't be deterministically identifie=
d by
>>>>>>     LSRs because the payload type is a context of the PW label, and t=
he
>>>>>>     PW label and its context are only known to the Edge routers (PEs/
>>>>>>     LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, C=
ES,
>>>>>>     etc).  Even if one restricts an LSP to only carry Ethernet PWs, t=
he
>>>>>>     LSRs dont have the knowledge of whether PW Control Word (CW) is
>>>>>>     present or not and therefore can not deterministically identify t=
he
>>>>>>     payload.
>>>>>>
>>>>>>     A generic method is defined in this document that does not requir=
e
>>>>>>     deep packet inspection at line rate, and can deterministically
>>>>>>     identify Timing messages.  This method can be used to detect Timi=
ng
>>>>>>     Messages in both one-step and two-step clock implementations of
>>>>>>     ordinary, boundary and transparent clocks.
>>>>>>
>>>>>> SB> Needs a ref and I am sure many MPLS specialists will not understa=
nd
>>>>>> SB> the msg types.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page=
 8]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 4.  Timing over MPLS Architecture
>>>>>>
>>>>>>     Timing messages are exchange between Timing ports on ordinary and
>>>>>>
>>>>>> SB> Have you defined a timing port?
>>>>>>
>>>>>>     boundary clocks.  Boundary clocks terminate the Timing messages a=
nd
>>>>>>     act as master for other boundary clocks or for slave clocks.  End=
-to-
>>>>>>     End Transparent clocks do not terminate the Timing messages but t=
hey
>>>>>>     do modify the contents of the Timing messages as they transit acr=
oss
>>>>>>     the transparent clock.
>>>>>>
>>>>>>     Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent C=
lock
>>>>>>
>>>>>>     (TC) could be implemented in either LERs or LSRs.
>>>>>>
>>>>>> SB> LER and LSR need to be expanded
>>>>>>
>>>>>>     An example is shown in Figure 1, where the LERs act as Ordinary C=
lock
>>>>>>     (OC) and are the initiating/terminating point for Timing messages=
.
>>>>>>     The ingress LER encapsulates the Timing messages in Timing LSP an=
d
>>>>>>     the Egress LER terminates the Timing LSP.  The LSRs act as
>>>>>>     Transparent Clock (TC) and just update the Timing field in the Ti=
ming
>>>>>>     messages.
>>>>>>
>>>>>>
>>>>>>        +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>        |Switch, |     |       |     |       |     |       |     |Swit=
ch, |
>>>>>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rou=
ter |
>>>>>>        |        |     |  OC   |     |  TC   |     |  OC   |     |   =
     |
>>>>>>        +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>                       /                                 \
>>>>>>        +-------+     /                                   \     +-----=
--+
>>>>>>        |  LER  |    /                                     \    |  LER=
  |
>>>>>>        | Master|---/                                       \---| Slav=
e |
>>>>>>        | Clock |                                               | Cloc=
k |
>>>>>>        +-------+                                               +-----=
--+
>>>>>>
>>>>>>       Figure (1) - Deployment example 1 of timing over MPLS network
>>>>>>
>>>>>>     Another example is shown in Figure2, where LERs terminate the Tim=
ing
>>>>>>     messages received from switch/routers that are outside of the MPL=
S
>>>>>>     network acting as OC or BC.  In this example LERs regenerate the
>>>>>>     clock and initiate timing messages encapsulated in Timing LSP tow=
ard
>>>>>>     the MPLS network, while the LSRs act as Transparent Clock (TC) an=
d
>>>>>>     just update the Timing field in the Timing messages, which are
>>>>>>     already encapsulated in Timing LSPs.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013               [Page=
 9]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>>       +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>       |Switch, |     |       |     |       |     |       |     |Switc=
h, |
>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
>>>>>>       | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/B=
C  |
>>>>>>       +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>
>>>>>>       Figure (2) - Deployment example 2 of timing over MPLS network
>>>>>>
>>>>>>
>>>>>>     Another example is shown in Figure 3, where LERs do not terminate=
 the
>>>>>>     Timing messages received from switch/routers that are outside of=
 the
>>>>>>     MPLS network acting as OC, TC or BC.  The LERs act as TC and upda=
te
>>>>>>     the Timing field in the Timing messages as they transit the LER,
>>>>>>     while encapsulating them in timing LSP.  The LSRs also act as
>>>>>>     Transparent Clock (TC) and just update the Timing field in the Ti=
ming
>>>>>>     messages which are already encapsulated in Timing LSPs.
>>>>>>
>>>>>>        +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>        |Switch, |     |       |     |       |     |       |     |Swit=
ch, |
>>>>>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rou=
ter |
>>>>>>        |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/T=
C/BC|
>>>>>>        +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>
>>>>>>      Figure (3) - Deployment example 3 of timing over MPLS network
>>>>>>
>>>>>>     Another example is shown in Figure 4, where LERs and LSRs support
>>>>>>     Boundary Clocks.  A single-hop LSP is created between two adjacen=
t
>>>>>>     LSRs engaged in BC operation.  Other methods such as PTP transpor=
t
>>>>>>     over Ethernet MAY be used for transporting timing messages if the
>>>>>>     link between the two routers is Ethernet.
>>>>>>
>>>>>>       +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>       |Switch, |     |       |     |       |     |       |     |Switc=
h, |
>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
>>>>>>       | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/B=
C  |
>>>>>>       +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>
>>>>>>     Figure (4) - Deployment example 3 of timing over MPLS network
>>>>>>
>>>>>>     An MPLS domain MAY serve multiple customers.  In these cases the
>>> MPLS
>>>>>>     domain (maintained by a service provider) may provide timing serv=
ices
>>>>>>     to multiple customers, each having their own Timing domain.
>>>>>>
>>>>>>     The Timing over MPLS architecture assumes full mesh of Timing LSP=
s
>>>>>>     between all LERs supporting this specification.
>>>>>>
>>>>>> SB> Note sure this is right - the salves surely do not need to
>>>>>> SB> exchange timing amongst themselves
>>>>>>
>>>>>>     It supports
>>>>>>     Point-to- point (VPWS) and Multipoint (VPLS) services.
>>>>>>
>>>>>> SB> What does that mean? You do not carry user data traffic?
>>>>>> SB> Maybe it's the ordering of the statemnets that is causing
>>>>>> SB> confusion.
>>>>>>
>>>>>>     This means
>>>>>>     that a customer may purchase a Point-to-point Timing service betw=
een
>>>>>>     two customer sites or a Multipoint Timing service between more th=
an
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 10]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>>     two customer sites.
>>>>>>
>>>>>>     The Timing over MPLS architecture supports P2P or P2MP Timing LSP=
s.
>>>>>>     This means that the Timing Multicast messages such as PTP Multica=
st
>>>>>>     event messages can be transported over P2MP Timing LSP or be
>>>>>>     replicated and transported over many P2P Timing LSPs.
>>>>>>
>>>>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>>>>>> SB> nor a P2MP PW, although we are close.
>>>>>>
>>>>>>     Timing messages, that do not require Time stamping or Correction
>>>>>>     Field update MAY be transported over Timing LSPs to simplify hard=
ware
>>>>>>     and software.
>>>>>>
>>>>>>     PTP Announce messages that determine the Timing LSP terminating
>>> point
>>>>>>     behavior such as BC/OC/TC SHOULD be transported over the Timing L=
SP
>>>>>>     to simplify hardware and software.
>>>>>>
>>>>>> SB> have you defined and referenced PTP announce msgs?
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 11]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 5.  Dedicated LSPs for Timing messages
>>>>>>
>>>>>>     Many methods have been considered for identifying the Timing
>>> messages
>>>>>>     when they are encapsulated in MPLS such as using GAL/G-ACH or a n=
ew
>>>>>>     reserved label.  These methods were not attractive since they eit=
her
>>>>>>     required deep packet inspection at line rate in the intermediate=
 LSRs
>>>>>>     or they required use of a scarce new reserved label.  Also one of=
 the
>>>>>>     goals was to reuse existing OAM mechanisms.
>>>>>>
>>>>>> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
>>>>>>
>>>>>>     The method defined in this document can be used by LER and LSRs t=
o
>>>>>>     identify Timing messages in MPLS tunnels by just looking at the t=
op
>>>>>>     label in the MPLS label stack, which only carry Timing messages a=
s
>>>>>>     well as OAM, but not data plane client traffic.
>>>>>>
>>>>>>     Compliant implementations MUST use dedicated LSPs to carry Timing
>>>>>>     messages over MPLS.
>>>>>>
>>>>>> SB> I think that we need a definition of the properies of these LSPs
>>>>>>
>>>>>>     These LSPs are herein referred to as "Timing
>>>>>>     LSPs" and the labels associated with these LSPs as "Timing LSP
>>>>>>     labels".  The Timing LSPs that runs between Ingress and Egress LE=
Rs
>>>>>>     MUST be co-routed.  Alternatively, a single bidirectional co-rout=
ed
>>>>>>     LSP can be used.
>>>>>>
>>>>>> SB> I though that you said you could use M2MP LSPs - these are not
>>>>>> SB> bidirectional.
>>>>>>
>>>>>>     Co-routing of the two directions is required to limit the differe=
nce
>>>>>>     in the delays in the Master clock to Slave clock direction compar=
ed
>>>>>>     to the Slave clock to Master clock direction.  The Timing LSP MAY=
 be
>>>>>>     MPLS/MPLS-TP LSP.
>>>>>>
>>>>>>     The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS=
.
>>>>>>     New Extensions to RSVP-TE/GMPLS TLVs are required; however they a=
re
>>>>>>     outside the scope of this document.
>>>>>>
>>>>>>     The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such=
 as
>>>>>>     BFD and LSP Ping but the LSP data plane client plane traffic MUST=
 be
>>>>>>     Timing packets only.
>>>>>>
>>>>>> SB> Why?
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 12]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 6.  Timing over LSP Encapsulation
>>>>>>
>>>>>> The encapsulations is not LSP is it?
>>>>>>
>>>>>>     This document defines two methods for carrying Timing messages ov=
er
>>>>>>     MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>>>>>     messages over Timing LSPs, and the second method, is carrying
>>>>>>     Ethernet encapsulated Timing messages over Ethernet PWs inside
>>> Timing
>>>>>>     LSPs.
>>>>>>
>>>>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>>>>>
>>>>>>     The simplest method of transporting Timing messages over MPLS is=
 to
>>>>>>     encapsulate Timing PDUs in UDP/IP and then encapsulate them in
>>> Timing
>>>>>>     LSP.  This format is shown in Figure 4.
>>>>>>
>>>>>>
>>>>>>                      +----------------------+
>>>>>>                      |   Timing LSP Label   |
>>>>>>                      +----------------------+
>>>>>>                      |        IPv4/6        |
>>>>>>                      +----------------------+
>>>>>>                      |         UDP          |
>>>>>>                      +----------------------+
>>>>>>                      |     Timing PDU       |
>>>>>>                      +----------------------+
>>>>>>
>>>>>>        Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>>>>>
>>>>>>
>>>>>>     This encapsulation is very simple and is useful when the network
>>>>>>     between Timing Master Clock and Slave Clock is MPLS network.
>>>>>>
>>>>>> SB> Simple is a judgement call
>>>>>>
>>>>>>     In order for an LER/LSR to process Timing messages, the Timing LS=
P
>>>>>>     Label must be at the top label of the label stack.  The LER/LSR M=
UST
>>>>>>     know that the Timing LSP Label is used for carrying Timing messag=
es.
>>>>>>     This can be accomplished via static configuration or via RSVP-TE
>>>>>>     signaling.
>>>>>>
>>>>>>     The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>>>>>     [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>>>>>>     [RFC5905].
>>>>>>
>>>>>> 6.2.  Timing over PW Encapsulation
>>>>>>
>>>>>>     Another method of transporting Timing over MPLS networks is by
>>>>>>     encapsulating Timing PDUs in PW which in turn is transported over
>>>>>>     Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448]=
,
>>>>>>     shown in Fig 5(A) MUST be used and the Ethernet encapsulation of=
 PTP
>>>>>>     MUST follow Annex F of [IEEE-1588].
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 13]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>>     The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
>>> the
>>>>>>     Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>>>>>>     Timing over PW encapsulation MUST use the Control Word (CW) as
>>>>>>     specified in [RFC4448] to ensure proper detection of PTP messages
>>>>>>     inside the MPLS packets for Timing over LSP and Timing over PW
>>>>>>     encapsulation.
>>>>>>
>>>>>> SB> That needs explanation
>>>>>>
>>>>>>     The use of Sequence Number in the CW is optional.
>>>>>>
>>>>>> SB> Given that s/n are never in practice deployed, you could probably
>>>>>> SB> simplify things by sayig that they are not used.
>>>>>>
>>>>>>     Timing over PW encapsulation for NTP MUST use NTP over UDP/IP ove=
r
>>> PW
>>>>>>     (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>>>>>
>>>>>>                      +----------------+  +----------------+
>>>>>>                       |Timing LSP Label|  |Timing LSP Label|
>>>>>>                       +----------------+  +----------------+
>>>>>>                       |    PW Label    |  |    PW Label    |
>>>>>>                       +----------------+  +----------------+
>>>>>>                       |  Control Word  |  |      IP        |
>>>>>>                       +----------------+  +----------------+
>>>>>>                       |    Ethernet    |  |      UDP       |
>>>>>>                       |     Header     |  +----------------+
>>>>>>                       +----------------+  |   Timing PDU   |
>>>>>>                       |S-VLAN(Optional)|  |                |
>>>>>>                       +----------------+  +----------------+
>>>>>>                       |C-VLAN(Optional)|        (B)
>>>>>>                       +----------------+
>>>>>>                       |   Timing PDU   |
>>>>>>                       |                |
>>>>>>                       +----------------+
>>>>>>                              (A)
>>>>>>
>>>>>>                Figure (5) - Timing over PW Encapsulations
>>>>>>
>>>>>>     In order for an LSR to process PTP messages, the top label of the
>>>>>>     label stack (the Tunnel Label) MUST be a Timing label.
>>>>>>
>>>>>> S> You said that before.
>>>>>>
>>>>>> 6.3.  Other Timing Encapsulation methods
>>>>>>
>>>>>>     In future other timing encapsulation methods may be introduced, s=
uch
>>>>>>     as a new shim header after the Bottom of Stack to carry the Timin=
g
>>>>>>     information.  Such new encapsulations are outside the scope of th=
is
>>>>>>     document.
>>>>>>
>>>>>>
>>>>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>>>>>> SB> out of the definition of the LSP
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 14]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> SB> I think we need a section on LSP processing
>>>>>>
>>>>>> 7.  Timing message Processing
>>>>>>
>>>>>>     Each Timing protocol such as PTP and NTP, define their set of Tim=
ing
>>>>>>     messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>>>>>>     FOLLOW_UP, etc messages.
>>>>>>
>>>>>>     Some of the Timing messages require time stamping or correction f=
ield
>>>>>>     update at port level and some dont.  It is the job of the LER/LSR=
 to
>>>>>>     parse the timing message and find out the type of the Timing mess=
age
>>>>>>     and decide whether and how to Time- stamp it (e.g., BC) or update
>>>>>>     correction field(e.g., TC).
>>>>>>
>>>>>>
>>>>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
>>>>>> SB> function rather than the LER?
>>>>>>
>>>>>>     For example the following PTP messages (called Event messages)
>>>>>>     require time-stamping or correction field update:
>>>>>>
>>>>>>     o  SYNC
>>>>>>
>>>>>>     o  DELAY_REQ (Delay Request)
>>>>>>
>>>>>>     o  PDELAY_REQ (Peer Delay Request)
>>>>>>
>>>>>>     o  PDELAY_RESP (Peer Delay Response)
>>>>>>
>>>>>>     SYNC and DELAY_REQ are exchanged between Master Clock and Slave
>>> Clock
>>>>>>     and MUST be transported over PTP LSPs.  PDELAY_REQ and
>>> PDELAY_RESP
>>>>>>     are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>>>>>>     Boundary, or Transparent) and SHOULD be transported over single h=
op
>>>>>>     PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP=
,
>>>>>>     and PDELAY_RESP_FOLLOW_UP messages MUST also be transported
>>> over
>>>>>> the
>>>>>>     PTP LSPs.
>>>>>>
>>>>>>     For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
>>>>>>     transported over two PTP LSPs that are in opposite directions.  T=
hese
>>>>>>     PTP LSPs, which are in opposite directions MUST be congruent and=
 co-
>>>>>>     routed.  Alternatively, a single bidirectional co-routed LSP can=
 be
>>>>>>     used.
>>>>>>
>>>>>>     Except as indicated above for the two-step PTP clocks, Non-Event=
 PTP
>>>>>>     message types do not need to be processed by intermediate routers=
.
>>>>>>     These message types MAY be carried in PTP Tunnel LSPs.
>>>>>>
>>>>>> SB> Are you saying that a timing P router has to be msg type sensitiv=
e?
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 15]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 8.  Protection and Redundancy
>>>>>>
>>>>>>
>>>>>> SB> This is a bit of a jump - I don't know how the LSP itself works y=
et!
>>>>>>
>>>>>>     In order to ensure continuous uninterrupted operation of slave
>>>>>>     clocks, usually as a general practice, slave clocks (or ports) tr=
ack
>>>>>>     redundant master clocks.
>>>>>>
>>>>>>     It is the responsibility of the network operator to ensure that
>>>>>>     physically disjoint Timing LSPs are established between a slave c=
lock
>>>>>>     (or port) and redundant master clocks (or ports).
>>>>>>
>>>>>>     When a slave clock (or port) listens to redundant master clocks o=
r
>>>>>>     ports, any prolonged Timing LSP outage will trigger the slave clo=
ck
>>>>>>     or port to switch to a redundant master clock or port.
>>>>>>
>>>>>>     LSP/PW protection such as Linear protection Switching (1:1, 1+1),
>>>>>>     Ring protection switching or MPLS Fast Reroute (FRR) generally sw=
itch
>>>>>>     alternative path that usually cause a change in delay, which if
>>>>>>     undetected by slave clock can reduce accuracy of the slave clock.
>>>>>>
>>>>>>     Therefore protection switching MAY be used, as long as phase jump=
s
>>>>>>     upon switchover due to differences in path latency are detected a=
nd
>>>>>>     compensated for (such compensation not being required if BCs or p=
eer-
>>>>>>     peer TCs are used throughout).
>>>>>>
>>>>>>     Note that any protection or reroute mechanism that adds additiona=
l
>>>>>>     MPLS label to the label stack, such as Facility Backup Fast Rerou=
te,
>>>>>>     MUST ensure that the pushed label is also a Timing Label to ensur=
e
>>>>>>     recognition of the MPLS frame as containing Timing messages, as i=
t
>>>>>>     transits the backup path.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 16]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 9.  ECMP
>>>>>>
>>>>>>     To ensure the optimal operation of slave clocks and avoid error
>>>>>>     introduced by forward and reverse path delay asymmetry, the physi=
cal
>>>>>>     path for Timing messages from master clock to slave Clock and vic=
e
>>>>>>     versa must be the same for all Event Timing messages listed in
>>>>>>     section 7.
>>>>>>
>>>>>>     Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
>>>>>>     Multipath).
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 17]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 10.  PHP
>>>>>>
>>>>>>     To ensure that the label on the top of the label stack is the Tim=
ing
>>>>>>     LSP Label, PHP MUST not be used.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 18]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 11.  Entropy
>>>>>>
>>>>>>     To ensure all Timing messages in a Timing LSP take the same path,
>>>>>>     Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>>>>>     Entropy Label MUST NOT be used for the PWs that are carried insid=
e
>>>>>>     Timing LSP [RFC6391].
>>>>>>
>>>>>> SB> This is incorrect - you mean that all msgs of the same timing
>>>>>> SB> flow need to have the same EL value.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 19]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 12.  OAM, Control and Management
>>>>>>
>>>>>>     In order to monitor Timing LSPs and their encapsulated PWs, they=
 MUST
>>>>>>     be able to carry OAM and management messages.  These management
>>>>>>     messages MUST be differentiated from Timing messages via already
>>>>>>     defined IETF methods.
>>>>>>
>>>>>>     For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY r=
un
>>>>>>     over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>>>>>>     Management protocols can easily be identified by the UDP Destinat=
ion
>>>>>>     Port number or by GAL/G-ACH respectively.
>>>>>>
>>>>>>     Also BFD, LSP-Ping and other management messages MAY run over the
>>> PWs
>>>>>>     encapsulated in Timing LSP via one of the defined VCCVs (Type 1,=
 3 or
>>>>>>     4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is g=
oing
>>>>>>     to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1=
) or
>>>>>>     GAL-ACH are used to identify such management messages.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 20]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 13.  QoS Considerations
>>>>>>
>>>>>>     In network deployments where not every LSR/LER is Timing-aware, i=
t is
>>>>>>     important to reduce the impact of the non-Timing-aware LSR/LERs o=
n
>>>>>>     the timing recovery in the slave clock.  The Timing messages are=
 time
>>>>>>     critical and must be treated with the highest priority.  Therefor=
e
>>>>>>     Timing over MPLS messages must be treated with the highest priori=
ty
>>>>>>     in the routers.  This can be achieved by proper setup of Timing L=
SPs.
>>>>>>
>>>>>>     It is recommended that the Timing LSPs are setup or configured
>>>>>>     properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC26=
97]
>>>>>>     for drop eligibility.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 21]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 14.  FCS and Checksum Recalculation
>>>>>>
>>>>>>     When time-stamp generation and timing packet adjustment is perfor=
med
>>>>>>     near the physical port hardware, the process MUST include
>>>>>>     recalculation of the Ethernet FCS.
>>>>>>
>>>>>> SB> The above is confusing - an LSR always recomputes the link layer
>>>>>> SB> CRC which may or may not be Ethernet.
>>>>>>
>>>>>>     Also FCS retention for the
>>>>>>     payload Ethernet described in [RFC4720] MUST NOT be used.
>>>>>>
>>>>>>     For UDP/IP encapsulation mode of Timing over MPLS, the UDP checks=
um
>>>>>>     may be required as per UDP transport standards.
>>>>>>
>>>>>> SB> You really need to be working on getting the IPv6 C?S computation
>>>>>> SB> removed from PTP msgs.
>>>>>>
>>>>>>     When UDP checksum is used, each Timing-aware LER/LSR must either
>>>>>>     incrementally update the UDP checksum after Time stamping or
>>>>>>     Correction Field update or verify the UDP checksum on reception f=
rom
>>>>>>     upstream and recalculate the checksum completely on transmission=
 to
>>>>>>     downstream node after Time stamping or Correction Field update.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 22]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 15.  Behavior of LER/LSR
>>>>>>
>>>>>>     Timing-capable/aware LERs and LSRs are routers that have one or m=
ore
>>>>>>
>>>>>> SB> You mean physical interfaces?
>>>>>>
>>>>>>     interfaces that can perform Timing operations (OC/BC/TC) on Timin=
g
>>>>>>     packets and are configured to do so.  Timing-capable/aware LERs a=
nd
>>>>>>     LSRs can advertise their Timing-capability per-interface via cont=
rol
>>>>>>     plane such as OSPF or IS-IS.
>>>>>> SB> ISIS and OSPF are routing protocols.
>>>>>>
>>>>>>    The Timing-capable/aware LERs can then
>>>>>>     signals Timing LSPs via RSVP-TE signaling.  Alternatively the Tim=
ing
>>>>>>     capability of LER and LSRs may be configured in a centralized
>>>>>>     controller and the Timing LSP may be setup using manual configura=
tion
>>>>>>     or other methods such as SDN.
>>>>>>
>>>>>> SB> it can also be configured individually rather then through
>>>>>> SB> a cebtral controllwe
>>>>>>
>>>>>> 15.1.  Behavior of Timing-capable/aware LER
>>>>>>
>>>>>>     When a Timing-capable/aware LER behaves as a Transparent clock an=
d
>>>>>>     receives a Timing message from a Timing-capable/aware non-MPLS
>>>>>>     interface, the LER updates the Correction Field (CF) and encapsul=
ates
>>>>>>     and forwards the timing message over previously established Timin=
g
>>>>>>     LSP.
>>>>>>
>>>>>> SB> You need to call out the details so that people properly
>>>>>> SB> understand the definition of the new LSP.
>>>>>>
>>>>>>     Also when a Timing message is received from a Timing-capable/
>>>>>>     aware MPLS interface, LER updates the Correction Filed (CF) and
>>>>>>     decapsulates the MPLS encapsulation and forwards the timing messa=
ge
>>>>>>     to a non-MPLS interface.
>>>>>>
>>>>>>     When a Timing-capable/aware LER behaves as a Boundary clock and
>>>>>>     receives a Timing message from a Timing-capable/aware non MPLS
>>>>>>     interface, the LER Timestamps the Timing packet and sends it to t=
he
>>>>>>     LERs Boundary clock processing module.  Also when a Timing messag=
e is
>>>>>>     received from a Timing- capable/aware MPLS interface, the LER
>>>>>>     Timestamps the Timing packet and sends it to the LERs Boundary cl=
ock
>>>>>>     processing module.
>>>>>>
>>>>>>     When a Timing-capable/aware LER behaves as an Ordinary Clock towa=
rd
>>>>>>     the MPLS network, and receives a Timing message from a Timing-
>>>>>>     capable/aware MPLS interface, the LER Timestamps the Timing packe=
t
>>>>>>     and sends it to the LERs Ordinary clock processing module.
>>>>>>
>>>>>> 15.2.  Behavior of Timing-capable/aware LSR
>>>>>>
>>>>>>     When a Timing-capable/aware LSR behaves as a Transparent clock an=
d
>>>>>>     receives a Timing message from a Timing-capable/aware MPLS
>>> interface,
>>>>>>     The LSR updates the Correction Filed (CF) and forwards the timing
>>>>>>     message over another MPLS interface.
>>>>>>
>>>>>>     When a Timing-capable/aware LSR behaves as a Boundary clock and
>>>>>>     receives a Timing message from a Timing-capable/aware MPLS
>>> interface.
>>>>>>     The LSR performs the functions of a Boundary Clock in terminating=
 the
>>>>>>     received Timing message and re-generating a new timing message ov=
er
>>>>>>     another (or the same) MPLS interface.
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 23]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>>>>>
>>>>>>     It is most beneficial when all LSRs in the path of a Timing LSP b=
e
>>>>>>     timing-Capable/aware LSRs.  This would ensure the highest quality
>>>>>>     time and clock synchronization by Timing Slave Clocks.  However,=
 this
>>>>>>     specification does not mandate that all LSRs in path of a Timing=
 LSP
>>>>>>     be Timing- capable/aware.
>>>>>>
>>>>>>     Non-Timing-capable/aware LSRs just switch the packets encapsulate=
d in
>>>>>>     Timing LSPs and dont perform any Timing operation (TC or BC).
>>>>>>     However as explained in QoS section the Timing over MPLS packets
>>> MUST
>>>>>>     be still be treated with the highest priority based on their Traf=
fic
>>>>>>     Class (TC) marking.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 24]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 16.  Other considerations
>>>>>>
>>>>>>     [IEEE-1588] defines an optional peer-to-peer Transparent clocking
>>>>>>     that requires peer delay measurement between two adjacent Timing-
>>>>>>     capable/ aware routers/switches.  Peer delay measurement messages
>>>>>>     need to be time stamped and terminated by the Timing-capable/awar=
e
>>>>>>     routers/ switches.  This means that two adjacent LSRs may be enga=
ged
>>>>>>     in a peer delay measurement.
>>>>>>
>>>>>>     For transporting such peer delay measurement messages a single-ho=
p
>>>>>>     LSP SHOULD to be created between the two adjacent LSRs engaged in
>>>>>>     peer delay measurement to carry peer delay measurement messages.
>>>>>>     Other methods such as PTP transport over Ethernet MAY be used for
>>>>>>     transporting peer delay measurement messages if the link between=
 the
>>>>>>     two routers is Ethernet.
>>>>>>
>>>>>>     In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/=
 ware
>>>>>>     routers/switches MUST maintain a list of all the neighbors it nee=
ds
>>>>>>     to send a PDelay_Req to, where each neighbor corresponds to a tim=
ing
>>>>>>     LSP.
>>>>>>
>>>>>>     The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as=
 long
>>>>>>     as either the Explicit Null label is the bottom of stack label
>>>>>>     (applicable only to UDP/IP encapsulation) or the label below the
>>>>>>     Explicit Null label is a PTP label.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 25]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 17.  Security Considerations
>>>>>>
>>>>>>     MPLS PW security considerations in general are discussed in [RFC3=
985]
>>>>>>     and [RFC4447],and those considerations also apply to this documen=
t.
>>>>>>
>>>>>>     An experimental security protocol is defined in [IEEE-1588].The P=
TP
>>>>>>     security extension and protocol provides group source authenticat=
ion,
>>>>>>     message integrity, and replay attack protection for PTP messages.
>>>>>>
>>>>>>     When the MPLS network (provider network) serves multiple customer=
s,
>>>>>>     it is important to maintain and process each customers clock and
>>>>>>     Timing messages separately from other customers to ensure there i=
s no
>>>>>>     cross- customer effect.  For example if an LER BC is synchronized=
 to
>>>>>>     a specific grandmaster, belonging to customer A, then the LER MUS=
T
>>>>>>     use that BC clock only for customer A to ensure that customer A
>>>>>>     cannot attack other customers by manipulating its time.
>>>>>>
>>>>>>     Timing messages MAY be encrypted or authenticated, provided that=
 the
>>>>>>     LERs/LSRs that are Timing capable/aware can authenticate/ decrypt=
 the
>>>>>>     timing messages.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 26]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 18.  Acknowledgements
>>>>>>
>>>>>>     The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizr=
ahi,
>>>>>>     Stefano Ruffini, Peter Meyer, and other members of IETF for revie=
wing
>>>>>>     and providing feedback on this draft.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 27]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 19.  IANA Considerations
>>>>>>
>>>>>>     There are no IANA requirements in this specification.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 28]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 20.  References
>>>>>>
>>>>>> 20.1.  Normative References
>>>>>>
>>>>>>     [IEEE-1588]
>>>>>>                IEEE 1588-2008, "IEEE Standard for a Precision Clock
>>>>>>                Synchronization Protocol for Networked Measurement and
>>>>>>                Control Systems".
>>>>>>
>>>>>>     [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>>>>>                Requirement Levels", BCP 14, RFC 2119, March 1997.
>>>>>>
>>>>>>     [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to=
-
>>>>>>                Edge (PWE3) Architecture", RFC 3985, March 2005.
>>>>>>
>>>>>>     [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discov=
ery
>>>>>>                Proxies (ND Proxy)", RFC 4389, April 2006.
>>>>>>
>>>>>>     [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G=
.
>>>>>>                Heron, "Pseudowire Setup and Maintenance Using the Lab=
el
>>>>>>                Distribution Protocol (LDP)", RFC 4447, April 2006.
>>>>>>
>>>>>>     [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>>>>>                "Encapsulation Methods for Transport of Ethernet over=
 MPLS
>>>>>>                Networks", RFC 4448, April 2006.
>>>>>>
>>>>>>     [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>>>>>                Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>>>>>                Retention", RFC 4720, November 2006.
>>>>>>
>>>>>>     [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circu=
it
>>>>>>                Connectivity Verification (VCCV): A Control Channel fo=
r
>>>>>>                Pseudowires", RFC 5085, December 2007.
>>>>>>
>>>>>>     [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detect=
ion
>>>>>>                (BFD)", RFC 5880, June 2010.
>>>>>>
>>>>>>     [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow=
,
>>>>>>                "Bidirectional Forwarding Detection (BFD) for MPLS Lab=
el
>>>>>>                Switched Paths (LSPs)", RFC 5884, June 2010.
>>>>>>
>>>>>> 20.2.  Informative References
>>>>>>
>>>>>>     [I-D.ietf-pwe3-fat-pw]
>>>>>>                Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Reg=
an,
>>>>>>                J., and S. Amante, "Flow Aware Transport of Pseudowire=
s
>>>>>>                over an MPLS Packet Switched Network",
>>>>>>                draft-ietf-pwe3-fat-pw-07 (work in progress), July 201=
1.
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 29]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>>     [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermedia=
te
>>>>>>                system routeing information exchange protocol for use=
 in
>>>>>>                conjunction with the Protocol for providing the
>>>>>>                Connectionless-mode Network Service (ISO 8473)".
>>>>>>
>>>>>>     [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP an=
d
>>>>>>                dual environments", RFC 1195, December 1990.
>>>>>>
>>>>>>     [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 199=
8.
>>>>>>
>>>>>>     [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
>>>>>>                Marker", RFC 2697, September 1999.
>>>>>>
>>>>>>     [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boud=
ec,
>>>>>>                J., Courtney, W., Davari, S., Firoiu, V., and D.
>>>>>>                Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>>>>>                Behavior)", RFC 3246, March 2002.
>>>>>>
>>>>>>     [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Enginee=
ring
>>>>>>                (TE) Extensions to OSPF Version 2", RFC 3630,
>>>>>>                September 2003.
>>>>>>
>>>>>>     [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermedia=
te
>>>>>>                System (IS-IS) Extensions for Traffic Engineering (TE)=
",
>>>>>>                RFC 3784, June 2004.
>>>>>>
>>>>>>     [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and=
 S.
>>>>>>                Shaffer, "Extensions to OSPF for Advertising Optional
>>>>>>                Router Capabilities", RFC 4970, July 2007.
>>>>>>
>>>>>>     [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
>>>>>>                System to Intermediate System (IS-IS) Extensions for
>>>>>>                Advertising Router Information", RFC 4971, July 2007.
>>>>>>
>>>>>>     [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
>>>>>>                Topology (MT) Routing in Intermediate System to
>>>>>>                Intermediate Systems (IS-ISs)", RFC 5120, February 200=
8.
>>>>>>
>>>>>>     [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>>>>>                Engineering", RFC 5305, October 2008.
>>>>>>
>>>>>>     [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>>>>>                "Traffic Engineering Extensions to OSPF Version 3",
>>>>>>                RFC 5329, September 2008.
>>>>>>
>>>>>>     [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSP=
F
>>>>>>                for IPv6", RFC 5340, July 2008.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 30]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>>     [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Net=
work
>>>>>>                Time Protocol Version 4: Protocol and Algorithms
>>>>>>                Specification", RFC 5905, June 2010.
>>>>>>
>>>>>>     [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Reg=
an,
>>>>>>                J., and S. Amante, "Flow-Aware Transport of Pseudowire=
s
>>>>>>                over an MPLS Packet Switched Network", RFC 6391,
>>>>>>                November 2011.
>>>>>>
>>>>>>     [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., a=
nd
>>>>>>                L. Yong, "The Use of Entropy Labels in MPLS Forwarding=
",
>>>>>>                RFC 6790, November 2012.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 31]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 1.  Routing extensions for Timing-aware Routers
>>>>>>
>>>>>>     MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340]=
 and
>>>>>>     IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (=
TE)
>>>>>>     link information used for constraint-based routing.
>>>>>>
>>>>>>     Indeed, it is useful to advertise data plane TE router link
>>>>>>     capabilities, such as the capability for a router to be Timing-aw=
are.
>>>>>>     This capability MUST then be taken into account during path
>>>>>>     computation to prefer or even require links that advertise themse=
lves
>>>>>>     as Timing-aware.  In this way the path can ensure the entry and e=
xit
>>>>>>     points into the LERs and, if desired, the links into the LSRs are
>>>>>>     able to perform port based time-stamping thus minimizing their im=
pact
>>>>>>     on the performance of the slave clock.
>>>>>>
>>>>>>     extensions are required to OSPF and IS-IS in order to advertise
>>>>>>     Timing-aware capabilities of a link.  Such extensions are outside=
 the
>>>>>>     scope of this document; however such extension SHOULD be able to
>>>>>>     signal the following information per Router Link:
>>>>>>
>>>>>>     o  Capable of processing PTP, NTP or other Timing flows
>>>>>>
>>>>>>     o  Capable of performing Transparent Clock operation
>>>>>>
>>>>>>     o  Capable of performing Boundary Clock operation
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 32]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> 2.  Signaling Extensions for Creating Timing LSPs
>>>>>>
>>>>>>     RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSV=
P-TE
>>>>>>     is used to setup Timing LSPs, some information that indicates tha=
t
>>>>>>     the LSP is carrying Timing flows MUST be included in the new
>>>>>>     Extensions to RSVP-TE:
>>>>>>
>>>>>>     The following information MAY also be included in the new Extensi=
ons
>>>>>>     to RSVP-TE:
>>>>>>
>>>>>>     o  Offset from Bottom of Stack (BoS) to the start of the Time-sta=
mp
>>>>>>        field
>>>>>>
>>>>>>     o  Number of VLANs in case of PW encapsulation
>>>>>>
>>>>>>     o  Timestamp field Type
>>>>>>
>>>>>>        *  Correction Field, Timestamp
>>>>>>
>>>>>>     o  Timestamp Field format
>>>>>>
>>>>>>        *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
>>>>>>           NTP, etc.
>>>>>>
>>>>>>     Note that in case the above optional information is signaled with
>>>>>>     RSVP-TE for a Timing LSP, all the Timing packets carried in that=
 LSP
>>>>>>     must have the same signaled characteristics.  For example if
>>>>>>     Timestamp format is signaled as 64-bit PTPv1, then all Timing pac=
kets
>>>>>>     must use 64-bit PTPv1 time-stamp.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 33]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>> Authors' Addresses
>>>>>>
>>>>>>     Shahram Davari
>>>>>>     Broadcom Corp.
>>>>>>     San Jose, CA  95134
>>>>>>     USA
>>>>>>
>>>>>>     Email: davari@broadcom.com
>>>>>>
>>>>>>
>>>>>>     Amit Oren
>>>>>>     Broadcom Corp.
>>>>>>     San Jose, CA  95134
>>>>>>     USA
>>>>>>
>>>>>>     Email: amito@broadcom.com
>>>>>>
>>>>>>
>>>>>>     Manav Bhatia
>>>>>>     Alcatel-Lucent
>>>>>>     Bangalore,
>>>>>>     India
>>>>>>
>>>>>>     Email: manav.bhatia@alcatel-lucent.com
>>>>>>
>>>>>>
>>>>>>     Peter Roberts
>>>>>>     Alcatel-Lucent
>>>>>>     Kanata,
>>>>>>     Canada
>>>>>>
>>>>>>     Email: peter.roberts@alcatel-lucent.com
>>>>>>
>>>>>>
>>>>>>     Laurent Montini
>>>>>>     Cisco Systems
>>>>>>     San Jose CA
>>>>>>     USA
>>>>>>
>>>>>>     Email: lmontini@cisco.com
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 34]
>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>
>>>>>>
>>>>>>     Luca
>>>>>>     Cisco Systems
>>>>>>     San Jose CA
>>>>>>     USA
>>>>>>
>>>>>>     Email: lmartini@cisco.com
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 35]
>>>>>> --
>>>>>> For corporate legal information go to:
>>>>>>
>>>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>>>
>>>>>> _______________________________________________
>>>>>> TICTOC mailing list
>>>>>> TICTOC@ietf.org
>>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>> This e-mail message is intended for the recipient only and contains
>>> information which is CONFIDENTIAL and which may be proprietary to ECI
>>> Telecom. If you have received this transmission in error, please inform=
 us by e-
>>> mail, phone or fax, and then delete the original and all copies thereof.
>>>>> _______________________________________________
>>>>> mpls mailing list
>>>>> mpls@ietf.org
>>>>> https://www.ietf.org/mailman/listinfo/mpls
>>>> .
>>>
>>> --
>>> For corporate legal information go to:
>>>
>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>
>>> _______________________________________________
>>> TICTOC mailing list
>>> TICTOC@ietf.org
>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>
>>>
>>> _______________________________________________
>>> TICTOC mailing list
>>> TICTOC@ietf.org
>>> https://www.ietf.org/mailman/listinfo/tictoc
>>
>> This e-mail message is intended for the recipient only and contains infor=
mation which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If=
 you have received this transmission in error, please inform us by e-mail, p=
hone or fax, and then delete the original and all copies thereof.
>>
> .
>


--
For corporate legal information go to:

http://www.cisco.com/web/about/doing_business/legal/cri/index.html

This e-mail message is intended for the recipient only and contains informat=
ion which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If yo=
u have received this transmission in error, please inform us by e-mail, phon=
e or fax, and then delete the original and all copies thereof.


From davarish@yahoo.com  Tue Aug 20 10:01:05 2013
Return-Path: <davarish@yahoo.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 863DC11E80E4 for <tictoc@ietfa.amsl.com>; Tue, 20 Aug 2013 10:01:05 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.203
X-Spam-Level: 
X-Spam-Status: No, score=-1.203 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, MIME_QP_LONG_LINE=1.396]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Av8Xyn7A8-Ty for <tictoc@ietfa.amsl.com>; Tue, 20 Aug 2013 10:01:00 -0700 (PDT)
Received: from nm18-vm3.bullet.mail.ne1.yahoo.com (nm18-vm3.bullet.mail.ne1.yahoo.com [98.138.91.148]) by ietfa.amsl.com (Postfix) with ESMTP id 249B121F8EB5 for <tictoc@ietf.org>; Tue, 20 Aug 2013 10:01:00 -0700 (PDT)
Received: from [98.138.90.49] by nm18.bullet.mail.ne1.yahoo.com with NNFMP; 20 Aug 2013 17:00:57 -0000
Received: from [98.138.226.60] by tm2.bullet.mail.ne1.yahoo.com with NNFMP; 20 Aug 2013 17:00:57 -0000
Received: from [127.0.0.1] by smtp211.mail.ne1.yahoo.com with NNFMP; 20 Aug 2013 17:00:57 -0000
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yahoo.com; s=s1024; t=1377018057; bh=M0bf6LnX77wu2l/Yao9cgZiMVg5N6yW9OnOa9AClIhU=; h=X-Yahoo-Newman-Id:X-Yahoo-Newman-Property:X-YMail-OSG:X-Yahoo-SMTP:X-Rocket-Received:References:Mime-Version:In-Reply-To:Content-Type:Content-Transfer-Encoding:Message-Id:Cc:X-Mailer:From:Subject:Date:To; b=pmwnsOf7H3U8mEDAY16GhCQm5kHWl9m2Iai1kzAJBFlRw9X0z5NhIZHd8XfHerukmF0p9D6HPVLTb+fbAahIvaQ2nJKC2Id/FvemKGpKLng2+9IQR51mEnY4SK2+v3uu56/EdSVqHTuIXvdmrVwQF83be4mTn2B/kTVs++1Rv08=
X-Yahoo-Newman-Id: 221381.57402.bm@smtp211.mail.ne1.yahoo.com
X-Yahoo-Newman-Property: ymail-3
X-YMail-OSG: ojNBIeAVM1mcvuHGcrWrxotzD54cneuv3UZd_vPT1alj2ZB GSnPkLIkrPuKO8tnrHADdIE9b253rOzbxIj37zkwhCGrt44Oz.jlbs5FKKZq NhDsscpKYFtFZyGInk1EJmJZdosIkoh26m.vnhunr29YPlWt1l_DYGYs9rcL N4_4nXmnejDcJvhN8aqcsXBIne5pUUy5N2Mm9klrPb2nKlFgsYX0bbtBbMlX diZQsPiQtD77Pzj.rC2hthEvBRY2LagIxFBhDMZfNQ1Tm5SdzG5wqp1aQcqp 7l3qQVyXFEqtMp.qzZOJt.bIDkEpbwX5E_EmSyU_alQSQQaqRT_qKvo4hZZm uHicQYiv8ry.paTje5AElFocLYXh35VhX2jDtPzDfcsUYFT05M2ypUuKCSHG uMncv6nt5dJaMke_n2Zpgnx5YD8ANv1XlUsbqF5Fe7Gucb_3haeeyCtwhAi. IC3g7oe4cxvaJjf40wgpJEiMRhULa.6HZ0vs3er5P4lKDP5flMGtUBmPAAwd xS8wn9vcRpGwreBBCNxWDjcQEwi7dsJUA0z8k.S0g2dZ0tipLhR1hgQ--
X-Yahoo-SMTP: ygPrP9CswBCWPbPtKJlJyLY0KMlg
X-Rocket-Received: from [192.168.0.103] (davarish@98.248.42.143 with ) by smtp211.mail.ne1.yahoo.com with SMTP; 20 Aug 2013 10:00:57 -0700 PDT
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com> <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com> <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com> <52139903.8040801@cisco.com> <F9336571731ADE42A5397FC831CEAA021513A083@ILPTWPVEXMB01.ecitele.com>
Mime-Version: 1.0 (1.0)
In-Reply-To: <F9336571731ADE42A5397FC831CEAA021513A083@ILPTWPVEXMB01.ecitele.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: quoted-printable
Message-Id: <B7A656C4-6D0E-4719-9E42-57689CDAEFA9@yahoo.com>
X-Mailer: iPhone Mail (10A403)
From: "S. Davari" <davarish@yahoo.com>
Date: Tue, 20 Aug 2013 10:00:41 -0700
To: Alexander Vainshtein <Alexander.Vainshtein@ecitele.com>
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 20 Aug 2013 17:01:05 -0000

Sasha

In (2) are you proposing to create multiple segment LSPs, where each if thos=
e segment LSPs span between two Timing aware LSR? So if I have 10 timing awa=
re LSR in the path I would need 9 LSPs?

