
From RBhatt@ixiacom.com  Tue Dec  1 00:20:26 2009
Return-Path: <RBhatt@ixiacom.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 3094028C1C9 for <l3vpn@core3.amsl.com>; Tue,  1 Dec 2009 00:20:26 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.598
X-Spam-Level: 
X-Spam-Status: No, score=-2.598 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HTML_MESSAGE=0.001]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id u4Jf0HuqUf27 for <l3vpn@core3.amsl.com>; Tue,  1 Dec 2009 00:20:24 -0800 (PST)
Received: from ixqw-mail-out.ixiacom.com (ixqw-mail-out.ixiacom.com [66.77.12.12]) by core3.amsl.com (Postfix) with ESMTP id 0081628C1C4 for <l3vpn@ietf.org>; Tue,  1 Dec 2009 00:20:23 -0800 (PST)
Received: from ixcaexch07.ixiacom.com ([fe80:0000:0000:0000:0000:5efe:10.200.2.124]) by IXCA-HC2.ixiacom.com ([10.200.2.51]) with mapi; Tue, 1 Dec 2009 00:20:16 -0800
From: Rakesh Bhatt <RBhatt@ixiacom.com>
To: "l3vpn@ietf.org" <l3vpn@ietf.org>
Date: Tue, 1 Dec 2009 00:20:15 -0800
Subject: Query on : draft-ietf-l3vpn-2547bis-mcast-07.txt
Thread-Topic: Query on : draft-ietf-l3vpn-2547bis-mcast-07.txt
Thread-Index: AcpyXxxcsxivyz3bSkaktwMiyUAzLw==
Message-ID: <716209EC190CA740BA799AC4ACCBFB5D1650B6F6AA@IXCAEXCH07.ixiacom.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
Content-Type: multipart/alternative; boundary="_000_716209EC190CA740BA799AC4ACCBFB5D1650B6F6AAIXCAEXCH07ixi_"
MIME-Version: 1.0
X-Mailman-Approved-At: Tue, 01 Dec 2009 04:38:16 -0800
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 01 Dec 2009 08:20:26 -0000

--_000_716209EC190CA740BA799AC4ACCBFB5D1650B6F6AAIXCAEXCH07ixi_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Hi Team

One Question on MVPN -p2mp

What should be extended community attribute ( route target ) value  for Lea=
f AD Route ( type 4 route ).


Thanks
Rakesh Bhatt

--_000_716209EC190CA740BA799AC4ACCBFB5D1650B6F6AAIXCAEXCH07ixi_
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=3D"http://www.w3.org/TR/REC-html40">

<head>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; charset=3Dus-ascii"=
>
<meta name=3DGenerator content=3D"Microsoft Word 11 (filtered medium)">
<style>
<!--
 /* Font Definitions */
 @font-face
	{font-family:Century;
	panose-1:2 4 6 4 5 5 5 2 3 4;}
 /* Style Definitions */
 p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0in;
	margin-bottom:.0001pt;
	font-size:12.0pt;
	font-family:Century;}
a:link, span.MsoHyperlink
	{color:blue;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{color:purple;
	text-decoration:underline;}
span.EmailStyle17
	{mso-style-type:personal-compose;
	font-family:Arial;
	color:windowtext;}
@page Section1
	{size:8.5in 11.0in;
	margin:1.0in 1.25in 1.0in 1.25in;}
div.Section1
	{page:Section1;}
-->
</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=3DEN-US link=3Dblue vlink=3Dpurple>

<div class=3DSection1>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span style=3D'font-size:1=
0.0pt;
font-family:Arial'>Hi Team <o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span style=3D'font-size:1=
0.0pt;
font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span style=3D'font-size:1=
0.0pt;
font-family:Arial'>One Question on MVPN &#8211;p2mp <o:p></o:p></span></fon=
t></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span style=3D'font-size:1=
0.0pt;
font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span style=3D'font-size:1=
0.0pt;
font-family:Arial'>What should be extended community attribute ( route targ=
et )
value &nbsp;for Leaf AD Route ( type 4 route ).<o:p></o:p></span></font></p=
>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span style=3D'font-size:1=
0.0pt;
font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span style=3D'font-size:1=
0.0pt;
font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span style=3D'font-size:1=
0.0pt;
font-family:Arial'>Thanks<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span style=3D'font-size:1=
0.0pt;
font-family:Arial'>Rakesh Bhatt<o:p></o:p></span></font></p>

</div>

</body>

</html>

--_000_716209EC190CA740BA799AC4ACCBFB5D1650B6F6AAIXCAEXCH07ixi_--

From yakov@juniper.net  Tue Dec  1 05:34:03 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 67C073A6934 for <l3vpn@core3.amsl.com>; Tue,  1 Dec 2009 05:34:03 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.549
X-Spam-Level: 
X-Spam-Status: No, score=-6.549 tagged_above=-999 required=5 tests=[AWL=0.050,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id hOee8irIiKnF for <l3vpn@core3.amsl.com>; Tue,  1 Dec 2009 05:34:02 -0800 (PST)
Received: from exprod7og117.obsmtp.com (exprod7og117.obsmtp.com [64.18.2.6]) by core3.amsl.com (Postfix) with ESMTP id 44B6028C238 for <l3vpn@ietf.org>; Tue,  1 Dec 2009 05:26:32 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob117.postini.com ([64.18.6.12]) with SMTP ID DSNKSxUZgES78GtHWXP209uvOVkdJws/fAec@postini.com; Tue, 01 Dec 2009 05:26:25 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Tue, 1 Dec 2009 05:24:37 -0800
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 1 Dec 2009 05:24:37 -0800
Received: from emailsmtp55.jnpr.net ([172.24.18.132]) by p-emlb01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 1 Dec 2009 05:24:36