Regards,
Shahram


On Aug 20, 2013, at 9:53 AM, Alexander Vainshtein <Alexander.Vainshtein@ecit=
ele.com> wrote:

> Stewart, Shahram and all,
>=20
> A couple of (hopefully, last) comments:
>=20
> 1. Supporting a "Timing Alert" label at the top of the label stack introdu=
ces a backward compatibility problem. We differ in our estimate of the impor=
tance of this problem, but we all agree that it exists.
>=20
> 2. On the other hand, this problem does not exist if the "Timing Alert" la=
bel is inserted at the bottom of the label stack. At the same time, HW that a=
nalyzes the label stack hopefully could analyze the bottom label (almost) as=
 easily as it would analyze the top label. Thus we could both provide backwa=
rd compatibility (incompatible transit LSRs simply would not notice the "Tim=
ing Alert" label") and avoid the need for dedicated "Timing-only" LSPs. In p=
articular, all the MPLS OAM tools (including MPLS-TP) would work without any=
 changes.=20
>=20
> 3. A "timing shim header" (between the bottom of the label stack and the b=
eginning of the timing  packet) has been mentioned by Stewart as an already p=
roposed mechanism. IMHO and FWIW its usage would eliminate the need to under=
stand format of the specific timing distribution protocol in the transit LSR=
s.  It would further solve the problem of secure transport of timing packets=
 since security mechanisms would be applied to its original encapsulation (o=
ver IP or over Ethernet) without involving the transit LSRs.
>=20
> Did I miss something substantial?
>=20
> Regards, and lots of thanks in advance,
>     Sasha
>=20
>=20
> ________________________________________
> From: Stewart Bryant [stbryant@cisco.com]
> Sent: Tuesday, August 20, 2013 6:27 PM
> To: S. Davari
> Cc: Alexander Vainshtein; Shahram Davari; mpls@ietf.org; tictoc@ietf.org
> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>=20
> On 06/08/2013 14:35, S. Davari wrote:
>> Sasha,
>>=20
>> If a router is not 1588 aware we want it to switch the packet with minimu=
m jitter and delay via high priority queue.
> Yes, but you can explicitly tunnel across such island of incompatibility
>=20
> Stewart
>=20
>>=20
>>=20
>> Sending it to CPU creates large jitter and delay since the CPUs are gener=
ally busy doing other things. Time stamping or updating CF in the CPU won't h=
elp since it does not cover the variable delay that is incurred from the tim=
e the packet enters the router to the time the packet gets processed by CPU.=

>>=20
>> I wish it was that easy!
>>=20
>> Regards,
>> Shahram
>>=20
>>=20
>> On Aug 6, 2013, at 12:38 AM, Alexander Vainshtein <Alexander.Vainshtein@e=
citele.com> wrote:
>>=20
>>> Shahram,
>>> I agree with you that the routers that do not recognize a Timing Alert l=
abel would send the packet to the CPU. (This would be easy to arrange since w=
e are speaking about a reserved label.)
>>>=20
>>> I do not, however, see this as a serious issue, because the SW running o=
n this CPU would (hopefully) be upgradable to handle the packet correctly i.=
e., to forward it to where it should be forwarded and, if it is forwarded as=
 a labeled packet, to prepend the Timing Alert Label on top of its label sta=
ck. It could even record the residence time (as observed by the CPU in the p=
roper place in the packet. This would mean that introducing additional error=
 to whatever timing information is associated with this packet - but this is=
 what you should anyway expect if there are non-compliant routers on your pa=
th, right?
>>>=20
>>> Regards,
>>>     Sasha
>>>=20
>>>> -----Original Message-----
>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behal=
f Of
>>>> Shahram Davari
>>>> Sent: Monday, August 05, 2013 11:29 PM
>>>> To: stbryant@cisco.com; S. Davari
>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls=

>>>>=20
>>>> Hi Stewart,
>>>>=20
>>>> Your suggestion of " I suggested an LSP type that has the properties "t=
imestamp
>>>> and pass to application", is a subset of the existing draft and should w=
ork. It
>>>> basically limits the time stamping and correction field update to LERs (=
while the
>>>> draft supports time stamping at LER and LSR).
>>>>=20
>>>> However Sasha's suggestion of using a reserved label (RAL, GAL or any o=
ther
>>>> reserved label) does not satisfy one of the major requirements, which i=
s
>>>> backward compatibility. Routers that don't understand this reserved lab=
el will
>>>> drop or copy to CPU such packets.
>>>>=20
>>>> Regards,
>>>> Shahram
>>>>=20
>>>> -----Original Message-----
>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behal=
f Of
>>>> Stewart Bryant
>>>> Sent: Monday, August 05, 2013 2:29 AM
>>>> To: S. Davari
>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls=

>>>>=20
>>>> My concern is that we are creating a new LSP type in MPLS
>>>> (which is a architectural change and thus needs to be very carefully
>>>> considered) without asking the question "how can I do this
>>>> in the most general way, to maximize flexibility/reuse"
>>>>=20
>>>> I suggested an LSP type that has the properties "timestamp
>>>> and pass to application", but I think that Sasha raises a good
>>>> point about using router alert. However RA has no implicit
>>>> timestamp, so maybe we need a new type of RA that has the
>>>> properties "timestamp, and pass top application indicated by
>>>> the next label". That would be quite useful in a number
>>>> of OAM applications. There is possibly some GAL variant
>>>> of that design that should also be considered.
>>>>=20
>>>> The application could worry about the path and where it
>>>> was necessary to skip some hops a hierarchical LSP would
>>>> accomplish that, with the specific benefit that the application
>>>> would consciously do this, and may be able to apply some
>>>> form of compensation within the network.
>>>>=20
>>>> - Stewart
>>>>=20
>>>> On 04/08/2013 10:40, S. Davari wrote:
>>>>> Hi Sasha
>>>>>=20
>>>>> Perhaps you have not understood the draft well. The main reason that t=
he
>>>> draft chose to use a dedicated LSP that is signaled via RSVPTE indicati=
ng it is a
>>>> timing LSP is to be backward compatible with non-1588 nodes. Such nodes=

>>>> simply just switch the packet.
>>>>> Your proposal had been considered and was rejected since it was not
>>>> backward compatible. Nodes receiving alert label or TTL =3D 1 send the p=
acket to
>>>> CPU. You can refer to the meeting notes.
>>>>> Regards,
>>>>> Shahram
>>>>>=20
>>>>>=20
>>>>> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein
>>>> <Alexander.Vainshtein@ecitele.com> wrote:
>>>>>> Stewart and all,
>>>>>> I concur with Stewart's statement that the draft does not define "ful=
l
>>>> interaction with MPLS architecture".
>>>>>> E.g., one of the objectives of the draft is to provide a technique th=
at would
>>>> be backward-compatible with old LSRs that cannot provide on-path suppor=
t for
>>>> timing distribution, while the other objective is to make every LSR on t=
he path
>>>> aware that some MPLS packets are carrying timing-related messages and h=
ence
>>>> require on-path support.
>>>>>> IMHO and FWIW, the MPLS data plane architecture allows just two metho=
ds
>>>> for making a transit LSR to provide special processing to a labeled pac=
ket:
>>>>>> - It would carry some kind of an "alert label" on top of the label st=
ack and,
>>>> specifically, on top of any labels used for actual forwarding
>>>>>> OR,
>>>>>> - The TTL in the top label stack entry has been set to 1.
>>>>>>=20
>>>>>>=20
>>>>>> The draft does not follow any of these approaches.
>>>>>>=20
>>>>>> I must also admit that Section 12 "OAM, Control and Management" of th=
e
>>>> draft looks somewhat in
>>>>>> E.g., I could not understand from the text whether BFD, LSP-Ping etc.=
 OAM
>>>> messages would be subjected to on-path support procedures for timing
>>>> messages or not.
>>>>>> My 2c,
>>>>>>     Sasha
>>>>>>=20
>>>>>>> -----Original Message-----
>>>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Be=
half
>>>> Of
>>>>>>> Stewart Bryant
>>>>>>> Sent: Friday, August 02, 2013 12:01 PM
>>>>>>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
>>>>>>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.o=
rg;
>>>> draft-
>>>>>>> ietf-tictoc-1588overmpls@tools.ietf.org
>>>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>>>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
>>>>>>>=20
>>>>>>> SB> This draft does not seem to provide a precise definition
>>>>>>> SB> the properties of the new LSP type that it wishes to
>>>>>>> SB> define, in particular it does it define the PHB of those
>>>>>>> SB> LSPs, nor the full interaction with the MPLS
>>>>>>> SB> architecture.
>>>>>>> SB>
>>>>>>> SB> I have not tracked TICTOC for a while but I thought that
>>>>>>> SB> the original plan was to define the concept of an offset
>>>>>>> SB> into a packet to do the correction.
>>>>>>> SB>
>>>>>>> SB> It is disappointing that the opportunity was not taken
>>>>>>> SB> to define a timing shim inside the timing LSP so that
>>>>>>> SB> a time correction could be added to any packet such that
>>>>>>> SB> the MPLS system was isolated from the details of the
>>>>>>> SB> complexity of the particular time transfer type.
>>>>>>>=20
>>>>>>> SB> I think that much more clarify is needed in terms of
>>>>>>> SB> definition of the new LSP type, since it is unclear
>>>>>>> SB> from this text how to implement one.
>>>>>>> SB>
>>>>>>> SB> There are a lot of other MPLS services such as
>>>>>>> SB> LSP ping that need to be considered.
>>>>>>> SB>
>>>>>>> SB> Please see inline for more comments. However these
>>>>>>> SB> comments are made in the context of the text as written
>>>>>>> SB> whilst I have a fundamental concern that this approach
>>>>>>> SB> lacks an MPLS architectural soundness that need
>>>>>>> SB> greater thought with significant impact on the
>>>>>>> SB> draft.
>>>>>>>=20
>>>>>>> - Stewart
>>>>>>>=20
>>>>>>>=20
>>>>>>> TICTOC Working Group                                           S. Da=
vari
>>>>>>> Internet-Draft                                                   A. O=
ren
>>>>>>> Intended status: Standards Track                          Broadcom C=
orp.
>>>>>>> Expires: December 17, 2013                                     M. Bh=
atia
>>>>>>>                                                               P. Rob=
erts
>>>>>>>                                                           Alcatel-Lu=
cent
>>>>>>>                                                               L. Mon=
tini
>>>>>>>                                                               L. Mar=
tini
>>>>>>>                                                            Cisco Sys=
tems
>>>>>>>                                                            June 15, 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>             Transporting Timing messages over MPLS Networks
>>>>>>>                    draft-ietf-tictoc-1588overmpls-05
>>>>>>>=20
>>>>>>> Abstract
>>>>>>>=20
>>>>>>>    This document defines the method for transporting Timing messages=

>>>>>>>    such as PTP and NTP over an MPLS network.  The method allows for t=
he
>>>>>>>    easy identification of these PDUs at the port level to allow for p=
ort
>>>>>>>=20
>>>>>>> SB> What is a port
>>>>>>>=20
>>>>>>>    level processing of these PDUs in both LERs and LSRs.
>>>>>>>=20
>>>>>>>    The basic idea is to transport Timing messages inside dedicated M=
PLS
>>>>>>>    LSPs.  These LSPs only carry Timing messages and possibly Control=
 and
>>>>>>>    Management packets, but they do not carry customer traffic.
>>>>>>>=20
>>>>>>> SB> More specifically they only carry traffic associated with the
>>>>>>> SB> timing service and its support.
>>>>>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
>>>>>>> SB> also it gets carried in a structure that causes it to get
>>>>>>> SB> timestamped.
>>>>>>>=20
>>>>>>>    Two methods for transporting Timing messages over MPLS are define=
d.
>>>>>>>=20
>>>>>>> SB> Perhaps the right approach is to define the new LSP type and the=
n
>>>>>>> SB> seperately to define  the mapping of the various timing services=

>>>>>>> SB> over that LSP type.
>>>>>>>=20
>>>>>>>    The first method is to transport Timing messages directly over th=
e
>>>>>>>    dedicated MPLS LSP via UDP/IP encapsulation, which is suitable fo=
r
>>>>>>>    MPLS networks.  The second method is to transport Timing messages=

>>>>>>>    inside a PW via Ethernet encapsulation.
>>>>>>>=20
>>>>>>> SB> I think that we should note that there are some
>>>>>>> SB> h/w reasons for this preference. A clean sheet approach
>>>>>>> SB> would have been to use PTP over MPLS with no intermediate
>>>>>>> SB> layers.
>>>>>>>=20
>>>>>>> Status of this Memo
>>>>>>>=20
>>>>>>>    This Internet-Draft is submitted in full conformance with the
>>>>>>>    provisions of BCP 78 and BCP 79.
>>>>>>>=20
>>>>>>>    Internet-Drafts are working documents of the Internet Engineering=

>>>>>>>    Task Force (IETF).  Note that other groups may also distribute
>>>>>>>    working documents as Internet-Drafts.  The list of current Intern=
et-
>>>>>>>    Drafts is at http://datatracker.ietf.org/drafts/current/.
>>>>>>>=20
>>>>>>>    Internet-Drafts are draft documents valid for a maximum of six mo=
nths
>>>>>>>    and may be updated, replaced, or obsoleted by other documents at a=
ny
>>>>>>>    time.  It is inappropriate to use Internet-Drafts as reference
>>>>>>>    material or to cite them other than as "work in progress."
>>>>>>>=20
>>>>>>>    This Internet-Draft will expire on December 17, 2013.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 1]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> Copyright Notice
>>>>>>>=20
>>>>>>>    Copyright (c) 2013 IETF Trust and the persons identified as the
>>>>>>>    document authors.  All rights reserved.
>>>>>>>=20
>>>>>>>    This document is subject to BCP 78 and the IETF Trust's Legal
>>>>>>>    Provisions Relating to IETF Documents
>>>>>>>    (http://trustee.ietf.org/license-info) in effect on the date of
>>>>>>>    publication of this document.  Please review these documents
>>>>>>>    carefully, as they describe your rights and restrictions with res=
pect
>>>>>>>    to this document.  Code Components extracted from this document m=
ust
>>>>>>>    include Simplified BSD License text as described in Section 4.e o=
f
>>>>>>>    the Trust Legal Provisions and are provided without warranty as
>>>>>>>    described in the Simplified BSD License.
>>>>>>>=20
>>>>>>>=20
>>>>>>> Table of Contents
>>>>>>>=20
>>>>>>>    1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .=
  5
>>>>>>>=20
>>>>>>>    2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .=
  7
>>>>>>>=20
>>>>>>>    3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .=
  8
>>>>>>>=20
>>>>>>>    4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .=
  9
>>>>>>>=20
>>>>>>>    5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . .=
 12
>>>>>>>=20
>>>>>>>    6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . .=
 13
>>>>>>>      6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . .=
 13
>>>>>>>      6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . .=
 13
>>>>>>>      6.3.  Other Timing Encapsulation methods . . . . . . . . . . . .=
 14
>>>>>>>=20
>>>>>>>    7.  Timing message Processing  . . . . . . . . . . . . . . . . . .=
 15
>>>>>>>=20
>>>>>>>    8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . .=
 16
>>>>>>>=20
>>>>>>>    9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 17
>>>>>>>=20
>>>>>>>    10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 18
>>>>>>>=20
>>>>>>>    11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 19
>>>>>>>=20
>>>>>>>    12. OAM, Control and Management  . . . . . . . . . . . . . . . . .=
 20
>>>>>>>=20
>>>>>>>    13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . .=
 21
>>>>>>>=20
>>>>>>>    14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . .=
 22
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 2]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . .=
 23
>>>>>>>      15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . .=
 23
>>>>>>>      15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . .=
 23
>>>>>>>      15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . .=
 24
>>>>>>>=20
>>>>>>>    16. Other considerations . . . . . . . . . . . . . . . . . . . . .=
 25
>>>>>>>=20
>>>>>>>    17. Security Considerations  . . . . . . . . . . . . . . . . . . .=
 26
>>>>>>>=20
>>>>>>>    18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . .=
 27
>>>>>>>=20
>>>>>>>    19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . .=
 28
>>>>>>>=20
>>>>>>>    20. References . . . . . . . . . . . . . . . . . . . . . . . . . .=
 29
>>>>>>>      20.1. Normative References . . . . . . . . . . . . . . . . . . .=
 29
>>>>>>>      20.2. Informative References . . . . . . . . . . . . . . . . . .=
 29
>>>>>>>=20
>>>>>>>    Appendix 1.  Routing extensions for Timing-aware Routers . . . . .=
 32
>>>>>>>=20
>>>>>>>    Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . .=
 33
>>>>>>>=20
>>>>>>>    Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . .=
 34
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 3]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
>>>> NOT",
>>>>>>>    "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL"
>>>> in
>>>>>>> this
>>>>>>>    document are to be interpreted as described in RFC2119 [RFC2119].=

>>>>>>>=20
>>>>>>>    When used in lower case, these words convey their typical use in
>>>>>>>    common language, and are not to be interpreted as described in
>>>>>>>    RFC2119 [RFC2119].
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 4]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 1.  Introduction
>>>>>>>=20
>>>>>>>    The objective of Precision Time Protocol (PTP) and Network Timing=

>>>>>>>    Protocol (NTP) are to synchronize independent clocks running on
>>>>>>>    separate nodes of a distributed system.
>>>>>>>=20
>>>>>>>    [IEEE-1588] defines PTP messages for frequency, phase and time
>>>>>>>    synchronization.  The PTP messages include PTP PDUs over UDP/IP
>>>>>>>    (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex =
F of
>>>>>>>    [IEEE-1588]).
>>>>>>>=20
>>>>>>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
>>>>>>> SB> of other PTP mappings if they provide better optimisation.
>>>>>>>=20
>>>>>>>    This document defines mapping and transport of the PTP
>>>>>>>    messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>>>>>>>    defines several clock types: ordinary clocks, boundary clocks, en=
d-
>>>>>>>    to-end transparent clocks, and peer-to-peer transparent clocks.
>>>>>>>    Transparent clocks require intermediate nodes to update correctio=
n
>>>>>>>    field inside PTP message that reflects the transit time in the no=
de.
>>>>>>>=20
>>>>>>>    [RFC5905] defines NTP messages for clock and time synchronization=
.
>>>>>>>    The PTP messages (PDUs) are transported over UDP/IP.  This docume=
nt
>>>>>>> SB> Should that be NTP messages?
>>>>>>> SB> It needs to be made clear as soon as you introduce NTP that
>>>>>>> SB> they use different time representations.
>>>>>>>=20
>>>>>>>    defines mapping and transport of the NTP messages defined in
>>>>>>>    [RFC5905] over MPLS networks.
>>>>>>>=20
>>>>>>>    One key attribute of all of these Timing messages is that the Tim=
e
>>>>>>>    stamp processing should occur as close as possible to the actual
>>>>>>>    transmission and reception at the physical port interface.  This
>>>>>>>    targets optimal time and/or frequency recovery by avoiding variab=
le
>>>>>>>    delay introduced by queues internal to the clocks.
>>>>>>>=20
>>>>>>> SB> As I recall NTP has no epoch point defined, and I am not sure
>>>>>>> SB> where that point is in the case of PTP in this mapping
>>>>>>> SB> Hopefully this will get defined in due course.
>>>>>>>=20
>>>>>>>    To facilitate the fast and efficient recognition of Timing messag=
es
>>>>>>>    at the port level when the Timing messages are carried over MPLS
>>>>>>>    LSPs,
>>>>>>>=20
>>>>>>> SB> Over a new LSP type with time optimied characteristics
>>>>>>>=20
>>>>>>>    this document defines the specific encapsulations that should
>>>>>>>    be used.
>>>>>>> SB> Hopefully it will also define the PHP
>>>>>>>=20
>>>>>>>    In addition, it can be expected that there will exist LSR/
>>>>>>>    LERs where only a subset of the physical ports will have the port=
-
>>>>>>>    based Timing message processing capabilities.
>>>>>>> SB> Do you need to clarify that this only works at base and not in
>>>>>>> SB> a label heirarchy.
>>>>>>>=20
>>>>>>>=20
>>>>>>>    In order to ensure
>>>>>>>    that the LSPs carrying Timing packets always enter and exit ports=

>>>>>>>    with this capability, routing extensions are defined to advertise=

>>>>>>>    this capability on a port basis and to allow for the establishmen=
t of
>>>>>>>    LSPs that only transit such ports.  While this path establishment=

>>>>>>>    restriction may be applied only at the LER Ingress and/or egress
>>>>>>>    ports, it becomes more important when using transparent clock cap=
able
>>>>>>>    LSRs in the path.
>>>>>>> SB> I do not understand the implications of the last
>>>>>>> SB> sentences - starting ", it becomes"
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Port based Timing message processing involves Timing message
>>>>>>>    recognition.  Once the Timing messages are recognized they can be=

>>>>>>>    modified based on the reception or transmission Time-stamp.
>>>>>>>=20
>>>>>>>    This document provides two methods for transporting Timing messag=
es
>>>>>>>    over MPLS.  One is applicable to MPLS environment and the other o=
ne
>>>>>>>    is applicable to MPLS/MPLS-TP environment
>>>>>>>=20
>>>>>>> SB> I think the sentence is incomplete.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 5]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    The solution involves transporting Timing messages over dedicated=

>>>>>>>    LSPs called Timing LSPs.  These LSPs carry Timing messages and MA=
Y
>>>>>>>    carry Management and control messages, but not data plane client
>>>>>>>    traffic.
>>>>>>>=20
>>>>>>> SB> It is not clear why this restriction applies.
>>>>>>>=20
>>>>>>>    Timing LSPs can be established statically or via signaling.
>>>>>>> SB> s/statically/by provisioning/network management/
>>>>>>>=20
>>>>>>>    Extensions to control plane (OSPF, ISIS, etc.) is required to ena=
ble
>>>>>>>    routers to distribute their Timing processing capabilities over M=
PLS
>>>>>>>    to other routers.  However such extensions are outside the scope o=
f
>>>>>>>    this document.
>>>>>>>=20
>>>>>>>    When signaling is used to setup the PTP LSP, Extensions to signal=
ing
>>>>>>> SB> is it a PTP LSP or a Timing LSP?
>>>>>>>=20
>>>>>>>    protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.=

>>>>>>>    However such extensions are outside the scope of this document.
>>>>>>>=20
>>>>>>> SB> for mpls-tp GMPLS is the signalling protocol
>>>>>>>=20
>>>>>>>    While the techniques included herein allow for the establishment o=
f
>>>>>>>    paths optimized to include Time-stamping capable links, the
>>>>>>>    performance of the Slave clocks is outside the scope of this
>>>>>>>    document.
>>>>>>>=20
>>>>>>>    At the time of publishing this specification, Transparent Clockin=
g
>>>>>>>    (TC) is only defined for PTP.  Therefore at this time any part of=

>>>>>>>    this specification that talks about Transparent Clocking applies o=
nly
>>>>>>>    to PTP.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 6]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 2.  Terminology
>>>>>>>=20
>>>>>>>    1588: The timing and synchronization as defined by IEEE 1588.
>>>>>>>=20
>>>>>>> SB> I think that there is a more formal name for the 1588 group
>>>>>>> SB> that needs to be used here.
>>>>>>> SB> Also do we need to talk about 1588-200? as there is
>>>>>>> SB> an update in progress
>>>>>>>=20
>>>>>>>    NTP: The timing and synchronization protocol defined by IETF RFC-=
1305
>>>>>>>    and RFC-5905.
>>>>>>>=20
>>>>>>>    PTP: The timing and synchronization protocol used by 1588.
>>>>>>> SB> need the proper name for 1588
>>>>>>>=20
>>>>>>>    Master Clock: The source of 1588 timing to a set of slave clocks.=

>>>>>>>=20
>>>>>>>    Master Port: A port on a ordinary or boundary clock that is in Ma=
ster
>>>>>>>    state.  This is the source of timing toward slave ports.
>>>>>>>=20
>>>>>>> SB> I am not sure the reader knows what a port is
>>>>>>>=20
>>>>>>>    Slave Clock: A receiver of 1588 timing from a master clock.
>>>>>>>=20
>>>>>>>    Slave Port: A port on a boundary clock or ordinary clock that is
>>>>>>>    receiving timing from a master clock.
>>>>>>>=20
>>>>>>>    Ordinary Clock: A device with a single PTP port.
>>>>>>>=20
>>>>>>>    Transparent Clock.  A device that measures the time taken for a P=
TP
>>>>>>>    event message to transit the device and then updates the
>>>>>>>    correctionField of the message with this transit time.
>>>>>>>=20
>>>>>>>    Boundary Clock: A device with more than one PTP port.  Generally
>>>>>>>    boundary clocks will have one port in slave state to receive timi=
ng
>>>>>>>    and then other ports in master state to re-distribute the timing.=

>>>>>>>=20
>>>>>>>    PTP LSP: An LSP dedicated to carry PTP messages
>>>>>>>=20
>>>>>>> SB> PTP or timing?
>>>>>>>=20
>>>>>>>=20
>>>>>>>    PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>>>>>>    messages.
>>>>>>>=20
>>>>>>> SB> Ah I don't think that PWE3 know what one of these is
>>>>>>>=20
>>>>>>>    CW: Pseudowire Control Word
>>>>>>>=20
>>>>>>>    LAG: Link Aggregation
>>>>>>>=20
>>>>>>>    ECMP: Equal Cost Multipath
>>>>>>>=20
>>>>>>>    CF: Correction Field, a field inside certain PTP messages (messag=
e
>>>>>>>    type 0-3)that holds the accumulative transit time inside intermed=
iate
>>>>>>>    switches
>>>>>>>=20
>>>>>>>    Timing messages: Timing Protocol messages that are exchanged
>>>> between
>>>>>>>    routers in order to establish a synchronized clock.
>>>>>>>=20
>>>>>>> SB> A number of these definitions look like copies of IEEE1588
>>>>>>> SB> definitions. We need to provide references and note the
>>>>>>> SB> priority of the IEEE base reference.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 7]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 3.  Problem Statement
>>>>>>>=20
>>>>>>>    [IEEE-1588] has defined methods for transporting PTP messages ove=
r
>>>>>>>    Ethernet and IP networks.  [RFC5905] has defined the method of
>>>>>>>    transporting NTP messages over IP networks.  There is a need to
>>>>>>>    transport Timing messages over MPLS networks while supporting the=

>>>>>>>    Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (O=
C)
>>>>>>>    functionality in the LER and LSRs in the MPLS network.
>>>>>>>=20
>>>>>>>    There are multiple ways of transporting Timing over MPLS.  Howeve=
r,
>>>>>>>    there is a requirement to limit the possible encapsulation option=
s to
>>>>>>>    simplify the Timing message identification and processing require=
d at
>>>>>>>    the port level.
>>>>>>>=20
>>>>>>>    When Timing-awareness is needed, Timing messages should not be
>>>>>>>    transported over LSPs or PWs that are carrying customer traffic
>>>>>>>    because LSRs perform Label switching based on the top label in th=
e
>>>>>>>    stack.
>>>>>>>=20
>>>>>>> SB> Have you explained why?
>>>>>>>=20
>>>>>>>    To detect Timing messages inside such LSPs require special
>>>>>>>    hardware to do deep packet inspection at line rate.  Even if such=

>>>>>>>    hardware exists, the payload can't be deterministically identifie=
d by
>>>>>>>    LSRs because the payload type is a context of the PW label, and t=
he
>>>>>>>    PW label and its context are only known to the Edge routers (PEs/=

>>>>>>>    LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, C=
ES,
>>>>>>>    etc).  Even if one restricts an LSP to only carry Ethernet PWs, t=
he
>>>>>>>    LSRs dont have the knowledge of whether PW Control Word (CW) is
>>>>>>>    present or not and therefore can not deterministically identify t=
he
>>>>>>>    payload.
>>>>>>>=20
>>>>>>>    A generic method is defined in this document that does not requir=
e
>>>>>>>    deep packet inspection at line rate, and can deterministically
>>>>>>>    identify Timing messages.  This method can be used to detect Timi=
ng
>>>>>>>    Messages in both one-step and two-step clock implementations of
>>>>>>>    ordinary, boundary and transparent clocks.
>>>>>>>=20
>>>>>>> SB> Needs a ref and I am sure many MPLS specialists will not underst=
and
>>>>>>> SB> the msg types.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 8]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 4.  Timing over MPLS Architecture
>>>>>>>=20
>>>>>>>    Timing messages are exchange between Timing ports on ordinary and=

>>>>>>>=20
>>>>>>> SB> Have you defined a timing port?
>>>>>>>=20
>>>>>>>    boundary clocks.  Boundary clocks terminate the Timing messages a=
nd
>>>>>>>    act as master for other boundary clocks or for slave clocks.  End=
-to-
>>>>>>>    End Transparent clocks do not terminate the Timing messages but t=
hey
>>>>>>>    do modify the contents of the Timing messages as they transit acr=
oss
>>>>>>>    the transparent clock.
>>>>>>>=20
>>>>>>>    Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent C=
lock
>>>>>>>=20
>>>>>>>    (TC) could be implemented in either LERs or LSRs.
>>>>>>>=20
>>>>>>> SB> LER and LSR need to be expanded
>>>>>>>=20
>>>>>>>    An example is shown in Figure 1, where the LERs act as Ordinary C=
lock
>>>>>>>    (OC) and are the initiating/terminating point for Timing messages=
.
>>>>>>>    The ingress LER encapsulates the Timing messages in Timing LSP an=
d
>>>>>>>    the Egress LER terminates the Timing LSP.  The LSRs act as
>>>>>>>    Transparent Clock (TC) and just update the Timing field in the Ti=
ming
>>>>>>>    messages.
>>>>>>>=20
>>>>>>>=20
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>       |Switch, |     |       |     |       |     |       |     |Swit=
ch, |
>>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rou=
ter |
>>>>>>>       |        |     |  OC   |     |  TC   |     |  OC   |     |    =
    |
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>                      /                                 \
>>>>>>>       +-------+     /                                   \     +-----=
--+
>>>>>>>       |  LER  |    /                                     \    |  LER=
  |
>>>>>>>       | Master|---/                                       \---| Slav=
e |
>>>>>>>       | Clock |                                               | Cloc=
k |
>>>>>>>       +-------+                                               +-----=
--+
>>>>>>>=20
>>>>>>>      Figure (1) - Deployment example 1 of timing over MPLS network
>>>>>>>=20
>>>>>>>    Another example is shown in Figure2, where LERs terminate the Tim=
ing
>>>>>>>    messages received from switch/routers that are outside of the MPL=
S
>>>>>>>    network acting as OC or BC.  In this example LERs regenerate the
>>>>>>>    clock and initiate timing messages encapsulated in Timing LSP tow=
ard
>>>>>>>    the MPLS network, while the LSRs act as Transparent Clock (TC) an=
d
>>>>>>>    just update the Timing field in the Timing messages, which are
>>>>>>>    already encapsulated in Timing LSPs.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 9]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>      |Switch, |     |       |     |       |     |       |     |Switc=
h, |
>>>>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
>>>>>>>      | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/B=
C  |
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>=20
>>>>>>>      Figure (2) - Deployment example 2 of timing over MPLS network
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Another example is shown in Figure 3, where LERs do not terminate=
 the
>>>>>>>    Timing messages received from switch/routers that are outside of t=
he
>>>>>>>    MPLS network acting as OC, TC or BC.  The LERs act as TC and upda=
te
>>>>>>>    the Timing field in the Timing messages as they transit the LER,
>>>>>>>    while encapsulating them in timing LSP.  The LSRs also act as
>>>>>>>    Transparent Clock (TC) and just update the Timing field in the Ti=
ming
>>>>>>>    messages which are already encapsulated in Timing LSPs.
>>>>>>>=20
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>       |Switch, |     |       |     |       |     |       |     |Swit=
ch, |
>>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rou=
ter |
>>>>>>>       |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/T=
C/BC|
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>=20
>>>>>>>     Figure (3) - Deployment example 3 of timing over MPLS network
>>>>>>>=20
>>>>>>>    Another example is shown in Figure 4, where LERs and LSRs support=

>>>>>>>    Boundary Clocks.  A single-hop LSP is created between two adjacen=
t
>>>>>>>    LSRs engaged in BC operation.  Other methods such as PTP transpor=
t
>>>>>>>    over Ethernet MAY be used for transporting timing messages if the=

>>>>>>>    link between the two routers is Ethernet.
>>>>>>>=20
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>      |Switch, |     |       |     |       |     |       |     |Switc=
h, |
>>>>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
>>>>>>>      | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/B=
C  |
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>=20
>>>>>>>    Figure (4) - Deployment example 3 of timing over MPLS network
>>>>>>>=20
>>>>>>>    An MPLS domain MAY serve multiple customers.  In these cases the
>>>> MPLS
>>>>>>>    domain (maintained by a service provider) may provide timing serv=
ices
>>>>>>>    to multiple customers, each having their own Timing domain.
>>>>>>>=20
>>>>>>>    The Timing over MPLS architecture assumes full mesh of Timing LSP=
s
>>>>>>>    between all LERs supporting this specification.
>>>>>>>=20
>>>>>>> SB> Note sure this is right - the salves surely do not need to
>>>>>>> SB> exchange timing amongst themselves
>>>>>>>=20
>>>>>>>    It supports
>>>>>>>    Point-to- point (VPWS) and Multipoint (VPLS) services.
>>>>>>>=20
>>>>>>> SB> What does that mean? You do not carry user data traffic?
>>>>>>> SB> Maybe it's the ordering of the statemnets that is causing
>>>>>>> SB> confusion.
>>>>>>>=20
>>>>>>>    This means
>>>>>>>    that a customer may purchase a Point-to-point Timing service betw=
een
>>>>>>>    two customer sites or a Multipoint Timing service between more th=
an
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 10]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    two customer sites.
>>>>>>>=20
>>>>>>>    The Timing over MPLS architecture supports P2P or P2MP Timing LSP=
s.
>>>>>>>    This means that the Timing Multicast messages such as PTP Multica=
st
>>>>>>>    event messages can be transported over P2MP Timing LSP or be
>>>>>>>    replicated and transported over many P2P Timing LSPs.
>>>>>>>=20
>>>>>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>>>>>>> SB> nor a P2MP PW, although we are close.
>>>>>>>=20
>>>>>>>    Timing messages, that do not require Time stamping or Correction
>>>>>>>    Field update MAY be transported over Timing LSPs to simplify hard=
ware
>>>>>>>    and software.
>>>>>>>=20
>>>>>>>    PTP Announce messages that determine the Timing LSP terminating
>>>> point
>>>>>>>    behavior such as BC/OC/TC SHOULD be transported over the Timing L=
SP
>>>>>>>    to simplify hardware and software.
>>>>>>>=20
>>>>>>> SB> have you defined and referenced PTP announce msgs?
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 11]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 5.  Dedicated LSPs for Timing messages
>>>>>>>=20
>>>>>>>    Many methods have been considered for identifying the Timing
>>>> messages
>>>>>>>    when they are encapsulated in MPLS such as using GAL/G-ACH or a n=
ew
>>>>>>>    reserved label.  These methods were not attractive since they eit=
her
>>>>>>>    required deep packet inspection at line rate in the intermediate L=
SRs
>>>>>>>    or they required use of a scarce new reserved label.  Also one of=
 the
>>>>>>>    goals was to reuse existing OAM mechanisms.
>>>>>>>=20
>>>>>>> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
>>>>>>>=20
>>>>>>>    The method defined in this document can be used by LER and LSRs t=
o
>>>>>>>    identify Timing messages in MPLS tunnels by just looking at the t=
op
>>>>>>>    label in the MPLS label stack, which only carry Timing messages a=
s
>>>>>>>    well as OAM, but not data plane client traffic.
>>>>>>>=20
>>>>>>>    Compliant implementations MUST use dedicated LSPs to carry Timing=

>>>>>>>    messages over MPLS.
>>>>>>>=20
>>>>>>> SB> I think that we need a definition of the properies of these LSPs=

>>>>>>>=20
>>>>>>>    These LSPs are herein referred to as "Timing
>>>>>>>    LSPs" and the labels associated with these LSPs as "Timing LSP
>>>>>>>    labels".  The Timing LSPs that runs between Ingress and Egress LE=
Rs
>>>>>>>    MUST be co-routed.  Alternatively, a single bidirectional co-rout=
ed
>>>>>>>    LSP can be used.
>>>>>>>=20
>>>>>>> SB> I though that you said you could use M2MP LSPs - these are not
>>>>>>> SB> bidirectional.
>>>>>>>=20
>>>>>>>    Co-routing of the two directions is required to limit the differe=
nce
>>>>>>>    in the delays in the Master clock to Slave clock direction compar=
ed
>>>>>>>    to the Slave clock to Master clock direction.  The Timing LSP MAY=
 be
>>>>>>>    MPLS/MPLS-TP LSP.
>>>>>>>=20
>>>>>>>    The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS=
.
>>>>>>>    New Extensions to RSVP-TE/GMPLS TLVs are required; however they a=
re
>>>>>>>    outside the scope of this document.
>>>>>>>=20
>>>>>>>    The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such=
 as
>>>>>>>    BFD and LSP Ping but the LSP data plane client plane traffic MUST=
 be
>>>>>>>    Timing packets only.
>>>>>>>=20
>>>>>>> SB> Why?
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 12]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 6.  Timing over LSP Encapsulation
>>>>>>>=20
>>>>>>> The encapsulations is not LSP is it?
>>>>>>>=20
>>>>>>>    This document defines two methods for carrying Timing messages ov=
er
>>>>>>>    MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>>>>>>    messages over Timing LSPs, and the second method, is carrying
>>>>>>>    Ethernet encapsulated Timing messages over Ethernet PWs inside
>>>> Timing
>>>>>>>    LSPs.
>>>>>>>=20
>>>>>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>>>>>>=20
>>>>>>>    The simplest method of transporting Timing messages over MPLS is t=
o
>>>>>>>    encapsulate Timing PDUs in UDP/IP and then encapsulate them in
>>>> Timing
>>>>>>>    LSP.  This format is shown in Figure 4.
>>>>>>>=20
>>>>>>>=20
>>>>>>>                     +----------------------+
>>>>>>>                     |   Timing LSP Label   |
>>>>>>>                     +----------------------+
>>>>>>>                     |        IPv4/6        |
>>>>>>>                     +----------------------+
>>>>>>>                     |         UDP          |
>>>>>>>                     +----------------------+
>>>>>>>                     |     Timing PDU       |
>>>>>>>                     +----------------------+
>>>>>>>=20
>>>>>>>       Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>>>>>>=20
>>>>>>>=20
>>>>>>>    This encapsulation is very simple and is useful when the network
>>>>>>>    between Timing Master Clock and Slave Clock is MPLS network.
>>>>>>>=20
>>>>>>> SB> Simple is a judgement call
>>>>>>>=20
>>>>>>>    In order for an LER/LSR to process Timing messages, the Timing LS=
P
>>>>>>>    Label must be at the top label of the label stack.  The LER/LSR M=
UST
>>>>>>>    know that the Timing LSP Label is used for carrying Timing messag=
es.
>>>>>>>    This can be accomplished via static configuration or via RSVP-TE
>>>>>>>    signaling.
>>>>>>>=20
>>>>>>>    The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>>>>>>    [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>>>>>>>    [RFC5905].
>>>>>>>=20
>>>>>>> 6.2.  Timing over PW Encapsulation
>>>>>>>=20
>>>>>>>    Another method of transporting Timing over MPLS networks is by
>>>>>>>    encapsulating Timing PDUs in PW which in turn is transported over=

>>>>>>>    Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448]=
,
>>>>>>>    shown in Fig 5(A) MUST be used and the Ethernet encapsulation of P=
TP
>>>>>>>    MUST follow Annex F of [IEEE-1588].
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 13]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
>>>> the
>>>>>>>    Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>>>>>>>    Timing over PW encapsulation MUST use the Control Word (CW) as
>>>>>>>    specified in [RFC4448] to ensure proper detection of PTP messages=

>>>>>>>    inside the MPLS packets for Timing over LSP and Timing over PW
>>>>>>>    encapsulation.
>>>>>>>=20
>>>>>>> SB> That needs explanation
>>>>>>>=20
>>>>>>>    The use of Sequence Number in the CW is optional.
>>>>>>>=20
>>>>>>> SB> Given that s/n are never in practice deployed, you could probabl=
y
>>>>>>> SB> simplify things by sayig that they are not used.
>>>>>>>=20
>>>>>>>    Timing over PW encapsulation for NTP MUST use NTP over UDP/IP ove=
r
>>>> PW
>>>>>>>    (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>>>>>>=20
>>>>>>>                     +----------------+  +----------------+
>>>>>>>                      |Timing LSP Label|  |Timing LSP Label|
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |    PW Label    |  |    PW Label    |
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |  Control Word  |  |      IP        |
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |    Ethernet    |  |      UDP       |
>>>>>>>                      |     Header     |  +----------------+
>>>>>>>                      +----------------+  |   Timing PDU   |
>>>>>>>                      |S-VLAN(Optional)|  |                |
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |C-VLAN(Optional)|        (B)
>>>>>>>                      +----------------+
>>>>>>>                      |   Timing PDU   |
>>>>>>>                      |                |
>>>>>>>                      +----------------+
>>>>>>>                             (A)
>>>>>>>=20
>>>>>>>               Figure (5) - Timing over PW Encapsulations
>>>>>>>=20
>>>>>>>    In order for an LSR to process PTP messages, the top label of the=

>>>>>>>    label stack (the Tunnel Label) MUST be a Timing label.
>>>>>>>=20
>>>>>>> S> You said that before.
>>>>>>>=20
>>>>>>> 6.3.  Other Timing Encapsulation methods
>>>>>>>=20
>>>>>>>    In future other timing encapsulation methods may be introduced, s=
uch
>>>>>>>    as a new shim header after the Bottom of Stack to carry the Timin=
g
>>>>>>>    information.  Such new encapsulations are outside the scope of th=
is
>>>>>>>    document.
>>>>>>>=20
>>>>>>>=20
>>>>>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>>>>>>> SB> out of the definition of the LSP
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 14]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> SB> I think we need a section on LSP processing
>>>>>>>=20
>>>>>>> 7.  Timing message Processing
>>>>>>>=20
>>>>>>>    Each Timing protocol such as PTP and NTP, define their set of Tim=
ing
>>>>>>>    messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>>>>>>>    FOLLOW_UP, etc messages.
>>>>>>>=20
>>>>>>>    Some of the Timing messages require time stamping or correction f=
ield
>>>>>>>    update at port level and some dont.  It is the job of the LER/LSR=
 to
>>>>>>>    parse the timing message and find out the type of the Timing mess=
age
>>>>>>>    and decide whether and how to Time- stamp it (e.g., BC) or update=

>>>>>>>    correction field(e.g., TC).
>>>>>>>=20
>>>>>>>=20
>>>>>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
>>>>>>> SB> function rather than the LER?
>>>>>>>=20
>>>>>>>    For example the following PTP messages (called Event messages)
>>>>>>>    require time-stamping or correction field update:
>>>>>>>=20
>>>>>>>    o  SYNC
>>>>>>>=20
>>>>>>>    o  DELAY_REQ (Delay Request)
>>>>>>>=20
>>>>>>>    o  PDELAY_REQ (Peer Delay Request)
>>>>>>>=20
>>>>>>>    o  PDELAY_RESP (Peer Delay Response)
>>>>>>>=20
>>>>>>>    SYNC and DELAY_REQ are exchanged between Master Clock and Slave
>>>> Clock
>>>>>>>    and MUST be transported over PTP LSPs.  PDELAY_REQ and
>>>> PDELAY_RESP
>>>>>>>    are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>>>>>>>    Boundary, or Transparent) and SHOULD be transported over single h=
op
>>>>>>>    PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP=
,
>>>>>>>    and PDELAY_RESP_FOLLOW_UP messages MUST also be transported
>>>> over
>>>>>>> the
>>>>>>>    PTP LSPs.
>>>>>>>=20
>>>>>>>    For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be=

>>>>>>>    transported over two PTP LSPs that are in opposite directions.  T=
hese
>>>>>>>    PTP LSPs, which are in opposite directions MUST be congruent and c=
o-
>>>>>>>    routed.  Alternatively, a single bidirectional co-routed LSP can b=
e
>>>>>>>    used.
>>>>>>>=20
>>>>>>>    Except as indicated above for the two-step PTP clocks, Non-Event P=
TP
>>>>>>>    message types do not need to be processed by intermediate routers=
.
>>>>>>>    These message types MAY be carried in PTP Tunnel LSPs.
>>>>>>>=20
>>>>>>> SB> Are you saying that a timing P router has to be msg type sensiti=
ve?
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 15]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 8.  Protection and Redundancy
>>>>>>>=20
>>>>>>>=20
>>>>>>> SB> This is a bit of a jump - I don't know how the LSP itself works y=
et!
>>>>>>>=20
>>>>>>>    In order to ensure continuous uninterrupted operation of slave
>>>>>>>    clocks, usually as a general practice, slave clocks (or ports) tr=
ack
>>>>>>>    redundant master clocks.
>>>>>>>=20
>>>>>>>    It is the responsibility of the network operator to ensure that
>>>>>>>    physically disjoint Timing LSPs are established between a slave c=
lock
>>>>>>>    (or port) and redundant master clocks (or ports).
>>>>>>>=20
>>>>>>>    When a slave clock (or port) listens to redundant master clocks o=
r
>>>>>>>    ports, any prolonged Timing LSP outage will trigger the slave clo=
ck
>>>>>>>    or port to switch to a redundant master clock or port.
>>>>>>>=20
>>>>>>>    LSP/PW protection such as Linear protection Switching (1:1, 1+1),=

>>>>>>>    Ring protection switching or MPLS Fast Reroute (FRR) generally sw=
itch
>>>>>>>    alternative path that usually cause a change in delay, which if
>>>>>>>    undetected by slave clock can reduce accuracy of the slave clock.=

>>>>>>>=20
>>>>>>>    Therefore protection switching MAY be used, as long as phase jump=
s
>>>>>>>    upon switchover due to differences in path latency are detected a=
nd
>>>>>>>    compensated for (such compensation not being required if BCs or p=
eer-
>>>>>>>    peer TCs are used throughout).
>>>>>>>=20
>>>>>>>    Note that any protection or reroute mechanism that adds additiona=
l
>>>>>>>    MPLS label to the label stack, such as Facility Backup Fast Rerou=
te,
>>>>>>>    MUST ensure that the pushed label is also a Timing Label to ensur=
e
>>>>>>>    recognition of the MPLS frame as containing Timing messages, as i=
t
>>>>>>>    transits the backup path.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 16]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 9.  ECMP
>>>>>>>=20
>>>>>>>    To ensure the optimal operation of slave clocks and avoid error
>>>>>>>    introduced by forward and reverse path delay asymmetry, the physi=
cal
>>>>>>>    path for Timing messages from master clock to slave Clock and vic=
e
>>>>>>>    versa must be the same for all Event Timing messages listed in
>>>>>>>    section 7.
>>>>>>>=20
>>>>>>>    Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost=

>>>>>>>    Multipath).
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 17]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 10.  PHP
>>>>>>>=20
>>>>>>>    To ensure that the label on the top of the label stack is the Tim=
ing
>>>>>>>    LSP Label, PHP MUST not be used.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 18]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 11.  Entropy
>>>>>>>=20
>>>>>>>    To ensure all Timing messages in a Timing LSP take the same path,=

>>>>>>>    Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>>>>>>    Entropy Label MUST NOT be used for the PWs that are carried insid=
e
>>>>>>>    Timing LSP [RFC6391].
>>>>>>>=20
>>>>>>> SB> This is incorrect - you mean that all msgs of the same timing
>>>>>>> SB> flow need to have the same EL value.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 19]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 12.  OAM, Control and Management
>>>>>>>=20
>>>>>>>    In order to monitor Timing LSPs and their encapsulated PWs, they M=
UST
>>>>>>>    be able to carry OAM and management messages.  These management
>>>>>>>    messages MUST be differentiated from Timing messages via already
>>>>>>>    defined IETF methods.
>>>>>>>=20
>>>>>>>    For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY r=
un
>>>>>>>    over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>>>>>>>    Management protocols can easily be identified by the UDP Destinat=
ion
>>>>>>>    Port number or by GAL/G-ACH respectively.
>>>>>>>=20
>>>>>>>    Also BFD, LSP-Ping and other management messages MAY run over the=

>>>> PWs
>>>>>>>    encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3=
 or
>>>>>>>    4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is g=
oing
>>>>>>>    to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1=
) or
>>>>>>>    GAL-ACH are used to identify such management messages.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 20]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 13.  QoS Considerations
>>>>>>>=20
>>>>>>>    In network deployments where not every LSR/LER is Timing-aware, i=
t is
>>>>>>>    important to reduce the impact of the non-Timing-aware LSR/LERs o=
n
>>>>>>>    the timing recovery in the slave clock.  The Timing messages are t=
ime
>>>>>>>    critical and must be treated with the highest priority.  Therefor=
e
>>>>>>>    Timing over MPLS messages must be treated with the highest priori=
ty
>>>>>>>    in the routers.  This can be achieved by proper setup of Timing L=
SPs.
>>>>>>>=20
>>>>>>>    It is recommended that the Timing LSPs are setup or configured
>>>>>>>    properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC26=
97]
>>>>>>>    for drop eligibility.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 21]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 14.  FCS and Checksum Recalculation
>>>>>>>=20
>>>>>>>    When time-stamp generation and timing packet adjustment is perfor=
med
>>>>>>>    near the physical port hardware, the process MUST include
>>>>>>>    recalculation of the Ethernet FCS.
>>>>>>>=20
>>>>>>> SB> The above is confusing - an LSR always recomputes the link layer=

>>>>>>> SB> CRC which may or may not be Ethernet.
>>>>>>>=20
>>>>>>>    Also FCS retention for the
>>>>>>>    payload Ethernet described in [RFC4720] MUST NOT be used.
>>>>>>>=20
>>>>>>>    For UDP/IP encapsulation mode of Timing over MPLS, the UDP checks=
um
>>>>>>>    may be required as per UDP transport standards.
>>>>>>>=20
>>>>>>> SB> You really need to be working on getting the IPv6 C?S computatio=
n
>>>>>>> SB> removed from PTP msgs.
>>>>>>>=20
>>>>>>>    When UDP checksum is used, each Timing-aware LER/LSR must either
>>>>>>>    incrementally update the UDP checksum after Time stamping or
>>>>>>>    Correction Field update or verify the UDP checksum on reception f=
rom
>>>>>>>    upstream and recalculate the checksum completely on transmission t=
o
>>>>>>>    downstream node after Time stamping or Correction Field update.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 22]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 15.  Behavior of LER/LSR
>>>>>>>=20
>>>>>>>    Timing-capable/aware LERs and LSRs are routers that have one or m=
ore
>>>>>>>=20
>>>>>>> SB> You mean physical interfaces?
>>>>>>>=20
>>>>>>>    interfaces that can perform Timing operations (OC/BC/TC) on Timin=
g
>>>>>>>    packets and are configured to do so.  Timing-capable/aware LERs a=
nd
>>>>>>>    LSRs can advertise their Timing-capability per-interface via cont=
rol
>>>>>>>    plane such as OSPF or IS-IS.
>>>>>>> SB> ISIS and OSPF are routing protocols.
>>>>>>>=20
>>>>>>>   The Timing-capable/aware LERs can then
>>>>>>>    signals Timing LSPs via RSVP-TE signaling.  Alternatively the Tim=
ing
>>>>>>>    capability of LER and LSRs may be configured in a centralized
>>>>>>>    controller and the Timing LSP may be setup using manual configura=
tion
>>>>>>>    or other methods such as SDN.
>>>>>>>=20
>>>>>>> SB> it can also be configured individually rather then through
>>>>>>> SB> a cebtral controllwe
>>>>>>>=20
>>>>>>> 15.1.  Behavior of Timing-capable/aware LER
>>>>>>>=20
>>>>>>>    When a Timing-capable/aware LER behaves as a Transparent clock an=
d
>>>>>>>    receives a Timing message from a Timing-capable/aware non-MPLS
>>>>>>>    interface, the LER updates the Correction Field (CF) and encapsul=
ates
>>>>>>>    and forwards the timing message over previously established Timin=
g
>>>>>>>    LSP.
>>>>>>>=20
>>>>>>> SB> You need to call out the details so that people properly
>>>>>>> SB> understand the definition of the new LSP.
>>>>>>>=20
>>>>>>>    Also when a Timing message is received from a Timing-capable/
>>>>>>>    aware MPLS interface, LER updates the Correction Filed (CF) and
>>>>>>>    decapsulates the MPLS encapsulation and forwards the timing messa=
ge
>>>>>>>    to a non-MPLS interface.
>>>>>>>=20
>>>>>>>    When a Timing-capable/aware LER behaves as a Boundary clock and
>>>>>>>    receives a Timing message from a Timing-capable/aware non MPLS
>>>>>>>    interface, the LER Timestamps the Timing packet and sends it to t=
he
>>>>>>>    LERs Boundary clock processing module.  Also when a Timing messag=
e is
>>>>>>>    received from a Timing- capable/aware MPLS interface, the LER
>>>>>>>    Timestamps the Timing packet and sends it to the LERs Boundary cl=
ock
>>>>>>>    processing module.
>>>>>>>=20
>>>>>>>    When a Timing-capable/aware LER behaves as an Ordinary Clock towa=
rd
>>>>>>>    the MPLS network, and receives a Timing message from a Timing-
>>>>>>>    capable/aware MPLS interface, the LER Timestamps the Timing packe=
t
>>>>>>>    and sends it to the LERs Ordinary clock processing module.
>>>>>>>=20
>>>>>>> 15.2.  Behavior of Timing-capable/aware LSR
>>>>>>>=20
>>>>>>>    When a Timing-capable/aware LSR behaves as a Transparent clock an=
d
>>>>>>>    receives a Timing message from a Timing-capable/aware MPLS
>>>> interface,
>>>>>>>    The LSR updates the Correction Filed (CF) and forwards the timing=

>>>>>>>    message over another MPLS interface.
>>>>>>>=20
>>>>>>>    When a Timing-capable/aware LSR behaves as a Boundary clock and
>>>>>>>    receives a Timing message from a Timing-capable/aware MPLS
>>>> interface.
>>>>>>>    The LSR performs the functions of a Boundary Clock in terminating=
 the
>>>>>>>    received Timing message and re-generating a new timing message ov=
er
>>>>>>>    another (or the same) MPLS interface.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 23]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>>>>>>=20
>>>>>>>    It is most beneficial when all LSRs in the path of a Timing LSP b=
e
>>>>>>>    timing-Capable/aware LSRs.  This would ensure the highest quality=

>>>>>>>    time and clock synchronization by Timing Slave Clocks.  However, t=
his
>>>>>>>    specification does not mandate that all LSRs in path of a Timing L=
SP
>>>>>>>    be Timing- capable/aware.
>>>>>>>=20
>>>>>>>    Non-Timing-capable/aware LSRs just switch the packets encapsulate=
d in
>>>>>>>    Timing LSPs and dont perform any Timing operation (TC or BC).
>>>>>>>    However as explained in QoS section the Timing over MPLS packets
>>>> MUST
>>>>>>>    be still be treated with the highest priority based on their Traf=
fic
>>>>>>>    Class (TC) marking.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 24]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 16.  Other considerations
>>>>>>>=20
>>>>>>>    [IEEE-1588] defines an optional peer-to-peer Transparent clocking=

>>>>>>>    that requires peer delay measurement between two adjacent Timing-=

>>>>>>>    capable/ aware routers/switches.  Peer delay measurement messages=

>>>>>>>    need to be time stamped and terminated by the Timing-capable/awar=
e
>>>>>>>    routers/ switches.  This means that two adjacent LSRs may be enga=
ged
>>>>>>>    in a peer delay measurement.
>>>>>>>=20
>>>>>>>    For transporting such peer delay measurement messages a single-ho=
p
>>>>>>>    LSP SHOULD to be created between the two adjacent LSRs engaged in=

>>>>>>>    peer delay measurement to carry peer delay measurement messages.
>>>>>>>    Other methods such as PTP transport over Ethernet MAY be used for=

>>>>>>>    transporting peer delay measurement messages if the link between t=
he
>>>>>>>    two routers is Ethernet.
>>>>>>>=20
>>>>>>>    In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ w=
are
>>>>>>>    routers/switches MUST maintain a list of all the neighbors it nee=
ds
>>>>>>>    to send a PDelay_Req to, where each neighbor corresponds to a tim=
ing
>>>>>>>    LSP.
>>>>>>>=20
>>>>>>>    The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as=
 long
>>>>>>>    as either the Explicit Null label is the bottom of stack label
>>>>>>>    (applicable only to UDP/IP encapsulation) or the label below the
>>>>>>>    Explicit Null label is a PTP label.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 25]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 17.  Security Considerations
>>>>>>>=20
>>>>>>>    MPLS PW security considerations in general are discussed in [RFC3=
985]
>>>>>>>    and [RFC4447],and those considerations also apply to this documen=
t.
>>>>>>>=20
>>>>>>>    An experimental security protocol is defined in [IEEE-1588].The P=
TP
>>>>>>>    security extension and protocol provides group source authenticat=
ion,
>>>>>>>    message integrity, and replay attack protection for PTP messages.=

>>>>>>>=20
>>>>>>>    When the MPLS network (provider network) serves multiple customer=
s,
>>>>>>>    it is important to maintain and process each customers clock and
>>>>>>>    Timing messages separately from other customers to ensure there i=
s no
>>>>>>>    cross- customer effect.  For example if an LER BC is synchronized=
 to
>>>>>>>    a specific grandmaster, belonging to customer A, then the LER MUS=
T
>>>>>>>    use that BC clock only for customer A to ensure that customer A
>>>>>>>    cannot attack other customers by manipulating its time.
>>>>>>>=20
>>>>>>>    Timing messages MAY be encrypted or authenticated, provided that t=
he
>>>>>>>    LERs/LSRs that are Timing capable/aware can authenticate/ decrypt=
 the
>>>>>>>    timing messages.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 26]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 18.  Acknowledgements
>>>>>>>=20
>>>>>>>    The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizr=
ahi,
>>>>>>>    Stefano Ruffini, Peter Meyer, and other members of IETF for revie=
wing
>>>>>>>    and providing feedback on this draft.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 27]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 19.  IANA Considerations
>>>>>>>=20
>>>>>>>    There are no IANA requirements in this specification.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 28]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 20.  References
>>>>>>>=20
>>>>>>> 20.1.  Normative References
>>>>>>>=20
>>>>>>>    [IEEE-1588]
>>>>>>>               IEEE 1588-2008, "IEEE Standard for a Precision Clock
>>>>>>>               Synchronization Protocol for Networked Measurement and=

>>>>>>>               Control Systems".
>>>>>>>=20
>>>>>>>    [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>>>>>>               Requirement Levels", BCP 14, RFC 2119, March 1997.
>>>>>>>=20
>>>>>>>    [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to=
-
>>>>>>>               Edge (PWE3) Architecture", RFC 3985, March 2005.
>>>>>>>=20
>>>>>>>    [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discov=
ery
>>>>>>>               Proxies (ND Proxy)", RFC 4389, April 2006.
>>>>>>>=20
>>>>>>>    [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G=
.
>>>>>>>               Heron, "Pseudowire Setup and Maintenance Using the Lab=
el
>>>>>>>               Distribution Protocol (LDP)", RFC 4447, April 2006.
>>>>>>>=20
>>>>>>>    [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>>>>>>               "Encapsulation Methods for Transport of Ethernet over M=
PLS
>>>>>>>               Networks", RFC 4448, April 2006.
>>>>>>>=20
>>>>>>>    [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>>>>>>               Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>>>>>>               Retention", RFC 4720, November 2006.
>>>>>>>=20
>>>>>>>    [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circu=
it
>>>>>>>               Connectivity Verification (VCCV): A Control Channel fo=
r
>>>>>>>               Pseudowires", RFC 5085, December 2007.
>>>>>>>=20
>>>>>>>    [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detect=
ion
>>>>>>>               (BFD)", RFC 5880, June 2010.
>>>>>>>=20
>>>>>>>    [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow=
,
>>>>>>>               "Bidirectional Forwarding Detection (BFD) for MPLS Lab=
el
>>>>>>>               Switched Paths (LSPs)", RFC 5884, June 2010.
>>>>>>>=20
>>>>>>> 20.2.  Informative References
>>>>>>>=20
>>>>>>>    [I-D.ietf-pwe3-fat-pw]
>>>>>>>               Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Reg=
an,
>>>>>>>               J., and S. Amante, "Flow Aware Transport of Pseudowire=
s
>>>>>>>               over an MPLS Packet Switched Network",
>>>>>>>               draft-ietf-pwe3-fat-pw-07 (work in progress), July 201=
1.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 29]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermedia=
te
>>>>>>>               system routeing information exchange protocol for use i=
n
>>>>>>>               conjunction with the Protocol for providing the
>>>>>>>               Connectionless-mode Network Service (ISO 8473)".
>>>>>>>=20
>>>>>>>    [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP an=
d
>>>>>>>               dual environments", RFC 1195, December 1990.
>>>>>>>=20
>>>>>>>    [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 199=
8.
>>>>>>>=20
>>>>>>>    [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color=

>>>>>>>               Marker", RFC 2697, September 1999.
>>>>>>>=20
>>>>>>>    [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boud=
ec,
>>>>>>>               J., Courtney, W., Davari, S., Firoiu, V., and D.
>>>>>>>               Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>>>>>>               Behavior)", RFC 3246, March 2002.
>>>>>>>=20
>>>>>>>    [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Enginee=
ring
>>>>>>>               (TE) Extensions to OSPF Version 2", RFC 3630,
>>>>>>>               September 2003.
>>>>>>>=20
>>>>>>>    [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermedia=
te
>>>>>>>               System (IS-IS) Extensions for Traffic Engineering (TE)=
",
>>>>>>>               RFC 3784, June 2004.
>>>>>>>=20
>>>>>>>    [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S=
.
>>>>>>>               Shaffer, "Extensions to OSPF for Advertising Optional
>>>>>>>               Router Capabilities", RFC 4970, July 2007.
>>>>>>>=20
>>>>>>>    [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate=

>>>>>>>               System to Intermediate System (IS-IS) Extensions for
>>>>>>>               Advertising Router Information", RFC 4971, July 2007.
>>>>>>>=20
>>>>>>>    [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi=

>>>>>>>               Topology (MT) Routing in Intermediate System to
>>>>>>>               Intermediate Systems (IS-ISs)", RFC 5120, February 200=
8.
>>>>>>>=20
>>>>>>>    [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>>>>>>               Engineering", RFC 5305, October 2008.
>>>>>>>=20
>>>>>>>    [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>>>>>>               "Traffic Engineering Extensions to OSPF Version 3",
>>>>>>>               RFC 5329, September 2008.
>>>>>>>=20
>>>>>>>    [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSP=
F
>>>>>>>               for IPv6", RFC 5340, July 2008.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 30]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Net=
work
>>>>>>>               Time Protocol Version 4: Protocol and Algorithms
>>>>>>>               Specification", RFC 5905, June 2010.
>>>>>>>=20
>>>>>>>    [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Reg=
an,
>>>>>>>               J., and S. Amante, "Flow-Aware Transport of Pseudowire=
s
>>>>>>>               over an MPLS Packet Switched Network", RFC 6391,
>>>>>>>               November 2011.
>>>>>>>=20
>>>>>>>    [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., a=
nd
>>>>>>>               L. Yong, "The Use of Entropy Labels in MPLS Forwarding=
",
>>>>>>>               RFC 6790, November 2012.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 31]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 1.  Routing extensions for Timing-aware Routers
>>>>>>>=20
>>>>>>>    MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] a=
nd
>>>>>>>    IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (=
TE)
>>>>>>>    link information used for constraint-based routing.
>>>>>>>=20
>>>>>>>    Indeed, it is useful to advertise data plane TE router link
>>>>>>>    capabilities, such as the capability for a router to be Timing-aw=
are.
>>>>>>>    This capability MUST then be taken into account during path
>>>>>>>    computation to prefer or even require links that advertise themse=
lves
>>>>>>>    as Timing-aware.  In this way the path can ensure the entry and e=
xit
>>>>>>>    points into the LERs and, if desired, the links into the LSRs are=

>>>>>>>    able to perform port based time-stamping thus minimizing their im=
pact
>>>>>>>    on the performance of the slave clock.
>>>>>>>=20
>>>>>>>    extensions are required to OSPF and IS-IS in order to advertise
>>>>>>>    Timing-aware capabilities of a link.  Such extensions are outside=
 the
>>>>>>>    scope of this document; however such extension SHOULD be able to
>>>>>>>    signal the following information per Router Link:
>>>>>>>=20
>>>>>>>    o  Capable of processing PTP, NTP or other Timing flows
>>>>>>>=20
>>>>>>>    o  Capable of performing Transparent Clock operation
>>>>>>>=20
>>>>>>>    o  Capable of performing Boundary Clock operation
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 32]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> 2.  Signaling Extensions for Creating Timing LSPs
>>>>>>>=20
>>>>>>>    RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSV=
P-TE
>>>>>>>    is used to setup Timing LSPs, some information that indicates tha=
t
>>>>>>>    the LSP is carrying Timing flows MUST be included in the new
>>>>>>>    Extensions to RSVP-TE:
>>>>>>>=20
>>>>>>>    The following information MAY also be included in the new Extensi=
ons
>>>>>>>    to RSVP-TE:
>>>>>>>=20
>>>>>>>    o  Offset from Bottom of Stack (BoS) to the start of the Time-sta=
mp
>>>>>>>       field
>>>>>>>=20
>>>>>>>    o  Number of VLANs in case of PW encapsulation
>>>>>>>=20
>>>>>>>    o  Timestamp field Type
>>>>>>>=20
>>>>>>>       *  Correction Field, Timestamp
>>>>>>>=20
>>>>>>>    o  Timestamp Field format
>>>>>>>=20
>>>>>>>       *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit=

>>>>>>>          NTP, etc.
>>>>>>>=20
>>>>>>>    Note that in case the above optional information is signaled with=

>>>>>>>    RSVP-TE for a Timing LSP, all the Timing packets carried in that L=
SP
>>>>>>>    must have the same signaled characteristics.  For example if
>>>>>>>    Timestamp format is signaled as 64-bit PTPv1, then all Timing pac=
kets
>>>>>>>    must use 64-bit PTPv1 time-stamp.
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 33]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>> Authors' Addresses
>>>>>>>=20
>>>>>>>    Shahram Davari
>>>>>>>    Broadcom Corp.
>>>>>>>    San Jose, CA  95134
>>>>>>>    USA
>>>>>>>=20
>>>>>>>    Email: davari@broadcom.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Amit Oren
>>>>>>>    Broadcom Corp.
>>>>>>>    San Jose, CA  95134
>>>>>>>    USA
>>>>>>>=20
>>>>>>>    Email: amito@broadcom.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Manav Bhatia
>>>>>>>    Alcatel-Lucent
>>>>>>>    Bangalore,
>>>>>>>    India
>>>>>>>=20
>>>>>>>    Email: manav.bhatia@alcatel-lucent.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Peter Roberts
>>>>>>>    Alcatel-Lucent
>>>>>>>    Kanata,
>>>>>>>    Canada
>>>>>>>=20
>>>>>>>    Email: peter.roberts@alcatel-lucent.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Laurent Montini
>>>>>>>    Cisco Systems
>>>>>>>    San Jose CA
>>>>>>>    USA
>>>>>>>=20
>>>>>>>    Email: lmontini@cisco.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 34]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2=
013
>>>>>>>=20
>>>>>>>=20
>>>>>>>    Luca
>>>>>>>    Cisco Systems
>>>>>>>    San Jose CA
>>>>>>>    USA
>>>>>>>=20
>>>>>>>    Email: lmartini@cisco.com
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>>=20
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 35]
>>>>>>> --
>>>>>>> For corporate legal information go to:
>>>>>>>=20
>>>>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>>>>=20
>>>>>>> _______________________________________________
>>>>>>> TICTOC mailing list
>>>>>>> TICTOC@ietf.org
>>>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>>> This e-mail message is intended for the recipient only and contains
>>>> information which is CONFIDENTIAL and which may be proprietary to ECI
>>>> Telecom. If you have received this transmission in error, please inform=
 us by e-
>>>> mail, phone or fax, and then delete the original and all copies thereof=
.
>>>>>> _______________________________________________
>>>>>> mpls mailing list
>>>>>> mpls@ietf.org
>>>>>> https://www.ietf.org/mailman/listinfo/mpls
>>>>> .
>>>>=20
>>>> --
>>>> For corporate legal information go to:
>>>>=20
>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>=20
>>>> _______________________________________________
>>>> TICTOC mailing list
>>>> TICTOC@ietf.org
>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>=20
>>>>=20
>>>> _______________________________________________
>>>> TICTOC mailing list
>>>> TICTOC@ietf.org
>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>=20
>>> This e-mail message is intended for the recipient only and contains info=
rmation which is CONFIDENTIAL and which may be proprietary to ECI Telecom. I=
f you have received this transmission in error, please inform us by e-mail, p=
hone or fax, and then delete the original and all copies thereof.
>> .
>=20
>=20
> --
> For corporate legal information go to:
>=20
> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>=20
> This e-mail message is intended for the recipient only and contains inform=
ation which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If y=
ou have received this transmission in error, please inform us by e-mail, pho=
ne or fax, and then delete the original and all copies thereof.
>=20

From Alexander.Vainshtein@ecitele.com  Tue Aug 20 11:54:39 2013
Return-Path: <Alexander.Vainshtein@ecitele.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 397D611E8202 for <tictoc@ietfa.amsl.com>; Tue, 20 Aug 2013 11:54:39 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.202
X-Spam-Level: 
X-Spam-Status: No, score=-2.202 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, MIME_QP_LONG_LINE=1.396, RCVD_IN_DNSWL_LOW=-1,  UNPARSEABLE_RELAY=0.001]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id xVHSyQ85Z63H for <tictoc@ietfa.amsl.com>; Tue, 20 Aug 2013 11:54:34 -0700 (PDT)
Received: from mail1.bemta3.messagelabs.com (mail1.bemta3.messagelabs.com [195.245.230.166]) by ietfa.amsl.com (Postfix) with ESMTP id 686F811E814E for <tictoc@ietf.org>; Tue, 20 Aug 2013 11:54:24 -0700 (PDT)
Received: from [85.158.137.68:41310] by server-6.bemta-3.messagelabs.com id 24/51-32441-D5BB3125; Tue, 20 Aug 2013 18:54:21 +0000
X-Env-Sender: Alexander.Vainshtein@ecitele.com
X-Msg-Ref: server-5.tower-31.messagelabs.com!1377024858!2840732!3
X-Originating-IP: [147.234.242.234]
X-StarScan-Received: 
X-StarScan-Version: 6.9.11; banners=-,-,-
X-VirusChecked: Checked
Received: (qmail 25190 invoked from network); 20 Aug 2013 18:54:20 -0000
Received: from ilptbmg01-out.ecitele.com (HELO ilptbmg01-out.ecitele.com) (147.234.242.234) by server-5.tower-31.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 20 Aug 2013 18:54:20 -0000
X-AuditID: 93eaf2e7-b7f0e6d000004f79-06-5213bac3d386
Received: from ILPTWPVEXCA02.ecitele.com ( [172.31.244.232]) by ilptbmg01-out.ecitele.com (Symantec Messaging Gateway) with SMTP id D9.D4.20345.3CAB3125; Tue, 20 Aug 2013 21:51:48 +0300 (IDT)
Received: from ILPTWPVEXMB01.ecitele.com ([fe80::f152:8eaf:8fb0:a5da]) by ILPTWPVEXCA02.ecitele.com ([fe80::c473:490d:3a7e:e34a%12]) with mapi id 14.03.0123.003; Tue, 20 Aug 2013 21:51:47 +0300
From: Alexander Vainshtein <Alexander.Vainshtein@ecitele.com>
To: "S. Davari" <davarish@yahoo.com>
Thread-Topic: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
Thread-Index: AQHOkhqJ+XHALQ/MzkWnZzr9BEl1M5mHyTyggBaXBuSAAAM16f//06OAgABOFGw=
Date: Tue, 20 Aug 2013 18:51:46 +0000
Message-ID: <F9336571731ADE42A5397FC831CEAA021513A0C6@ILPTWPVEXMB01.ecitele.com>
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com> <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com> <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com> <52139903.8040801@cisco.com> <F9336571731ADE42A5397FC831CEAA021513A083@ILPTWPVEXMB01.ecitele.com>, <B7A656C4-6D0E-4719-9E42-57689CDAEFA9@yahoo.com>
In-Reply-To: <B7A656C4-6D0E-4719-9E42-57689CDAEFA9@yahoo.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-originating-ip: [147.234.1.1]
Content-Type: text/plain; charset="us-ascii"
content-transfer-encoding: quoted-printable
MIME-Version: 1.0
X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFvrLIsWRmVeSWpSXmKPExsUy+dWnV7pRu4WDDHa9FLJY3+tp8be5h92B yWPW/bNsHkuW/GQKYIpqYLRJzMvLL0ksSVVISS1OtlUKKMosS0yuVFLITLFVMlRSKMhJTE7N Tc0rsVVKLChIzUtRsuNSwAA2QGWZeQqpecn5KZl56bZKnsH+uhYWppa6hkp2asqGxtZcIRmZ xQqpurmJmTkKuanFxYnpqQpAkYQtzBl355UX/FzGWtF04RZ7A+Pev8xdjJwcEgImEs+7e6Fs MYkL99azdTFycQgJHGSU+PTgBSuEc5RRYk7fdxaQKjYBW4lNq++ygdgiAioSS7cvBYszC8wF 6vjt38XIwSEs4C5x+IcfRImHROfU84wQtp/EvouzwZaxCKhK/P69lR3E5hUIkNh2+z4LxK7Z LBK/eh+DzecE2jX1wl+wZkag676fWsMEsUtc4taT+UwQVwtILNlzHuoDUYmXj/+xQtjyEhc/ PICq15FYsPsTG4StLbFs4WtmiMWCEidnPmGZwCg2C8nYWUhaZiFpmYWkZQEjyypG0cycgpKk 3HQDQ73U5MyS1JxUveT83E2MkITxfAfjr/kqhxhdgZ6dyCzFnZwPTDh5JfHGBga4OUrivMsb wv2FBNKBqSU7NbUgtSi+qDQntfgQIxMHp1QD44VWDx2eI3M+vkuMzSuMklx455mKtdqG0yvm ONqcLNzudV9qYY/1hWbXGTULJqRteSTzX9jlcSgDi9mhpV8cIw8tTp5oeevQ3wfbTfuvC/ta u37eb7vJ6LQy96MFLy8vEWd44SYtYDY3xt9vfsOEXzcuS185wpZd7rp85+bDzVs8Tpwqvm3S elCJpTgj0VCLuag4EQCGWe9aEAMAAA==
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 20 Aug 2013 18:54:39 -0000

Shahram,
No I am not proposing anything of the kind.

At the MPLS data plane level the only the head-end and the tail-end LERs hav=
e to be aware of the BoS Timing Alert labels. Transit LSRs simply will ignor=
e them.

At the on-path assistance level each transit LSR (and the tail-end LER) can=
 do one of the following:
- It may ignore the Timing Alert label at the BoS and just forward the packe=
t based the labels at the top of the stack.
- It may identify the packet as a timing one based on the Timing Alert label=
 at the bottom of the label stack, and perform the required operations. The=
 result of these operations will be stored in the "timing shim header" and n=
ot in the packet itself. (E.g., an E2E TC clock would note the packet reside=
nce time and add it to the value already carried in the shim header). 

Non-timing packets would not have the Timing Alert label at the BoS and henc=
e would travel thru he LSP travel along the LSP in the usual way.

The tail-end LER would take the reception timestamp and forward this timesta=
mp, the protocol packet and the shim header to the entity that processes the=
 appropriate timing distribution protocol based on the Timing Alert label. 

Hopefully this clarifies my proposal. My gut feeling is that similar ideas h=
ave been proposed by many people earlier...

Regards,
     Sasha


________________________________________
From: S. Davari [davarish@yahoo.com]
Sent: Tuesday, August 20, 2013 7:00 PM
To: Alexander Vainshtein
Cc: stbryant@cisco.com; Shahram Davari; mpls@ietf.org; tictoc@ietf.org
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls

Sasha

In (2) are you proposing to create multiple segment LSPs, where each if thos=
e segment LSPs span between two Timing aware LSR? So if I have 10 timing awa=
re LSR in the path I would need 9 LSPs?

Regards,
Shahram


On Aug 20, 2013, at 9:53 AM, Alexander Vainshtein <Alexander.Vainshtein@ecit=
ele.com> wrote:

> Stewart, Shahram and all,
>
> A couple of (hopefully, last) comments:
>
> 1. Supporting a "Timing Alert" label at the top of the label stack introdu=
ces a backward compatibility problem. We differ in our estimate of the impor=
tance of this problem, but we all agree that it exists.
>
> 2. On the other hand, this problem does not exist if the "Timing Alert" la=
bel is inserted at the bottom of the label stack. At the same time, HW that=
 analyzes the label stack hopefully could analyze the bottom label (almost)=
 as easily as it would analyze the top label. Thus we could both provide bac=
kward compatibility (incompatible transit LSRs simply would not notice the "=
Timing Alert" label") and avoid the need for dedicated "Timing-only" LSPs. I=
n particular, all the MPLS OAM tools (including MPLS-TP) would work without=
 any changes.
>
> 3. A "timing shim header" (between the bottom of the label stack and the b=
eginning of the timing  packet) has been mentioned by Stewart as an already=
 proposed mechanism. IMHO and FWIW its usage would eliminate the need to und=
erstand format of the specific timing distribution protocol in the transit L=
SRs.  It would further solve the problem of secure transport of timing packe=
ts since security mechanisms would be applied to its original encapsulation=
 (over IP or over Ethernet) without involving the transit LSRs.
>
> Did I miss something substantial?
>
> Regards, and lots of thanks in advance,
>     Sasha
>
>
> ________________________________________
> From: Stewart Bryant [stbryant@cisco.com]
> Sent: Tuesday, August 20, 2013 6:27 PM
> To: S. Davari
> Cc: Alexander Vainshtein; Shahram Davari; mpls@ietf.org; tictoc@ietf.org
> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>
> On 06/08/2013 14:35, S. Davari wrote:
>> Sasha,
>>
>> If a router is not 1588 aware we want it to switch the packet with minimu=
m jitter and delay via high priority queue.
> Yes, but you can explicitly tunnel across such island of incompatibility
>
> Stewart
>
>>
>>
>> Sending it to CPU creates large jitter and delay since the CPUs are gener=
ally busy doing other things. Time stamping or updating CF in the CPU won't=
 help since it does not cover the variable delay that is incurred from the t=
ime the packet enters the router to the time the packet gets processed by CP=
U.
>>
>> I wish it was that easy!
>>
>> Regards,
>> Shahram
>>
>>
>> On Aug 6, 2013, at 12:38 AM, Alexander Vainshtein <Alexander.Vainshtein@e=
citele.com> wrote:
>>
>>> Shahram,
>>> I agree with you that the routers that do not recognize a Timing Alert l=
abel would send the packet to the CPU. (This would be easy to arrange since=
 we are speaking about a reserved label.)
>>>
>>> I do not, however, see this as a serious issue, because the SW running o=
n this CPU would (hopefully) be upgradable to handle the packet correctly i.=
e., to forward it to where it should be forwarded and, if it is forwarded as=
 a labeled packet, to prepend the Timing Alert Label on top of its label sta=
ck. It could even record the residence time (as observed by the CPU in the p=
roper place in the packet. This would mean that introducing additional error=
 to whatever timing information is associated with this packet - but this is=
 what you should anyway expect if there are non-compliant routers on your pa=
th, right?
>>>
>>> Regards,
>>>     Sasha
>>>
>>>> -----Original Message-----
>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behal=
f Of
>>>> Shahram Davari
>>>> Sent: Monday, August 05, 2013 11:29 PM
>>>> To: stbryant@cisco.com; S. Davari
>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>>>
>>>> Hi Stewart,
>>>>
>>>> Your suggestion of " I suggested an LSP type that has the properties "t=
imestamp
>>>> and pass to application", is a subset of the existing draft and should=
 work. It
>>>> basically limits the time stamping and correction field update to LERs=
 (while the
>>>> draft supports time stamping at LER and LSR).
>>>>
>>>> However Sasha's suggestion of using a reserved label (RAL, GAL or any o=
ther
>>>> reserved label) does not satisfy one of the major requirements, which i=
s
>>>> backward compatibility. Routers that don't understand this reserved lab=
el will
>>>> drop or copy to CPU such packets.
>>>>
>>>> Regards,
>>>> Shahram
>>>>
>>>> -----Original Message-----
>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behal=
f Of
>>>> Stewart Bryant
>>>> Sent: Monday, August 05, 2013 2:29 AM
>>>> To: S. Davari
>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>>>
>>>> My concern is that we are creating a new LSP type in MPLS
>>>> (which is a architectural change and thus needs to be very carefully
>>>> considered) without asking the question "how can I do this
>>>> in the most general way, to maximize flexibility/reuse"
>>>>
>>>> I suggested an LSP type that has the properties "timestamp
>>>> and pass to application", but I think that Sasha raises a good
>>>> point about using router alert. However RA has no implicit
>>>> timestamp, so maybe we need a new type of RA that has the
>>>> properties "timestamp, and pass top application indicated by
>>>> the next label". That would be quite useful in a number
>>>> of OAM applications. There is possibly some GAL variant
>>>> of that design that should also be considered.
>>>>
>>>> The application could worry about the path and where it
>>>> was necessary to skip some hops a hierarchical LSP would
>>>> accomplish that, with the specific benefit that the application
>>>> would consciously do this, and may be able to apply some
>>>> form of compensation within the network.
>>>>
>>>> - Stewart
>>>>
>>>> On 04/08/2013 10:40, S. Davari wrote:
>>>>> Hi Sasha
>>>>>
>>>>> Perhaps you have not understood the draft well. The main reason that t=
he
>>>> draft chose to use a dedicated LSP that is signaled via RSVPTE indicati=
ng it is a
>>>> timing LSP is to be backward compatible with non-1588 nodes. Such nodes
>>>> simply just switch the packet.
>>>>> Your proposal had been considered and was rejected since it was not
>>>> backward compatible. Nodes receiving alert label or TTL =3D 1 send the=
 packet to
>>>> CPU. You can refer to the meeting notes.
>>>>> Regards,
>>>>> Shahram
>>>>>
>>>>>
>>>>> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein
>>>> <Alexander.Vainshtein@ecitele.com> wrote:
>>>>>> Stewart and all,
>>>>>> I concur with Stewart's statement that the draft does not define "ful=
l
>>>> interaction with MPLS architecture".
>>>>>> E.g., one of the objectives of the draft is to provide a technique th=
at would
>>>> be backward-compatible with old LSRs that cannot provide on-path suppor=
t for
>>>> timing distribution, while the other objective is to make every LSR on=
 the path
>>>> aware that some MPLS packets are carrying timing-related messages and h=
ence
>>>> require on-path support.
>>>>>> IMHO and FWIW, the MPLS data plane architecture allows just two metho=
ds
>>>> for making a transit LSR to provide special processing to a labeled pac=
ket:
>>>>>> - It would carry some kind of an "alert label" on top of the label st=
ack and,
>>>> specifically, on top of any labels used for actual forwarding
>>>>>> OR,
>>>>>> - The TTL in the top label stack entry has been set to 1.
>>>>>>
>>>>>>
>>>>>> The draft does not follow any of these approaches.
>>>>>>
>>>>>> I must also admit that Section 12 "OAM, Control and Management" of th=
e
>>>> draft looks somewhat in
>>>>>> E.g., I could not understand from the text whether BFD, LSP-Ping etc.=
 OAM
>>>> messages would be subjected to on-path support procedures for timing
>>>> messages or not.
>>>>>> My 2c,
>>>>>>     Sasha
>>>>>>
>>>>>>> -----Original Message-----
>>>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Be=
half
>>>> Of
>>>>>>> Stewart Bryant
>>>>>>> Sent: Friday, August 02, 2013 12:01 PM
>>>>>>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
>>>>>>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.o=
rg;
>>>> draft-
>>>>>>> ietf-tictoc-1588overmpls@tools.ietf.org
>>>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>>>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
>>>>>>>
>>>>>>> SB> This draft does not seem to provide a precise definition
>>>>>>> SB> the properties of the new LSP type that it wishes to
>>>>>>> SB> define, in particular it does it define the PHB of those
>>>>>>> SB> LSPs, nor the full interaction with the MPLS
>>>>>>> SB> architecture.
>>>>>>> SB>
>>>>>>> SB> I have not tracked TICTOC for a while but I thought that
>>>>>>> SB> the original plan was to define the concept of an offset
>>>>>>> SB> into a packet to do the correction.
>>>>>>> SB>
>>>>>>> SB> It is disappointing that the opportunity was not taken
>>>>>>> SB> to define a timing shim inside the timing LSP so that
>>>>>>> SB> a time correction could be added to any packet such that
>>>>>>> SB> the MPLS system was isolated from the details of the
>>>>>>> SB> complexity of the particular time transfer type.
>>>>>>>
>>>>>>> SB> I think that much more clarify is needed in terms of
>>>>>>> SB> definition of the new LSP type, since it is unclear
>>>>>>> SB> from this text how to implement one.
>>>>>>> SB>
>>>>>>> SB> There are a lot of other MPLS services such as
>>>>>>> SB> LSP ping that need to be considered.
>>>>>>> SB>
>>>>>>> SB> Please see inline for more comments. However these
>>>>>>> SB> comments are made in the context of the text as written
>>>>>>> SB> whilst I have a fundamental concern that this approach
>>>>>>> SB> lacks an MPLS architectural soundness that need
>>>>>>> SB> greater thought with significant impact on the
>>>>>>> SB> draft.
>>>>>>>
>>>>>>> - Stewart
>>>>>>>
>>>>>>>
>>>>>>> TICTOC Working Group                                           S. Da=
vari
>>>>>>> Internet-Draft                                                   A.=
 Oren
>>>>>>> Intended status: Standards Track                          Broadcom C=
orp.
>>>>>>> Expires: December 17, 2013                                     M. Bh=
atia
>>>>>>>                                                               P. Rob=
erts
>>>>>>>                                                           Alcatel-Lu=
cent
>>>>>>>                                                               L. Mon=
tini
>>>>>>>                                                               L. Mar=
tini
>>>>>>>                                                            Cisco Sys=
tems
>>>>>>>                                                            June 15,=
 2013
>>>>>>>
>>>>>>>
>>>>>>>             Transporting Timing messages over MPLS Networks
>>>>>>>                    draft-ietf-tictoc-1588overmpls-05
>>>>>>>
>>>>>>> Abstract
>>>>>>>
>>>>>>>    This document defines the method for transporting Timing messages
>>>>>>>    such as PTP and NTP over an MPLS network.  The method allows for=
 the
>>>>>>>    easy identification of these PDUs at the port level to allow for=
 port
>>>>>>>
>>>>>>> SB> What is a port
>>>>>>>
>>>>>>>    level processing of these PDUs in both LERs and LSRs.
>>>>>>>
>>>>>>>    The basic idea is to transport Timing messages inside dedicated M=
PLS
>>>>>>>    LSPs.  These LSPs only carry Timing messages and possibly Control=
 and
>>>>>>>    Management packets, but they do not carry customer traffic.
>>>>>>>
>>>>>>> SB> More specifically they only carry traffic associated with the
>>>>>>> SB> timing service and its support.
>>>>>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
>>>>>>> SB> also it gets carried in a structure that causes it to get
>>>>>>> SB> timestamped.
>>>>>>>
>>>>>>>    Two methods for transporting Timing messages over MPLS are define=
d.
>>>>>>>
>>>>>>> SB> Perhaps the right approach is to define the new LSP type and the=
n
>>>>>>> SB> seperately to define  the mapping of the various timing services
>>>>>>> SB> over that LSP type.
>>>>>>>
>>>>>>>    The first method is to transport Timing messages directly over th=
e
>>>>>>>    dedicated MPLS LSP via UDP/IP encapsulation, which is suitable fo=
r
>>>>>>>    MPLS networks.  The second method is to transport Timing messages
>>>>>>>    inside a PW via Ethernet encapsulation.
>>>>>>>
>>>>>>> SB> I think that we should note that there are some
>>>>>>> SB> h/w reasons for this preference. A clean sheet approach
>>>>>>> SB> would have been to use PTP over MPLS with no intermediate
>>>>>>> SB> layers.
>>>>>>>
>>>>>>> Status of this Memo
>>>>>>>
>>>>>>>    This Internet-Draft is submitted in full conformance with the
>>>>>>>    provisions of BCP 78 and BCP 79.
>>>>>>>
>>>>>>>    Internet-Drafts are working documents of the Internet Engineering
>>>>>>>    Task Force (IETF).  Note that other groups may also distribute
>>>>>>>    working documents as Internet-Drafts.  The list of current Intern=
et-
>>>>>>>    Drafts is at http://datatracker.ietf.org/drafts/current/.
>>>>>>>
>>>>>>>    Internet-Drafts are draft documents valid for a maximum of six mo=
nths
>>>>>>>    and may be updated, replaced, or obsoleted by other documents at=
 any
>>>>>>>    time.  It is inappropriate to use Internet-Drafts as reference
>>>>>>>    material or to cite them other than as "work in progress."
>>>>>>>
>>>>>>>    This Internet-Draft will expire on December 17, 2013.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 1]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> Copyright Notice
>>>>>>>
>>>>>>>    Copyright (c) 2013 IETF Trust and the persons identified as the
>>>>>>>    document authors.  All rights reserved.
>>>>>>>
>>>>>>>    This document is subject to BCP 78 and the IETF Trust's Legal
>>>>>>>    Provisions Relating to IETF Documents
>>>>>>>    (http://trustee.ietf.org/license-info) in effect on the date of
>>>>>>>    publication of this document.  Please review these documents
>>>>>>>    carefully, as they describe your rights and restrictions with res=
pect
>>>>>>>    to this document.  Code Components extracted from this document m=
ust
>>>>>>>    include Simplified BSD License text as described in Section 4.e o=
f
>>>>>>>    the Trust Legal Provisions and are provided without warranty as
>>>>>>>    described in the Simplified BSD License.
>>>>>>>
>>>>>>>
>>>>>>> Table of Contents
>>>>>>>
>>>>>>>    1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . .=
 .  5
>>>>>>>
>>>>>>>    2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . .=
 .  7
>>>>>>>
>>>>>>>    3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . .=
 .  8
>>>>>>>
>>>>>>>    4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . .=
 .  9
>>>>>>>
>>>>>>>    5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . .=
 . 12
>>>>>>>
>>>>>>>    6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . .=
 . 13
>>>>>>>      6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . .=
 . 13
>>>>>>>      6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . .=
 . 13
>>>>>>>      6.3.  Other Timing Encapsulation methods . . . . . . . . . . .=
 . 14
>>>>>>>
>>>>>>>    7.  Timing message Processing  . . . . . . . . . . . . . . . . .=
 . 15
>>>>>>>
>>>>>>>    8.  Protection and Redundancy  . . . . . . . . . . . . . . . . .=
 . 16
>>>>>>>
>>>>>>>    9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 17
>>>>>>>
>>>>>>>    10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 18
>>>>>>>
>>>>>>>    11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 19
>>>>>>>
>>>>>>>    12. OAM, Control and Management  . . . . . . . . . . . . . . . .=
 . 20
>>>>>>>
>>>>>>>    13. QoS Considerations . . . . . . . . . . . . . . . . . . . . .=
 . 21
>>>>>>>
>>>>>>>    14. FCS and Checksum Recalculation . . . . . . . . . . . . . . .=
 . 22
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 2]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>>    15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . .=
 . 23
>>>>>>>      15.1. Behavior of Timing-capable/aware LER . . . . . . . . . .=
 . 23
>>>>>>>      15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . .=
 . 23
>>>>>>>      15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . .=
 . 24
>>>>>>>
>>>>>>>    16. Other considerations . . . . . . . . . . . . . . . . . . . .=
 . 25
>>>>>>>
>>>>>>>    17. Security Considerations  . . . . . . . . . . . . . . . . . .=
 . 26
>>>>>>>
>>>>>>>    18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . .=
 . 27
>>>>>>>
>>>>>>>    19. IANA Considerations  . . . . . . . . . . . . . . . . . . . .=
 . 28
>>>>>>>
>>>>>>>    20. References . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 29
>>>>>>>      20.1. Normative References . . . . . . . . . . . . . . . . . .=
 . 29
>>>>>>>      20.2. Informative References . . . . . . . . . . . . . . . . .=
 . 29
>>>>>>>
>>>>>>>    Appendix 1.  Routing extensions for Timing-aware Routers . . . .=
 . 32
>>>>>>>
>>>>>>>    Appendix 2.  Signaling Extensions for Creating Timing LSPs . . .=
 . 33
>>>>>>>
>>>>>>>    Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . .=
 . 34
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 3]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>>    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
>>>> NOT",
>>>>>>>    "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL"
>>>> in
>>>>>>> this
>>>>>>>    document are to be interpreted as described in RFC2119 [RFC2119].
>>>>>>>
>>>>>>>    When used in lower case, these words convey their typical use in
>>>>>>>    common language, and are not to be interpreted as described in
>>>>>>>    RFC2119 [RFC2119].
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 4]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 1.  Introduction
>>>>>>>
>>>>>>>    The objective of Precision Time Protocol (PTP) and Network Timing
>>>>>>>    Protocol (NTP) are to synchronize independent clocks running on
>>>>>>>    separate nodes of a distributed system.
>>>>>>>
>>>>>>>    [IEEE-1588] defines PTP messages for frequency, phase and time
>>>>>>>    synchronization.  The PTP messages include PTP PDUs over UDP/IP
>>>>>>>    (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex=
 F of
>>>>>>>    [IEEE-1588]).
>>>>>>>
>>>>>>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
>>>>>>> SB> of other PTP mappings if they provide better optimisation.
>>>>>>>
>>>>>>>    This document defines mapping and transport of the PTP
>>>>>>>    messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>>>>>>>    defines several clock types: ordinary clocks, boundary clocks, en=
d-
>>>>>>>    to-end transparent clocks, and peer-to-peer transparent clocks.
>>>>>>>    Transparent clocks require intermediate nodes to update correctio=
n
>>>>>>>    field inside PTP message that reflects the transit time in the no=
de.
>>>>>>>
>>>>>>>    [RFC5905] defines NTP messages for clock and time synchronization=
.
>>>>>>>    The PTP messages (PDUs) are transported over UDP/IP.  This docume=
nt
>>>>>>> SB> Should that be NTP messages?
>>>>>>> SB> It needs to be made clear as soon as you introduce NTP that
>>>>>>> SB> they use different time representations.
>>>>>>>
>>>>>>>    defines mapping and transport of the NTP messages defined in
>>>>>>>    [RFC5905] over MPLS networks.
>>>>>>>
>>>>>>>    One key attribute of all of these Timing messages is that the Tim=
e
>>>>>>>    stamp processing should occur as close as possible to the actual
>>>>>>>    transmission and reception at the physical port interface.  This
>>>>>>>    targets optimal time and/or frequency recovery by avoiding variab=
le
>>>>>>>    delay introduced by queues internal to the clocks.
>>>>>>>
>>>>>>> SB> As I recall NTP has no epoch point defined, and I am not sure
>>>>>>> SB> where that point is in the case of PTP in this mapping
>>>>>>> SB> Hopefully this will get defined in due course.
>>>>>>>
>>>>>>>    To facilitate the fast and efficient recognition of Timing messag=
es
>>>>>>>    at the port level when the Timing messages are carried over MPLS
>>>>>>>    LSPs,
>>>>>>>
>>>>>>> SB> Over a new LSP type with time optimied characteristics
>>>>>>>
>>>>>>>    this document defines the specific encapsulations that should
>>>>>>>    be used.
>>>>>>> SB> Hopefully it will also define the PHP
>>>>>>>
>>>>>>>    In addition, it can be expected that there will exist LSR/
>>>>>>>    LERs where only a subset of the physical ports will have the port=
-
>>>>>>>    based Timing message processing capabilities.
>>>>>>> SB> Do you need to clarify that this only works at base and not in
>>>>>>> SB> a label heirarchy.
>>>>>>>
>>>>>>>
>>>>>>>    In order to ensure
>>>>>>>    that the LSPs carrying Timing packets always enter and exit ports
>>>>>>>    with this capability, routing extensions are defined to advertise
>>>>>>>    this capability on a port basis and to allow for the establishmen=
t of
>>>>>>>    LSPs that only transit such ports.  While this path establishment
>>>>>>>    restriction may be applied only at the LER Ingress and/or egress
>>>>>>>    ports, it becomes more important when using transparent clock cap=
able
>>>>>>>    LSRs in the path.
>>>>>>> SB> I do not understand the implications of the last
>>>>>>> SB> sentences - starting ", it becomes"
>>>>>>>
>>>>>>>
>>>>>>>    Port based Timing message processing involves Timing message
>>>>>>>    recognition.  Once the Timing messages are recognized they can be
>>>>>>>    modified based on the reception or transmission Time-stamp.
>>>>>>>
>>>>>>>    This document provides two methods for transporting Timing messag=
es
>>>>>>>    over MPLS.  One is applicable to MPLS environment and the other o=
ne
>>>>>>>    is applicable to MPLS/MPLS-TP environment
>>>>>>>
>>>>>>> SB> I think the sentence is incomplete.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 5]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>>    The solution involves transporting Timing messages over dedicated
>>>>>>>    LSPs called Timing LSPs.  These LSPs carry Timing messages and MA=
Y
>>>>>>>    carry Management and control messages, but not data plane client
>>>>>>>    traffic.
>>>>>>>
>>>>>>> SB> It is not clear why this restriction applies.
>>>>>>>
>>>>>>>    Timing LSPs can be established statically or via signaling.
>>>>>>> SB> s/statically/by provisioning/network management/
>>>>>>>
>>>>>>>    Extensions to control plane (OSPF, ISIS, etc.) is required to ena=
ble
>>>>>>>    routers to distribute their Timing processing capabilities over M=
PLS
>>>>>>>    to other routers.  However such extensions are outside the scope=
 of
>>>>>>>    this document.
>>>>>>>
>>>>>>>    When signaling is used to setup the PTP LSP, Extensions to signal=
ing
>>>>>>> SB> is it a PTP LSP or a Timing LSP?
>>>>>>>
>>>>>>>    protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
>>>>>>>    However such extensions are outside the scope of this document.
>>>>>>>
>>>>>>> SB> for mpls-tp GMPLS is the signalling protocol
>>>>>>>
>>>>>>>    While the techniques included herein allow for the establishment=
 of
>>>>>>>    paths optimized to include Time-stamping capable links, the
>>>>>>>    performance of the Slave clocks is outside the scope of this
>>>>>>>    document.
>>>>>>>
>>>>>>>    At the time of publishing this specification, Transparent Clockin=
g
>>>>>>>    (TC) is only defined for PTP.  Therefore at this time any part of
>>>>>>>    this specification that talks about Transparent Clocking applies=
 only
>>>>>>>    to PTP.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 6]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 2.  Terminology
>>>>>>>
>>>>>>>    1588: The timing and synchronization as defined by IEEE 1588.
>>>>>>>
>>>>>>> SB> I think that there is a more formal name for the 1588 group
>>>>>>> SB> that needs to be used here.
>>>>>>> SB> Also do we need to talk about 1588-200? as there is
>>>>>>> SB> an update in progress
>>>>>>>
>>>>>>>    NTP: The timing and synchronization protocol defined by IETF RFC-=
1305
>>>>>>>    and RFC-5905.
>>>>>>>
>>>>>>>    PTP: The timing and synchronization protocol used by 1588.
>>>>>>> SB> need the proper name for 1588
>>>>>>>
>>>>>>>    Master Clock: The source of 1588 timing to a set of slave clocks.
>>>>>>>
>>>>>>>    Master Port: A port on a ordinary or boundary clock that is in Ma=
ster
>>>>>>>    state.  This is the source of timing toward slave ports.
>>>>>>>
>>>>>>> SB> I am not sure the reader knows what a port is
>>>>>>>
>>>>>>>    Slave Clock: A receiver of 1588 timing from a master clock.
>>>>>>>
>>>>>>>    Slave Port: A port on a boundary clock or ordinary clock that is
>>>>>>>    receiving timing from a master clock.
>>>>>>>
>>>>>>>    Ordinary Clock: A device with a single PTP port.
>>>>>>>
>>>>>>>    Transparent Clock.  A device that measures the time taken for a P=
TP
>>>>>>>    event message to transit the device and then updates the
>>>>>>>    correctionField of the message with this transit time.
>>>>>>>
>>>>>>>    Boundary Clock: A device with more than one PTP port.  Generally
>>>>>>>    boundary clocks will have one port in slave state to receive timi=
ng
>>>>>>>    and then other ports in master state to re-distribute the timing.
>>>>>>>
>>>>>>>    PTP LSP: An LSP dedicated to carry PTP messages
>>>>>>>
>>>>>>> SB> PTP or timing?
>>>>>>>
>>>>>>>
>>>>>>>    PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>>>>>>    messages.
>>>>>>>
>>>>>>> SB> Ah I don't think that PWE3 know what one of these is
>>>>>>>
>>>>>>>    CW: Pseudowire Control Word
>>>>>>>
>>>>>>>    LAG: Link Aggregation
>>>>>>>
>>>>>>>    ECMP: Equal Cost Multipath
>>>>>>>
>>>>>>>    CF: Correction Field, a field inside certain PTP messages (messag=
e
>>>>>>>    type 0-3)that holds the accumulative transit time inside intermed=
iate
>>>>>>>    switches
>>>>>>>
>>>>>>>    Timing messages: Timing Protocol messages that are exchanged
>>>> between
>>>>>>>    routers in order to establish a synchronized clock.
>>>>>>>
>>>>>>> SB> A number of these definitions look like copies of IEEE1588
>>>>>>> SB> definitions. We need to provide references and note the
>>>>>>> SB> priority of the IEEE base reference.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 7]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 3.  Problem Statement
>>>>>>>
>>>>>>>    [IEEE-1588] has defined methods for transporting PTP messages ove=
r
>>>>>>>    Ethernet and IP networks.  [RFC5905] has defined the method of
>>>>>>>    transporting NTP messages over IP networks.  There is a need to
>>>>>>>    transport Timing messages over MPLS networks while supporting the
>>>>>>>    Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (O=
C)
>>>>>>>    functionality in the LER and LSRs in the MPLS network.
>>>>>>>
>>>>>>>    There are multiple ways of transporting Timing over MPLS.  Howeve=
r,
>>>>>>>    there is a requirement to limit the possible encapsulation option=
s to
>>>>>>>    simplify the Timing message identification and processing require=
d at
>>>>>>>    the port level.
>>>>>>>
>>>>>>>    When Timing-awareness is needed, Timing messages should not be
>>>>>>>    transported over LSPs or PWs that are carrying customer traffic
>>>>>>>    because LSRs perform Label switching based on the top label in th=
e
>>>>>>>    stack.
>>>>>>>
>>>>>>> SB> Have you explained why?
>>>>>>>
>>>>>>>    To detect Timing messages inside such LSPs require special
>>>>>>>    hardware to do deep packet inspection at line rate.  Even if such
>>>>>>>    hardware exists, the payload can't be deterministically identifie=
d by
>>>>>>>    LSRs because the payload type is a context of the PW label, and t=
he
>>>>>>>    PW label and its context are only known to the Edge routers (PEs/
>>>>>>>    LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, C=
ES,
>>>>>>>    etc).  Even if one restricts an LSP to only carry Ethernet PWs, t=
he
>>>>>>>    LSRs dont have the knowledge of whether PW Control Word (CW) is
>>>>>>>    present or not and therefore can not deterministically identify t=
he
>>>>>>>    payload.
>>>>>>>
>>>>>>>    A generic method is defined in this document that does not requir=
e
>>>>>>>    deep packet inspection at line rate, and can deterministically
>>>>>>>    identify Timing messages.  This method can be used to detect Timi=
ng
>>>>>>>    Messages in both one-step and two-step clock implementations of
>>>>>>>    ordinary, boundary and transparent clocks.
>>>>>>>
>>>>>>> SB> Needs a ref and I am sure many MPLS specialists will not underst=
and
>>>>>>> SB> the msg types.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 8]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 4.  Timing over MPLS Architecture
>>>>>>>
>>>>>>>    Timing messages are exchange between Timing ports on ordinary and
>>>>>>>
>>>>>>> SB> Have you defined a timing port?
>>>>>>>
>>>>>>>    boundary clocks.  Boundary clocks terminate the Timing messages a=
nd
>>>>>>>    act as master for other boundary clocks or for slave clocks.  End=
-to-
>>>>>>>    End Transparent clocks do not terminate the Timing messages but t=
hey
>>>>>>>    do modify the contents of the Timing messages as they transit acr=
oss
>>>>>>>    the transparent clock.
>>>>>>>
>>>>>>>    Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent C=
lock
>>>>>>>
>>>>>>>    (TC) could be implemented in either LERs or LSRs.
>>>>>>>
>>>>>>> SB> LER and LSR need to be expanded
>>>>>>>
>>>>>>>    An example is shown in Figure 1, where the LERs act as Ordinary C=
lock
>>>>>>>    (OC) and are the initiating/terminating point for Timing messages=
.
>>>>>>>    The ingress LER encapsulates the Timing messages in Timing LSP an=
d
>>>>>>>    the Egress LER terminates the Timing LSP.  The LSRs act as
>>>>>>>    Transparent Clock (TC) and just update the Timing field in the Ti=
ming
>>>>>>>    messages.
>>>>>>>
>>>>>>>
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>       |Switch, |     |       |     |       |     |       |     |Swit=
ch, |
>>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rou=
ter |
>>>>>>>       |        |     |  OC   |     |  TC   |     |  OC   |     |   =
     |
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>                      /                                 \
>>>>>>>       +-------+     /                                   \     +-----=
--+
>>>>>>>       |  LER  |    /                                     \    |  LER=
  |
>>>>>>>       | Master|---/                                       \---| Slav=
e |
>>>>>>>       | Clock |                                               | Cloc=
k |
>>>>>>>       +-------+                                               +-----=
--+
>>>>>>>
>>>>>>>      Figure (1) - Deployment example 1 of timing over MPLS network
>>>>>>>
>>>>>>>    Another example is shown in Figure2, where LERs terminate the Tim=
ing
>>>>>>>    messages received from switch/routers that are outside of the MPL=
S
>>>>>>>    network acting as OC or BC.  In this example LERs regenerate the
>>>>>>>    clock and initiate timing messages encapsulated in Timing LSP tow=
ard
>>>>>>>    the MPLS network, while the LSRs act as Transparent Clock (TC) an=
d
>>>>>>>    just update the Timing field in the Timing messages, which are
>>>>>>>    already encapsulated in Timing LSPs.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013               [Pag=
e 9]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>      |Switch, |     |       |     |       |     |       |     |Switc=
h, |
>>>>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
>>>>>>>      | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/B=
C  |
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>
>>>>>>>      Figure (2) - Deployment example 2 of timing over MPLS network
>>>>>>>
>>>>>>>
>>>>>>>    Another example is shown in Figure 3, where LERs do not terminate=
 the
>>>>>>>    Timing messages received from switch/routers that are outside of=
 the
>>>>>>>    MPLS network acting as OC, TC or BC.  The LERs act as TC and upda=
te
>>>>>>>    the Timing field in the Timing messages as they transit the LER,
>>>>>>>    while encapsulating them in timing LSP.  The LSRs also act as
>>>>>>>    Transparent Clock (TC) and just update the Timing field in the Ti=
ming
>>>>>>>    messages which are already encapsulated in Timing LSPs.
>>>>>>>
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>       |Switch, |     |       |     |       |     |       |     |Swit=
ch, |
>>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rou=
ter |
>>>>>>>       |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/T=
C/BC|
>>>>>>>       +--------+     +-------+     +-------+     +-------+     +----=
----+
>>>>>>>
>>>>>>>     Figure (3) - Deployment example 3 of timing over MPLS network
>>>>>>>
>>>>>>>    Another example is shown in Figure 4, where LERs and LSRs support
>>>>>>>    Boundary Clocks.  A single-hop LSP is created between two adjacen=
t
>>>>>>>    LSRs engaged in BC operation.  Other methods such as PTP transpor=
t
>>>>>>>    over Ethernet MAY be used for transporting timing messages if the
>>>>>>>    link between the two routers is Ethernet.
>>>>>>>
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>      |Switch, |     |       |     |       |     |       |     |Switc=
h, |
>>>>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Rout=
er |
>>>>>>>      | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/B=
C  |
>>>>>>>      +--------+     +-------+     +-------+     +-------+     +-----=
---+
>>>>>>>
>>>>>>>    Figure (4) - Deployment example 3 of timing over MPLS network
>>>>>>>
>>>>>>>    An MPLS domain MAY serve multiple customers.  In these cases the
>>>> MPLS
>>>>>>>    domain (maintained by a service provider) may provide timing serv=
ices
>>>>>>>    to multiple customers, each having their own Timing domain.
>>>>>>>
>>>>>>>    The Timing over MPLS architecture assumes full mesh of Timing LSP=
s
>>>>>>>    between all LERs supporting this specification.
>>>>>>>
>>>>>>> SB> Note sure this is right - the salves surely do not need to
>>>>>>> SB> exchange timing amongst themselves
>>>>>>>
>>>>>>>    It supports
>>>>>>>    Point-to- point (VPWS) and Multipoint (VPLS) services.
>>>>>>>
>>>>>>> SB> What does that mean? You do not carry user data traffic?
>>>>>>> SB> Maybe it's the ordering of the statemnets that is causing
>>>>>>> SB> confusion.
>>>>>>>
>>>>>>>    This means
>>>>>>>    that a customer may purchase a Point-to-point Timing service betw=
een
>>>>>>>    two customer sites or a Multipoint Timing service between more th=
an
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 10]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>>    two customer sites.
>>>>>>>
>>>>>>>    The Timing over MPLS architecture supports P2P or P2MP Timing LSP=
s.
>>>>>>>    This means that the Timing Multicast messages such as PTP Multica=
st
>>>>>>>    event messages can be transported over P2MP Timing LSP or be
>>>>>>>    replicated and transported over many P2P Timing LSPs.
>>>>>>>
>>>>>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>>>>>>> SB> nor a P2MP PW, although we are close.
>>>>>>>
>>>>>>>    Timing messages, that do not require Time stamping or Correction
>>>>>>>    Field update MAY be transported over Timing LSPs to simplify hard=
ware
>>>>>>>    and software.
>>>>>>>
>>>>>>>    PTP Announce messages that determine the Timing LSP terminating
>>>> point
>>>>>>>    behavior such as BC/OC/TC SHOULD be transported over the Timing L=
SP
>>>>>>>    to simplify hardware and software.
>>>>>>>
>>>>>>> SB> have you defined and referenced PTP announce msgs?
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 11]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 5.  Dedicated LSPs for Timing messages
>>>>>>>
>>>>>>>    Many methods have been considered for identifying the Timing
>>>> messages
>>>>>>>    when they are encapsulated in MPLS such as using GAL/G-ACH or a n=
ew
>>>>>>>    reserved label.  These methods were not attractive since they eit=
her
>>>>>>>    required deep packet inspection at line rate in the intermediate=
 LSRs
>>>>>>>    or they required use of a scarce new reserved label.  Also one of=
 the
>>>>>>>    goals was to reuse existing OAM mechanisms.
>>>>>>>
>>>>>>> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
>>>>>>>
>>>>>>>    The method defined in this document can be used by LER and LSRs t=
o
>>>>>>>    identify Timing messages in MPLS tunnels by just looking at the t=
op
>>>>>>>    label in the MPLS label stack, which only carry Timing messages a=
s
>>>>>>>    well as OAM, but not data plane client traffic.
>>>>>>>
>>>>>>>    Compliant implementations MUST use dedicated LSPs to carry Timing
>>>>>>>    messages over MPLS.
>>>>>>>
>>>>>>> SB> I think that we need a definition of the properies of these LSPs
>>>>>>>
>>>>>>>    These LSPs are herein referred to as "Timing
>>>>>>>    LSPs" and the labels associated with these LSPs as "Timing LSP
>>>>>>>    labels".  The Timing LSPs that runs between Ingress and Egress LE=
Rs
>>>>>>>    MUST be co-routed.  Alternatively, a single bidirectional co-rout=
ed
>>>>>>>    LSP can be used.
>>>>>>>
>>>>>>> SB> I though that you said you could use M2MP LSPs - these are not
>>>>>>> SB> bidirectional.
>>>>>>>
>>>>>>>    Co-routing of the two directions is required to limit the differe=
nce
>>>>>>>    in the delays in the Master clock to Slave clock direction compar=
ed
>>>>>>>    to the Slave clock to Master clock direction.  The Timing LSP MAY=
 be
>>>>>>>    MPLS/MPLS-TP LSP.
>>>>>>>
>>>>>>>    The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS=
.
>>>>>>>    New Extensions to RSVP-TE/GMPLS TLVs are required; however they a=
re
>>>>>>>    outside the scope of this document.
>>>>>>>
>>>>>>>    The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such=
 as
>>>>>>>    BFD and LSP Ping but the LSP data plane client plane traffic MUST=
 be
>>>>>>>    Timing packets only.
>>>>>>>
>>>>>>> SB> Why?
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 12]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 6.  Timing over LSP Encapsulation
>>>>>>>
>>>>>>> The encapsulations is not LSP is it?
>>>>>>>
>>>>>>>    This document defines two methods for carrying Timing messages ov=
er
>>>>>>>    MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>>>>>>    messages over Timing LSPs, and the second method, is carrying
>>>>>>>    Ethernet encapsulated Timing messages over Ethernet PWs inside
>>>> Timing
>>>>>>>    LSPs.
>>>>>>>
>>>>>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>>>>>>
>>>>>>>    The simplest method of transporting Timing messages over MPLS is=
 to
>>>>>>>    encapsulate Timing PDUs in UDP/IP and then encapsulate them in
>>>> Timing
>>>>>>>    LSP.  This format is shown in Figure 4.
>>>>>>>
>>>>>>>
>>>>>>>                     +----------------------+
>>>>>>>                     |   Timing LSP Label   |
>>>>>>>                     +----------------------+
>>>>>>>                     |        IPv4/6        |
>>>>>>>                     +----------------------+
>>>>>>>                     |         UDP          |
>>>>>>>                     +----------------------+
>>>>>>>                     |     Timing PDU       |
>>>>>>>                     +----------------------+
>>>>>>>
>>>>>>>       Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>>>>>>
>>>>>>>
>>>>>>>    This encapsulation is very simple and is useful when the network
>>>>>>>    between Timing Master Clock and Slave Clock is MPLS network.
>>>>>>>
>>>>>>> SB> Simple is a judgement call
>>>>>>>
>>>>>>>    In order for an LER/LSR to process Timing messages, the Timing LS=
P
>>>>>>>    Label must be at the top label of the label stack.  The LER/LSR M=
UST
>>>>>>>    know that the Timing LSP Label is used for carrying Timing messag=
es.
>>>>>>>    This can be accomplished via static configuration or via RSVP-TE
>>>>>>>    signaling.
>>>>>>>
>>>>>>>    The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>>>>>>    [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>>>>>>>    [RFC5905].
>>>>>>>
>>>>>>> 6.2.  Timing over PW Encapsulation
>>>>>>>
>>>>>>>    Another method of transporting Timing over MPLS networks is by
>>>>>>>    encapsulating Timing PDUs in PW which in turn is transported over
>>>>>>>    Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448]=
,
>>>>>>>    shown in Fig 5(A) MUST be used and the Ethernet encapsulation of=
 PTP
>>>>>>>    MUST follow Annex F of [IEEE-1588].
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 13]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>>    The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
>>>> the
>>>>>>>    Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>>>>>>>    Timing over PW encapsulation MUST use the Control Word (CW) as
>>>>>>>    specified in [RFC4448] to ensure proper detection of PTP messages
>>>>>>>    inside the MPLS packets for Timing over LSP and Timing over PW
>>>>>>>    encapsulation.
>>>>>>>
>>>>>>> SB> That needs explanation
>>>>>>>
>>>>>>>    The use of Sequence Number in the CW is optional.
>>>>>>>
>>>>>>> SB> Given that s/n are never in practice deployed, you could probabl=
y
>>>>>>> SB> simplify things by sayig that they are not used.
>>>>>>>
>>>>>>>    Timing over PW encapsulation for NTP MUST use NTP over UDP/IP ove=
r
>>>> PW
>>>>>>>    (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>>>>>>
>>>>>>>                     +----------------+  +----------------+
>>>>>>>                      |Timing LSP Label|  |Timing LSP Label|
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |    PW Label    |  |    PW Label    |
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |  Control Word  |  |      IP        |
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |    Ethernet    |  |      UDP       |
>>>>>>>                      |     Header     |  +----------------+
>>>>>>>                      +----------------+  |   Timing PDU   |
>>>>>>>                      |S-VLAN(Optional)|  |                |
>>>>>>>                      +----------------+  +----------------+
>>>>>>>                      |C-VLAN(Optional)|        (B)
>>>>>>>                      +----------------+
>>>>>>>                      |   Timing PDU   |
>>>>>>>                      |                |
>>>>>>>                      +----------------+
>>>>>>>                             (A)
>>>>>>>
>>>>>>>               Figure (5) - Timing over PW Encapsulations
>>>>>>>
>>>>>>>    In order for an LSR to process PTP messages, the top label of the
>>>>>>>    label stack (the Tunnel Label) MUST be a Timing label.
>>>>>>>
>>>>>>> S> You said that before.
>>>>>>>
>>>>>>> 6.3.  Other Timing Encapsulation methods
>>>>>>>
>>>>>>>    In future other timing encapsulation methods may be introduced, s=
uch
>>>>>>>    as a new shim header after the Bottom of Stack to carry the Timin=
g
>>>>>>>    information.  Such new encapsulations are outside the scope of th=
is
>>>>>>>    document.
>>>>>>>
>>>>>>>
>>>>>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>>>>>>> SB> out of the definition of the LSP
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 14]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> SB> I think we need a section on LSP processing
>>>>>>>
>>>>>>> 7.  Timing message Processing
>>>>>>>
>>>>>>>    Each Timing protocol such as PTP and NTP, define their set of Tim=
ing
>>>>>>>    messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>>>>>>>    FOLLOW_UP, etc messages.
>>>>>>>
>>>>>>>    Some of the Timing messages require time stamping or correction f=
ield
>>>>>>>    update at port level and some dont.  It is the job of the LER/LSR=
 to
>>>>>>>    parse the timing message and find out the type of the Timing mess=
age
>>>>>>>    and decide whether and how to Time- stamp it (e.g., BC) or update
>>>>>>>    correction field(e.g., TC).
>>>>>>>
>>>>>>>
>>>>>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
>>>>>>> SB> function rather than the LER?
>>>>>>>
>>>>>>>    For example the following PTP messages (called Event messages)
>>>>>>>    require time-stamping or correction field update:
>>>>>>>
>>>>>>>    o  SYNC
>>>>>>>
>>>>>>>    o  DELAY_REQ (Delay Request)
>>>>>>>
>>>>>>>    o  PDELAY_REQ (Peer Delay Request)
>>>>>>>
>>>>>>>    o  PDELAY_RESP (Peer Delay Response)
>>>>>>>
>>>>>>>    SYNC and DELAY_REQ are exchanged between Master Clock and Slave
>>>> Clock
>>>>>>>    and MUST be transported over PTP LSPs.  PDELAY_REQ and
>>>> PDELAY_RESP
>>>>>>>    are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>>>>>>>    Boundary, or Transparent) and SHOULD be transported over single h=
op
>>>>>>>    PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP=
,
>>>>>>>    and PDELAY_RESP_FOLLOW_UP messages MUST also be transported
>>>> over
>>>>>>> the
>>>>>>>    PTP LSPs.
>>>>>>>
>>>>>>>    For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
>>>>>>>    transported over two PTP LSPs that are in opposite directions.  T=
hese
>>>>>>>    PTP LSPs, which are in opposite directions MUST be congruent and=
 co-
>>>>>>>    routed.  Alternatively, a single bidirectional co-routed LSP can=
 be
>>>>>>>    used.
>>>>>>>
>>>>>>>    Except as indicated above for the two-step PTP clocks, Non-Event=
 PTP
>>>>>>>    message types do not need to be processed by intermediate routers=
.
>>>>>>>    These message types MAY be carried in PTP Tunnel LSPs.
>>>>>>>
>>>>>>> SB> Are you saying that a timing P router has to be msg type sensiti=
ve?
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 15]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 8.  Protection and Redundancy
>>>>>>>
>>>>>>>
>>>>>>> SB> This is a bit of a jump - I don't know how the LSP itself works=
 yet!
>>>>>>>
>>>>>>>    In order to ensure continuous uninterrupted operation of slave
>>>>>>>    clocks, usually as a general practice, slave clocks (or ports) tr=
ack
>>>>>>>    redundant master clocks.
>>>>>>>
>>>>>>>    It is the responsibility of the network operator to ensure that
>>>>>>>    physically disjoint Timing LSPs are established between a slave c=
lock
>>>>>>>    (or port) and redundant master clocks (or ports).
>>>>>>>
>>>>>>>    When a slave clock (or port) listens to redundant master clocks o=
r
>>>>>>>    ports, any prolonged Timing LSP outage will trigger the slave clo=
ck
>>>>>>>    or port to switch to a redundant master clock or port.
>>>>>>>
>>>>>>>    LSP/PW protection such as Linear protection Switching (1:1, 1+1),
>>>>>>>    Ring protection switching or MPLS Fast Reroute (FRR) generally sw=
itch
>>>>>>>    alternative path that usually cause a change in delay, which if
>>>>>>>    undetected by slave clock can reduce accuracy of the slave clock.
>>>>>>>
>>>>>>>    Therefore protection switching MAY be used, as long as phase jump=
s
>>>>>>>    upon switchover due to differences in path latency are detected a=
nd
>>>>>>>    compensated for (such compensation not being required if BCs or p=
eer-
>>>>>>>    peer TCs are used throughout).
>>>>>>>
>>>>>>>    Note that any protection or reroute mechanism that adds additiona=
l
>>>>>>>    MPLS label to the label stack, such as Facility Backup Fast Rerou=
te,
>>>>>>>    MUST ensure that the pushed label is also a Timing Label to ensur=
e
>>>>>>>    recognition of the MPLS frame as containing Timing messages, as i=
t
>>>>>>>    transits the backup path.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 16]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 9.  ECMP
>>>>>>>
>>>>>>>    To ensure the optimal operation of slave clocks and avoid error
>>>>>>>    introduced by forward and reverse path delay asymmetry, the physi=
cal
>>>>>>>    path for Timing messages from master clock to slave Clock and vic=
e
>>>>>>>    versa must be the same for all Event Timing messages listed in
>>>>>>>    section 7.
>>>>>>>
>>>>>>>    Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
>>>>>>>    Multipath).
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 17]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 10.  PHP
>>>>>>>
>>>>>>>    To ensure that the label on the top of the label stack is the Tim=
ing
>>>>>>>    LSP Label, PHP MUST not be used.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 18]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 11.  Entropy
>>>>>>>
>>>>>>>    To ensure all Timing messages in a Timing LSP take the same path,
>>>>>>>    Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>>>>>>    Entropy Label MUST NOT be used for the PWs that are carried insid=
e
>>>>>>>    Timing LSP [RFC6391].
>>>>>>>
>>>>>>> SB> This is incorrect - you mean that all msgs of the same timing
>>>>>>> SB> flow need to have the same EL value.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 19]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 12.  OAM, Control and Management
>>>>>>>
>>>>>>>    In order to monitor Timing LSPs and their encapsulated PWs, they=
 MUST
>>>>>>>    be able to carry OAM and management messages.  These management
>>>>>>>    messages MUST be differentiated from Timing messages via already
>>>>>>>    defined IETF methods.
>>>>>>>
>>>>>>>    For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY r=
un
>>>>>>>    over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>>>>>>>    Management protocols can easily be identified by the UDP Destinat=
ion
>>>>>>>    Port number or by GAL/G-ACH respectively.
>>>>>>>
>>>>>>>    Also BFD, LSP-Ping and other management messages MAY run over the
>>>> PWs
>>>>>>>    encapsulated in Timing LSP via one of the defined VCCVs (Type 1,=
 3 or
>>>>>>>    4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is g=
oing
>>>>>>>    to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D1=
) or
>>>>>>>    GAL-ACH are used to identify such management messages.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 20]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 13.  QoS Considerations
>>>>>>>
>>>>>>>    In network deployments where not every LSR/LER is Timing-aware, i=
t is
>>>>>>>    important to reduce the impact of the non-Timing-aware LSR/LERs o=
n
>>>>>>>    the timing recovery in the slave clock.  The Timing messages are=
 time
>>>>>>>    critical and must be treated with the highest priority.  Therefor=
e
>>>>>>>    Timing over MPLS messages must be treated with the highest priori=
ty
>>>>>>>    in the routers.  This can be achieved by proper setup of Timing L=
SPs.
>>>>>>>
>>>>>>>    It is recommended that the Timing LSPs are setup or configured
>>>>>>>    properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC26=
97]
>>>>>>>    for drop eligibility.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 21]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 14.  FCS and Checksum Recalculation
>>>>>>>
>>>>>>>    When time-stamp generation and timing packet adjustment is perfor=
med
>>>>>>>    near the physical port hardware, the process MUST include
>>>>>>>    recalculation of the Ethernet FCS.
>>>>>>>
>>>>>>> SB> The above is confusing - an LSR always recomputes the link layer
>>>>>>> SB> CRC which may or may not be Ethernet.
>>>>>>>
>>>>>>>    Also FCS retention for the
>>>>>>>    payload Ethernet described in [RFC4720] MUST NOT be used.
>>>>>>>
>>>>>>>    For UDP/IP encapsulation mode of Timing over MPLS, the UDP checks=
um
>>>>>>>    may be required as per UDP transport standards.
>>>>>>>
>>>>>>> SB> You really need to be working on getting the IPv6 C?S computatio=
n
>>>>>>> SB> removed from PTP msgs.
>>>>>>>
>>>>>>>    When UDP checksum is used, each Timing-aware LER/LSR must either
>>>>>>>    incrementally update the UDP checksum after Time stamping or
>>>>>>>    Correction Field update or verify the UDP checksum on reception f=
rom
>>>>>>>    upstream and recalculate the checksum completely on transmission=
 to
>>>>>>>    downstream node after Time stamping or Correction Field update.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 22]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 15.  Behavior of LER/LSR
>>>>>>>
>>>>>>>    Timing-capable/aware LERs and LSRs are routers that have one or m=
ore
>>>>>>>
>>>>>>> SB> You mean physical interfaces?
>>>>>>>
>>>>>>>    interfaces that can perform Timing operations (OC/BC/TC) on Timin=
g
>>>>>>>    packets and are configured to do so.  Timing-capable/aware LERs a=
nd
>>>>>>>    LSRs can advertise their Timing-capability per-interface via cont=
rol
>>>>>>>    plane such as OSPF or IS-IS.
>>>>>>> SB> ISIS and OSPF are routing protocols.
>>>>>>>
>>>>>>>   The Timing-capable/aware LERs can then
>>>>>>>    signals Timing LSPs via RSVP-TE signaling.  Alternatively the Tim=
ing
>>>>>>>    capability of LER and LSRs may be configured in a centralized
>>>>>>>    controller and the Timing LSP may be setup using manual configura=
tion
>>>>>>>    or other methods such as SDN.
>>>>>>>
>>>>>>> SB> it can also be configured individually rather then through
>>>>>>> SB> a cebtral controllwe
>>>>>>>
>>>>>>> 15.1.  Behavior of Timing-capable/aware LER
>>>>>>>
>>>>>>>    When a Timing-capable/aware LER behaves as a Transparent clock an=
d
>>>>>>>    receives a Timing message from a Timing-capable/aware non-MPLS
>>>>>>>    interface, the LER updates the Correction Field (CF) and encapsul=
ates
>>>>>>>    and forwards the timing message over previously established Timin=
g
>>>>>>>    LSP.
>>>>>>>
>>>>>>> SB> You need to call out the details so that people properly
>>>>>>> SB> understand the definition of the new LSP.
>>>>>>>
>>>>>>>    Also when a Timing message is received from a Timing-capable/
>>>>>>>    aware MPLS interface, LER updates the Correction Filed (CF) and
>>>>>>>    decapsulates the MPLS encapsulation and forwards the timing messa=
ge
>>>>>>>    to a non-MPLS interface.
>>>>>>>
>>>>>>>    When a Timing-capable/aware LER behaves as a Boundary clock and
>>>>>>>    receives a Timing message from a Timing-capable/aware non MPLS
>>>>>>>    interface, the LER Timestamps the Timing packet and sends it to t=
he
>>>>>>>    LERs Boundary clock processing module.  Also when a Timing messag=
e is
>>>>>>>    received from a Timing- capable/aware MPLS interface, the LER
>>>>>>>    Timestamps the Timing packet and sends it to the LERs Boundary cl=
ock
>>>>>>>    processing module.
>>>>>>>
>>>>>>>    When a Timing-capable/aware LER behaves as an Ordinary Clock towa=
rd
>>>>>>>    the MPLS network, and receives a Timing message from a Timing-
>>>>>>>    capable/aware MPLS interface, the LER Timestamps the Timing packe=
t
>>>>>>>    and sends it to the LERs Ordinary clock processing module.
>>>>>>>
>>>>>>> 15.2.  Behavior of Timing-capable/aware LSR
>>>>>>>
>>>>>>>    When a Timing-capable/aware LSR behaves as a Transparent clock an=
d
>>>>>>>    receives a Timing message from a Timing-capable/aware MPLS
>>>> interface,
>>>>>>>    The LSR updates the Correction Filed (CF) and forwards the timing
>>>>>>>    message over another MPLS interface.
>>>>>>>
>>>>>>>    When a Timing-capable/aware LSR behaves as a Boundary clock and
>>>>>>>    receives a Timing message from a Timing-capable/aware MPLS
>>>> interface.
>>>>>>>    The LSR performs the functions of a Boundary Clock in terminating=
 the
>>>>>>>    received Timing message and re-generating a new timing message ov=
er
>>>>>>>    another (or the same) MPLS interface.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 23]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>>>>>>
>>>>>>>    It is most beneficial when all LSRs in the path of a Timing LSP b=
e
>>>>>>>    timing-Capable/aware LSRs.  This would ensure the highest quality
>>>>>>>    time and clock synchronization by Timing Slave Clocks.  However,=
 this
>>>>>>>    specification does not mandate that all LSRs in path of a Timing=
 LSP
>>>>>>>    be Timing- capable/aware.
>>>>>>>
>>>>>>>    Non-Timing-capable/aware LSRs just switch the packets encapsulate=
d in
>>>>>>>    Timing LSPs and dont perform any Timing operation (TC or BC).
>>>>>>>    However as explained in QoS section the Timing over MPLS packets
>>>> MUST
>>>>>>>    be still be treated with the highest priority based on their Traf=
fic
>>>>>>>    Class (TC) marking.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 24]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 16.  Other considerations
>>>>>>>
>>>>>>>    [IEEE-1588] defines an optional peer-to-peer Transparent clocking
>>>>>>>    that requires peer delay measurement between two adjacent Timing-
>>>>>>>    capable/ aware routers/switches.  Peer delay measurement messages
>>>>>>>    need to be time stamped and terminated by the Timing-capable/awar=
e
>>>>>>>    routers/ switches.  This means that two adjacent LSRs may be enga=
ged
>>>>>>>    in a peer delay measurement.
>>>>>>>
>>>>>>>    For transporting such peer delay measurement messages a single-ho=
p
>>>>>>>    LSP SHOULD to be created between the two adjacent LSRs engaged in
>>>>>>>    peer delay measurement to carry peer delay measurement messages.
>>>>>>>    Other methods such as PTP transport over Ethernet MAY be used for
>>>>>>>    transporting peer delay measurement messages if the link between=
 the
>>>>>>>    two routers is Ethernet.
>>>>>>>
>>>>>>>    In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/=
 ware
>>>>>>>    routers/switches MUST maintain a list of all the neighbors it nee=
ds
>>>>>>>    to send a PDelay_Req to, where each neighbor corresponds to a tim=
ing
>>>>>>>    LSP.
>>>>>>>
>>>>>>>    The use of Explicit Null Label (Label=3D 0 or 2) is acceptable as=
 long
>>>>>>>    as either the Explicit Null label is the bottom of stack label
>>>>>>>    (applicable only to UDP/IP encapsulation) or the label below the
>>>>>>>    Explicit Null label is a PTP label.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 25]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 17.  Security Considerations
>>>>>>>
>>>>>>>    MPLS PW security considerations in general are discussed in [RFC3=
985]
>>>>>>>    and [RFC4447],and those considerations also apply to this documen=
t.
>>>>>>>
>>>>>>>    An experimental security protocol is defined in [IEEE-1588].The P=
TP
>>>>>>>    security extension and protocol provides group source authenticat=
ion,
>>>>>>>    message integrity, and replay attack protection for PTP messages.
>>>>>>>
>>>>>>>    When the MPLS network (provider network) serves multiple customer=
s,
>>>>>>>    it is important to maintain and process each customers clock and
>>>>>>>    Timing messages separately from other customers to ensure there i=
s no
>>>>>>>    cross- customer effect.  For example if an LER BC is synchronized=
 to
>>>>>>>    a specific grandmaster, belonging to customer A, then the LER MUS=
T
>>>>>>>    use that BC clock only for customer A to ensure that customer A
>>>>>>>    cannot attack other customers by manipulating its time.
>>>>>>>
>>>>>>>    Timing messages MAY be encrypted or authenticated, provided that=
 the
>>>>>>>    LERs/LSRs that are Timing capable/aware can authenticate/ decrypt=
 the
>>>>>>>    timing messages.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 26]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 18.  Acknowledgements
>>>>>>>
>>>>>>>    The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizr=
ahi,
>>>>>>>    Stefano Ruffini, Peter Meyer, and other members of IETF for revie=
wing
>>>>>>>    and providing feedback on this draft.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 27]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 19.  IANA Considerations
>>>>>>>
>>>>>>>    There are no IANA requirements in this specification.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 28]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 20.  References
>>>>>>>
>>>>>>> 20.1.  Normative References
>>>>>>>
>>>>>>>    [IEEE-1588]
>>>>>>>               IEEE 1588-2008, "IEEE Standard for a Precision Clock
>>>>>>>               Synchronization Protocol for Networked Measurement and
>>>>>>>               Control Systems".
>>>>>>>
>>>>>>>    [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>>>>>>               Requirement Levels", BCP 14, RFC 2119, March 1997.
>>>>>>>
>>>>>>>    [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to=
-
>>>>>>>               Edge (PWE3) Architecture", RFC 3985, March 2005.
>>>>>>>
>>>>>>>    [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discov=
ery
>>>>>>>               Proxies (ND Proxy)", RFC 4389, April 2006.
>>>>>>>
>>>>>>>    [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G=
.
>>>>>>>               Heron, "Pseudowire Setup and Maintenance Using the Lab=
el
>>>>>>>               Distribution Protocol (LDP)", RFC 4447, April 2006.
>>>>>>>
>>>>>>>    [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>>>>>>               "Encapsulation Methods for Transport of Ethernet over=
 MPLS
>>>>>>>               Networks", RFC 4448, April 2006.
>>>>>>>
>>>>>>>    [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>>>>>>               Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>>>>>>               Retention", RFC 4720, November 2006.
>>>>>>>
>>>>>>>    [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circu=
it
>>>>>>>               Connectivity Verification (VCCV): A Control Channel fo=
r
>>>>>>>               Pseudowires", RFC 5085, December 2007.
>>>>>>>
>>>>>>>    [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detect=
ion
>>>>>>>               (BFD)", RFC 5880, June 2010.
>>>>>>>
>>>>>>>    [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow=
,
>>>>>>>               "Bidirectional Forwarding Detection (BFD) for MPLS Lab=
el
>>>>>>>               Switched Paths (LSPs)", RFC 5884, June 2010.
>>>>>>>
>>>>>>> 20.2.  Informative References
>>>>>>>
>>>>>>>    [I-D.ietf-pwe3-fat-pw]
>>>>>>>               Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Reg=
an,
>>>>>>>               J., and S. Amante, "Flow Aware Transport of Pseudowire=
s
>>>>>>>               over an MPLS Packet Switched Network",
>>>>>>>               draft-ietf-pwe3-fat-pw-07 (work in progress), July 201=
1.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 29]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>>    [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermedia=
te
>>>>>>>               system routeing information exchange protocol for use=
 in
>>>>>>>               conjunction with the Protocol for providing the
>>>>>>>               Connectionless-mode Network Service (ISO 8473)".
>>>>>>>
>>>>>>>    [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP an=
d
>>>>>>>               dual environments", RFC 1195, December 1990.
>>>>>>>
>>>>>>>    [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 199=
8.
>>>>>>>
>>>>>>>    [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
>>>>>>>               Marker", RFC 2697, September 1999.
>>>>>>>
>>>>>>>    [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boud=
ec,
>>>>>>>               J., Courtney, W., Davari, S., Firoiu, V., and D.
>>>>>>>               Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>>>>>>               Behavior)", RFC 3246, March 2002.
>>>>>>>
>>>>>>>    [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Enginee=
ring
>>>>>>>               (TE) Extensions to OSPF Version 2", RFC 3630,
>>>>>>>               September 2003.
>>>>>>>
>>>>>>>    [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermedia=
te
>>>>>>>               System (IS-IS) Extensions for Traffic Engineering (TE)=
",
>>>>>>>               RFC 3784, June 2004.
>>>>>>>
>>>>>>>    [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and=
 S.
>>>>>>>               Shaffer, "Extensions to OSPF for Advertising Optional
>>>>>>>               Router Capabilities", RFC 4970, July 2007.
>>>>>>>
>>>>>>>    [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
>>>>>>>               System to Intermediate System (IS-IS) Extensions for
>>>>>>>               Advertising Router Information", RFC 4971, July 2007.
>>>>>>>
>>>>>>>    [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
>>>>>>>               Topology (MT) Routing in Intermediate System to
>>>>>>>               Intermediate Systems (IS-ISs)", RFC 5120, February 200=
8.
>>>>>>>
>>>>>>>    [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>>>>>>               Engineering", RFC 5305, October 2008.
>>>>>>>
>>>>>>>    [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>>>>>>               "Traffic Engineering Extensions to OSPF Version 3",
>>>>>>>               RFC 5329, September 2008.
>>>>>>>
>>>>>>>    [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSP=
F
>>>>>>>               for IPv6", RFC 5340, July 2008.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 30]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>>    [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Net=
work
>>>>>>>               Time Protocol Version 4: Protocol and Algorithms
>>>>>>>               Specification", RFC 5905, June 2010.
>>>>>>>
>>>>>>>    [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Reg=
an,
>>>>>>>               J., and S. Amante, "Flow-Aware Transport of Pseudowire=
s
>>>>>>>               over an MPLS Packet Switched Network", RFC 6391,
>>>>>>>               November 2011.
>>>>>>>
>>>>>>>    [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., a=
nd
>>>>>>>               L. Yong, "The Use of Entropy Labels in MPLS Forwarding=
",
>>>>>>>               RFC 6790, November 2012.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 31]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 1.  Routing extensions for Timing-aware Routers
>>>>>>>
>>>>>>>    MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340]=
 and
>>>>>>>    IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (=
TE)
>>>>>>>    link information used for constraint-based routing.
>>>>>>>
>>>>>>>    Indeed, it is useful to advertise data plane TE router link
>>>>>>>    capabilities, such as the capability for a router to be Timing-aw=
are.
>>>>>>>    This capability MUST then be taken into account during path
>>>>>>>    computation to prefer or even require links that advertise themse=
lves
>>>>>>>    as Timing-aware.  In this way the path can ensure the entry and e=
xit
>>>>>>>    points into the LERs and, if desired, the links into the LSRs are
>>>>>>>    able to perform port based time-stamping thus minimizing their im=
pact
>>>>>>>    on the performance of the slave clock.
>>>>>>>
>>>>>>>    extensions are required to OSPF and IS-IS in order to advertise
>>>>>>>    Timing-aware capabilities of a link.  Such extensions are outside=
 the
>>>>>>>    scope of this document; however such extension SHOULD be able to
>>>>>>>    signal the following information per Router Link:
>>>>>>>
>>>>>>>    o  Capable of processing PTP, NTP or other Timing flows
>>>>>>>
>>>>>>>    o  Capable of performing Transparent Clock operation
>>>>>>>
>>>>>>>    o  Capable of performing Boundary Clock operation
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 32]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> 2.  Signaling Extensions for Creating Timing LSPs
>>>>>>>
>>>>>>>    RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSV=
P-TE
>>>>>>>    is used to setup Timing LSPs, some information that indicates tha=
t
>>>>>>>    the LSP is carrying Timing flows MUST be included in the new
>>>>>>>    Extensions to RSVP-TE:
>>>>>>>
>>>>>>>    The following information MAY also be included in the new Extensi=
ons
>>>>>>>    to RSVP-TE:
>>>>>>>
>>>>>>>    o  Offset from Bottom of Stack (BoS) to the start of the Time-sta=
mp
>>>>>>>       field
>>>>>>>
>>>>>>>    o  Number of VLANs in case of PW encapsulation
>>>>>>>
>>>>>>>    o  Timestamp field Type
>>>>>>>
>>>>>>>       *  Correction Field, Timestamp
>>>>>>>
>>>>>>>    o  Timestamp Field format
>>>>>>>
>>>>>>>       *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
>>>>>>>          NTP, etc.
>>>>>>>
>>>>>>>    Note that in case the above optional information is signaled with
>>>>>>>    RSVP-TE for a Timing LSP, all the Timing packets carried in that=
 LSP
>>>>>>>    must have the same signaled characteristics.  For example if
>>>>>>>    Timestamp format is signaled as 64-bit PTPv1, then all Timing pac=
kets
>>>>>>>    must use 64-bit PTPv1 time-stamp.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 33]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>> Authors' Addresses
>>>>>>>
>>>>>>>    Shahram Davari
>>>>>>>    Broadcom Corp.
>>>>>>>    San Jose, CA  95134
>>>>>>>    USA
>>>>>>>
>>>>>>>    Email: davari@broadcom.com
>>>>>>>
>>>>>>>
>>>>>>>    Amit Oren
>>>>>>>    Broadcom Corp.
>>>>>>>    San Jose, CA  95134
>>>>>>>    USA
>>>>>>>
>>>>>>>    Email: amito@broadcom.com
>>>>>>>
>>>>>>>
>>>>>>>    Manav Bhatia
>>>>>>>    Alcatel-Lucent
>>>>>>>    Bangalore,
>>>>>>>    India
>>>>>>>
>>>>>>>    Email: manav.bhatia@alcatel-lucent.com
>>>>>>>
>>>>>>>
>>>>>>>    Peter Roberts
>>>>>>>    Alcatel-Lucent
>>>>>>>    Kanata,
>>>>>>>    Canada
>>>>>>>
>>>>>>>    Email: peter.roberts@alcatel-lucent.com
>>>>>>>
>>>>>>>
>>>>>>>    Laurent Montini
>>>>>>>    Cisco Systems
>>>>>>>    San Jose CA
>>>>>>>    USA
>>>>>>>
>>>>>>>    Email: lmontini@cisco.com
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 34]
>>>>>>> Internet-Draft        Transporting Timing over MPLS            June=
 2013
>>>>>>>
>>>>>>>
>>>>>>>    Luca
>>>>>>>    Cisco Systems
>>>>>>>    San Jose CA
>>>>>>>    USA
>>>>>>>
>>>>>>>    Email: lmartini@cisco.com
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Davari, et al.          Expires December 17, 2013              [Page=
 35]
>>>>>>> --
>>>>>>> For corporate legal information go to:
>>>>>>>
>>>>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> TICTOC mailing list
>>>>>>> TICTOC@ietf.org
>>>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>>> This e-mail message is intended for the recipient only and contains
>>>> information which is CONFIDENTIAL and which may be proprietary to ECI
>>>> Telecom. If you have received this transmission in error, please inform=
 us by e-
>>>> mail, phone or fax, and then delete the original and all copies thereof=
.
>>>>>> _______________________________________________
>>>>>> mpls mailing list
>>>>>> mpls@ietf.org
>>>>>> https://www.ietf.org/mailman/listinfo/mpls
>>>>> .
>>>>
>>>> --
>>>> For corporate legal information go to:
>>>>
>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>
>>>> _______________________________________________
>>>> TICTOC mailing list
>>>> TICTOC@ietf.org
>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>
>>>>
>>>> _______________________________________________
>>>> TICTOC mailing list
>>>> TICTOC@ietf.org
>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>
>>> This e-mail message is intended for the recipient only and contains info=
rmation which is CONFIDENTIAL and which may be proprietary to ECI Telecom. I=
f you have received this transmission in error, please inform us by e-mail,=
 phone or fax, and then delete the original and all copies thereof.
>> .
>
>
> --
> For corporate legal information go to:
>
> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>
> This e-mail message is intended for the recipient only and contains inform=
ation which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If=
 you have received this transmission in error, please inform us by e-mail, p=
hone or fax, and then delete the original and all copies thereof.
>

This e-mail message is intended for the recipient only and contains informat=
ion which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If yo=
u have received this transmission in error, please inform us by e-mail, phon=
e or fax, and then delete the original and all copies thereof.


From stbryant@cisco.com  Wed Aug 21 03:45:20 2013
Return-Path: <stbryant@cisco.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id A18D311E8390; Wed, 21 Aug 2013 03:45:20 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -110.599
X-Spam-Level: 
X-Spam-Status: No, score=-110.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id v7dxRr0p0tUt; Wed, 21 Aug 2013 03:45:15 -0700 (PDT)
Received: from ams-iport-4.cisco.com (ams-iport-4.cisco.com [144.254.224.147]) by ietfa.amsl.com (Postfix) with ESMTP id CB29111E838F; Wed, 21 Aug 2013 03:45:11 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=89361; q=dns/txt; s=iport; t=1377081912; x=1378291512; h=message-id:date:from:reply-to:mime-version:to:cc:subject: references:in-reply-to:content-transfer-encoding; bh=8ebJAqFUKBZYdHWJd3fdLX0OhMQBchgI7tOUQqu5s2E=; b=Qu9ZN26FbX4A8oLT3nV/abxN1UPqHj7V7wKcj9I2kiUxow8Kezha2+kf 8b15agE0XqWRfFiOVmjRZx452DrBTV0jHaJETj3OG2CGL7bNe19MFrml9 K4J0W7X4rFucLvYEu32xx2p3wAM650tkx562HajKa77riI08VZaLQzZRx w=;
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AgYFAB6ZFFKQ/khN/2dsb2JhbABQCoMFNa1ukjyBJRZ0giQBAQEEAQEBFwEdLgEHCgEMBAsRAQMBAQEJDAoIBwkDAgECARUeAQMGCAYKAwEFAgEBF4djAw8MrXqNXYEkAgkGBQWBFyIHBgSECgOMRIcIQYFzgWWBLYVxhRCFKIMdgWgIFw
X-IronPort-AV: E=Sophos;i="4.89,927,1367971200"; d="scan'208";a="17355467"
Received: from ams-core-4.cisco.com ([144.254.72.77]) by ams-iport-4.cisco.com with ESMTP; 21 Aug 2013 10:45:07 +0000
Received: from cisco.com (mrwint.cisco.com [64.103.70.36]) by ams-core-4.cisco.com (8.14.5/8.14.5) with ESMTP id r7LAj5ja023053 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Aug 2013 10:45:05 GMT
Received: from [IPv6:::1] (localhost [127.0.0.1]) by cisco.com (8.14.4+Sun/8.8.8) with ESMTP id r7LAj1Eg020042; Wed, 21 Aug 2013 11:45:04 +0100 (BST)
Message-ID: <52149A2D.3050109@cisco.com>
Date: Wed, 21 Aug 2013 11:45:01 +0100
From: Stewart Bryant <stbryant@cisco.com>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:17.0) Gecko/20130620 Thunderbird/17.0.7
MIME-Version: 1.0
To: Alexander Vainshtein <Alexander.Vainshtein@ecitele.com>
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com> <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com> <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com> <52139903.8040801@cisco.com> <F9336571731ADE42A5397FC831CEAA021513A083@ILPTWPVEXMB01.ecitele.com>, <B7A656C4-6D0E-4719-9E42-57689CDAEFA9@yahoo.com> <F9336571731ADE42A5397FC831CEAA021513A0C6@ILPTWPVEXMB01.ecitele.com>
In-Reply-To: <F9336571731ADE42A5397FC831CEAA021513A0C6@ILPTWPVEXMB01.ecitele.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>, "S. Davari" <davarish@yahoo.com>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
Reply-To: stbryant@cisco.com
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 21 Aug 2013 10:45:20 -0000

Sasha

Note that if the next hop understands the alert, you can put the alert 
label at the ToS.

If it does not understand, you would want to tunnel through it and hence 
there would be a label on top or the alert label.

Shahram, there seems to be some confusion about stitching vs hierarchy, 
and I have been proposing hierarchy to step over the nodes that are 
unable to understand the alert label.

Stewart



On 20/08/2013 19:51, Alexander Vainshtein wrote:
> Shahram,
> No I am not proposing anything of the kind.
>
> At the MPLS data plane level the only the head-end and the tail-end LERs have to be aware of the BoS Timing Alert labels. Transit LSRs simply will ignore them.
>
> At the on-path assistance level each transit LSR (and the tail-end LER) can do one of the following:
> - It may ignore the Timing Alert label at the BoS and just forward the packet based the labels at the top of the stack.
> - It may identify the packet as a timing one based on the Timing Alert label at the bottom of the label stack, and perform the required operations. The result of these operations will be stored in the "timing shim header" and not in the packet itself. (E.g., an E2E TC clock would note the packet residence time and add it to the value already carried in the shim header).
>
> Non-timing packets would not have the Timing Alert label at the BoS and hence would travel thru he LSP travel along the LSP in the usual way.
>
> The tail-end LER would take the reception timestamp and forward this timestamp, the protocol packet and the shim header to the entity that processes the appropriate timing distribution protocol based on the Timing Alert label.
>
> Hopefully this clarifies my proposal. My gut feeling is that similar ideas have been proposed by many people earlier...
>
> Regards,
>       Sasha
>
>
> ________________________________________
> From: S. Davari [davarish@yahoo.com]
> Sent: Tuesday, August 20, 2013 7:00 PM
> To: Alexander Vainshtein
> Cc: stbryant@cisco.com; Shahram Davari; mpls@ietf.org; tictoc@ietf.org
> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>
> Sasha
>
> In (2) are you proposing to create multiple segment LSPs, where each if those segment LSPs span between two Timing aware LSR? So if I have 10 timing aware LSR in the path I would need 9 LSPs?
>
> Regards,
> Shahram
>
>
> On Aug 20, 2013, at 9:53 AM, Alexander Vainshtein <Alexander.Vainshtein@ecitele.com> wrote:
>
>> Stewart, Shahram and all,
>>
>> A couple of (hopefully, last) comments:
>>
>> 1. Supporting a "Timing Alert" label at the top of the label stack introduces a backward compatibility problem. We differ in our estimate of the importance of this problem, but we all agree that it exists.
>>
>> 2. On the other hand, this problem does not exist if the "Timing Alert" label is inserted at the bottom of the label stack. At the same time, HW that analyzes the label stack hopefully could analyze the bottom label (almost) as easily as it would analyze the top label. Thus we could both provide backward compatibility (incompatible transit LSRs simply would not notice the "Timing Alert" label") and avoid the need for dedicated "Timing-only" LSPs. In particular, all the MPLS OAM tools (including MPLS-TP) would work without any changes.
>>
>> 3. A "timing shim header" (between the bottom of the label stack and the beginning of the timing  packet) has been mentioned by Stewart as an already proposed mechanism. IMHO and FWIW its usage would eliminate the need to understand format of the specific timing distribution protocol in the transit LSRs.  It would further solve the problem of secure transport of timing packets since security mechanisms would be applied to its original encapsulation (over IP or over Ethernet) without involving the transit LSRs.
>>
>> Did I miss something substantial?
>>
>> Regards, and lots of thanks in advance,
>>      Sasha
>>
>>
>> ________________________________________
>> From: Stewart Bryant [stbryant@cisco.com]
>> Sent: Tuesday, August 20, 2013 6:27 PM
>> To: S. Davari
>> Cc: Alexander Vainshtein; Shahram Davari; mpls@ietf.org; tictoc@ietf.org
>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>
>> On 06/08/2013 14:35, S. Davari wrote:
>>> Sasha,
>>>
>>> If a router is not 1588 aware we want it to switch the packet with minimum jitter and delay via high priority queue.
>> Yes, but you can explicitly tunnel across such island of incompatibility
>>
>> Stewart
>>
>>>
>>> Sending it to CPU creates large jitter and delay since the CPUs are generally busy doing other things. Time stamping or updating CF in the CPU won't help since it does not cover the variable delay that is incurred from the time the packet enters the router to the time the packet gets processed by CPU.
>>>
>>> I wish it was that easy!
>>>
>>> Regards,
>>> Shahram
>>>
>>>
>>> On Aug 6, 2013, at 12:38 AM, Alexander Vainshtein <Alexander.Vainshtein@ecitele.com> wrote:
>>>
>>>> Shahram,
>>>> I agree with you that the routers that do not recognize a Timing Alert label would send the packet to the CPU. (This would be easy to arrange since we are speaking about a reserved label.)
>>>>
>>>> I do not, however, see this as a serious issue, because the SW running on this CPU would (hopefully) be upgradable to handle the packet correctly i.e., to forward it to where it should be forwarded and, if it is forwarded as a labeled packet, to prepend the Timing Alert Label on top of its label stack. It could even record the residence time (as observed by the CPU in the proper place in the packet. This would mean that introducing additional error to whatever timing information is associated with this packet - but this is what you should anyway expect if there are non-compliant routers on your path, right?
>>>>
>>>> Regards,
>>>>      Sasha
>>>>
>>>>> -----Original Message-----
>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf Of
>>>>> Shahram Davari
>>>>> Sent: Monday, August 05, 2013 11:29 PM
>>>>> To: stbryant@cisco.com; S. Davari
>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>>>>
>>>>> Hi Stewart,
>>>>>
>>>>> Your suggestion of " I suggested an LSP type that has the properties "timestamp
>>>>> and pass to application", is a subset of the existing draft and should work. It
>>>>> basically limits the time stamping and correction field update to LERs (while the
>>>>> draft supports time stamping at LER and LSR).
>>>>>
>>>>> However Sasha's suggestion of using a reserved label (RAL, GAL or any other
>>>>> reserved label) does not satisfy one of the major requirements, which is
>>>>> backward compatibility. Routers that don't understand this reserved label will
>>>>> drop or copy to CPU such packets.
>>>>>
>>>>> Regards,
>>>>> Shahram
>>>>>
>>>>> -----Original Message-----
>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf Of
>>>>> Stewart Bryant
>>>>> Sent: Monday, August 05, 2013 2:29 AM
>>>>> To: S. Davari
>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>>>>
>>>>> My concern is that we are creating a new LSP type in MPLS
>>>>> (which is a architectural change and thus needs to be very carefully
>>>>> considered) without asking the question "how can I do this
>>>>> in the most general way, to maximize flexibility/reuse"
>>>>>
>>>>> I suggested an LSP type that has the properties "timestamp
>>>>> and pass to application", but I think that Sasha raises a good
>>>>> point about using router alert. However RA has no implicit
>>>>> timestamp, so maybe we need a new type of RA that has the
>>>>> properties "timestamp, and pass top application indicated by
>>>>> the next label". That would be quite useful in a number
>>>>> of OAM applications. There is possibly some GAL variant
>>>>> of that design that should also be considered.
>>>>>
>>>>> The application could worry about the path and where it
>>>>> was necessary to skip some hops a hierarchical LSP would
>>>>> accomplish that, with the specific benefit that the application
>>>>> would consciously do this, and may be able to apply some
>>>>> form of compensation within the network.
>>>>>
>>>>> - Stewart
>>>>>
>>>>> On 04/08/2013 10:40, S. Davari wrote:
>>>>>> Hi Sasha
>>>>>>
>>>>>> Perhaps you have not understood the draft well. The main reason that the
>>>>> draft chose to use a dedicated LSP that is signaled via RSVPTE indicating it is a
>>>>> timing LSP is to be backward compatible with non-1588 nodes. Such nodes
>>>>> simply just switch the packet.
>>>>>> Your proposal had been considered and was rejected since it was not
>>>>> backward compatible. Nodes receiving alert label or TTL = 1 send the packet to
>>>>> CPU. You can refer to the meeting notes.
>>>>>> Regards,
>>>>>> Shahram
>>>>>>
>>>>>>
>>>>>> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein
>>>>> <Alexander.Vainshtein@ecitele.com> wrote:
>>>>>>> Stewart and all,
>>>>>>> I concur with Stewart's statement that the draft does not define "full
>>>>> interaction with MPLS architecture".
>>>>>>> E.g., one of the objectives of the draft is to provide a technique that would
>>>>> be backward-compatible with old LSRs that cannot provide on-path support for
>>>>> timing distribution, while the other objective is to make every LSR on the path
>>>>> aware that some MPLS packets are carrying timing-related messages and hence
>>>>> require on-path support.
>>>>>>> IMHO and FWIW, the MPLS data plane architecture allows just two methods
>>>>> for making a transit LSR to provide special processing to a labeled packet:
>>>>>>> - It would carry some kind of an "alert label" on top of the label stack and,
>>>>> specifically, on top of any labels used for actual forwarding
>>>>>>> OR,
>>>>>>> - The TTL in the top label stack entry has been set to 1.
>>>>>>>
>>>>>>>
>>>>>>> The draft does not follow any of these approaches.
>>>>>>>
>>>>>>> I must also admit that Section 12 "OAM, Control and Management" of the
>>>>> draft looks somewhat in
>>>>>>> E.g., I could not understand from the text whether BFD, LSP-Ping etc. OAM
>>>>> messages would be subjected to on-path support procedures for timing
>>>>> messages or not.
>>>>>>> My 2c,
>>>>>>>      Sasha
>>>>>>>
>>>>>>>> -----Original Message-----
>>>>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Behalf
>>>>> Of
>>>>>>>> Stewart Bryant
>>>>>>>> Sent: Friday, August 02, 2013 12:01 PM
>>>>>>>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-
>>>>>>>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf.org;
>>>>> draft-
>>>>>>>> ietf-tictoc-1588overmpls@tools.ietf.org
>>>>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>>>>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
>>>>>>>>
>>>>>>>> SB> This draft does not seem to provide a precise definition
>>>>>>>> SB> the properties of the new LSP type that it wishes to
>>>>>>>> SB> define, in particular it does it define the PHB of those
>>>>>>>> SB> LSPs, nor the full interaction with the MPLS
>>>>>>>> SB> architecture.
>>>>>>>> SB>
>>>>>>>> SB> I have not tracked TICTOC for a while but I thought that
>>>>>>>> SB> the original plan was to define the concept of an offset
>>>>>>>> SB> into a packet to do the correction.
>>>>>>>> SB>
>>>>>>>> SB> It is disappointing that the opportunity was not taken
>>>>>>>> SB> to define a timing shim inside the timing LSP so that
>>>>>>>> SB> a time correction could be added to any packet such that
>>>>>>>> SB> the MPLS system was isolated from the details of the
>>>>>>>> SB> complexity of the particular time transfer type.
>>>>>>>>
>>>>>>>> SB> I think that much more clarify is needed in terms of
>>>>>>>> SB> definition of the new LSP type, since it is unclear
>>>>>>>> SB> from this text how to implement one.
>>>>>>>> SB>
>>>>>>>> SB> There are a lot of other MPLS services such as
>>>>>>>> SB> LSP ping that need to be considered.
>>>>>>>> SB>
>>>>>>>> SB> Please see inline for more comments. However these
>>>>>>>> SB> comments are made in the context of the text as written
>>>>>>>> SB> whilst I have a fundamental concern that this approach
>>>>>>>> SB> lacks an MPLS architectural soundness that need
>>>>>>>> SB> greater thought with significant impact on the
>>>>>>>> SB> draft.
>>>>>>>>
>>>>>>>> - Stewart
>>>>>>>>
>>>>>>>>
>>>>>>>> TICTOC Working Group                                           S. Davari
>>>>>>>> Internet-Draft                                                   A. Oren
>>>>>>>> Intended status: Standards Track                          Broadcom Corp.
>>>>>>>> Expires: December 17, 2013                                     M. Bhatia
>>>>>>>>                                                                P. Roberts
>>>>>>>>                                                            Alcatel-Lucent
>>>>>>>>                                                                L. Montini
>>>>>>>>                                                                L. Martini
>>>>>>>>                                                             Cisco Systems
>>>>>>>>                                                             June 15, 2013
>>>>>>>>
>>>>>>>>
>>>>>>>>              Transporting Timing messages over MPLS Networks
>>>>>>>>                     draft-ietf-tictoc-1588overmpls-05
>>>>>>>>
>>>>>>>> Abstract
>>>>>>>>
>>>>>>>>     This document defines the method for transporting Timing messages
>>>>>>>>     such as PTP and NTP over an MPLS network.  The method allows for the
>>>>>>>>     easy identification of these PDUs at the port level to allow for port
>>>>>>>>
>>>>>>>> SB> What is a port
>>>>>>>>
>>>>>>>>     level processing of these PDUs in both LERs and LSRs.
>>>>>>>>
>>>>>>>>     The basic idea is to transport Timing messages inside dedicated MPLS
>>>>>>>>     LSPs.  These LSPs only carry Timing messages and possibly Control and
>>>>>>>>     Management packets, but they do not carry customer traffic.
>>>>>>>>
>>>>>>>> SB> More specifically they only carry traffic associated with the
>>>>>>>> SB> timing service and its support.
>>>>>>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely
>>>>>>>> SB> also it gets carried in a structure that causes it to get
>>>>>>>> SB> timestamped.
>>>>>>>>
>>>>>>>>     Two methods for transporting Timing messages over MPLS are defined.
>>>>>>>>
>>>>>>>> SB> Perhaps the right approach is to define the new LSP type and then
>>>>>>>> SB> seperately to define  the mapping of the various timing services
>>>>>>>> SB> over that LSP type.
>>>>>>>>
>>>>>>>>     The first method is to transport Timing messages directly over the
>>>>>>>>     dedicated MPLS LSP via UDP/IP encapsulation, which is suitable for
>>>>>>>>     MPLS networks.  The second method is to transport Timing messages
>>>>>>>>     inside a PW via Ethernet encapsulation.
>>>>>>>>
>>>>>>>> SB> I think that we should note that there are some
>>>>>>>> SB> h/w reasons for this preference. A clean sheet approach
>>>>>>>> SB> would have been to use PTP over MPLS with no intermediate
>>>>>>>> SB> layers.
>>>>>>>>
>>>>>>>> Status of this Memo
>>>>>>>>
>>>>>>>>     This Internet-Draft is submitted in full conformance with the
>>>>>>>>     provisions of BCP 78 and BCP 79.
>>>>>>>>
>>>>>>>>     Internet-Drafts are working documents of the Internet Engineering
>>>>>>>>     Task Force (IETF).  Note that other groups may also distribute
>>>>>>>>     working documents as Internet-Drafts.  The list of current Internet-
>>>>>>>>     Drafts is at http://datatracker.ietf.org/drafts/current/.
>>>>>>>>
>>>>>>>>     Internet-Drafts are draft documents valid for a maximum of six months
>>>>>>>>     and may be updated, replaced, or obsoleted by other documents at any
>>>>>>>>     time.  It is inappropriate to use Internet-Drafts as reference
>>>>>>>>     material or to cite them other than as "work in progress."
>>>>>>>>
>>>>>>>>     This Internet-Draft will expire on December 17, 2013.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013               [Page 1]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> Copyright Notice
>>>>>>>>
>>>>>>>>     Copyright (c) 2013 IETF Trust and the persons identified as the
>>>>>>>>     document authors.  All rights reserved.
>>>>>>>>
>>>>>>>>     This document is subject to BCP 78 and the IETF Trust's Legal
>>>>>>>>     Provisions Relating to IETF Documents
>>>>>>>>     (http://trustee.ietf.org/license-info) in effect on the date of
>>>>>>>>     publication of this document.  Please review these documents
>>>>>>>>     carefully, as they describe your rights and restrictions with respect
>>>>>>>>     to this document.  Code Components extracted from this document must
>>>>>>>>     include Simplified BSD License text as described in Section 4.e of
>>>>>>>>     the Trust Legal Provisions and are provided without warranty as
>>>>>>>>     described in the Simplified BSD License.
>>>>>>>>
>>>>>>>>
>>>>>>>> Table of Contents
>>>>>>>>
>>>>>>>>     1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  5
>>>>>>>>
>>>>>>>>     2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .  7
>>>>>>>>
>>>>>>>>     3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . . .  8
>>>>>>>>
>>>>>>>>     4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . . .  9
>>>>>>>>
>>>>>>>>     5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . . . 12
>>>>>>>>
>>>>>>>>     6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . . . 13
>>>>>>>>       6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . . . 13
>>>>>>>>       6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . . . 13
>>>>>>>>       6.3.  Other Timing Encapsulation methods . . . . . . . . . . . . 14
>>>>>>>>
>>>>>>>>     7.  Timing message Processing  . . . . . . . . . . . . . . . . . . 15
>>>>>>>>
>>>>>>>>     8.  Protection and Redundancy  . . . . . . . . . . . . . . . . . . 16
>>>>>>>>
>>>>>>>>     9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
>>>>>>>>
>>>>>>>>     10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
>>>>>>>>
>>>>>>>>     11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
>>>>>>>>
>>>>>>>>     12. OAM, Control and Management  . . . . . . . . . . . . . . . . . 20
>>>>>>>>
>>>>>>>>     13. QoS Considerations . . . . . . . . . . . . . . . . . . . . . . 21
>>>>>>>>
>>>>>>>>     14. FCS and Checksum Recalculation . . . . . . . . . . . . . . . . 22
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013               [Page 2]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>>     15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . . . 23
>>>>>>>>       15.1. Behavior of Timing-capable/aware LER . . . . . . . . . . . 23
>>>>>>>>       15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . . . 23
>>>>>>>>       15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . . . 24
>>>>>>>>
>>>>>>>>     16. Other considerations . . . . . . . . . . . . . . . . . . . . . 25
>>>>>>>>
>>>>>>>>     17. Security Considerations  . . . . . . . . . . . . . . . . . . . 26
>>>>>>>>
>>>>>>>>     18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 27
>>>>>>>>
>>>>>>>>     19. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 28
>>>>>>>>
>>>>>>>>     20. References . . . . . . . . . . . . . . . . . . . . . . . . . . 29
>>>>>>>>       20.1. Normative References . . . . . . . . . . . . . . . . . . . 29
>>>>>>>>       20.2. Informative References . . . . . . . . . . . . . . . . . . 29
>>>>>>>>
>>>>>>>>     Appendix 1.  Routing extensions for Timing-aware Routers . . . . . 32
>>>>>>>>
>>>>>>>>     Appendix 2.  Signaling Extensions for Creating Timing LSPs . . . . 33
>>>>>>>>
>>>>>>>>     Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 34
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013               [Page 3]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>>     The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
>>>>> NOT",
>>>>>>>>     "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL"
>>>>> in
>>>>>>>> this
>>>>>>>>     document are to be interpreted as described in RFC2119 [RFC2119].
>>>>>>>>
>>>>>>>>     When used in lower case, these words convey their typical use in
>>>>>>>>     common language, and are not to be interpreted as described in
>>>>>>>>     RFC2119 [RFC2119].
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013               [Page 4]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 1.  Introduction
>>>>>>>>
>>>>>>>>     The objective of Precision Time Protocol (PTP) and Network Timing
>>>>>>>>     Protocol (NTP) are to synchronize independent clocks running on
>>>>>>>>     separate nodes of a distributed system.
>>>>>>>>
>>>>>>>>     [IEEE-1588] defines PTP messages for frequency, phase and time
>>>>>>>>     synchronization.  The PTP messages include PTP PDUs over UDP/IP
>>>>>>>>     (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Annex F of
>>>>>>>>     [IEEE-1588]).
>>>>>>>>
>>>>>>>> SB> Sure BUT it is acknowledged that IEEE is open to the definition
>>>>>>>> SB> of other PTP mappings if they provide better optimisation.
>>>>>>>>
>>>>>>>>     This document defines mapping and transport of the PTP
>>>>>>>>     messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PTP
>>>>>>>>     defines several clock types: ordinary clocks, boundary clocks, end-
>>>>>>>>     to-end transparent clocks, and peer-to-peer transparent clocks.
>>>>>>>>     Transparent clocks require intermediate nodes to update correction
>>>>>>>>     field inside PTP message that reflects the transit time in the node.
>>>>>>>>
>>>>>>>>     [RFC5905] defines NTP messages for clock and time synchronization.
>>>>>>>>     The PTP messages (PDUs) are transported over UDP/IP.  This document
>>>>>>>> SB> Should that be NTP messages?
>>>>>>>> SB> It needs to be made clear as soon as you introduce NTP that
>>>>>>>> SB> they use different time representations.
>>>>>>>>
>>>>>>>>     defines mapping and transport of the NTP messages defined in
>>>>>>>>     [RFC5905] over MPLS networks.
>>>>>>>>
>>>>>>>>     One key attribute of all of these Timing messages is that the Time
>>>>>>>>     stamp processing should occur as close as possible to the actual
>>>>>>>>     transmission and reception at the physical port interface.  This
>>>>>>>>     targets optimal time and/or frequency recovery by avoiding variable
>>>>>>>>     delay introduced by queues internal to the clocks.
>>>>>>>>
>>>>>>>> SB> As I recall NTP has no epoch point defined, and I am not sure
>>>>>>>> SB> where that point is in the case of PTP in this mapping
>>>>>>>> SB> Hopefully this will get defined in due course.
>>>>>>>>
>>>>>>>>     To facilitate the fast and efficient recognition of Timing messages
>>>>>>>>     at the port level when the Timing messages are carried over MPLS
>>>>>>>>     LSPs,
>>>>>>>>
>>>>>>>> SB> Over a new LSP type with time optimied characteristics
>>>>>>>>
>>>>>>>>     this document defines the specific encapsulations that should
>>>>>>>>     be used.
>>>>>>>> SB> Hopefully it will also define the PHP
>>>>>>>>
>>>>>>>>     In addition, it can be expected that there will exist LSR/
>>>>>>>>     LERs where only a subset of the physical ports will have the port-
>>>>>>>>     based Timing message processing capabilities.
>>>>>>>> SB> Do you need to clarify that this only works at base and not in
>>>>>>>> SB> a label heirarchy.
>>>>>>>>
>>>>>>>>
>>>>>>>>     In order to ensure
>>>>>>>>     that the LSPs carrying Timing packets always enter and exit ports
>>>>>>>>     with this capability, routing extensions are defined to advertise
>>>>>>>>     this capability on a port basis and to allow for the establishment of
>>>>>>>>     LSPs that only transit such ports.  While this path establishment
>>>>>>>>     restriction may be applied only at the LER Ingress and/or egress
>>>>>>>>     ports, it becomes more important when using transparent clock capable
>>>>>>>>     LSRs in the path.
>>>>>>>> SB> I do not understand the implications of the last
>>>>>>>> SB> sentences - starting ", it becomes"
>>>>>>>>
>>>>>>>>
>>>>>>>>     Port based Timing message processing involves Timing message
>>>>>>>>     recognition.  Once the Timing messages are recognized they can be
>>>>>>>>     modified based on the reception or transmission Time-stamp.
>>>>>>>>
>>>>>>>>     This document provides two methods for transporting Timing messages
>>>>>>>>     over MPLS.  One is applicable to MPLS environment and the other one
>>>>>>>>     is applicable to MPLS/MPLS-TP environment
>>>>>>>>
>>>>>>>> SB> I think the sentence is incomplete.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013               [Page 5]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>>     The solution involves transporting Timing messages over dedicated
>>>>>>>>     LSPs called Timing LSPs.  These LSPs carry Timing messages and MAY
>>>>>>>>     carry Management and control messages, but not data plane client
>>>>>>>>     traffic.
>>>>>>>>
>>>>>>>> SB> It is not clear why this restriction applies.
>>>>>>>>
>>>>>>>>     Timing LSPs can be established statically or via signaling.
>>>>>>>> SB> s/statically/by provisioning/network management/
>>>>>>>>
>>>>>>>>     Extensions to control plane (OSPF, ISIS, etc.) is required to enable
>>>>>>>>     routers to distribute their Timing processing capabilities over MPLS
>>>>>>>>     to other routers.  However such extensions are outside the scope of
>>>>>>>>     this document.
>>>>>>>>
>>>>>>>>     When signaling is used to setup the PTP LSP, Extensions to signaling
>>>>>>>> SB> is it a PTP LSP or a Timing LSP?
>>>>>>>>
>>>>>>>>     protocols (e.g., RSVP-TE) are required for establishing PTP LSPs.
>>>>>>>>     However such extensions are outside the scope of this document.
>>>>>>>>
>>>>>>>> SB> for mpls-tp GMPLS is the signalling protocol
>>>>>>>>
>>>>>>>>     While the techniques included herein allow for the establishment of
>>>>>>>>     paths optimized to include Time-stamping capable links, the
>>>>>>>>     performance of the Slave clocks is outside the scope of this
>>>>>>>>     document.
>>>>>>>>
>>>>>>>>     At the time of publishing this specification, Transparent Clocking
>>>>>>>>     (TC) is only defined for PTP.  Therefore at this time any part of
>>>>>>>>     this specification that talks about Transparent Clocking applies only
>>>>>>>>     to PTP.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013               [Page 6]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 2.  Terminology
>>>>>>>>
>>>>>>>>     1588: The timing and synchronization as defined by IEEE 1588.
>>>>>>>>
>>>>>>>> SB> I think that there is a more formal name for the 1588 group
>>>>>>>> SB> that needs to be used here.
>>>>>>>> SB> Also do we need to talk about 1588-200? as there is
>>>>>>>> SB> an update in progress
>>>>>>>>
>>>>>>>>     NTP: The timing and synchronization protocol defined by IETF RFC-1305
>>>>>>>>     and RFC-5905.
>>>>>>>>
>>>>>>>>     PTP: The timing and synchronization protocol used by 1588.
>>>>>>>> SB> need the proper name for 1588
>>>>>>>>
>>>>>>>>     Master Clock: The source of 1588 timing to a set of slave clocks.
>>>>>>>>
>>>>>>>>     Master Port: A port on a ordinary or boundary clock that is in Master
>>>>>>>>     state.  This is the source of timing toward slave ports.
>>>>>>>>
>>>>>>>> SB> I am not sure the reader knows what a port is
>>>>>>>>
>>>>>>>>     Slave Clock: A receiver of 1588 timing from a master clock.
>>>>>>>>
>>>>>>>>     Slave Port: A port on a boundary clock or ordinary clock that is
>>>>>>>>     receiving timing from a master clock.
>>>>>>>>
>>>>>>>>     Ordinary Clock: A device with a single PTP port.
>>>>>>>>
>>>>>>>>     Transparent Clock.  A device that measures the time taken for a PTP
>>>>>>>>     event message to transit the device and then updates the
>>>>>>>>     correctionField of the message with this transit time.
>>>>>>>>
>>>>>>>>     Boundary Clock: A device with more than one PTP port.  Generally
>>>>>>>>     boundary clocks will have one port in slave state to receive timing
>>>>>>>>     and then other ports in master state to re-distribute the timing.
>>>>>>>>
>>>>>>>>     PTP LSP: An LSP dedicated to carry PTP messages
>>>>>>>>
>>>>>>>> SB> PTP or timing?
>>>>>>>>
>>>>>>>>
>>>>>>>>     PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>>>>>>>     messages.
>>>>>>>>
>>>>>>>> SB> Ah I don't think that PWE3 know what one of these is
>>>>>>>>
>>>>>>>>     CW: Pseudowire Control Word
>>>>>>>>
>>>>>>>>     LAG: Link Aggregation
>>>>>>>>
>>>>>>>>     ECMP: Equal Cost Multipath
>>>>>>>>
>>>>>>>>     CF: Correction Field, a field inside certain PTP messages (message
>>>>>>>>     type 0-3)that holds the accumulative transit time inside intermediate
>>>>>>>>     switches
>>>>>>>>
>>>>>>>>     Timing messages: Timing Protocol messages that are exchanged
>>>>> between
>>>>>>>>     routers in order to establish a synchronized clock.
>>>>>>>>
>>>>>>>> SB> A number of these definitions look like copies of IEEE1588
>>>>>>>> SB> definitions. We need to provide references and note the
>>>>>>>> SB> priority of the IEEE base reference.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013               [Page 7]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 3.  Problem Statement
>>>>>>>>
>>>>>>>>     [IEEE-1588] has defined methods for transporting PTP messages over
>>>>>>>>     Ethernet and IP networks.  [RFC5905] has defined the method of
>>>>>>>>     transporting NTP messages over IP networks.  There is a need to
>>>>>>>>     transport Timing messages over MPLS networks while supporting the
>>>>>>>>     Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (OC)
>>>>>>>>     functionality in the LER and LSRs in the MPLS network.
>>>>>>>>
>>>>>>>>     There are multiple ways of transporting Timing over MPLS.  However,
>>>>>>>>     there is a requirement to limit the possible encapsulation options to
>>>>>>>>     simplify the Timing message identification and processing required at
>>>>>>>>     the port level.
>>>>>>>>
>>>>>>>>     When Timing-awareness is needed, Timing messages should not be
>>>>>>>>     transported over LSPs or PWs that are carrying customer traffic
>>>>>>>>     because LSRs perform Label switching based on the top label in the
>>>>>>>>     stack.
>>>>>>>>
>>>>>>>> SB> Have you explained why?
>>>>>>>>
>>>>>>>>     To detect Timing messages inside such LSPs require special
>>>>>>>>     hardware to do deep packet inspection at line rate.  Even if such
>>>>>>>>     hardware exists, the payload can't be deterministically identified by
>>>>>>>>     LSRs because the payload type is a context of the PW label, and the
>>>>>>>>     PW label and its context are only known to the Edge routers (PEs/
>>>>>>>>     LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR, CES,
>>>>>>>>     etc).  Even if one restricts an LSP to only carry Ethernet PWs, the
>>>>>>>>     LSRs dont have the knowledge of whether PW Control Word (CW) is
>>>>>>>>     present or not and therefore can not deterministically identify the
>>>>>>>>     payload.
>>>>>>>>
>>>>>>>>     A generic method is defined in this document that does not require
>>>>>>>>     deep packet inspection at line rate, and can deterministically
>>>>>>>>     identify Timing messages.  This method can be used to detect Timing
>>>>>>>>     Messages in both one-step and two-step clock implementations of
>>>>>>>>     ordinary, boundary and transparent clocks.
>>>>>>>>
>>>>>>>> SB> Needs a ref and I am sure many MPLS specialists will not understand
>>>>>>>> SB> the msg types.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013               [Page 8]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 4.  Timing over MPLS Architecture
>>>>>>>>
>>>>>>>>     Timing messages are exchange between Timing ports on ordinary and
>>>>>>>>
>>>>>>>> SB> Have you defined a timing port?
>>>>>>>>
>>>>>>>>     boundary clocks.  Boundary clocks terminate the Timing messages and
>>>>>>>>     act as master for other boundary clocks or for slave clocks.  End-to-
>>>>>>>>     End Transparent clocks do not terminate the Timing messages but they
>>>>>>>>     do modify the contents of the Timing messages as they transit across
>>>>>>>>     the transparent clock.
>>>>>>>>
>>>>>>>>     Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent Clock
>>>>>>>>
>>>>>>>>     (TC) could be implemented in either LERs or LSRs.
>>>>>>>>
>>>>>>>> SB> LER and LSR need to be expanded
>>>>>>>>
>>>>>>>>     An example is shown in Figure 1, where the LERs act as Ordinary Clock
>>>>>>>>     (OC) and are the initiating/terminating point for Timing messages.
>>>>>>>>     The ingress LER encapsulates the Timing messages in Timing LSP and
>>>>>>>>     the Egress LER terminates the Timing LSP.  The LSRs act as
>>>>>>>>     Transparent Clock (TC) and just update the Timing field in the Timing
>>>>>>>>     messages.
>>>>>>>>
>>>>>>>>
>>>>>>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>>>        |Switch, |     |       |     |       |     |       |     |Switch, |
>>>>>>>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>>>>>>        |        |     |  OC   |     |  TC   |     |  OC   |     |        |
>>>>>>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>>>                       /                                 \
>>>>>>>>        +-------+     /                                   \     +-------+
>>>>>>>>        |  LER  |    /                                     \    |  LER  |
>>>>>>>>        | Master|---/                                       \---| Slave |
>>>>>>>>        | Clock |                                               | Clock |
>>>>>>>>        +-------+                                               +-------+
>>>>>>>>
>>>>>>>>       Figure (1) - Deployment example 1 of timing over MPLS network
>>>>>>>>
>>>>>>>>     Another example is shown in Figure2, where LERs terminate the Timing
>>>>>>>>     messages received from switch/routers that are outside of the MPLS
>>>>>>>>     network acting as OC or BC.  In this example LERs regenerate the
>>>>>>>>     clock and initiate timing messages encapsulated in Timing LSP toward
>>>>>>>>     the MPLS network, while the LSRs act as Transparent Clock (TC) and
>>>>>>>>     just update the Timing field in the Timing messages, which are
>>>>>>>>     already encapsulated in Timing LSPs.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013               [Page 9]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>>>       |Switch, |     |       |     |       |     |       |     |Switch, |
>>>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>>>>>>       | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC/BC  |
>>>>>>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>>>
>>>>>>>>       Figure (2) - Deployment example 2 of timing over MPLS network
>>>>>>>>
>>>>>>>>
>>>>>>>>     Another example is shown in Figure 3, where LERs do not terminate the
>>>>>>>>     Timing messages received from switch/routers that are outside of the
>>>>>>>>     MPLS network acting as OC, TC or BC.  The LERs act as TC and update
>>>>>>>>     the Timing field in the Timing messages as they transit the LER,
>>>>>>>>     while encapsulating them in timing LSP.  The LSRs also act as
>>>>>>>>     Transparent Clock (TC) and just update the Timing field in the Timing
>>>>>>>>     messages which are already encapsulated in Timing LSPs.
>>>>>>>>
>>>>>>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>>>        |Switch, |     |       |     |       |     |       |     |Switch, |
>>>>>>>>        | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>>>>>>        |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC/TC/BC|
>>>>>>>>        +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>>>
>>>>>>>>      Figure (3) - Deployment example 3 of timing over MPLS network
>>>>>>>>
>>>>>>>>     Another example is shown in Figure 4, where LERs and LSRs support
>>>>>>>>     Boundary Clocks.  A single-hop LSP is created between two adjacent
>>>>>>>>     LSRs engaged in BC operation.  Other methods such as PTP transport
>>>>>>>>     over Ethernet MAY be used for transporting timing messages if the
>>>>>>>>     link between the two routers is Ethernet.
>>>>>>>>
>>>>>>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>>>       |Switch, |     |       |     |       |     |       |     |Switch, |
>>>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Router |
>>>>>>>>       | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC/BC  |
>>>>>>>>       +--------+     +-------+     +-------+     +-------+     +--------+
>>>>>>>>
>>>>>>>>     Figure (4) - Deployment example 3 of timing over MPLS network
>>>>>>>>
>>>>>>>>     An MPLS domain MAY serve multiple customers.  In these cases the
>>>>> MPLS
>>>>>>>>     domain (maintained by a service provider) may provide timing services
>>>>>>>>     to multiple customers, each having their own Timing domain.
>>>>>>>>
>>>>>>>>     The Timing over MPLS architecture assumes full mesh of Timing LSPs
>>>>>>>>     between all LERs supporting this specification.
>>>>>>>>
>>>>>>>> SB> Note sure this is right - the salves surely do not need to
>>>>>>>> SB> exchange timing amongst themselves
>>>>>>>>
>>>>>>>>     It supports
>>>>>>>>     Point-to- point (VPWS) and Multipoint (VPLS) services.
>>>>>>>>
>>>>>>>> SB> What does that mean? You do not carry user data traffic?
>>>>>>>> SB> Maybe it's the ordering of the statemnets that is causing
>>>>>>>> SB> confusion.
>>>>>>>>
>>>>>>>>     This means
>>>>>>>>     that a customer may purchase a Point-to-point Timing service between
>>>>>>>>     two customer sites or a Multipoint Timing service between more than
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 10]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>>     two customer sites.
>>>>>>>>
>>>>>>>>     The Timing over MPLS architecture supports P2P or P2MP Timing LSPs.
>>>>>>>>     This means that the Timing Multicast messages such as PTP Multicast
>>>>>>>>     event messages can be transported over P2MP Timing LSP or be
>>>>>>>>     replicated and transported over many P2P Timing LSPs.
>>>>>>>>
>>>>>>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>>>>>>>> SB> nor a P2MP PW, although we are close.
>>>>>>>>
>>>>>>>>     Timing messages, that do not require Time stamping or Correction
>>>>>>>>     Field update MAY be transported over Timing LSPs to simplify hardware
>>>>>>>>     and software.
>>>>>>>>
>>>>>>>>     PTP Announce messages that determine the Timing LSP terminating
>>>>> point
>>>>>>>>     behavior such as BC/OC/TC SHOULD be transported over the Timing LSP
>>>>>>>>     to simplify hardware and software.
>>>>>>>>
>>>>>>>> SB> have you defined and referenced PTP announce msgs?
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 11]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 5.  Dedicated LSPs for Timing messages
>>>>>>>>
>>>>>>>>     Many methods have been considered for identifying the Timing
>>>>> messages
>>>>>>>>     when they are encapsulated in MPLS such as using GAL/G-ACH or a new
>>>>>>>>     reserved label.  These methods were not attractive since they either
>>>>>>>>     required deep packet inspection at line rate in the intermediate LSRs
>>>>>>>>     or they required use of a scarce new reserved label.  Also one of the
>>>>>>>>     goals was to reuse existing OAM mechanisms.
>>>>>>>>
>>>>>>>> SB> RLs = SPLs are not so rare now. In any case needs a ref.
>>>>>>>>
>>>>>>>>     The method defined in this document can be used by LER and LSRs to
>>>>>>>>     identify Timing messages in MPLS tunnels by just looking at the top
>>>>>>>>     label in the MPLS label stack, which only carry Timing messages as
>>>>>>>>     well as OAM, but not data plane client traffic.
>>>>>>>>
>>>>>>>>     Compliant implementations MUST use dedicated LSPs to carry Timing
>>>>>>>>     messages over MPLS.
>>>>>>>>
>>>>>>>> SB> I think that we need a definition of the properies of these LSPs
>>>>>>>>
>>>>>>>>     These LSPs are herein referred to as "Timing
>>>>>>>>     LSPs" and the labels associated with these LSPs as "Timing LSP
>>>>>>>>     labels".  The Timing LSPs that runs between Ingress and Egress LERs
>>>>>>>>     MUST be co-routed.  Alternatively, a single bidirectional co-routed
>>>>>>>>     LSP can be used.
>>>>>>>>
>>>>>>>> SB> I though that you said you could use M2MP LSPs - these are not
>>>>>>>> SB> bidirectional.
>>>>>>>>
>>>>>>>>     Co-routing of the two directions is required to limit the difference
>>>>>>>>     in the delays in the Master clock to Slave clock direction compared
>>>>>>>>     to the Slave clock to Master clock direction.  The Timing LSP MAY be
>>>>>>>>     MPLS/MPLS-TP LSP.
>>>>>>>>
>>>>>>>>     The Timing LSPs could be configured or signaled via RSVP-TE/GMPLS.
>>>>>>>>     New Extensions to RSVP-TE/GMPLS TLVs are required; however they are
>>>>>>>>     outside the scope of this document.
>>>>>>>>
>>>>>>>>     The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic such as
>>>>>>>>     BFD and LSP Ping but the LSP data plane client plane traffic MUST be
>>>>>>>>     Timing packets only.
>>>>>>>>
>>>>>>>> SB> Why?
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 12]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 6.  Timing over LSP Encapsulation
>>>>>>>>
>>>>>>>> The encapsulations is not LSP is it?
>>>>>>>>
>>>>>>>>     This document defines two methods for carrying Timing messages over
>>>>>>>>     MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>>>>>>>     messages over Timing LSPs, and the second method, is carrying
>>>>>>>>     Ethernet encapsulated Timing messages over Ethernet PWs inside
>>>>> Timing
>>>>>>>>     LSPs.
>>>>>>>>
>>>>>>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>>>>>>>
>>>>>>>>     The simplest method of transporting Timing messages over MPLS is to
>>>>>>>>     encapsulate Timing PDUs in UDP/IP and then encapsulate them in
>>>>> Timing
>>>>>>>>     LSP.  This format is shown in Figure 4.
>>>>>>>>
>>>>>>>>
>>>>>>>>                      +----------------------+
>>>>>>>>                      |   Timing LSP Label   |
>>>>>>>>                      +----------------------+
>>>>>>>>                      |        IPv4/6        |
>>>>>>>>                      +----------------------+
>>>>>>>>                      |         UDP          |
>>>>>>>>                      +----------------------+
>>>>>>>>                      |     Timing PDU       |
>>>>>>>>                      +----------------------+
>>>>>>>>
>>>>>>>>        Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>>>>>>>
>>>>>>>>
>>>>>>>>     This encapsulation is very simple and is useful when the network
>>>>>>>>     between Timing Master Clock and Slave Clock is MPLS network.
>>>>>>>>
>>>>>>>> SB> Simple is a judgement call
>>>>>>>>
>>>>>>>>     In order for an LER/LSR to process Timing messages, the Timing LSP
>>>>>>>>     Label must be at the top label of the label stack.  The LER/LSR MUST
>>>>>>>>     know that the Timing LSP Label is used for carrying Timing messages.
>>>>>>>>     This can be accomplished via static configuration or via RSVP-TE
>>>>>>>>     signaling.
>>>>>>>>
>>>>>>>>     The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>>>>>>>     [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow
>>>>>>>>     [RFC5905].
>>>>>>>>
>>>>>>>> 6.2.  Timing over PW Encapsulation
>>>>>>>>
>>>>>>>>     Another method of transporting Timing over MPLS networks is by
>>>>>>>>     encapsulating Timing PDUs in PW which in turn is transported over
>>>>>>>>     Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC4448],
>>>>>>>>     shown in Fig 5(A) MUST be used and the Ethernet encapsulation of PTP
>>>>>>>>     MUST follow Annex F of [IEEE-1588].
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 13]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>>     The RAW mode or Tagged mode defined in [RFC4448] MAY be used and
>>>>> the
>>>>>>>>     Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  The
>>>>>>>>     Timing over PW encapsulation MUST use the Control Word (CW) as
>>>>>>>>     specified in [RFC4448] to ensure proper detection of PTP messages
>>>>>>>>     inside the MPLS packets for Timing over LSP and Timing over PW
>>>>>>>>     encapsulation.
>>>>>>>>
>>>>>>>> SB> That needs explanation
>>>>>>>>
>>>>>>>>     The use of Sequence Number in the CW is optional.
>>>>>>>>
>>>>>>>> SB> Given that s/n are never in practice deployed, you could probably
>>>>>>>> SB> simplify things by sayig that they are not used.
>>>>>>>>
>>>>>>>>     Timing over PW encapsulation for NTP MUST use NTP over UDP/IP over
>>>>> PW
>>>>>>>>     (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>>>>>>>
>>>>>>>>                      +----------------+  +----------------+
>>>>>>>>                       |Timing LSP Label|  |Timing LSP Label|
>>>>>>>>                       +----------------+  +----------------+
>>>>>>>>                       |    PW Label    |  |    PW Label    |
>>>>>>>>                       +----------------+  +----------------+
>>>>>>>>                       |  Control Word  |  |      IP        |
>>>>>>>>                       +----------------+  +----------------+
>>>>>>>>                       |    Ethernet    |  |      UDP       |
>>>>>>>>                       |     Header     |  +----------------+
>>>>>>>>                       +----------------+  |   Timing PDU   |
>>>>>>>>                       |S-VLAN(Optional)|  |                |
>>>>>>>>                       +----------------+  +----------------+
>>>>>>>>                       |C-VLAN(Optional)|        (B)
>>>>>>>>                       +----------------+
>>>>>>>>                       |   Timing PDU   |
>>>>>>>>                       |                |
>>>>>>>>                       +----------------+
>>>>>>>>                              (A)
>>>>>>>>
>>>>>>>>                Figure (5) - Timing over PW Encapsulations
>>>>>>>>
>>>>>>>>     In order for an LSR to process PTP messages, the top label of the
>>>>>>>>     label stack (the Tunnel Label) MUST be a Timing label.
>>>>>>>>
>>>>>>>> S> You said that before.
>>>>>>>>
>>>>>>>> 6.3.  Other Timing Encapsulation methods
>>>>>>>>
>>>>>>>>     In future other timing encapsulation methods may be introduced, such
>>>>>>>>     as a new shim header after the Bottom of Stack to carry the Timing
>>>>>>>>     information.  Such new encapsulations are outside the scope of this
>>>>>>>>     document.
>>>>>>>>
>>>>>>>>
>>>>>>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>>>>>>>> SB> out of the definition of the LSP
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 14]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> SB> I think we need a section on LSP processing
>>>>>>>>
>>>>>>>> 7.  Timing message Processing
>>>>>>>>
>>>>>>>>     Each Timing protocol such as PTP and NTP, define their set of Timing
>>>>>>>>     messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,
>>>>>>>>     FOLLOW_UP, etc messages.
>>>>>>>>
>>>>>>>>     Some of the Timing messages require time stamping or correction field
>>>>>>>>     update at port level and some dont.  It is the job of the LER/LSR to
>>>>>>>>     parse the timing message and find out the type of the Timing message
>>>>>>>>     and decide whether and how to Time- stamp it (e.g., BC) or update
>>>>>>>>     correction field(e.g., TC).
>>>>>>>>
>>>>>>>>
>>>>>>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing
>>>>>>>> SB> function rather than the LER?
>>>>>>>>
>>>>>>>>     For example the following PTP messages (called Event messages)
>>>>>>>>     require time-stamping or correction field update:
>>>>>>>>
>>>>>>>>     o  SYNC
>>>>>>>>
>>>>>>>>     o  DELAY_REQ (Delay Request)
>>>>>>>>
>>>>>>>>     o  PDELAY_REQ (Peer Delay Request)
>>>>>>>>
>>>>>>>>     o  PDELAY_RESP (Peer Delay Response)
>>>>>>>>
>>>>>>>>     SYNC and DELAY_REQ are exchanged between Master Clock and Slave
>>>>> Clock
>>>>>>>>     and MUST be transported over PTP LSPs.  PDELAY_REQ and
>>>>> PDELAY_RESP
>>>>>>>>     are exchanged between adjacent PTP clocks (i.e.  Master, Slave,
>>>>>>>>     Boundary, or Transparent) and SHOULD be transported over single hop
>>>>>>>>     PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_UP,
>>>>>>>>     and PDELAY_RESP_FOLLOW_UP messages MUST also be transported
>>>>> over
>>>>>>>> the
>>>>>>>>     PTP LSPs.
>>>>>>>>
>>>>>>>>     For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST be
>>>>>>>>     transported over two PTP LSPs that are in opposite directions.  These
>>>>>>>>     PTP LSPs, which are in opposite directions MUST be congruent and co-
>>>>>>>>     routed.  Alternatively, a single bidirectional co-routed LSP can be
>>>>>>>>     used.
>>>>>>>>
>>>>>>>>     Except as indicated above for the two-step PTP clocks, Non-Event PTP
>>>>>>>>     message types do not need to be processed by intermediate routers.
>>>>>>>>     These message types MAY be carried in PTP Tunnel LSPs.
>>>>>>>>
>>>>>>>> SB> Are you saying that a timing P router has to be msg type sensitive?
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 15]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 8.  Protection and Redundancy
>>>>>>>>
>>>>>>>>
>>>>>>>> SB> This is a bit of a jump - I don't know how the LSP itself works yet!
>>>>>>>>
>>>>>>>>     In order to ensure continuous uninterrupted operation of slave
>>>>>>>>     clocks, usually as a general practice, slave clocks (or ports) track
>>>>>>>>     redundant master clocks.
>>>>>>>>
>>>>>>>>     It is the responsibility of the network operator to ensure that
>>>>>>>>     physically disjoint Timing LSPs are established between a slave clock
>>>>>>>>     (or port) and redundant master clocks (or ports).
>>>>>>>>
>>>>>>>>     When a slave clock (or port) listens to redundant master clocks or
>>>>>>>>     ports, any prolonged Timing LSP outage will trigger the slave clock
>>>>>>>>     or port to switch to a redundant master clock or port.
>>>>>>>>
>>>>>>>>     LSP/PW protection such as Linear protection Switching (1:1, 1+1),
>>>>>>>>     Ring protection switching or MPLS Fast Reroute (FRR) generally switch
>>>>>>>>     alternative path that usually cause a change in delay, which if
>>>>>>>>     undetected by slave clock can reduce accuracy of the slave clock.
>>>>>>>>
>>>>>>>>     Therefore protection switching MAY be used, as long as phase jumps
>>>>>>>>     upon switchover due to differences in path latency are detected and
>>>>>>>>     compensated for (such compensation not being required if BCs or peer-
>>>>>>>>     peer TCs are used throughout).
>>>>>>>>
>>>>>>>>     Note that any protection or reroute mechanism that adds additional
>>>>>>>>     MPLS label to the label stack, such as Facility Backup Fast Reroute,
>>>>>>>>     MUST ensure that the pushed label is also a Timing Label to ensure
>>>>>>>>     recognition of the MPLS frame as containing Timing messages, as it
>>>>>>>>     transits the backup path.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 16]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 9.  ECMP
>>>>>>>>
>>>>>>>>     To ensure the optimal operation of slave clocks and avoid error
>>>>>>>>     introduced by forward and reverse path delay asymmetry, the physical
>>>>>>>>     path for Timing messages from master clock to slave Clock and vice
>>>>>>>>     versa must be the same for all Event Timing messages listed in
>>>>>>>>     section 7.
>>>>>>>>
>>>>>>>>     Therefore the Timing LSPs MUST not be subject to ECMP (Equal Cost
>>>>>>>>     Multipath).
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 17]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 10.  PHP
>>>>>>>>
>>>>>>>>     To ensure that the label on the top of the label stack is the Timing
>>>>>>>>     LSP Label, PHP MUST not be used.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 18]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 11.  Entropy
>>>>>>>>
>>>>>>>>     To ensure all Timing messages in a Timing LSP take the same path,
>>>>>>>>     Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>>>>>>>     Entropy Label MUST NOT be used for the PWs that are carried inside
>>>>>>>>     Timing LSP [RFC6391].
>>>>>>>>
>>>>>>>> SB> This is incorrect - you mean that all msgs of the same timing
>>>>>>>> SB> flow need to have the same EL value.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 19]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 12.  OAM, Control and Management
>>>>>>>>
>>>>>>>>     In order to monitor Timing LSPs and their encapsulated PWs, they MUST
>>>>>>>>     be able to carry OAM and management messages.  These management
>>>>>>>>     messages MUST be differentiated from Timing messages via already
>>>>>>>>     defined IETF methods.
>>>>>>>>
>>>>>>>>     For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY run
>>>>>>>>     over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These
>>>>>>>>     Management protocols can easily be identified by the UDP Destination
>>>>>>>>     Port number or by GAL/G-ACH respectively.
>>>>>>>>
>>>>>>>>     Also BFD, LSP-Ping and other management messages MAY run over the
>>>>> PWs
>>>>>>>>     encapsulated in Timing LSP via one of the defined VCCVs (Type 1, 3 or
>>>>>>>>     4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is going
>>>>>>>>     to be deprecated by IETF).  In this case G-ACH, PW label (TTL=1) or
>>>>>>>>     GAL-ACH are used to identify such management messages.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 20]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 13.  QoS Considerations
>>>>>>>>
>>>>>>>>     In network deployments where not every LSR/LER is Timing-aware, it is
>>>>>>>>     important to reduce the impact of the non-Timing-aware LSR/LERs on
>>>>>>>>     the timing recovery in the slave clock.  The Timing messages are time
>>>>>>>>     critical and must be treated with the highest priority.  Therefore
>>>>>>>>     Timing over MPLS messages must be treated with the highest priority
>>>>>>>>     in the routers.  This can be achieved by proper setup of Timing LSPs.
>>>>>>>>
>>>>>>>>     It is recommended that the Timing LSPs are setup or configured
>>>>>>>>     properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC2697]
>>>>>>>>     for drop eligibility.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 21]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 14.  FCS and Checksum Recalculation
>>>>>>>>
>>>>>>>>     When time-stamp generation and timing packet adjustment is performed
>>>>>>>>     near the physical port hardware, the process MUST include
>>>>>>>>     recalculation of the Ethernet FCS.
>>>>>>>>
>>>>>>>> SB> The above is confusing - an LSR always recomputes the link layer
>>>>>>>> SB> CRC which may or may not be Ethernet.
>>>>>>>>
>>>>>>>>     Also FCS retention for the
>>>>>>>>     payload Ethernet described in [RFC4720] MUST NOT be used.
>>>>>>>>
>>>>>>>>     For UDP/IP encapsulation mode of Timing over MPLS, the UDP checksum
>>>>>>>>     may be required as per UDP transport standards.
>>>>>>>>
>>>>>>>> SB> You really need to be working on getting the IPv6 C?S computation
>>>>>>>> SB> removed from PTP msgs.
>>>>>>>>
>>>>>>>>     When UDP checksum is used, each Timing-aware LER/LSR must either
>>>>>>>>     incrementally update the UDP checksum after Time stamping or
>>>>>>>>     Correction Field update or verify the UDP checksum on reception from
>>>>>>>>     upstream and recalculate the checksum completely on transmission to
>>>>>>>>     downstream node after Time stamping or Correction Field update.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 22]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 15.  Behavior of LER/LSR
>>>>>>>>
>>>>>>>>     Timing-capable/aware LERs and LSRs are routers that have one or more
>>>>>>>>
>>>>>>>> SB> You mean physical interfaces?
>>>>>>>>
>>>>>>>>     interfaces that can perform Timing operations (OC/BC/TC) on Timing
>>>>>>>>     packets and are configured to do so.  Timing-capable/aware LERs and
>>>>>>>>     LSRs can advertise their Timing-capability per-interface via control
>>>>>>>>     plane such as OSPF or IS-IS.
>>>>>>>> SB> ISIS and OSPF are routing protocols.
>>>>>>>>
>>>>>>>>    The Timing-capable/aware LERs can then
>>>>>>>>     signals Timing LSPs via RSVP-TE signaling.  Alternatively the Timing
>>>>>>>>     capability of LER and LSRs may be configured in a centralized
>>>>>>>>     controller and the Timing LSP may be setup using manual configuration
>>>>>>>>     or other methods such as SDN.
>>>>>>>>
>>>>>>>> SB> it can also be configured individually rather then through
>>>>>>>> SB> a cebtral controllwe
>>>>>>>>
>>>>>>>> 15.1.  Behavior of Timing-capable/aware LER
>>>>>>>>
>>>>>>>>     When a Timing-capable/aware LER behaves as a Transparent clock and
>>>>>>>>     receives a Timing message from a Timing-capable/aware non-MPLS
>>>>>>>>     interface, the LER updates the Correction Field (CF) and encapsulates
>>>>>>>>     and forwards the timing message over previously established Timing
>>>>>>>>     LSP.
>>>>>>>>
>>>>>>>> SB> You need to call out the details so that people properly
>>>>>>>> SB> understand the definition of the new LSP.
>>>>>>>>
>>>>>>>>     Also when a Timing message is received from a Timing-capable/
>>>>>>>>     aware MPLS interface, LER updates the Correction Filed (CF) and
>>>>>>>>     decapsulates the MPLS encapsulation and forwards the timing message
>>>>>>>>     to a non-MPLS interface.
>>>>>>>>
>>>>>>>>     When a Timing-capable/aware LER behaves as a Boundary clock and
>>>>>>>>     receives a Timing message from a Timing-capable/aware non MPLS
>>>>>>>>     interface, the LER Timestamps the Timing packet and sends it to the
>>>>>>>>     LERs Boundary clock processing module.  Also when a Timing message is
>>>>>>>>     received from a Timing- capable/aware MPLS interface, the LER
>>>>>>>>     Timestamps the Timing packet and sends it to the LERs Boundary clock
>>>>>>>>     processing module.
>>>>>>>>
>>>>>>>>     When a Timing-capable/aware LER behaves as an Ordinary Clock toward
>>>>>>>>     the MPLS network, and receives a Timing message from a Timing-
>>>>>>>>     capable/aware MPLS interface, the LER Timestamps the Timing packet
>>>>>>>>     and sends it to the LERs Ordinary clock processing module.
>>>>>>>>
>>>>>>>> 15.2.  Behavior of Timing-capable/aware LSR
>>>>>>>>
>>>>>>>>     When a Timing-capable/aware LSR behaves as a Transparent clock and
>>>>>>>>     receives a Timing message from a Timing-capable/aware MPLS
>>>>> interface,
>>>>>>>>     The LSR updates the Correction Filed (CF) and forwards the timing
>>>>>>>>     message over another MPLS interface.
>>>>>>>>
>>>>>>>>     When a Timing-capable/aware LSR behaves as a Boundary clock and
>>>>>>>>     receives a Timing message from a Timing-capable/aware MPLS
>>>>> interface.
>>>>>>>>     The LSR performs the functions of a Boundary Clock in terminating the
>>>>>>>>     received Timing message and re-generating a new timing message over
>>>>>>>>     another (or the same) MPLS interface.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 23]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>>>>>>>
>>>>>>>>     It is most beneficial when all LSRs in the path of a Timing LSP be
>>>>>>>>     timing-Capable/aware LSRs.  This would ensure the highest quality
>>>>>>>>     time and clock synchronization by Timing Slave Clocks.  However, this
>>>>>>>>     specification does not mandate that all LSRs in path of a Timing LSP
>>>>>>>>     be Timing- capable/aware.
>>>>>>>>
>>>>>>>>     Non-Timing-capable/aware LSRs just switch the packets encapsulated in
>>>>>>>>     Timing LSPs and dont perform any Timing operation (TC or BC).
>>>>>>>>     However as explained in QoS section the Timing over MPLS packets
>>>>> MUST
>>>>>>>>     be still be treated with the highest priority based on their Traffic
>>>>>>>>     Class (TC) marking.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 24]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 16.  Other considerations
>>>>>>>>
>>>>>>>>     [IEEE-1588] defines an optional peer-to-peer Transparent clocking
>>>>>>>>     that requires peer delay measurement between two adjacent Timing-
>>>>>>>>     capable/ aware routers/switches.  Peer delay measurement messages
>>>>>>>>     need to be time stamped and terminated by the Timing-capable/aware
>>>>>>>>     routers/ switches.  This means that two adjacent LSRs may be engaged
>>>>>>>>     in a peer delay measurement.
>>>>>>>>
>>>>>>>>     For transporting such peer delay measurement messages a single-hop
>>>>>>>>     LSP SHOULD to be created between the two adjacent LSRs engaged in
>>>>>>>>     peer delay measurement to carry peer delay measurement messages.
>>>>>>>>     Other methods such as PTP transport over Ethernet MAY be used for
>>>>>>>>     transporting peer delay measurement messages if the link between the
>>>>>>>>     two routers is Ethernet.
>>>>>>>>
>>>>>>>>     In Peer-to-peer transparent clocking (P2P TC), a Timing-capable/ ware
>>>>>>>>     routers/switches MUST maintain a list of all the neighbors it needs
>>>>>>>>     to send a PDelay_Req to, where each neighbor corresponds to a timing
>>>>>>>>     LSP.
>>>>>>>>
>>>>>>>>     The use of Explicit Null Label (Label= 0 or 2) is acceptable as long
>>>>>>>>     as either the Explicit Null label is the bottom of stack label
>>>>>>>>     (applicable only to UDP/IP encapsulation) or the label below the
>>>>>>>>     Explicit Null label is a PTP label.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 25]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 17.  Security Considerations
>>>>>>>>
>>>>>>>>     MPLS PW security considerations in general are discussed in [RFC3985]
>>>>>>>>     and [RFC4447],and those considerations also apply to this document.
>>>>>>>>
>>>>>>>>     An experimental security protocol is defined in [IEEE-1588].The PTP
>>>>>>>>     security extension and protocol provides group source authentication,
>>>>>>>>     message integrity, and replay attack protection for PTP messages.
>>>>>>>>
>>>>>>>>     When the MPLS network (provider network) serves multiple customers,
>>>>>>>>     it is important to maintain and process each customers clock and
>>>>>>>>     Timing messages separately from other customers to ensure there is no
>>>>>>>>     cross- customer effect.  For example if an LER BC is synchronized to
>>>>>>>>     a specific grandmaster, belonging to customer A, then the LER MUST
>>>>>>>>     use that BC clock only for customer A to ensure that customer A
>>>>>>>>     cannot attack other customers by manipulating its time.
>>>>>>>>
>>>>>>>>     Timing messages MAY be encrypted or authenticated, provided that the
>>>>>>>>     LERs/LSRs that are Timing capable/aware can authenticate/ decrypt the
>>>>>>>>     timing messages.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 26]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 18.  Acknowledgements
>>>>>>>>
>>>>>>>>     The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mizrahi,
>>>>>>>>     Stefano Ruffini, Peter Meyer, and other members of IETF for reviewing
>>>>>>>>     and providing feedback on this draft.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 27]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 19.  IANA Considerations
>>>>>>>>
>>>>>>>>     There are no IANA requirements in this specification.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 28]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 20.  References
>>>>>>>>
>>>>>>>> 20.1.  Normative References
>>>>>>>>
>>>>>>>>     [IEEE-1588]
>>>>>>>>                IEEE 1588-2008, "IEEE Standard for a Precision Clock
>>>>>>>>                Synchronization Protocol for Networked Measurement and
>>>>>>>>                Control Systems".
>>>>>>>>
>>>>>>>>     [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>>>>>>>                Requirement Levels", BCP 14, RFC 2119, March 1997.
>>>>>>>>
>>>>>>>>     [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to-
>>>>>>>>                Edge (PWE3) Architecture", RFC 3985, March 2005.
>>>>>>>>
>>>>>>>>     [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
>>>>>>>>                Proxies (ND Proxy)", RFC 4389, April 2006.
>>>>>>>>
>>>>>>>>     [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G.
>>>>>>>>                Heron, "Pseudowire Setup and Maintenance Using the Label
>>>>>>>>                Distribution Protocol (LDP)", RFC 4447, April 2006.
>>>>>>>>
>>>>>>>>     [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>>>>>>>                "Encapsulation Methods for Transport of Ethernet over MPLS
>>>>>>>>                Networks", RFC 4448, April 2006.
>>>>>>>>
>>>>>>>>     [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>>>>>>>                Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>>>>>>>                Retention", RFC 4720, November 2006.
>>>>>>>>
>>>>>>>>     [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit
>>>>>>>>                Connectivity Verification (VCCV): A Control Channel for
>>>>>>>>                Pseudowires", RFC 5085, December 2007.
>>>>>>>>
>>>>>>>>     [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Detection
>>>>>>>>                (BFD)", RFC 5880, June 2010.
>>>>>>>>
>>>>>>>>     [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow,
>>>>>>>>                "Bidirectional Forwarding Detection (BFD) for MPLS Label
>>>>>>>>                Switched Paths (LSPs)", RFC 5884, June 2010.
>>>>>>>>
>>>>>>>> 20.2.  Informative References
>>>>>>>>
>>>>>>>>     [I-D.ietf-pwe3-fat-pw]
>>>>>>>>                Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>>>>>>>>                J., and S. Amante, "Flow Aware Transport of Pseudowires
>>>>>>>>                over an MPLS Packet Switched Network",
>>>>>>>>                draft-ietf-pwe3-fat-pw-07 (work in progress), July 2011.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 29]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>>     [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermediate
>>>>>>>>                system routeing information exchange protocol for use in
>>>>>>>>                conjunction with the Protocol for providing the
>>>>>>>>                Connectionless-mode Network Service (ISO 8473)".
>>>>>>>>
>>>>>>>>     [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP and
>>>>>>>>                dual environments", RFC 1195, December 1990.
>>>>>>>>
>>>>>>>>     [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1998.
>>>>>>>>
>>>>>>>>     [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Color
>>>>>>>>                Marker", RFC 2697, September 1999.
>>>>>>>>
>>>>>>>>     [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Boudec,
>>>>>>>>                J., Courtney, W., Davari, S., Firoiu, V., and D.
>>>>>>>>                Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>>>>>>>                Behavior)", RFC 3246, March 2002.
>>>>>>>>
>>>>>>>>     [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engineering
>>>>>>>>                (TE) Extensions to OSPF Version 2", RFC 3630,
>>>>>>>>                September 2003.
>>>>>>>>
>>>>>>>>     [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermediate
>>>>>>>>                System (IS-IS) Extensions for Traffic Engineering (TE)",
>>>>>>>>                RFC 3784, June 2004.
>>>>>>>>
>>>>>>>>     [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., and S.
>>>>>>>>                Shaffer, "Extensions to OSPF for Advertising Optional
>>>>>>>>                Router Capabilities", RFC 4970, July 2007.
>>>>>>>>
>>>>>>>>     [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermediate
>>>>>>>>                System to Intermediate System (IS-IS) Extensions for
>>>>>>>>                Advertising Router Information", RFC 4971, July 2007.
>>>>>>>>
>>>>>>>>     [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi
>>>>>>>>                Topology (MT) Routing in Intermediate System to
>>>>>>>>                Intermediate Systems (IS-ISs)", RFC 5120, February 2008.
>>>>>>>>
>>>>>>>>     [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>>>>>>>                Engineering", RFC 5305, October 2008.
>>>>>>>>
>>>>>>>>     [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>>>>>>>                "Traffic Engineering Extensions to OSPF Version 3",
>>>>>>>>                RFC 5329, September 2008.
>>>>>>>>
>>>>>>>>     [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF
>>>>>>>>                for IPv6", RFC 5340, July 2008.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 30]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>>     [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network
>>>>>>>>                Time Protocol Version 4: Protocol and Algorithms
>>>>>>>>                Specification", RFC 5905, June 2010.
>>>>>>>>
>>>>>>>>     [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan,
>>>>>>>>                J., and S. Amante, "Flow-Aware Transport of Pseudowires
>>>>>>>>                over an MPLS Packet Switched Network", RFC 6391,
>>>>>>>>                November 2011.
>>>>>>>>
>>>>>>>>     [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W., and
>>>>>>>>                L. Yong, "The Use of Entropy Labels in MPLS Forwarding",
>>>>>>>>                RFC 6790, November 2012.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 31]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 1.  Routing extensions for Timing-aware Routers
>>>>>>>>
>>>>>>>>     MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340] and
>>>>>>>>     IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering (TE)
>>>>>>>>     link information used for constraint-based routing.
>>>>>>>>
>>>>>>>>     Indeed, it is useful to advertise data plane TE router link
>>>>>>>>     capabilities, such as the capability for a router to be Timing-aware.
>>>>>>>>     This capability MUST then be taken into account during path
>>>>>>>>     computation to prefer or even require links that advertise themselves
>>>>>>>>     as Timing-aware.  In this way the path can ensure the entry and exit
>>>>>>>>     points into the LERs and, if desired, the links into the LSRs are
>>>>>>>>     able to perform port based time-stamping thus minimizing their impact
>>>>>>>>     on the performance of the slave clock.
>>>>>>>>
>>>>>>>>     extensions are required to OSPF and IS-IS in order to advertise
>>>>>>>>     Timing-aware capabilities of a link.  Such extensions are outside the
>>>>>>>>     scope of this document; however such extension SHOULD be able to
>>>>>>>>     signal the following information per Router Link:
>>>>>>>>
>>>>>>>>     o  Capable of processing PTP, NTP or other Timing flows
>>>>>>>>
>>>>>>>>     o  Capable of performing Transparent Clock operation
>>>>>>>>
>>>>>>>>     o  Capable of performing Boundary Clock operation
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 32]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> 2.  Signaling Extensions for Creating Timing LSPs
>>>>>>>>
>>>>>>>>     RSVP-TE signaling MAY be used to setup the timing LSPs.  When RSVP-TE
>>>>>>>>     is used to setup Timing LSPs, some information that indicates that
>>>>>>>>     the LSP is carrying Timing flows MUST be included in the new
>>>>>>>>     Extensions to RSVP-TE:
>>>>>>>>
>>>>>>>>     The following information MAY also be included in the new Extensions
>>>>>>>>     to RSVP-TE:
>>>>>>>>
>>>>>>>>     o  Offset from Bottom of Stack (BoS) to the start of the Time-stamp
>>>>>>>>        field
>>>>>>>>
>>>>>>>>     o  Number of VLANs in case of PW encapsulation
>>>>>>>>
>>>>>>>>     o  Timestamp field Type
>>>>>>>>
>>>>>>>>        *  Correction Field, Timestamp
>>>>>>>>
>>>>>>>>     o  Timestamp Field format
>>>>>>>>
>>>>>>>>        *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-bit
>>>>>>>>           NTP, etc.
>>>>>>>>
>>>>>>>>     Note that in case the above optional information is signaled with
>>>>>>>>     RSVP-TE for a Timing LSP, all the Timing packets carried in that LSP
>>>>>>>>     must have the same signaled characteristics.  For example if
>>>>>>>>     Timestamp format is signaled as 64-bit PTPv1, then all Timing packets
>>>>>>>>     must use 64-bit PTPv1 time-stamp.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 33]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>> Authors' Addresses
>>>>>>>>
>>>>>>>>     Shahram Davari
>>>>>>>>     Broadcom Corp.
>>>>>>>>     San Jose, CA  95134
>>>>>>>>     USA
>>>>>>>>
>>>>>>>>     Email: davari@broadcom.com
>>>>>>>>
>>>>>>>>
>>>>>>>>     Amit Oren
>>>>>>>>     Broadcom Corp.
>>>>>>>>     San Jose, CA  95134
>>>>>>>>     USA
>>>>>>>>
>>>>>>>>     Email: amito@broadcom.com
>>>>>>>>
>>>>>>>>
>>>>>>>>     Manav Bhatia
>>>>>>>>     Alcatel-Lucent
>>>>>>>>     Bangalore,
>>>>>>>>     India
>>>>>>>>
>>>>>>>>     Email: manav.bhatia@alcatel-lucent.com
>>>>>>>>
>>>>>>>>
>>>>>>>>     Peter Roberts
>>>>>>>>     Alcatel-Lucent
>>>>>>>>     Kanata,
>>>>>>>>     Canada
>>>>>>>>
>>>>>>>>     Email: peter.roberts@alcatel-lucent.com
>>>>>>>>
>>>>>>>>
>>>>>>>>     Laurent Montini
>>>>>>>>     Cisco Systems
>>>>>>>>     San Jose CA
>>>>>>>>     USA
>>>>>>>>
>>>>>>>>     Email: lmontini@cisco.com
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 34]
>>>>>>>> Internet-Draft        Transporting Timing over MPLS            June 2013
>>>>>>>>
>>>>>>>>
>>>>>>>>     Luca
>>>>>>>>     Cisco Systems
>>>>>>>>     San Jose CA
>>>>>>>>     USA
>>>>>>>>
>>>>>>>>     Email: lmartini@cisco.com
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Davari, et al.          Expires December 17, 2013              [Page 35]
>>>>>>>> --
>>>>>>>> For corporate legal information go to:
>>>>>>>>
>>>>>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> TICTOC mailing list
>>>>>>>> TICTOC@ietf.org
>>>>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>>>> This e-mail message is intended for the recipient only and contains
>>>>> information which is CONFIDENTIAL and which may be proprietary to ECI
>>>>> Telecom. If you have received this transmission in error, please inform us by e-
>>>>> mail, phone or fax, and then delete the original and all copies thereof.
>>>>>>> _______________________________________________
>>>>>>> mpls mailing list
>>>>>>> mpls@ietf.org
>>>>>>> https://www.ietf.org/mailman/listinfo/mpls
>>>>>> .
>>>>> --
>>>>> For corporate legal information go to:
>>>>>
>>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>>
>>>>> _______________________________________________
>>>>> TICTOC mailing list
>>>>> TICTOC@ietf.org
>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> TICTOC mailing list
>>>>> TICTOC@ietf.org
>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>> This e-mail message is intended for the recipient only and contains information which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If you have received this transmission in error, please inform us by e-mail, phone or fax, and then delete the original and all copies thereof.
>>> .
>>
>> --
>> For corporate legal information go to:
>>
>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>
>> This e-mail message is intended for the recipient only and contains information which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If you have received this transmission in error, please inform us by e-mail, phone or fax, and then delete the original and all copies thereof.
>>
> This e-mail message is intended for the recipient only and contains information which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If you have received this transmission in error, please inform us by e-mail, phone or fax, and then delete the original and all copies thereof.
>
> .
>


-- 
For corporate legal information go to:

http://www.cisco.com/web/about/doing_business/legal/cri/index.html


From davarish@yahoo.com  Wed Aug 21 07:19:15 2013
Return-Path: <davarish@yahoo.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CDA3811E80F1 for <tictoc@ietfa.amsl.com>; Wed, 21 Aug 2013 07:19:15 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.203
X-Spam-Level: 
X-Spam-Status: No, score=-1.203 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, MIME_QP_LONG_LINE=1.396]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id a5iMT25QX86Y for <tictoc@ietfa.amsl.com>; Wed, 21 Aug 2013 07:19:10 -0700 (PDT)
Received: from nm15-vm6.bullet.mail.gq1.yahoo.com (nm15-vm6.bullet.mail.gq1.yahoo.com [98.137.176.78]) by ietfa.amsl.com (Postfix) with ESMTP id AF0EE11E8105 for <tictoc@ietf.org>; Wed, 21 Aug 2013 07:19:10 -0700 (PDT)
Received: from [98.137.12.58] by nm15.bullet.mail.gq1.yahoo.com with NNFMP; 21 Aug 2013 14:19:09 -0000
Received: from [208.71.42.198] by tm3.bullet.mail.gq1.yahoo.com with NNFMP; 21 Aug 2013 14:19:09 -0000
Received: from [127.0.0.1] by smtp209.mail.gq1.yahoo.com with NNFMP; 21 Aug 2013 14:19:09 -0000
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yahoo.com; s=s1024; t=1377094749; bh=+/XGrO2Q9iYkcmgb1SavajMGa3p+IyREdHqZe0OCrw8=; h=X-Yahoo-Newman-Id:X-Yahoo-Newman-Property:X-YMail-OSG:X-Yahoo-SMTP:X-Rocket-Received:References:Mime-Version:In-Reply-To:Content-Type:Content-Transfer-Encoding:Message-Id:Cc:X-Mailer:From:Subject:Date:To; b=uiM0MS1zDN2FtQMy4IpOiwDybLPk+tnDsFpa/UfTX+j0SFDGzvOfEapGJQB39Nws1Of0b9SPfy8zpzdHBRU/TR9ORwSRHWsfoGa/wm0M6uL6fs3cQumVN+C7TjB8YPAMSP3r/LrtzfQElDzBg3inw2XkpbASrGJNurtG3Zb5aeg=
X-Yahoo-Newman-Id: 850883.41072.bm@smtp209.mail.gq1.yahoo.com
X-Yahoo-Newman-Property: ymail-3
X-YMail-OSG: .vQ2kHUVM1lVpRyW5gDOMnXQYbULYfgOP9MGF75ObXXAiHt Zncy26HZL.EmeJvr2j8s_31zyhif6RJYoQLJAMC4rbevGQOs0YpyN1.wi4Hu 0ULC5tMX2MYVfX8iPv01LL4QviubYpLe0CdOyZc9EFWrDv2Nh20WrTHq5FXo 9xw03PH8Et6.cc6Qdd4Dx0XjpVCgcBvI4afMgEMRZIvjEJF1q4y1FJQzKJ8B y_sO.cCp.wMM8dw70GgaAkbltIqDyOBk53y6jQuY4cERRntvet_D935OT9mY Y.0IKyQC_NjZDLOOpGul5lBOgsXwVuheBagwv2_tl35V145yFITn5acSfXon xhudT8jKXOw.vdhRAyGxW0LhxX9Vvf6BLqL7beQMcTiisLorCqgmiV95S6L_ jhP3Og0T.D_3_7xxeBq9hCABqOXCK4MgWJksGpuW24U0.JCSnNPq9YECadqu 0CcnbZ98Yna_pA0xPDiz0eJMoJbvnA2BBsmQ.stOwxNU_Ps2ukM6AJPqci.6 xJ8OSy9gsuDkfRfMWEWWIEy0V3lquwXiFps6GG6PhRqjCaP5zZtK_KW2z3Pd VDCPI_4g-
X-Yahoo-SMTP: ygPrP9CswBCWPbPtKJlJyLY0KMlg
X-Rocket-Received: from [192.168.0.103] (davarish@98.248.42.143 with ) by smtp209.mail.gq1.yahoo.com with SMTP; 21 Aug 2013 07:19:09 -0700 PDT
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com> <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com> <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com> <52139903.8040801@cisco.com> <F9336571731ADE42A5397FC831CEAA021513A083@ILPTWPVEXMB01.ecitele.com> <B7A656C4-6D0E-4719-9E42-57689CDAEFA9@yahoo.com> <F9336571731ADE42A5397FC831CEAA021513A0C6@ILPTWPVEXMB01.ecitele.com> <52149A2D.3050109@cisco.com>
Mime-Version: 1.0 (1.0)
In-Reply-To: <52149A2D.3050109@cisco.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: quoted-printable
Message-Id: <5ADB2697-BC53-427A-82DB-013B89029343@yahoo.com>
X-Mailer: iPhone Mail (10A403)
From: "S. Davari" <davarish@yahoo.com>
Date: Wed, 21 Aug 2013 07:19:02 -0700
To: "stbryant@cisco.com" <stbryant@cisco.com>
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 21 Aug 2013 14:19:15 -0000

Hi Stewart,

Your proposal does tunneling between timing aware nodes but I guess it stitc=
hes multiple of those tunnels to create and end-2-end connection. Right?

Sasha's BoS proposal seems simpler, since it requires a single end-2-end LSP=
.=20

Regards,
Shahram


On Aug 21, 2013, at 3:45 AM, Stewart Bryant <stbryant@cisco.com> wrote:

> Sasha
>=20
> Note that if the next hop understands the alert, you can put the alert lab=
el at the ToS.
>=20
> If it does not understand, you would want to tunnel through it and hence t=
here would be a label on top or the alert label.
>=20
> Shahram, there seems to be some confusion about stitching vs hierarchy, an=
d I have been proposing hierarchy to step over the nodes that are unable to u=
nderstand the alert label.
>=20
> Stewart
>=20
>=20
>=20
> On 20/08/2013 19:51, Alexander Vainshtein wrote:
>> Shahram,
>> No I am not proposing anything of the kind.
>>=20
>> At the MPLS data plane level the only the head-end and the tail-end LERs h=
ave to be aware of the BoS Timing Alert labels. Transit LSRs simply will ign=
ore them.
>>=20
>> At the on-path assistance level each transit LSR (and the tail-end LER) c=
an do one of the following:
>> - It may ignore the Timing Alert label at the BoS and just forward the pa=
cket based the labels at the top of the stack.
>> - It may identify the packet as a timing one based on the Timing Alert la=
bel at the bottom of the label stack, and perform the required operations. T=
he result of these operations will be stored in the "timing shim header" and=
 not in the packet itself. (E.g., an E2E TC clock would note the packet resi=
dence time and add it to the value already carried in the shim header).
>>=20
>> Non-timing packets would not have the Timing Alert label at the BoS and h=
ence would travel thru he LSP travel along the LSP in the usual way.
>>=20
>> The tail-end LER would take the reception timestamp and forward this time=
stamp, the protocol packet and the shim header to the entity that processes t=
he appropriate timing distribution protocol based on the Timing Alert label.=

>>=20
>> Hopefully this clarifies my proposal. My gut feeling is that similar idea=
s have been proposed by many people earlier...
>>=20
>> Regards,
>>      Sasha
>>=20
>>=20
>> ________________________________________
>> From: S. Davari [davarish@yahoo.com]
>> Sent: Tuesday, August 20, 2013 7:00 PM
>> To: Alexander Vainshtein
>> Cc: stbryant@cisco.com; Shahram Davari; mpls@ietf.org; tictoc@ietf.org
>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>=20
>> Sasha
>>=20
>> In (2) are you proposing to create multiple segment LSPs, where each if t=
hose segment LSPs span between two Timing aware LSR? So if I have 10 timing a=
ware LSR in the path I would need 9 LSPs?
>>=20
>> Regards,
>> Shahram
>>=20
>>=20
>> On Aug 20, 2013, at 9:53 AM, Alexander Vainshtein <Alexander.Vainshtein@e=
citele.com> wrote:
>>=20
>>> Stewart, Shahram and all,
>>>=20
>>> A couple of (hopefully, last) comments:
>>>=20
>>> 1. Supporting a "Timing Alert" label at the top of the label stack intro=
duces a backward compatibility problem. We differ in our estimate of the imp=
ortance of this problem, but we all agree that it exists.
>>>=20
>>> 2. On the other hand, this problem does not exist if the "Timing Alert" l=
abel is inserted at the bottom of the label stack. At the same time, HW that=
 analyzes the label stack hopefully could analyze the bottom label (almost) a=
s easily as it would analyze the top label. Thus we could both provide backw=
ard compatibility (incompatible transit LSRs simply would not notice the "Ti=
ming Alert" label") and avoid the need for dedicated "Timing-only" LSPs. In p=
articular, all the MPLS OAM tools (including MPLS-TP) would work without any=
 changes.
>>>=20
>>> 3. A "timing shim header" (between the bottom of the label stack and the=
 beginning of the timing  packet) has been mentioned by Stewart as an alread=
y proposed mechanism. IMHO and FWIW its usage would eliminate the need to un=
derstand format of the specific timing distribution protocol in the transit L=
SRs.  It would further solve the problem of secure transport of timing packe=
ts since security mechanisms would be applied to its original encapsulation (=
over IP or over Ethernet) without involving the transit LSRs.
>>>=20
>>> Did I miss something substantial?
>>>=20
>>> Regards, and lots of thanks in advance,
>>>     Sasha
>>>=20
>>>=20
>>> ________________________________________
>>> From: Stewart Bryant [stbryant@cisco.com]
>>> Sent: Tuesday, August 20, 2013 6:27 PM
>>> To: S. Davari
>>> Cc: Alexander Vainshtein; Shahram Davari; mpls@ietf.org; tictoc@ietf.org=

>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
>>>=20
>>> On 06/08/2013 14:35, S. Davari wrote:
>>>> Sasha,
>>>>=20
>>>> If a router is not 1588 aware we want it to switch the packet with mini=
mum jitter and delay via high priority queue.
>>> Yes, but you can explicitly tunnel across such island of incompatibility=

>>>=20
>>> Stewart
>>>=20
>>>>=20
>>>> Sending it to CPU creates large jitter and delay since the CPUs are gen=
erally busy doing other things. Time stamping or updating CF in the CPU won'=
t help since it does not cover the variable delay that is incurred from the t=
ime the packet enters the router to the time the packet gets processed by CP=
U.
>>>>=20
>>>> I wish it was that easy!
>>>>=20
>>>> Regards,
>>>> Shahram
>>>>=20
>>>>=20
>>>> On Aug 6, 2013, at 12:38 AM, Alexander Vainshtein <Alexander.Vainshtein=
@ecitele.com> wrote:
>>>>=20
>>>>> Shahram,
>>>>> I agree with you that the routers that do not recognize a Timing Alert=
 label would send the packet to the CPU. (This would be easy to arrange sinc=
e we are speaking about a reserved label.)
>>>>>=20
>>>>> I do not, however, see this as a serious issue, because the SW running=
 on this CPU would (hopefully) be upgradable to handle the packet correctly i=
.e., to forward it to where it should be forwarded and, if it is forwarded a=
s a labeled packet, to prepend the Timing Alert Label on top of its label st=
ack. It could even record the residence time (as observed by the CPU in the p=
roper place in the packet. This would mean that introducing additional error=
 to whatever timing information is associated with this packet - but this is=
 what you should anyway expect if there are non-compliant routers on your pa=
th, right?
>>>>>=20
>>>>> Regards,
>>>>>     Sasha
>>>>>=20
>>>>>> -----Original Message-----
>>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Beh=
alf Of
>>>>>> Shahram Davari
>>>>>> Sent: Monday, August 05, 2013 11:29 PM
>>>>>> To: stbryant@cisco.com; S. Davari
>>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmp=
ls
>>>>>>=20
>>>>>> Hi Stewart,
>>>>>>=20
>>>>>> Your suggestion of " I suggested an LSP type that has the properties "=
timestamp
>>>>>> and pass to application", is a subset of the existing draft and shoul=
d work. It
>>>>>> basically limits the time stamping and correction field update to LER=
s (while the
>>>>>> draft supports time stamping at LER and LSR).
>>>>>>=20
>>>>>> However Sasha's suggestion of using a reserved label (RAL, GAL or any=
 other
>>>>>> reserved label) does not satisfy one of the major requirements, which=
 is
>>>>>> backward compatibility. Routers that don't understand this reserved l=
abel will
>>>>>> drop or copy to CPU such packets.
>>>>>>=20
>>>>>> Regards,
>>>>>> Shahram
>>>>>>=20
>>>>>> -----Original Message-----
>>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On Beh=
alf Of
>>>>>> Stewart Bryant
>>>>>> Sent: Monday, August 05, 2013 2:29 AM
>>>>>> To: S. Davari
>>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>>> Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmp=
ls
>>>>>>=20
>>>>>> My concern is that we are creating a new LSP type in MPLS
>>>>>> (which is a architectural change and thus needs to be very carefully
>>>>>> considered) without asking the question "how can I do this
>>>>>> in the most general way, to maximize flexibility/reuse"
>>>>>>=20
>>>>>> I suggested an LSP type that has the properties "timestamp
>>>>>> and pass to application", but I think that Sasha raises a good
>>>>>> point about using router alert. However RA has no implicit
>>>>>> timestamp, so maybe we need a new type of RA that has the
>>>>>> properties "timestamp, and pass top application indicated by
>>>>>> the next label". That would be quite useful in a number
>>>>>> of OAM applications. There is possibly some GAL variant
>>>>>> of that design that should also be considered.
>>>>>>=20
>>>>>> The application could worry about the path and where it
>>>>>> was necessary to skip some hops a hierarchical LSP would
>>>>>> accomplish that, with the specific benefit that the application
>>>>>> would consciously do this, and may be able to apply some
>>>>>> form of compensation within the network.
>>>>>>=20
>>>>>> - Stewart
>>>>>>=20
>>>>>> On 04/08/2013 10:40, S. Davari wrote:
>>>>>>> Hi Sasha
>>>>>>>=20
>>>>>>> Perhaps you have not understood the draft well. The main reason that=
 the
>>>>>> draft chose to use a dedicated LSP that is signaled via RSVPTE indica=
ting it is a
>>>>>> timing LSP is to be backward compatible with non-1588 nodes. Such nod=
es
>>>>>> simply just switch the packet.
>>>>>>> Your proposal had been considered and was rejected since it was not
>>>>>> backward compatible. Nodes receiving alert label or TTL =3D 1 send th=
e packet to
>>>>>> CPU. You can refer to the meeting notes.
>>>>>>> Regards,
>>>>>>> Shahram
>>>>>>>=20
>>>>>>>=20
>>>>>>> On Aug 4, 2013, at 1:21 AM, Alexander Vainshtein
>>>>>> <Alexander.Vainshtein@ecitele.com> wrote:
>>>>>>>> Stewart and all,
>>>>>>>> I concur with Stewart's statement that the draft does not define "f=
ull
>>>>>> interaction with MPLS architecture".
>>>>>>>> E.g., one of the objectives of the draft is to provide a technique t=
hat would
>>>>>> be backward-compatible with old LSRs that cannot provide on-path supp=
ort for
>>>>>> timing distribution, while the other objective is to make every LSR o=
n the path
>>>>>> aware that some MPLS packets are carrying timing-related messages and=
 hence
>>>>>> require on-path support.
>>>>>>>> IMHO and FWIW, the MPLS data plane architecture allows just two met=
hods
>>>>>> for making a transit LSR to provide special processing to a labeled p=
acket:
>>>>>>>> - It would carry some kind of an "alert label" on top of the label s=
tack and,
>>>>>> specifically, on top of any labels used for actual forwarding
>>>>>>>> OR,
>>>>>>>> - The TTL in the top label stack entry has been set to 1.
>>>>>>>>=20
>>>>>>>>=20
>>>>>>>> The draft does not follow any of these approaches.
>>>>>>>>=20
>>>>>>>> I must also admit that Section 12 "OAM, Control and Management" of t=
he
>>>>>> draft looks somewhat in
>>>>>>>> E.g., I could not understand from the text whether BFD, LSP-Ping et=
c. OAM
>>>>>> messages would be subjected to on-path support procedures for timing
>>>>>> messages or not.
>>>>>>>> My 2c,
>>>>>>>>     Sasha
>>>>>>>>=20
>>>>>>>>> -----Original Message-----
>>>>>>>>> From: tictoc-bounces@ietf.org [mailto:tictoc-bounces@ietf.org] On B=
ehalf
>>>>>> Of
>>>>>>>>> Stewart Bryant
>>>>>>>>> Sent: Friday, August 02, 2013 12:01 PM
>>>>>>>>> To: mpls-chairs@tools.ietf.org; tictoc-chairs@tools.ietf.org; int-=

>>>>>>>>> ads@tools.ietf.org; rtg-ads@tools.ietf.org; mpls-chairs@tools.ietf=
.org;
>>>>>> draft-
>>>>>>>>> ietf-tictoc-1588overmpls@tools.ietf.org
>>>>>>>>> Cc: mpls@ietf.org; tictoc@ietf.org
>>>>>>>>> Subject: [TICTOC] Comments on draft-ietf-tictoc-1588overmpls
>>>>>>>>>=20
>>>>>>>>> SB> This draft does not seem to provide a precise definition
>>>>>>>>> SB> the properties of the new LSP type that it wishes to
>>>>>>>>> SB> define, in particular it does it define the PHB of those
>>>>>>>>> SB> LSPs, nor the full interaction with the MPLS
>>>>>>>>> SB> architecture.
>>>>>>>>> SB>
>>>>>>>>> SB> I have not tracked TICTOC for a while but I thought that
>>>>>>>>> SB> the original plan was to define the concept of an offset
>>>>>>>>> SB> into a packet to do the correction.
>>>>>>>>> SB>
>>>>>>>>> SB> It is disappointing that the opportunity was not taken
>>>>>>>>> SB> to define a timing shim inside the timing LSP so that
>>>>>>>>> SB> a time correction could be added to any packet such that
>>>>>>>>> SB> the MPLS system was isolated from the details of the
>>>>>>>>> SB> complexity of the particular time transfer type.
>>>>>>>>>=20
>>>>>>>>> SB> I think that much more clarify is needed in terms of
>>>>>>>>> SB> definition of the new LSP type, since it is unclear
>>>>>>>>> SB> from this text how to implement one.
>>>>>>>>> SB>
>>>>>>>>> SB> There are a lot of other MPLS services such as
>>>>>>>>> SB> LSP ping that need to be considered.
>>>>>>>>> SB>
>>>>>>>>> SB> Please see inline for more comments. However these
>>>>>>>>> SB> comments are made in the context of the text as written
>>>>>>>>> SB> whilst I have a fundamental concern that this approach
>>>>>>>>> SB> lacks an MPLS architectural soundness that need
>>>>>>>>> SB> greater thought with significant impact on the
>>>>>>>>> SB> draft.
>>>>>>>>>=20
>>>>>>>>> - Stewart
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> TICTOC Working Group                                           S. D=
avari
>>>>>>>>> Internet-Draft                                                   A=
. Oren
>>>>>>>>> Intended status: Standards Track                          Broadcom=
 Corp.
>>>>>>>>> Expires: December 17, 2013                                     M. B=
hatia
>>>>>>>>>                                                               P. R=
oberts
>>>>>>>>>                                                           Alcatel-=
Lucent
>>>>>>>>>                                                               L. M=
ontini
>>>>>>>>>                                                               L. M=
artini
>>>>>>>>>                                                            Cisco S=
ystems
>>>>>>>>>                                                            June 15=
, 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>             Transporting Timing messages over MPLS Networks
>>>>>>>>>                    draft-ietf-tictoc-1588overmpls-05
>>>>>>>>>=20
>>>>>>>>> Abstract
>>>>>>>>>=20
>>>>>>>>>    This document defines the method for transporting Timing messag=
es
>>>>>>>>>    such as PTP and NTP over an MPLS network.  The method allows fo=
r the
>>>>>>>>>    easy identification of these PDUs at the port level to allow fo=
r port
>>>>>>>>>=20
>>>>>>>>> SB> What is a port
>>>>>>>>>=20
>>>>>>>>>    level processing of these PDUs in both LERs and LSRs.
>>>>>>>>>=20
>>>>>>>>>    The basic idea is to transport Timing messages inside dedicated=
 MPLS
>>>>>>>>>    LSPs.  These LSPs only carry Timing messages and possibly Contr=
ol and
>>>>>>>>>    Management packets, but they do not carry customer traffic.
>>>>>>>>>=20
>>>>>>>>> SB> More specifically they only carry traffic associated with the
>>>>>>>>> SB> timing service and its support.
>>>>>>>>> SB> The question arose WRT BFD - surely BFD is allowed, BUT surely=

>>>>>>>>> SB> also it gets carried in a structure that causes it to get
>>>>>>>>> SB> timestamped.
>>>>>>>>>=20
>>>>>>>>>    Two methods for transporting Timing messages over MPLS are defi=
ned.
>>>>>>>>>=20
>>>>>>>>> SB> Perhaps the right approach is to define the new LSP type and t=
hen
>>>>>>>>> SB> seperately to define  the mapping of the various timing servic=
es
>>>>>>>>> SB> over that LSP type.
>>>>>>>>>=20
>>>>>>>>>    The first method is to transport Timing messages directly over t=
he
>>>>>>>>>    dedicated MPLS LSP via UDP/IP encapsulation, which is suitable f=
or
>>>>>>>>>    MPLS networks.  The second method is to transport Timing messag=
es
>>>>>>>>>    inside a PW via Ethernet encapsulation.
>>>>>>>>>=20
>>>>>>>>> SB> I think that we should note that there are some
>>>>>>>>> SB> h/w reasons for this preference. A clean sheet approach
>>>>>>>>> SB> would have been to use PTP over MPLS with no intermediate
>>>>>>>>> SB> layers.
>>>>>>>>>=20
>>>>>>>>> Status of this Memo
>>>>>>>>>=20
>>>>>>>>>    This Internet-Draft is submitted in full conformance with the
>>>>>>>>>    provisions of BCP 78 and BCP 79.
>>>>>>>>>=20
>>>>>>>>>    Internet-Drafts are working documents of the Internet Engineeri=
ng
>>>>>>>>>    Task Force (IETF).  Note that other groups may also distribute
>>>>>>>>>    working documents as Internet-Drafts.  The list of current Inte=
rnet-
>>>>>>>>>    Drafts is at http://datatracker.ietf.org/drafts/current/.
>>>>>>>>>=20
>>>>>>>>>    Internet-Drafts are draft documents valid for a maximum of six m=
onths
>>>>>>>>>    and may be updated, replaced, or obsoleted by other documents a=
t any
>>>>>>>>>    time.  It is inappropriate to use Internet-Drafts as reference
>>>>>>>>>    material or to cite them other than as "work in progress."
>>>>>>>>>=20
>>>>>>>>>    This Internet-Draft will expire on December 17, 2013.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013               [P=
age 1]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Copyright Notice
>>>>>>>>>=20
>>>>>>>>>    Copyright (c) 2013 IETF Trust and the persons identified as the=

>>>>>>>>>    document authors.  All rights reserved.
>>>>>>>>>=20
>>>>>>>>>    This document is subject to BCP 78 and the IETF Trust's Legal
>>>>>>>>>    Provisions Relating to IETF Documents
>>>>>>>>>    (http://trustee.ietf.org/license-info) in effect on the date of=

>>>>>>>>>    publication of this document.  Please review these documents
>>>>>>>>>    carefully, as they describe your rights and restrictions with r=
espect
>>>>>>>>>    to this document.  Code Components extracted from this document=
 must
>>>>>>>>>    include Simplified BSD License text as described in Section 4.e=
 of
>>>>>>>>>    the Trust Legal Provisions and are provided without warranty as=

>>>>>>>>>    described in the Simplified BSD License.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Table of Contents
>>>>>>>>>=20
>>>>>>>>>    1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . .=
 .  5
>>>>>>>>>=20
>>>>>>>>>    2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . .=
 .  7
>>>>>>>>>=20
>>>>>>>>>    3.  Problem Statement  . . . . . . . . . . . . . . . . . . . . .=
 .  8
>>>>>>>>>=20
>>>>>>>>>    4.  Timing over MPLS Architecture  . . . . . . . . . . . . . . .=
 .  9
>>>>>>>>>=20
>>>>>>>>>    5.  Dedicated LSPs for Timing messages . . . . . . . . . . . . .=
 . 12
>>>>>>>>>=20
>>>>>>>>>    6.  Timing over LSP Encapsulation  . . . . . . . . . . . . . . .=
 . 13
>>>>>>>>>      6.1.  Timing over UDP/IP over MPLS Encapsulation . . . . . . .=
 . 13
>>>>>>>>>      6.2.  Timing over PW Encapsulation . . . . . . . . . . . . . .=
 . 13
>>>>>>>>>      6.3.  Other Timing Encapsulation methods . . . . . . . . . . .=
 . 14
>>>>>>>>>=20
>>>>>>>>>    7.  Timing message Processing  . . . . . . . . . . . . . . . . .=
 . 15
>>>>>>>>>=20
>>>>>>>>>    8.  Protection and Redundancy  . . . . . . . . . . . . . . . . .=
 . 16
>>>>>>>>>=20
>>>>>>>>>    9.  ECMP . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 17
>>>>>>>>>=20
>>>>>>>>>    10. PHP  . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 18
>>>>>>>>>=20
>>>>>>>>>    11. Entropy  . . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 19
>>>>>>>>>=20
>>>>>>>>>    12. OAM, Control and Management  . . . . . . . . . . . . . . . .=
 . 20
>>>>>>>>>=20
>>>>>>>>>    13. QoS Considerations . . . . . . . . . . . . . . . . . . . . .=
 . 21
>>>>>>>>>=20
>>>>>>>>>    14. FCS and Checksum Recalculation . . . . . . . . . . . . . . .=
 . 22
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013               [P=
age 2]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    15. Behavior of LER/LSR  . . . . . . . . . . . . . . . . . . . .=
 . 23
>>>>>>>>>      15.1. Behavior of Timing-capable/aware LER . . . . . . . . . .=
 . 23
>>>>>>>>>      15.2. Behavior of Timing-capable/aware LSR . . . . . . . . . .=
 . 23
>>>>>>>>>      15.3. Behavior of non-Timing-capable/aware LSR . . . . . . . .=
 . 24
>>>>>>>>>=20
>>>>>>>>>    16. Other considerations . . . . . . . . . . . . . . . . . . . .=
 . 25
>>>>>>>>>=20
>>>>>>>>>    17. Security Considerations  . . . . . . . . . . . . . . . . . .=
 . 26
>>>>>>>>>=20
>>>>>>>>>    18. Acknowledgements . . . . . . . . . . . . . . . . . . . . . .=
 . 27
>>>>>>>>>=20
>>>>>>>>>    19. IANA Considerations  . . . . . . . . . . . . . . . . . . . .=
 . 28
>>>>>>>>>=20
>>>>>>>>>    20. References . . . . . . . . . . . . . . . . . . . . . . . . .=
 . 29
>>>>>>>>>      20.1. Normative References . . . . . . . . . . . . . . . . . .=
 . 29
>>>>>>>>>      20.2. Informative References . . . . . . . . . . . . . . . . .=
 . 29
>>>>>>>>>=20
>>>>>>>>>    Appendix 1.  Routing extensions for Timing-aware Routers . . . .=
 . 32
>>>>>>>>>=20
>>>>>>>>>    Appendix 2.  Signaling Extensions for Creating Timing LSPs . . .=
 . 33
>>>>>>>>>=20
>>>>>>>>>    Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . .=
 . 34
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013               [P=
age 3]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
>>>>>> NOT",
>>>>>>>>>    "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL"
>>>>>> in
>>>>>>>>> this
>>>>>>>>>    document are to be interpreted as described in RFC2119 [RFC2119=
].
>>>>>>>>>=20
>>>>>>>>>    When used in lower case, these words convey their typical use i=
n
>>>>>>>>>    common language, and are not to be interpreted as described in
>>>>>>>>>    RFC2119 [RFC2119].
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013               [P=
age 4]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 1.  Introduction
>>>>>>>>>=20
>>>>>>>>>    The objective of Precision Time Protocol (PTP) and Network Timi=
ng
>>>>>>>>>    Protocol (NTP) are to synchronize independent clocks running on=

>>>>>>>>>    separate nodes of a distributed system.
>>>>>>>>>=20
>>>>>>>>>    [IEEE-1588] defines PTP messages for frequency, phase and time
>>>>>>>>>    synchronization.  The PTP messages include PTP PDUs over UDP/IP=

>>>>>>>>>    (Annex D and E of [IEEE-1588]) and PTP PDUs over Ethernet (Anne=
x F of
>>>>>>>>>    [IEEE-1588]).
>>>>>>>>>=20
>>>>>>>>> SB> Sure BUT it is acknowledged that IEEE is open to the definitio=
n
>>>>>>>>> SB> of other PTP mappings if they provide better optimisation.
>>>>>>>>>=20
>>>>>>>>>    This document defines mapping and transport of the PTP
>>>>>>>>>    messages defined in [IEEE-1588] over MPLS/MPLS-TP networks.  PT=
P
>>>>>>>>>    defines several clock types: ordinary clocks, boundary clocks, e=
nd-
>>>>>>>>>    to-end transparent clocks, and peer-to-peer transparent clocks.=

>>>>>>>>>    Transparent clocks require intermediate nodes to update correct=
ion
>>>>>>>>>    field inside PTP message that reflects the transit time in the n=
ode.
>>>>>>>>>=20
>>>>>>>>>    [RFC5905] defines NTP messages for clock and time synchronizati=
on.
>>>>>>>>>    The PTP messages (PDUs) are transported over UDP/IP.  This docu=
ment
>>>>>>>>> SB> Should that be NTP messages?
>>>>>>>>> SB> It needs to be made clear as soon as you introduce NTP that
>>>>>>>>> SB> they use different time representations.
>>>>>>>>>=20
>>>>>>>>>    defines mapping and transport of the NTP messages defined in
>>>>>>>>>    [RFC5905] over MPLS networks.
>>>>>>>>>=20
>>>>>>>>>    One key attribute of all of these Timing messages is that the T=
ime
>>>>>>>>>    stamp processing should occur as close as possible to the actua=
l
>>>>>>>>>    transmission and reception at the physical port interface.  Thi=
s
>>>>>>>>>    targets optimal time and/or frequency recovery by avoiding vari=
able
>>>>>>>>>    delay introduced by queues internal to the clocks.
>>>>>>>>>=20
>>>>>>>>> SB> As I recall NTP has no epoch point defined, and I am not sure
>>>>>>>>> SB> where that point is in the case of PTP in this mapping
>>>>>>>>> SB> Hopefully this will get defined in due course.
>>>>>>>>>=20
>>>>>>>>>    To facilitate the fast and efficient recognition of Timing mess=
ages
>>>>>>>>>    at the port level when the Timing messages are carried over MPL=
S
>>>>>>>>>    LSPs,
>>>>>>>>>=20
>>>>>>>>> SB> Over a new LSP type with time optimied characteristics
>>>>>>>>>=20
>>>>>>>>>    this document defines the specific encapsulations that should
>>>>>>>>>    be used.
>>>>>>>>> SB> Hopefully it will also define the PHP
>>>>>>>>>=20
>>>>>>>>>    In addition, it can be expected that there will exist LSR/
>>>>>>>>>    LERs where only a subset of the physical ports will have the po=
rt-
>>>>>>>>>    based Timing message processing capabilities.
>>>>>>>>> SB> Do you need to clarify that this only works at base and not in=

>>>>>>>>> SB> a label heirarchy.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    In order to ensure
>>>>>>>>>    that the LSPs carrying Timing packets always enter and exit por=
ts
>>>>>>>>>    with this capability, routing extensions are defined to adverti=
se
>>>>>>>>>    this capability on a port basis and to allow for the establishm=
ent of
>>>>>>>>>    LSPs that only transit such ports.  While this path establishme=
nt
>>>>>>>>>    restriction may be applied only at the LER Ingress and/or egres=
s
>>>>>>>>>    ports, it becomes more important when using transparent clock c=
apable
>>>>>>>>>    LSRs in the path.
>>>>>>>>> SB> I do not understand the implications of the last
>>>>>>>>> SB> sentences - starting ", it becomes"
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    Port based Timing message processing involves Timing message
>>>>>>>>>    recognition.  Once the Timing messages are recognized they can b=
e
>>>>>>>>>    modified based on the reception or transmission Time-stamp.
>>>>>>>>>=20
>>>>>>>>>    This document provides two methods for transporting Timing mess=
ages
>>>>>>>>>    over MPLS.  One is applicable to MPLS environment and the other=
 one
>>>>>>>>>    is applicable to MPLS/MPLS-TP environment
>>>>>>>>>=20
>>>>>>>>> SB> I think the sentence is incomplete.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013               [P=
age 5]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    The solution involves transporting Timing messages over dedicat=
ed
>>>>>>>>>    LSPs called Timing LSPs.  These LSPs carry Timing messages and M=
AY
>>>>>>>>>    carry Management and control messages, but not data plane clien=
t
>>>>>>>>>    traffic.
>>>>>>>>>=20
>>>>>>>>> SB> It is not clear why this restriction applies.
>>>>>>>>>=20
>>>>>>>>>    Timing LSPs can be established statically or via signaling.
>>>>>>>>> SB> s/statically/by provisioning/network management/
>>>>>>>>>=20
>>>>>>>>>    Extensions to control plane (OSPF, ISIS, etc.) is required to e=
nable
>>>>>>>>>    routers to distribute their Timing processing capabilities over=
 MPLS
>>>>>>>>>    to other routers.  However such extensions are outside the scop=
e of
>>>>>>>>>    this document.
>>>>>>>>>=20
>>>>>>>>>    When signaling is used to setup the PTP LSP, Extensions to sign=
aling
>>>>>>>>> SB> is it a PTP LSP or a Timing LSP?
>>>>>>>>>=20
>>>>>>>>>    protocols (e.g., RSVP-TE) are required for establishing PTP LSP=
s.
>>>>>>>>>    However such extensions are outside the scope of this document.=

>>>>>>>>>=20
>>>>>>>>> SB> for mpls-tp GMPLS is the signalling protocol
>>>>>>>>>=20
>>>>>>>>>    While the techniques included herein allow for the establishmen=
t of
>>>>>>>>>    paths optimized to include Time-stamping capable links, the
>>>>>>>>>    performance of the Slave clocks is outside the scope of this
>>>>>>>>>    document.
>>>>>>>>>=20
>>>>>>>>>    At the time of publishing this specification, Transparent Clock=
ing
>>>>>>>>>    (TC) is only defined for PTP.  Therefore at this time any part o=
f
>>>>>>>>>    this specification that talks about Transparent Clocking applie=
s only
>>>>>>>>>    to PTP.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013               [P=
age 6]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 2.  Terminology
>>>>>>>>>=20
>>>>>>>>>    1588: The timing and synchronization as defined by IEEE 1588.
>>>>>>>>>=20
>>>>>>>>> SB> I think that there is a more formal name for the 1588 group
>>>>>>>>> SB> that needs to be used here.
>>>>>>>>> SB> Also do we need to talk about 1588-200? as there is
>>>>>>>>> SB> an update in progress
>>>>>>>>>=20
>>>>>>>>>    NTP: The timing and synchronization protocol defined by IETF RFC=
-1305
>>>>>>>>>    and RFC-5905.
>>>>>>>>>=20
>>>>>>>>>    PTP: The timing and synchronization protocol used by 1588.
>>>>>>>>> SB> need the proper name for 1588
>>>>>>>>>=20
>>>>>>>>>    Master Clock: The source of 1588 timing to a set of slave clock=
s.
>>>>>>>>>=20
>>>>>>>>>    Master Port: A port on a ordinary or boundary clock that is in M=
aster
>>>>>>>>>    state.  This is the source of timing toward slave ports.
>>>>>>>>>=20
>>>>>>>>> SB> I am not sure the reader knows what a port is
>>>>>>>>>=20
>>>>>>>>>    Slave Clock: A receiver of 1588 timing from a master clock.
>>>>>>>>>=20
>>>>>>>>>    Slave Port: A port on a boundary clock or ordinary clock that i=
s
>>>>>>>>>    receiving timing from a master clock.
>>>>>>>>>=20
>>>>>>>>>    Ordinary Clock: A device with a single PTP port.
>>>>>>>>>=20
>>>>>>>>>    Transparent Clock.  A device that measures the time taken for a=
 PTP
>>>>>>>>>    event message to transit the device and then updates the
>>>>>>>>>    correctionField of the message with this transit time.
>>>>>>>>>=20
>>>>>>>>>    Boundary Clock: A device with more than one PTP port.  Generall=
y
>>>>>>>>>    boundary clocks will have one port in slave state to receive ti=
ming
>>>>>>>>>    and then other ports in master state to re-distribute the timin=
g.
>>>>>>>>>=20
>>>>>>>>>    PTP LSP: An LSP dedicated to carry PTP messages
>>>>>>>>>=20
>>>>>>>>> SB> PTP or timing?
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    PTP PW: A PW within a PTP LSP that is dedicated to carry PTP
>>>>>>>>>    messages.
>>>>>>>>>=20
>>>>>>>>> SB> Ah I don't think that PWE3 know what one of these is
>>>>>>>>>=20
>>>>>>>>>    CW: Pseudowire Control Word
>>>>>>>>>=20
>>>>>>>>>    LAG: Link Aggregation
>>>>>>>>>=20
>>>>>>>>>    ECMP: Equal Cost Multipath
>>>>>>>>>=20
>>>>>>>>>    CF: Correction Field, a field inside certain PTP messages (mess=
age
>>>>>>>>>    type 0-3)that holds the accumulative transit time inside interm=
ediate
>>>>>>>>>    switches
>>>>>>>>>=20
>>>>>>>>>    Timing messages: Timing Protocol messages that are exchanged
>>>>>> between
>>>>>>>>>    routers in order to establish a synchronized clock.
>>>>>>>>>=20
>>>>>>>>> SB> A number of these definitions look like copies of IEEE1588
>>>>>>>>> SB> definitions. We need to provide references and note the
>>>>>>>>> SB> priority of the IEEE base reference.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013               [P=
age 7]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 3.  Problem Statement
>>>>>>>>>=20
>>>>>>>>>    [IEEE-1588] has defined methods for transporting PTP messages o=
ver
>>>>>>>>>    Ethernet and IP networks.  [RFC5905] has defined the method of
>>>>>>>>>    transporting NTP messages over IP networks.  There is a need to=

>>>>>>>>>    transport Timing messages over MPLS networks while supporting t=
he
>>>>>>>>>    Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock (=
OC)
>>>>>>>>>    functionality in the LER and LSRs in the MPLS network.
>>>>>>>>>=20
>>>>>>>>>    There are multiple ways of transporting Timing over MPLS.  Howe=
ver,
>>>>>>>>>    there is a requirement to limit the possible encapsulation opti=
ons to
>>>>>>>>>    simplify the Timing message identification and processing requi=
red at
>>>>>>>>>    the port level.
>>>>>>>>>=20
>>>>>>>>>    When Timing-awareness is needed, Timing messages should not be
>>>>>>>>>    transported over LSPs or PWs that are carrying customer traffic=

>>>>>>>>>    because LSRs perform Label switching based on the top label in t=
he
>>>>>>>>>    stack.
>>>>>>>>>=20
>>>>>>>>> SB> Have you explained why?
>>>>>>>>>=20
>>>>>>>>>    To detect Timing messages inside such LSPs require special
>>>>>>>>>    hardware to do deep packet inspection at line rate.  Even if su=
ch
>>>>>>>>>    hardware exists, the payload can't be deterministically identif=
ied by
>>>>>>>>>    LSRs because the payload type is a context of the PW label, and=
 the
>>>>>>>>>    PW label and its context are only known to the Edge routers (PE=
s/
>>>>>>>>>    LERs); LSRs dont know what is a PWs payload (Ethernet, ATM, FR,=
 CES,
>>>>>>>>>    etc).  Even if one restricts an LSP to only carry Ethernet PWs,=
 the
>>>>>>>>>    LSRs dont have the knowledge of whether PW Control Word (CW) is=

>>>>>>>>>    present or not and therefore can not deterministically identify=
 the
>>>>>>>>>    payload.
>>>>>>>>>=20
>>>>>>>>>    A generic method is defined in this document that does not requ=
ire
>>>>>>>>>    deep packet inspection at line rate, and can deterministically
>>>>>>>>>    identify Timing messages.  This method can be used to detect Ti=
ming
>>>>>>>>>    Messages in both one-step and two-step clock implementations of=

>>>>>>>>>    ordinary, boundary and transparent clocks.
>>>>>>>>>=20
>>>>>>>>> SB> Needs a ref and I am sure many MPLS specialists will not under=
stand
>>>>>>>>> SB> the msg types.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013               [P=
age 8]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 4.  Timing over MPLS Architecture
>>>>>>>>>=20
>>>>>>>>>    Timing messages are exchange between Timing ports on ordinary a=
nd
>>>>>>>>>=20
>>>>>>>>> SB> Have you defined a timing port?
>>>>>>>>>=20
>>>>>>>>>    boundary clocks.  Boundary clocks terminate the Timing messages=
 and
>>>>>>>>>    act as master for other boundary clocks or for slave clocks.  E=
nd-to-
>>>>>>>>>    End Transparent clocks do not terminate the Timing messages but=
 they
>>>>>>>>>    do modify the contents of the Timing messages as they transit a=
cross
>>>>>>>>>    the transparent clock.
>>>>>>>>>=20
>>>>>>>>>    Master/Slave clocks (OCs), Boundary Clocks (BC) and Transparent=
 Clock
>>>>>>>>>=20
>>>>>>>>>    (TC) could be implemented in either LERs or LSRs.
>>>>>>>>>=20
>>>>>>>>> SB> LER and LSR need to be expanded
>>>>>>>>>=20
>>>>>>>>>    An example is shown in Figure 1, where the LERs act as Ordinary=
 Clock
>>>>>>>>>    (OC) and are the initiating/terminating point for Timing messag=
es.
>>>>>>>>>    The ingress LER encapsulates the Timing messages in Timing LSP a=
nd
>>>>>>>>>    the Egress LER terminates the Timing LSP.  The LSRs act as
>>>>>>>>>    Transparent Clock (TC) and just update the Timing field in the T=
iming
>>>>>>>>>    messages.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>       +--------+     +-------+     +-------+     +-------+     +--=
------+
>>>>>>>>>       |Switch, |     |       |     |       |     |       |     |Sw=
itch, |
>>>>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| R=
outer |
>>>>>>>>>       |        |     |  OC   |     |  TC   |     |  OC   |     |  =
      |
>>>>>>>>>       +--------+     +-------+     +-------+     +-------+     +--=
------+
>>>>>>>>>                      /                                 \
>>>>>>>>>       +-------+     /                                   \     +---=
----+
>>>>>>>>>       |  LER  |    /                                     \    |  L=
ER  |
>>>>>>>>>       | Master|---/                                       \---| Sl=
ave |
>>>>>>>>>       | Clock |                                               | Cl=
ock |
>>>>>>>>>       +-------+                                               +---=
----+
>>>>>>>>>=20
>>>>>>>>>      Figure (1) - Deployment example 1 of timing over MPLS network=

>>>>>>>>>=20
>>>>>>>>>    Another example is shown in Figure2, where LERs terminate the T=
iming
>>>>>>>>>    messages received from switch/routers that are outside of the M=
PLS
>>>>>>>>>    network acting as OC or BC.  In this example LERs regenerate th=
e
>>>>>>>>>    clock and initiate timing messages encapsulated in Timing LSP t=
oward
>>>>>>>>>    the MPLS network, while the LSRs act as Transparent Clock (TC) a=
nd
>>>>>>>>>    just update the Timing field in the Timing messages, which are
>>>>>>>>>    already encapsulated in Timing LSPs.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013               [P=
age 9]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>      +--------+     +-------+     +-------+     +-------+     +---=
-----+
>>>>>>>>>      |Switch, |     |       |     |       |     |       |     |Swi=
tch, |
>>>>>>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Ro=
uter |
>>>>>>>>>      | OC/BC  |     |  BC   |     |  TC   |     |  BC   |     | OC=
/BC  |
>>>>>>>>>      +--------+     +-------+     +-------+     +-------+     +---=
-----+
>>>>>>>>>=20
>>>>>>>>>      Figure (2) - Deployment example 2 of timing over MPLS network=

>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    Another example is shown in Figure 3, where LERs do not termina=
te the
>>>>>>>>>    Timing messages received from switch/routers that are outside o=
f the
>>>>>>>>>    MPLS network acting as OC, TC or BC.  The LERs act as TC and up=
date
>>>>>>>>>    the Timing field in the Timing messages as they transit the LER=
,
>>>>>>>>>    while encapsulating them in timing LSP.  The LSRs also act as
>>>>>>>>>    Transparent Clock (TC) and just update the Timing field in the T=
iming
>>>>>>>>>    messages which are already encapsulated in Timing LSPs.
>>>>>>>>>=20
>>>>>>>>>       +--------+     +-------+     +-------+     +-------+     +--=
------+
>>>>>>>>>       |Switch, |     |       |     |       |     |       |     |Sw=
itch, |
>>>>>>>>>       | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| R=
outer |
>>>>>>>>>       |OC/TC/BC|     |  TC   |     |  TC   |     |  TC   |     |OC=
/TC/BC|
>>>>>>>>>       +--------+     +-------+     +-------+     +-------+     +--=
------+
>>>>>>>>>=20
>>>>>>>>>     Figure (3) - Deployment example 3 of timing over MPLS network
>>>>>>>>>=20
>>>>>>>>>    Another example is shown in Figure 4, where LERs and LSRs suppo=
rt
>>>>>>>>>    Boundary Clocks.  A single-hop LSP is created between two adjac=
ent
>>>>>>>>>    LSRs engaged in BC operation.  Other methods such as PTP transp=
ort
>>>>>>>>>    over Ethernet MAY be used for transporting timing messages if t=
he
>>>>>>>>>    link between the two routers is Ethernet.
>>>>>>>>>=20
>>>>>>>>>      +--------+     +-------+     +-------+     +-------+     +---=
-----+
>>>>>>>>>      |Switch, |     |       |     |       |     |       |     |Swi=
tch, |
>>>>>>>>>      | Router |-----|  LER  |-----|  LSR  |-----|  LER  |-----| Ro=
uter |
>>>>>>>>>      | OC/BC  |     |  BC   |     |  BC   |     |  BC   |     | OC=
/BC  |
>>>>>>>>>      +--------+     +-------+     +-------+     +-------+     +---=
-----+
>>>>>>>>>=20
>>>>>>>>>    Figure (4) - Deployment example 3 of timing over MPLS network
>>>>>>>>>=20
>>>>>>>>>    An MPLS domain MAY serve multiple customers.  In these cases th=
e
>>>>>> MPLS
>>>>>>>>>    domain (maintained by a service provider) may provide timing se=
rvices
>>>>>>>>>    to multiple customers, each having their own Timing domain.
>>>>>>>>>=20
>>>>>>>>>    The Timing over MPLS architecture assumes full mesh of Timing L=
SPs
>>>>>>>>>    between all LERs supporting this specification.
>>>>>>>>>=20
>>>>>>>>> SB> Note sure this is right - the salves surely do not need to
>>>>>>>>> SB> exchange timing amongst themselves
>>>>>>>>>=20
>>>>>>>>>    It supports
>>>>>>>>>    Point-to- point (VPWS) and Multipoint (VPLS) services.
>>>>>>>>>=20
>>>>>>>>> SB> What does that mean? You do not carry user data traffic?
>>>>>>>>> SB> Maybe it's the ordering of the statemnets that is causing
>>>>>>>>> SB> confusion.
>>>>>>>>>=20
>>>>>>>>>    This means
>>>>>>>>>    that a customer may purchase a Point-to-point Timing service be=
tween
>>>>>>>>>    two customer sites or a Multipoint Timing service between more t=
han
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 10]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    two customer sites.
>>>>>>>>>=20
>>>>>>>>>    The Timing over MPLS architecture supports P2P or P2MP Timing L=
SPs.
>>>>>>>>>    This means that the Timing Multicast messages such as PTP Multi=
cast
>>>>>>>>>    event messages can be transported over P2MP Timing LSP or be
>>>>>>>>>    replicated and transported over many P2P Timing LSPs.
>>>>>>>>>=20
>>>>>>>>> SB> Note we do not yet have a definition of a P2MP mpls-tp LSP
>>>>>>>>> SB> nor a P2MP PW, although we are close.
>>>>>>>>>=20
>>>>>>>>>    Timing messages, that do not require Time stamping or Correctio=
n
>>>>>>>>>    Field update MAY be transported over Timing LSPs to simplify ha=
rdware
>>>>>>>>>    and software.
>>>>>>>>>=20
>>>>>>>>>    PTP Announce messages that determine the Timing LSP terminating=

>>>>>> point
>>>>>>>>>    behavior such as BC/OC/TC SHOULD be transported over the Timing=
 LSP
>>>>>>>>>    to simplify hardware and software.
>>>>>>>>>=20
>>>>>>>>> SB> have you defined and referenced PTP announce msgs?
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 11]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 5.  Dedicated LSPs for Timing messages
>>>>>>>>>=20
>>>>>>>>>    Many methods have been considered for identifying the Timing
>>>>>> messages
>>>>>>>>>    when they are encapsulated in MPLS such as using GAL/G-ACH or a=
 new
>>>>>>>>>    reserved label.  These methods were not attractive since they e=
ither
>>>>>>>>>    required deep packet inspection at line rate in the intermediat=
e LSRs
>>>>>>>>>    or they required use of a scarce new reserved label.  Also one o=
f the
>>>>>>>>>    goals was to reuse existing OAM mechanisms.
>>>>>>>>>=20
>>>>>>>>> SB> RLs =3D SPLs are not so rare now. In any case needs a ref.
>>>>>>>>>=20
>>>>>>>>>    The method defined in this document can be used by LER and LSRs=
 to
>>>>>>>>>    identify Timing messages in MPLS tunnels by just looking at the=
 top
>>>>>>>>>    label in the MPLS label stack, which only carry Timing messages=
 as
>>>>>>>>>    well as OAM, but not data plane client traffic.
>>>>>>>>>=20
>>>>>>>>>    Compliant implementations MUST use dedicated LSPs to carry Timi=
ng
>>>>>>>>>    messages over MPLS.
>>>>>>>>>=20
>>>>>>>>> SB> I think that we need a definition of the properies of these LS=
Ps
>>>>>>>>>=20
>>>>>>>>>    These LSPs are herein referred to as "Timing
>>>>>>>>>    LSPs" and the labels associated with these LSPs as "Timing LSP
>>>>>>>>>    labels".  The Timing LSPs that runs between Ingress and Egress L=
ERs
>>>>>>>>>    MUST be co-routed.  Alternatively, a single bidirectional co-ro=
uted
>>>>>>>>>    LSP can be used.
>>>>>>>>>=20
>>>>>>>>> SB> I though that you said you could use M2MP LSPs - these are not=

>>>>>>>>> SB> bidirectional.
>>>>>>>>>=20
>>>>>>>>>    Co-routing of the two directions is required to limit the diffe=
rence
>>>>>>>>>    in the delays in the Master clock to Slave clock direction comp=
ared
>>>>>>>>>    to the Slave clock to Master clock direction.  The Timing LSP M=
AY be
>>>>>>>>>    MPLS/MPLS-TP LSP.
>>>>>>>>>=20
>>>>>>>>>    The Timing LSPs could be configured or signaled via RSVP-TE/GMP=
LS.
>>>>>>>>>    New Extensions to RSVP-TE/GMPLS TLVs are required; however they=
 are
>>>>>>>>>    outside the scope of this document.
>>>>>>>>>=20
>>>>>>>>>    The Timing LSPs MAY carry essential MPLS/MPLS-TP OAM traffic su=
ch as
>>>>>>>>>    BFD and LSP Ping but the LSP data plane client plane traffic MU=
ST be
>>>>>>>>>    Timing packets only.
>>>>>>>>>=20
>>>>>>>>> SB> Why?
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 12]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 6.  Timing over LSP Encapsulation
>>>>>>>>>=20
>>>>>>>>> The encapsulations is not LSP is it?
>>>>>>>>>=20
>>>>>>>>>    This document defines two methods for carrying Timing messages o=
ver
>>>>>>>>>    MPLS.  The first method is carrying UDP/IP encapsulated Timing
>>>>>>>>>    messages over Timing LSPs, and the second method, is carrying
>>>>>>>>>    Ethernet encapsulated Timing messages over Ethernet PWs inside
>>>>>> Timing
>>>>>>>>>    LSPs.
>>>>>>>>>=20
>>>>>>>>> 6.1.  Timing over UDP/IP over MPLS Encapsulation
>>>>>>>>>=20
>>>>>>>>>    The simplest method of transporting Timing messages over MPLS i=
s to
>>>>>>>>>    encapsulate Timing PDUs in UDP/IP and then encapsulate them in
>>>>>> Timing
>>>>>>>>>    LSP.  This format is shown in Figure 4.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>                     +----------------------+
>>>>>>>>>                     |   Timing LSP Label   |
>>>>>>>>>                     +----------------------+
>>>>>>>>>                     |        IPv4/6        |
>>>>>>>>>                     +----------------------+
>>>>>>>>>                     |         UDP          |
>>>>>>>>>                     +----------------------+
>>>>>>>>>                     |     Timing PDU       |
>>>>>>>>>                     +----------------------+
>>>>>>>>>=20
>>>>>>>>>       Figure (4) - Timing over UDP/IP over MPLS Encapsulation
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    This encapsulation is very simple and is useful when the networ=
k
>>>>>>>>>    between Timing Master Clock and Slave Clock is MPLS network.
>>>>>>>>>=20
>>>>>>>>> SB> Simple is a judgement call
>>>>>>>>>=20
>>>>>>>>>    In order for an LER/LSR to process Timing messages, the Timing L=
SP
>>>>>>>>>    Label must be at the top label of the label stack.  The LER/LSR=
 MUST
>>>>>>>>>    know that the Timing LSP Label is used for carrying Timing mess=
ages.
>>>>>>>>>    This can be accomplished via static configuration or via RSVP-T=
E
>>>>>>>>>    signaling.
>>>>>>>>>=20
>>>>>>>>>    The UDP/IP encapsulation of PTP MUST follow Annex D and E of
>>>>>>>>>    [IEEE-1588].  While the UDP/IP encapsulation of NTP MUST follow=

>>>>>>>>>    [RFC5905].
>>>>>>>>>=20
>>>>>>>>> 6.2.  Timing over PW Encapsulation
>>>>>>>>>=20
>>>>>>>>>    Another method of transporting Timing over MPLS networks is by
>>>>>>>>>    encapsulating Timing PDUs in PW which in turn is transported ov=
er
>>>>>>>>>    Timing LSPs.  In case of PTP, Ethernet PW encapsulation [RFC444=
8],
>>>>>>>>>    shown in Fig 5(A) MUST be used and the Ethernet encapsulation o=
f PTP
>>>>>>>>>    MUST follow Annex F of [IEEE-1588].
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 13]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    The RAW mode or Tagged mode defined in [RFC4448] MAY be used an=
d
>>>>>> the
>>>>>>>>>    Payload MUST have 0, 1, or 2 VLAN tags (S-VLAN and C-VLAN).  Th=
e
>>>>>>>>>    Timing over PW encapsulation MUST use the Control Word (CW) as
>>>>>>>>>    specified in [RFC4448] to ensure proper detection of PTP messag=
es
>>>>>>>>>    inside the MPLS packets for Timing over LSP and Timing over PW
>>>>>>>>>    encapsulation.
>>>>>>>>>=20
>>>>>>>>> SB> That needs explanation
>>>>>>>>>=20
>>>>>>>>>    The use of Sequence Number in the CW is optional.
>>>>>>>>>=20
>>>>>>>>> SB> Given that s/n are never in practice deployed, you could proba=
bly
>>>>>>>>> SB> simplify things by sayig that they are not used.
>>>>>>>>>=20
>>>>>>>>>    Timing over PW encapsulation for NTP MUST use NTP over UDP/IP o=
ver
>>>>>> PW
>>>>>>>>>    (the IP PW discussed in [RFC4447]) shown in Fig 5(B).
>>>>>>>>>=20
>>>>>>>>>                     +----------------+  +----------------+
>>>>>>>>>                      |Timing LSP Label|  |Timing LSP Label|
>>>>>>>>>                      +----------------+  +----------------+
>>>>>>>>>                      |    PW Label    |  |    PW Label    |
>>>>>>>>>                      +----------------+  +----------------+
>>>>>>>>>                      |  Control Word  |  |      IP        |
>>>>>>>>>                      +----------------+  +----------------+
>>>>>>>>>                      |    Ethernet    |  |      UDP       |
>>>>>>>>>                      |     Header     |  +----------------+
>>>>>>>>>                      +----------------+  |   Timing PDU   |
>>>>>>>>>                      |S-VLAN(Optional)|  |                |
>>>>>>>>>                      +----------------+  +----------------+
>>>>>>>>>                      |C-VLAN(Optional)|        (B)
>>>>>>>>>                      +----------------+
>>>>>>>>>                      |   Timing PDU   |
>>>>>>>>>                      |                |
>>>>>>>>>                      +----------------+
>>>>>>>>>                             (A)
>>>>>>>>>=20
>>>>>>>>>               Figure (5) - Timing over PW Encapsulations
>>>>>>>>>=20
>>>>>>>>>    In order for an LSR to process PTP messages, the top label of t=
he
>>>>>>>>>    label stack (the Tunnel Label) MUST be a Timing label.
>>>>>>>>>=20
>>>>>>>>> S> You said that before.
>>>>>>>>>=20
>>>>>>>>> 6.3.  Other Timing Encapsulation methods
>>>>>>>>>=20
>>>>>>>>>    In future other timing encapsulation methods may be introduced,=
 such
>>>>>>>>>    as a new shim header after the Bottom of Stack to carry the Tim=
ing
>>>>>>>>>    information.  Such new encapsulations are outside the scope of t=
his
>>>>>>>>>    document.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> SB> Taking a pure MPLS pov, you can simplify a lot of the text
>>>>>>>>> SB> out of the definition of the LSP
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 14]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> SB> I think we need a section on LSP processing
>>>>>>>>>=20
>>>>>>>>> 7.  Timing message Processing
>>>>>>>>>=20
>>>>>>>>>    Each Timing protocol such as PTP and NTP, define their set of T=
iming
>>>>>>>>>    messages.  For example PTP defines SYNC, DELAY_REQ, DELAY_RESP,=

>>>>>>>>>    FOLLOW_UP, etc messages.
>>>>>>>>>=20
>>>>>>>>>    Some of the Timing messages require time stamping or correction=
 field
>>>>>>>>>    update at port level and some dont.  It is the job of the LER/L=
SR to
>>>>>>>>>    parse the timing message and find out the type of the Timing me=
ssage
>>>>>>>>>    and decide whether and how to Time- stamp it (e.g., BC) or upda=
te
>>>>>>>>>    correction field(e.g., TC).
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> SB> AAAAAAAAAAAH. Surely this is the function of the PTP prcessing=

>>>>>>>>> SB> function rather than the LER?
>>>>>>>>>=20
>>>>>>>>>    For example the following PTP messages (called Event messages)
>>>>>>>>>    require time-stamping or correction field update:
>>>>>>>>>=20
>>>>>>>>>    o  SYNC
>>>>>>>>>=20
>>>>>>>>>    o  DELAY_REQ (Delay Request)
>>>>>>>>>=20
>>>>>>>>>    o  PDELAY_REQ (Peer Delay Request)
>>>>>>>>>=20
>>>>>>>>>    o  PDELAY_RESP (Peer Delay Response)
>>>>>>>>>=20
>>>>>>>>>    SYNC and DELAY_REQ are exchanged between Master Clock and Slave=

>>>>>> Clock
>>>>>>>>>    and MUST be transported over PTP LSPs.  PDELAY_REQ and
>>>>>> PDELAY_RESP
>>>>>>>>>    are exchanged between adjacent PTP clocks (i.e.  Master, Slave,=

>>>>>>>>>    Boundary, or Transparent) and SHOULD be transported over single=
 hop
>>>>>>>>>    PTP LSPs.  If Two Step PTP clocks are present, then the FOLLOW_=
UP,
>>>>>>>>>    and PDELAY_RESP_FOLLOW_UP messages MUST also be transported
>>>>>> over
>>>>>>>>> the
>>>>>>>>>    PTP LSPs.
>>>>>>>>>=20
>>>>>>>>>    For a given instance of 1588 protocol, SYNC and DELAY_REQ MUST b=
e
>>>>>>>>>    transported over two PTP LSPs that are in opposite directions. =
 These
>>>>>>>>>    PTP LSPs, which are in opposite directions MUST be congruent an=
d co-
>>>>>>>>>    routed.  Alternatively, a single bidirectional co-routed LSP ca=
n be
>>>>>>>>>    used.
>>>>>>>>>=20
>>>>>>>>>    Except as indicated above for the two-step PTP clocks, Non-Even=
t PTP
>>>>>>>>>    message types do not need to be processed by intermediate route=
rs.
>>>>>>>>>    These message types MAY be carried in PTP Tunnel LSPs.
>>>>>>>>>=20
>>>>>>>>> SB> Are you saying that a timing P router has to be msg type sensi=
tive?
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 15]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 8.  Protection and Redundancy
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> SB> This is a bit of a jump - I don't know how the LSP itself work=
s yet!
>>>>>>>>>=20
>>>>>>>>>    In order to ensure continuous uninterrupted operation of slave
>>>>>>>>>    clocks, usually as a general practice, slave clocks (or ports) t=
rack
>>>>>>>>>    redundant master clocks.
>>>>>>>>>=20
>>>>>>>>>    It is the responsibility of the network operator to ensure that=

>>>>>>>>>    physically disjoint Timing LSPs are established between a slave=
 clock
>>>>>>>>>    (or port) and redundant master clocks (or ports).
>>>>>>>>>=20
>>>>>>>>>    When a slave clock (or port) listens to redundant master clocks=
 or
>>>>>>>>>    ports, any prolonged Timing LSP outage will trigger the slave c=
lock
>>>>>>>>>    or port to switch to a redundant master clock or port.
>>>>>>>>>=20
>>>>>>>>>    LSP/PW protection such as Linear protection Switching (1:1, 1+1=
),
>>>>>>>>>    Ring protection switching or MPLS Fast Reroute (FRR) generally s=
witch
>>>>>>>>>    alternative path that usually cause a change in delay, which if=

>>>>>>>>>    undetected by slave clock can reduce accuracy of the slave cloc=
k.
>>>>>>>>>=20
>>>>>>>>>    Therefore protection switching MAY be used, as long as phase ju=
mps
>>>>>>>>>    upon switchover due to differences in path latency are detected=
 and
>>>>>>>>>    compensated for (such compensation not being required if BCs or=
 peer-
>>>>>>>>>    peer TCs are used throughout).
>>>>>>>>>=20
>>>>>>>>>    Note that any protection or reroute mechanism that adds additio=
nal
>>>>>>>>>    MPLS label to the label stack, such as Facility Backup Fast Rer=
oute,
>>>>>>>>>    MUST ensure that the pushed label is also a Timing Label to ens=
ure
>>>>>>>>>    recognition of the MPLS frame as containing Timing messages, as=
 it
>>>>>>>>>    transits the backup path.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 16]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 9.  ECMP
>>>>>>>>>=20
>>>>>>>>>    To ensure the optimal operation of slave clocks and avoid error=

>>>>>>>>>    introduced by forward and reverse path delay asymmetry, the phy=
sical
>>>>>>>>>    path for Timing messages from master clock to slave Clock and v=
ice
>>>>>>>>>    versa must be the same for all Event Timing messages listed in
>>>>>>>>>    section 7.
>>>>>>>>>=20
>>>>>>>>>    Therefore the Timing LSPs MUST not be subject to ECMP (Equal Co=
st
>>>>>>>>>    Multipath).
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 17]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 10.  PHP
>>>>>>>>>=20
>>>>>>>>>    To ensure that the label on the top of the label stack is the T=
iming
>>>>>>>>>    LSP Label, PHP MUST not be used.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 18]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 11.  Entropy
>>>>>>>>>=20
>>>>>>>>>    To ensure all Timing messages in a Timing LSP take the same pat=
h,
>>>>>>>>>    Entropy Label MUST NOT be used for the Timing LSP[RFC6790] and
>>>>>>>>>    Entropy Label MUST NOT be used for the PWs that are carried ins=
ide
>>>>>>>>>    Timing LSP [RFC6391].
>>>>>>>>>=20
>>>>>>>>> SB> This is incorrect - you mean that all msgs of the same timing
>>>>>>>>> SB> flow need to have the same EL value.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 19]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 12.  OAM, Control and Management
>>>>>>>>>=20
>>>>>>>>>    In order to monitor Timing LSPs and their encapsulated PWs, the=
y MUST
>>>>>>>>>    be able to carry OAM and management messages.  These management=

>>>>>>>>>    messages MUST be differentiated from Timing messages via alread=
y
>>>>>>>>>    defined IETF methods.
>>>>>>>>>=20
>>>>>>>>>    For example BFD [RFC5880], [RFC5884] and LSP-Ping [RFC4389] MAY=
 run
>>>>>>>>>    over PTP LSPs via UDP/IP encapsulation or via GAL/G-ACH.  These=

>>>>>>>>>    Management protocols can easily be identified by the UDP Destin=
ation
>>>>>>>>>    Port number or by GAL/G-ACH respectively.
>>>>>>>>>=20
>>>>>>>>>    Also BFD, LSP-Ping and other management messages MAY run over t=
he
>>>>>> PWs
>>>>>>>>>    encapsulated in Timing LSP via one of the defined VCCVs (Type 1=
, 3 or
>>>>>>>>>    4) [RFC5085] (note that VCCV Type 2 using Router Alert Label is=
 going
>>>>>>>>>    to be deprecated by IETF).  In this case G-ACH, PW label (TTL=3D=
1) or
>>>>>>>>>    GAL-ACH are used to identify such management messages.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 20]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 13.  QoS Considerations
>>>>>>>>>=20
>>>>>>>>>    In network deployments where not every LSR/LER is Timing-aware,=
 it is
>>>>>>>>>    important to reduce the impact of the non-Timing-aware LSR/LERs=
 on
>>>>>>>>>    the timing recovery in the slave clock.  The Timing messages ar=
e time
>>>>>>>>>    critical and must be treated with the highest priority.  Theref=
ore
>>>>>>>>>    Timing over MPLS messages must be treated with the highest prio=
rity
>>>>>>>>>    in the routers.  This can be achieved by proper setup of Timing=
 LSPs.
>>>>>>>>>=20
>>>>>>>>>    It is recommended that the Timing LSPs are setup or configured
>>>>>>>>>    properly to indicate EF-PHB [RFC3246]for the CoS and Green [RFC=
2697]
>>>>>>>>>    for drop eligibility.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 21]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 14.  FCS and Checksum Recalculation
>>>>>>>>>=20
>>>>>>>>>    When time-stamp generation and timing packet adjustment is perf=
ormed
>>>>>>>>>    near the physical port hardware, the process MUST include
>>>>>>>>>    recalculation of the Ethernet FCS.
>>>>>>>>>=20
>>>>>>>>> SB> The above is confusing - an LSR always recomputes the link lay=
er
>>>>>>>>> SB> CRC which may or may not be Ethernet.
>>>>>>>>>=20
>>>>>>>>>    Also FCS retention for the
>>>>>>>>>    payload Ethernet described in [RFC4720] MUST NOT be used.
>>>>>>>>>=20
>>>>>>>>>    For UDP/IP encapsulation mode of Timing over MPLS, the UDP chec=
ksum
>>>>>>>>>    may be required as per UDP transport standards.
>>>>>>>>>=20
>>>>>>>>> SB> You really need to be working on getting the IPv6 C?S computat=
ion
>>>>>>>>> SB> removed from PTP msgs.
>>>>>>>>>=20
>>>>>>>>>    When UDP checksum is used, each Timing-aware LER/LSR must eithe=
r
>>>>>>>>>    incrementally update the UDP checksum after Time stamping or
>>>>>>>>>    Correction Field update or verify the UDP checksum on reception=
 from
>>>>>>>>>    upstream and recalculate the checksum completely on transmissio=
n to
>>>>>>>>>    downstream node after Time stamping or Correction Field update.=

>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 22]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 15.  Behavior of LER/LSR
>>>>>>>>>=20
>>>>>>>>>    Timing-capable/aware LERs and LSRs are routers that have one or=
 more
>>>>>>>>>=20
>>>>>>>>> SB> You mean physical interfaces?
>>>>>>>>>=20
>>>>>>>>>    interfaces that can perform Timing operations (OC/BC/TC) on Tim=
ing
>>>>>>>>>    packets and are configured to do so.  Timing-capable/aware LERs=
 and
>>>>>>>>>    LSRs can advertise their Timing-capability per-interface via co=
ntrol
>>>>>>>>>    plane such as OSPF or IS-IS.
>>>>>>>>> SB> ISIS and OSPF are routing protocols.
>>>>>>>>>=20
>>>>>>>>>   The Timing-capable/aware LERs can then
>>>>>>>>>    signals Timing LSPs via RSVP-TE signaling.  Alternatively the T=
iming
>>>>>>>>>    capability of LER and LSRs may be configured in a centralized
>>>>>>>>>    controller and the Timing LSP may be setup using manual configu=
ration
>>>>>>>>>    or other methods such as SDN.
>>>>>>>>>=20
>>>>>>>>> SB> it can also be configured individually rather then through
>>>>>>>>> SB> a cebtral controllwe
>>>>>>>>>=20
>>>>>>>>> 15.1.  Behavior of Timing-capable/aware LER
>>>>>>>>>=20
>>>>>>>>>    When a Timing-capable/aware LER behaves as a Transparent clock a=
nd
>>>>>>>>>    receives a Timing message from a Timing-capable/aware non-MPLS
>>>>>>>>>    interface, the LER updates the Correction Field (CF) and encaps=
ulates
>>>>>>>>>    and forwards the timing message over previously established Tim=
ing
>>>>>>>>>    LSP.
>>>>>>>>>=20
>>>>>>>>> SB> You need to call out the details so that people properly
>>>>>>>>> SB> understand the definition of the new LSP.
>>>>>>>>>=20
>>>>>>>>>    Also when a Timing message is received from a Timing-capable/
>>>>>>>>>    aware MPLS interface, LER updates the Correction Filed (CF) and=

>>>>>>>>>    decapsulates the MPLS encapsulation and forwards the timing mes=
sage
>>>>>>>>>    to a non-MPLS interface.
>>>>>>>>>=20
>>>>>>>>>    When a Timing-capable/aware LER behaves as a Boundary clock and=

>>>>>>>>>    receives a Timing message from a Timing-capable/aware non MPLS
>>>>>>>>>    interface, the LER Timestamps the Timing packet and sends it to=
 the
>>>>>>>>>    LERs Boundary clock processing module.  Also when a Timing mess=
age is
>>>>>>>>>    received from a Timing- capable/aware MPLS interface, the LER
>>>>>>>>>    Timestamps the Timing packet and sends it to the LERs Boundary c=
lock
>>>>>>>>>    processing module.
>>>>>>>>>=20
>>>>>>>>>    When a Timing-capable/aware LER behaves as an Ordinary Clock to=
ward
>>>>>>>>>    the MPLS network, and receives a Timing message from a Timing-
>>>>>>>>>    capable/aware MPLS interface, the LER Timestamps the Timing pac=
ket
>>>>>>>>>    and sends it to the LERs Ordinary clock processing module.
>>>>>>>>>=20
>>>>>>>>> 15.2.  Behavior of Timing-capable/aware LSR
>>>>>>>>>=20
>>>>>>>>>    When a Timing-capable/aware LSR behaves as a Transparent clock a=
nd
>>>>>>>>>    receives a Timing message from a Timing-capable/aware MPLS
>>>>>> interface,
>>>>>>>>>    The LSR updates the Correction Filed (CF) and forwards the timi=
ng
>>>>>>>>>    message over another MPLS interface.
>>>>>>>>>=20
>>>>>>>>>    When a Timing-capable/aware LSR behaves as a Boundary clock and=

>>>>>>>>>    receives a Timing message from a Timing-capable/aware MPLS
>>>>>> interface.
>>>>>>>>>    The LSR performs the functions of a Boundary Clock in terminati=
ng the
>>>>>>>>>    received Timing message and re-generating a new timing message o=
ver
>>>>>>>>>    another (or the same) MPLS interface.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 23]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 15.3.  Behavior of non-Timing-capable/aware LSR
>>>>>>>>>=20
>>>>>>>>>    It is most beneficial when all LSRs in the path of a Timing LSP=
 be
>>>>>>>>>    timing-Capable/aware LSRs.  This would ensure the highest quali=
ty
>>>>>>>>>    time and clock synchronization by Timing Slave Clocks.  However=
, this
>>>>>>>>>    specification does not mandate that all LSRs in path of a Timin=
g LSP
>>>>>>>>>    be Timing- capable/aware.
>>>>>>>>>=20
>>>>>>>>>    Non-Timing-capable/aware LSRs just switch the packets encapsula=
ted in
>>>>>>>>>    Timing LSPs and dont perform any Timing operation (TC or BC).
>>>>>>>>>    However as explained in QoS section the Timing over MPLS packet=
s
>>>>>> MUST
>>>>>>>>>    be still be treated with the highest priority based on their Tr=
affic
>>>>>>>>>    Class (TC) marking.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 24]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 16.  Other considerations
>>>>>>>>>=20
>>>>>>>>>    [IEEE-1588] defines an optional peer-to-peer Transparent clocki=
ng
>>>>>>>>>    that requires peer delay measurement between two adjacent Timin=
g-
>>>>>>>>>    capable/ aware routers/switches.  Peer delay measurement messag=
es
>>>>>>>>>    need to be time stamped and terminated by the Timing-capable/aw=
are
>>>>>>>>>    routers/ switches.  This means that two adjacent LSRs may be en=
gaged
>>>>>>>>>    in a peer delay measurement.
>>>>>>>>>=20
>>>>>>>>>    For transporting such peer delay measurement messages a single-=
hop
>>>>>>>>>    LSP SHOULD to be created between the two adjacent LSRs engaged i=
n
>>>>>>>>>    peer delay measurement to carry peer delay measurement messages=
.
>>>>>>>>>    Other methods such as PTP transport over Ethernet MAY be used f=
or
>>>>>>>>>    transporting peer delay measurement messages if the link betwee=
n the
>>>>>>>>>    two routers is Ethernet.
>>>>>>>>>=20
>>>>>>>>>    In Peer-to-peer transparent clocking (P2P TC), a Timing-capable=
/ ware
>>>>>>>>>    routers/switches MUST maintain a list of all the neighbors it n=
eeds
>>>>>>>>>    to send a PDelay_Req to, where each neighbor corresponds to a t=
iming
>>>>>>>>>    LSP.
>>>>>>>>>=20
>>>>>>>>>    The use of Explicit Null Label (Label=3D 0 or 2) is acceptable a=
s long
>>>>>>>>>    as either the Explicit Null label is the bottom of stack label
>>>>>>>>>    (applicable only to UDP/IP encapsulation) or the label below th=
e
>>>>>>>>>    Explicit Null label is a PTP label.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 25]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 17.  Security Considerations
>>>>>>>>>=20
>>>>>>>>>    MPLS PW security considerations in general are discussed in [RFC=
3985]
>>>>>>>>>    and [RFC4447],and those considerations also apply to this docum=
ent.
>>>>>>>>>=20
>>>>>>>>>    An experimental security protocol is defined in [IEEE-1588].The=
 PTP
>>>>>>>>>    security extension and protocol provides group source authentic=
ation,
>>>>>>>>>    message integrity, and replay attack protection for PTP message=
s.
>>>>>>>>>=20
>>>>>>>>>    When the MPLS network (provider network) serves multiple custom=
ers,
>>>>>>>>>    it is important to maintain and process each customers clock an=
d
>>>>>>>>>    Timing messages separately from other customers to ensure there=
 is no
>>>>>>>>>    cross- customer effect.  For example if an LER BC is synchroniz=
ed to
>>>>>>>>>    a specific grandmaster, belonging to customer A, then the LER M=
UST
>>>>>>>>>    use that BC clock only for customer A to ensure that customer A=

>>>>>>>>>    cannot attack other customers by manipulating its time.
>>>>>>>>>=20
>>>>>>>>>    Timing messages MAY be encrypted or authenticated, provided tha=
t the
>>>>>>>>>    LERs/LSRs that are Timing capable/aware can authenticate/ decry=
pt the
>>>>>>>>>    timing messages.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 26]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 18.  Acknowledgements
>>>>>>>>>=20
>>>>>>>>>    The authors would like to thank Ron Cohen, Yaakov Stein, Tal Mi=
zrahi,
>>>>>>>>>    Stefano Ruffini, Peter Meyer, and other members of IETF for rev=
iewing
>>>>>>>>>    and providing feedback on this draft.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 27]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 19.  IANA Considerations
>>>>>>>>>=20
>>>>>>>>>    There are no IANA requirements in this specification.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 28]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 20.  References
>>>>>>>>>=20
>>>>>>>>> 20.1.  Normative References
>>>>>>>>>=20
>>>>>>>>>    [IEEE-1588]
>>>>>>>>>               IEEE 1588-2008, "IEEE Standard for a Precision Clock=

>>>>>>>>>               Synchronization Protocol for Networked Measurement a=
nd
>>>>>>>>>               Control Systems".
>>>>>>>>>=20
>>>>>>>>>    [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
>>>>>>>>>               Requirement Levels", BCP 14, RFC 2119, March 1997.
>>>>>>>>>=20
>>>>>>>>>    [RFC3985]  Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-=
to-
>>>>>>>>>               Edge (PWE3) Architecture", RFC 3985, March 2005.
>>>>>>>>>=20
>>>>>>>>>    [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Disc=
overy
>>>>>>>>>               Proxies (ND Proxy)", RFC 4389, April 2006.
>>>>>>>>>=20
>>>>>>>>>    [RFC4447]  Martini, L., Rosen, E., El-Aawar, N., Smith, T., and=
 G.
>>>>>>>>>               Heron, "Pseudowire Setup and Maintenance Using the L=
abel
>>>>>>>>>               Distribution Protocol (LDP)", RFC 4447, April 2006.
>>>>>>>>>=20
>>>>>>>>>    [RFC4448]  Martini, L., Rosen, E., El-Aawar, N., and G. Heron,
>>>>>>>>>               "Encapsulation Methods for Transport of Ethernet ove=
r MPLS
>>>>>>>>>               Networks", RFC 4448, April 2006.
>>>>>>>>>=20
>>>>>>>>>    [RFC4720]  Malis, A., Allan, D., and N. Del Regno, "Pseudowire
>>>>>>>>>               Emulation Edge-to-Edge (PWE3) Frame Check Sequence
>>>>>>>>>               Retention", RFC 4720, November 2006.
>>>>>>>>>=20
>>>>>>>>>    [RFC5085]  Nadeau, T. and C. Pignataro, "Pseudowire Virtual Cir=
cuit
>>>>>>>>>               Connectivity Verification (VCCV): A Control Channel f=
or
>>>>>>>>>               Pseudowires", RFC 5085, December 2007.
>>>>>>>>>=20
>>>>>>>>>    [RFC5880]  Katz, D. and D. Ward, "Bidirectional Forwarding Dete=
ction
>>>>>>>>>               (BFD)", RFC 5880, June 2010.
>>>>>>>>>=20
>>>>>>>>>    [RFC5884]  Aggarwal, R., Kompella, K., Nadeau, T., and G. Swall=
ow,
>>>>>>>>>               "Bidirectional Forwarding Detection (BFD) for MPLS L=
abel
>>>>>>>>>               Switched Paths (LSPs)", RFC 5884, June 2010.
>>>>>>>>>=20
>>>>>>>>> 20.2.  Informative References
>>>>>>>>>=20
>>>>>>>>>    [I-D.ietf-pwe3-fat-pw]
>>>>>>>>>               Bryant, S., Filsfils, C., Drafz, U., Kompella, V., R=
egan,
>>>>>>>>>               J., and S. Amante, "Flow Aware Transport of Pseudowi=
res
>>>>>>>>>               over an MPLS Packet Switched Network",
>>>>>>>>>               draft-ietf-pwe3-fat-pw-07 (work in progress), July 2=
011.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 29]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    [ISO]      ISO/IEC 10589:1992, "Intermediate system to Intermed=
iate
>>>>>>>>>               system routeing information exchange protocol for us=
e in
>>>>>>>>>               conjunction with the Protocol for providing the
>>>>>>>>>               Connectionless-mode Network Service (ISO 8473)".
>>>>>>>>>=20
>>>>>>>>>    [RFC1195]  Callon, R., "Use of OSI IS-IS for routing in TCP/IP a=
nd
>>>>>>>>>               dual environments", RFC 1195, December 1990.
>>>>>>>>>=20
>>>>>>>>>    [RFC2328]  Moy, J., "OSPF Version 2", STD 54, RFC 2328, April 1=
998.
>>>>>>>>>=20
>>>>>>>>>    [RFC2697]  Heinanen, J. and R. Guerin, "A Single Rate Three Col=
or
>>>>>>>>>               Marker", RFC 2697, September 1999.
>>>>>>>>>=20
>>>>>>>>>    [RFC3246]  Davie, B., Charny, A., Bennet, J., Benson, K., Le Bo=
udec,
>>>>>>>>>               J., Courtney, W., Davari, S., Firoiu, V., and D.
>>>>>>>>>               Stiliadis, "An Expedited Forwarding PHB (Per-Hop
>>>>>>>>>               Behavior)", RFC 3246, March 2002.
>>>>>>>>>=20
>>>>>>>>>    [RFC3630]  Katz, D., Kompella, K., and D. Yeung, "Traffic Engin=
eering
>>>>>>>>>               (TE) Extensions to OSPF Version 2", RFC 3630,
>>>>>>>>>               September 2003.
>>>>>>>>>=20
>>>>>>>>>    [RFC3784]  Smit, H. and T. Li, "Intermediate System to Intermed=
iate
>>>>>>>>>               System (IS-IS) Extensions for Traffic Engineering (T=
E)",
>>>>>>>>>               RFC 3784, June 2004.
>>>>>>>>>=20
>>>>>>>>>    [RFC4970]  Lindem, A., Shen, N., Vasseur, JP., Aggarwal, R., an=
d S.
>>>>>>>>>               Shaffer, "Extensions to OSPF for Advertising Optiona=
l
>>>>>>>>>               Router Capabilities", RFC 4970, July 2007.
>>>>>>>>>=20
>>>>>>>>>    [RFC4971]  Vasseur, JP., Shen, N., and R. Aggarwal, "Intermedia=
te
>>>>>>>>>               System to Intermediate System (IS-IS) Extensions for=

>>>>>>>>>               Advertising Router Information", RFC 4971, July 2007=
.
>>>>>>>>>=20
>>>>>>>>>    [RFC5120]  Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Mul=
ti
>>>>>>>>>               Topology (MT) Routing in Intermediate System to
>>>>>>>>>               Intermediate Systems (IS-ISs)", RFC 5120, February 2=
008.
>>>>>>>>>=20
>>>>>>>>>    [RFC5305]  Li, T. and H. Smit, "IS-IS Extensions for Traffic
>>>>>>>>>               Engineering", RFC 5305, October 2008.
>>>>>>>>>=20
>>>>>>>>>    [RFC5329]  Ishiguro, K., Manral, V., Davey, A., and A. Lindem,
>>>>>>>>>               "Traffic Engineering Extensions to OSPF Version 3",
>>>>>>>>>               RFC 5329, September 2008.
>>>>>>>>>=20
>>>>>>>>>    [RFC5340]  Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "O=
SPF
>>>>>>>>>               for IPv6", RFC 5340, July 2008.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 30]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    [RFC5905]  Mills, D., Martin, J., Burbank, J., and W. Kasch, "N=
etwork
>>>>>>>>>               Time Protocol Version 4: Protocol and Algorithms
>>>>>>>>>               Specification", RFC 5905, June 2010.
>>>>>>>>>=20
>>>>>>>>>    [RFC6391]  Bryant, S., Filsfils, C., Drafz, U., Kompella, V., R=
egan,
>>>>>>>>>               J., and S. Amante, "Flow-Aware Transport of Pseudowi=
res
>>>>>>>>>               over an MPLS Packet Switched Network", RFC 6391,
>>>>>>>>>               November 2011.
>>>>>>>>>=20
>>>>>>>>>    [RFC6790]  Kompella, K., Drake, J., Amante, S., Henderickx, W.,=
 and
>>>>>>>>>               L. Yong, "The Use of Entropy Labels in MPLS Forwardi=
ng",
>>>>>>>>>               RFC 6790, November 2012.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 31]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 1.  Routing extensions for Timing-aware Routers
>>>>>>>>>=20
>>>>>>>>>    MPLS-TE routing relies on extensions to OSPF [RFC2328] [RFC5340=
] and
>>>>>>>>>    IS-IS [ISO] [RFC1195] in order to advertise Traffic Engineering=
 (TE)
>>>>>>>>>    link information used for constraint-based routing.
>>>>>>>>>=20
>>>>>>>>>    Indeed, it is useful to advertise data plane TE router link
>>>>>>>>>    capabilities, such as the capability for a router to be Timing-=
aware.
>>>>>>>>>    This capability MUST then be taken into account during path
>>>>>>>>>    computation to prefer or even require links that advertise them=
selves
>>>>>>>>>    as Timing-aware.  In this way the path can ensure the entry and=
 exit
>>>>>>>>>    points into the LERs and, if desired, the links into the LSRs a=
re
>>>>>>>>>    able to perform port based time-stamping thus minimizing their i=
mpact
>>>>>>>>>    on the performance of the slave clock.
>>>>>>>>>=20
>>>>>>>>>    extensions are required to OSPF and IS-IS in order to advertise=

>>>>>>>>>    Timing-aware capabilities of a link.  Such extensions are outsi=
de the
>>>>>>>>>    scope of this document; however such extension SHOULD be able t=
o
>>>>>>>>>    signal the following information per Router Link:
>>>>>>>>>=20
>>>>>>>>>    o  Capable of processing PTP, NTP or other Timing flows
>>>>>>>>>=20
>>>>>>>>>    o  Capable of performing Transparent Clock operation
>>>>>>>>>=20
>>>>>>>>>    o  Capable of performing Boundary Clock operation
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 32]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> 2.  Signaling Extensions for Creating Timing LSPs
>>>>>>>>>=20
>>>>>>>>>    RSVP-TE signaling MAY be used to setup the timing LSPs.  When R=
SVP-TE
>>>>>>>>>    is used to setup Timing LSPs, some information that indicates t=
hat
>>>>>>>>>    the LSP is carrying Timing flows MUST be included in the new
>>>>>>>>>    Extensions to RSVP-TE:
>>>>>>>>>=20
>>>>>>>>>    The following information MAY also be included in the new Exten=
sions
>>>>>>>>>    to RSVP-TE:
>>>>>>>>>=20
>>>>>>>>>    o  Offset from Bottom of Stack (BoS) to the start of the Time-s=
tamp
>>>>>>>>>       field
>>>>>>>>>=20
>>>>>>>>>    o  Number of VLANs in case of PW encapsulation
>>>>>>>>>=20
>>>>>>>>>    o  Timestamp field Type
>>>>>>>>>=20
>>>>>>>>>       *  Correction Field, Timestamp
>>>>>>>>>=20
>>>>>>>>>    o  Timestamp Field format
>>>>>>>>>=20
>>>>>>>>>       *  64-bit PTPv1, 80-bit PTPv2, 32-bit NTP, 64-bit NTP, 128-b=
it
>>>>>>>>>          NTP, etc.
>>>>>>>>>=20
>>>>>>>>>    Note that in case the above optional information is signaled wi=
th
>>>>>>>>>    RSVP-TE for a Timing LSP, all the Timing packets carried in tha=
t LSP
>>>>>>>>>    must have the same signaled characteristics.  For example if
>>>>>>>>>    Timestamp format is signaled as 64-bit PTPv1, then all Timing p=
ackets
>>>>>>>>>    must use 64-bit PTPv1 time-stamp.
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 33]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Authors' Addresses
>>>>>>>>>=20
>>>>>>>>>    Shahram Davari
>>>>>>>>>    Broadcom Corp.
>>>>>>>>>    San Jose, CA  95134
>>>>>>>>>    USA
>>>>>>>>>=20
>>>>>>>>>    Email: davari@broadcom.com
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    Amit Oren
>>>>>>>>>    Broadcom Corp.
>>>>>>>>>    San Jose, CA  95134
>>>>>>>>>    USA
>>>>>>>>>=20
>>>>>>>>>    Email: amito@broadcom.com
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    Manav Bhatia
>>>>>>>>>    Alcatel-Lucent
>>>>>>>>>    Bangalore,
>>>>>>>>>    India
>>>>>>>>>=20
>>>>>>>>>    Email: manav.bhatia@alcatel-lucent.com
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    Peter Roberts
>>>>>>>>>    Alcatel-Lucent
>>>>>>>>>    Kanata,
>>>>>>>>>    Canada
>>>>>>>>>=20
>>>>>>>>>    Email: peter.roberts@alcatel-lucent.com
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    Laurent Montini
>>>>>>>>>    Cisco Systems
>>>>>>>>>    San Jose CA
>>>>>>>>>    USA
>>>>>>>>>=20
>>>>>>>>>    Email: lmontini@cisco.com
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 34]
>>>>>>>>> Internet-Draft        Transporting Timing over MPLS            Jun=
e 2013
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>    Luca
>>>>>>>>>    Cisco Systems
>>>>>>>>>    San Jose CA
>>>>>>>>>    USA
>>>>>>>>>=20
>>>>>>>>>    Email: lmartini@cisco.com
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>>=20
>>>>>>>>> Davari, et al.          Expires December 17, 2013              [Pa=
ge 35]
>>>>>>>>> --
>>>>>>>>> For corporate legal information go to:
>>>>>>>>>=20
>>>>>>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html=

>>>>>>>>>=20
>>>>>>>>> _______________________________________________
>>>>>>>>> TICTOC mailing list
>>>>>>>>> TICTOC@ietf.org
>>>>>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>>>>> This e-mail message is intended for the recipient only and contains=

>>>>>> information which is CONFIDENTIAL and which may be proprietary to ECI=

>>>>>> Telecom. If you have received this transmission in error, please info=
rm us by e-
>>>>>> mail, phone or fax, and then delete the original and all copies there=
of.
>>>>>>>> _______________________________________________
>>>>>>>> mpls mailing list
>>>>>>>> mpls@ietf.org
>>>>>>>> https://www.ietf.org/mailman/listinfo/mpls
>>>>>>> .
>>>>>> --
>>>>>> For corporate legal information go to:
>>>>>>=20
>>>>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>>>>=20
>>>>>> _______________________________________________
>>>>>> TICTOC mailing list
>>>>>> TICTOC@ietf.org
>>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>>>=20
>>>>>>=20
>>>>>> _______________________________________________
>>>>>> TICTOC mailing list
>>>>>> TICTOC@ietf.org
>>>>>> https://www.ietf.org/mailman/listinfo/tictoc
>>>>> This e-mail message is intended for the recipient only and contains in=
formation which is CONFIDENTIAL and which may be proprietary to ECI Telecom.=
 If you have received this transmission in error, please inform us by e-mail=
, phone or fax, and then delete the original and all copies thereof.
>>>> .
>>>=20
>>> --
>>> For corporate legal information go to:
>>>=20
>>> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>>>=20
>>> This e-mail message is intended for the recipient only and contains info=
rmation which is CONFIDENTIAL and which may be proprietary to ECI Telecom. I=
f you have received this transmission in error, please inform us by e-mail, p=
hone or fax, and then delete the original and all copies thereof.
>> This e-mail message is intended for the recipient only and contains infor=
mation which is CONFIDENTIAL and which may be proprietary to ECI Telecom. If=
 you have received this transmission in error, please inform us by e-mail, p=
hone or fax, and then delete the original and all copies thereof.
>>=20
>> .
>=20
>=20
> --=20
> For corporate legal information go to:
>=20
> http://www.cisco.com/web/about/doing_business/legal/cri/index.html
>=20

From stbryant@cisco.com  Wed Aug 21 08:22:40 2013
Return-Path: <stbryant@cisco.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0764411E8112; Wed, 21 Aug 2013 08:22:40 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -110.599
X-Spam-Level: 
X-Spam-Status: No, score=-110.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id KJh89uP-Vdm6; Wed, 21 Aug 2013 08:22:34 -0700 (PDT)
Received: from ams-iport-1.cisco.com (ams-iport-1.cisco.com [144.254.224.140]) by ietfa.amsl.com (Postfix) with ESMTP id 0851521F9C00; Wed, 21 Aug 2013 08:22:33 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=362; q=dns/txt; s=iport; t=1377098554; x=1378308154; h=message-id:date:from:reply-to:mime-version:to:cc:subject: references:in-reply-to:content-transfer-encoding; bh=8atm9my2TbSk2JwQNT/n6LDv+MzBWZSGzQMIOKJgWmI=; b=mBHnIe+n+4Am4yLydb2kmEl/0cuRGV+A1r/SxXBF2NuJhxbJ5vntd54i 4JSvqYRx1q8Wm4iMCELOwOnTXQFfxeVF3P3j1bjcvT6nWbpbzNP2F7veA R9ekada7wgA8Eit8sYYEk2wbbdb+cj2qtXHSQE3z8nXnMVoOTtKbY/ktq U=;
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AgIFAHjaFFKQ/khM/2dsb2JhbABagwW9Z4J6gR8WdIIkAQEBBDhAARALGAkWDwkDAgECAUUGDQEHAQGIDK1ckFUHhBQDl2WRVoMd
X-IronPort-AV: E=Sophos;i="4.89,928,1367971200"; d="scan'208";a="158727268"
Received: from ams-core-3.cisco.com ([144.254.72.76]) by ams-iport-1.cisco.com with ESMTP; 21 Aug 2013 15:22:27 +0000
Received: from cisco.com (mrwint.cisco.com [64.103.70.36]) by ams-core-3.cisco.com (8.14.5/8.14.5) with ESMTP id r7LFMPsG019998 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Aug 2013 15:22:26 GMT
Received: from [IPv6:::1] (localhost [127.0.0.1]) by cisco.com (8.14.4+Sun/8.8.8) with ESMTP id r7LFMLEQ007322; Wed, 21 Aug 2013 16:22:21 +0100 (BST)
Message-ID: <5214DB2D.2000809@cisco.com>
Date: Wed, 21 Aug 2013 16:22:21 +0100
From: Stewart Bryant <stbryant@cisco.com>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:17.0) Gecko/20130620 Thunderbird/17.0.7
MIME-Version: 1.0
To: "S. Davari" <davarish@yahoo.com>
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com> <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com> <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com> <52139903.8040801@cisco.com> <F9336571731ADE42A5397FC831CEAA021513A083@ILPTWPVEXMB01.ecitele.com> <B7A656C4-6D0E-4719-9E42-57689CDAEFA9@yahoo.com> <F9336571731ADE42A5397FC831CEAA021513A0C6@ILPTWPVEXMB01.ecitele.com> <52149A2D.3050109@cisco.com> <5ADB2697-BC53-427A-82DB-013B89029343@yahoo.com>
In-Reply-To: <5ADB2697-BC53-427A-82DB-013B89029343@yahoo.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
Reply-To: stbryant@cisco.com
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 21 Aug 2013 15:22:40 -0000

yes, but they are one hop tunnels.

S

On 21/08/2013 15:19, S. Davari wrote:
> Hi Stewart,
>
> Your proposal does tunneling between timing aware nodes but I guess it stitches multiple of those tunnels to create and end-2-end connection. Right?
>
> Sasha's BoS proposal seems simpler, since it requires a single end-2-end LSP.
>
> Regards,
> Shahram


From stbryant@cisco.com  Wed Aug 21 08:27:06 2013
Return-Path: <stbryant@cisco.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 398E621F9D28; Wed, 21 Aug 2013 08:27:06 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -110.599
X-Spam-Level: 
X-Spam-Status: No, score=-110.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_HI=-8, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id XsiVkH9d-SEb; Wed, 21 Aug 2013 08:27:00 -0700 (PDT)
Received: from ams-iport-1.cisco.com (ams-iport-1.cisco.com [144.254.224.140]) by ietfa.amsl.com (Postfix) with ESMTP id 194AE11E8244; Wed, 21 Aug 2013 08:26:57 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=347; q=dns/txt; s=iport; t=1377098819; x=1378308419; h=message-id:date:from:reply-to:mime-version:to:cc:subject: references:in-reply-to:content-transfer-encoding; bh=y/+Ur8/yLSt/dJyTVLt0HyzK9KrG8bB9kXkE+bRbhgQ=; b=XmCnZt4U5A/XIJ3hz0LM2RDI7ymrWM5XrUbI4GesqQWA1fwLJ2ADVPAl lyIlDVuQrnBaTl2qs/ga5uXy8oF3ThtS5evHDysETHtcx0nLIfEq7q0zY tPQv3A6v78dJSmAMJHreJNpab298lNnUhgJYtVH322rPEVY+M53JW2rNc o=;
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AgIFAJzbFFKQ/khL/2dsb2JhbABagwW9Z4J6gR8WdIIkAQEBBDhAARALGAkWDwkDAgECAUUGDQEHAQGIDK1bkFUHhBQDl2WRVoMd
X-IronPort-AV: E=Sophos;i="4.89,928,1367971200"; d="scan'208";a="158727512"
Received: from ams-core-2.cisco.com ([144.254.72.75]) by ams-iport-1.cisco.com with ESMTP; 21 Aug 2013 15:26:56 +0000
Received: from cisco.com (mrwint.cisco.com [64.103.70.36]) by ams-core-2.cisco.com (8.14.5/8.14.5) with ESMTP id r7LFQshp018325 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Aug 2013 15:26:55 GMT
Received: from [IPv6:::1] (localhost [127.0.0.1]) by cisco.com (8.14.4+Sun/8.8.8) with ESMTP id r7LFQre5007754; Wed, 21 Aug 2013 16:26:54 +0100 (BST)
Message-ID: <5214DC3D.6050206@cisco.com>
Date: Wed, 21 Aug 2013 16:26:53 +0100
From: Stewart Bryant <stbryant@cisco.com>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:17.0) Gecko/20130620 Thunderbird/17.0.7
MIME-Version: 1.0
To: "S. Davari" <davarish@yahoo.com>
References: <51FB7543.801@cisco.com> <F9336571731ADE42A5397FC831CEAA0215124FCE@ILPTWPVEXMB02.ecitele.com> <C356618E-C662-4979-B70A-3277AAF4CD02@yahoo.com> <51FF7068.9080600@cisco.com> <4A6CE49E6084B141B15C0713B8993F281BEC731A@SJEXCHMB12.corp.ad.broadcom.com> <F9336571731ADE42A5397FC831CEAA0215132641@ILPTWPVEXMB01.ecitele.com> <D0E64B9E-D670-4419-AD5C-9E476F8DA0A8@yahoo.com> <52139903.8040801@cisco.com> <F9336571731ADE42A5397FC831CEAA021513A083@ILPTWPVEXMB01.ecitele.com> <B7A656C4-6D0E-4719-9E42-57689CDAEFA9@yahoo.com>
In-Reply-To: <B7A656C4-6D0E-4719-9E42-57689CDAEFA9@yahoo.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: "mpls@ietf.org" <mpls@ietf.org>, "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] [mpls] Comments on draft-ietf-tictoc-1588overmpls
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
Reply-To: stbryant@cisco.com
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 21 Aug 2013 15:27:06 -0000

On 20/08/2013 18:00, S. Davari wrote:
> Sasha
>
> In (2) are you proposing to create multiple segment LSPs, where each if those segment LSPs span between two Timing aware LSR? So if I have 10 timing aware LSR in the path I would need 9 LSPs?
>
> Regards,
> Shahram

Maybe, or maybe the path is held in the application payload.

Stewart

From yaakov_s@rad.com  Sat Aug 24 23:24:07 2013
Return-Path: <yaakov_s@rad.com>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B476111E821D for <tictoc@ietfa.amsl.com>; Sat, 24 Aug 2013 23:24:07 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -99.997
X-Spam-Level: 
X-Spam-Status: No, score=-99.997 tagged_above=-999 required=5 tests=[BAYES_50=0.001, HTML_MESSAGE=0.001, UNPARSEABLE_RELAY=0.001, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([12.22.58.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id QE43Jot+1e6X for <tictoc@ietfa.amsl.com>; Sat, 24 Aug 2013 23:24:02 -0700 (PDT)
Received: from rad.co.il (mailrelay02.rad.co.il [62.0.23.237]) by ietfa.amsl.com (Postfix) with ESMTP id 3909411E8219 for <tictoc@ietf.org>; Sat, 24 Aug 2013 23:23:57 -0700 (PDT)
Received: from Internal Mail-Server by MailRelay02 (envelope-from yaakov?s@rad.com) with AES128-SHA encrypted SMTP; 25 Aug 2013 09:26:36 +0300
Received: from EXRAD5.ad.rad.co.il ([192.114.24.28]) by EXRAD5.ad.rad.co.il ([192.114.24.28]) with mapi id 14.03.0146.000; Sun, 25 Aug 2013 09:23:53 +0300
From: Yaakov Stein <yaakov_s@rad.com>
To: "tictoc@ietf.org" <tictoc@ietf.org>
Thread-Topic: NIST Ytterbium Atomic Clocks Set Record for Stability
Thread-Index: Ac6hW6r1TYhFCkPOSxeOINXTBl+oXA==
Date: Sun, 25 Aug 2013 06:23:52 +0000
Message-ID: <07F7D7DED63154409F13298786A2ADC904E774C1@EXRAD5.ad.rad.co.il>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-originating-ip: [207.232.33.112]
Content-Type: multipart/alternative; boundary="_000_07F7D7DED63154409F13298786A2ADC904E774C1EXRAD5adradcoil_"
MIME-Version: 1.0
X-Commtouch-Refid: str=0001.0A0C0204.5219A2FA.0099,ss=1,fgs=0
Subject: [TICTOC] NIST Ytterbium Atomic Clocks Set Record for Stability
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 25 Aug 2013 06:24:07 -0000

--_000_07F7D7DED63154409F13298786A2ADC904E774C1EXRAD5adradcoil_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

This announcement may be of interest to readers of this list:

http://www.nist.gov/pml/div688/clock-082213.cfm

Y(J)S


--_000_07F7D7DED63154409F13298786A2ADC904E774C1EXRAD5adradcoil_
Content-Type: text/html; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<html xmlns:v=3D"urn:schemas-microsoft-com:vml" xmlns:o=3D"urn:schemas-micr=
osoft-com:office:office" xmlns:w=3D"urn:schemas-microsoft-com:office:word" =
xmlns:m=3D"http://schemas.microsoft.com/office/2004/12/omml" xmlns=3D"http:=
//www.w3.org/TR/REC-html40">
<head>
<meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3Dus-ascii"=
>
<meta name=3D"Generator" content=3D"Microsoft Word 12 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
@font-face
	{font-family:Tahoma;
	panose-1:2 11 6 4 3 5 4 4 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0in;
	margin-bottom:.0001pt;
	font-size:12.0pt;
	font-family:"Times New Roman","serif";}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:blue;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{mso-style-priority:99;
	color:purple;
	text-decoration:underline;}
p.MsoAcetate, li.MsoAcetate, div.MsoAcetate
	{mso-style-priority:99;
	mso-style-link:"Balloon Text Char";
	margin:0in;
	margin-bottom:.0001pt;
	font-size:8.0pt;
	font-family:"Tahoma","sans-serif";}
span.EmailStyle17
	{mso-style-type:personal-compose;
	font-family:"Calibri","sans-serif";
	color:windowtext;}
span.BalloonTextChar
	{mso-style-name:"Balloon Text Char";
	mso-style-priority:99;
	mso-style-link:"Balloon Text";
	font-family:"Tahoma","sans-serif";}
.MsoChpDefault
	{mso-style-type:export-only;}
@page WordSection1
	{size:8.5in 11.0in;
	margin:1.0in 1.25in 1.0in 1.25in;}
div.WordSection1
	{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext=3D"edit" spidmax=3D"1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext=3D"edit">
<o:idmap v:ext=3D"edit" data=3D"1" />
</o:shapelayout></xml><![endif]-->
</head>
<body lang=3D"EN-US" link=3D"blue" vlink=3D"purple">
<div class=3D"WordSection1">
<p class=3D"MsoNormal"><span style=3D"font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:black">This announcement may be of interest to read=
ers of this list:<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:13.5pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:black"><o:p>&nbsp;</o:p></span></p=
>
<p class=3D"MsoNormal"><span style=3D"font-size:13.5pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:black"><a href=3D"http://www.nist.=
gov/pml/div688/clock-082213.cfm">http://www.nist.gov/pml/div688/clock-08221=
3.cfm</a><o:p></o:p></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:13.5pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;;color:black"><o:p>&nbsp;</o:p></span></p=
>
<p class=3D"MsoNormal"><span style=3D"font-family:&quot;Calibri&quot;,&quot=
;sans-serif&quot;;color:black">Y(J)S<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,&quot;sans-serif&quot;"><o:p>&nbsp;</o:p></span></p>
</div>
</body>
</html>

--_000_07F7D7DED63154409F13298786A2ADC904E774C1EXRAD5adradcoil_--
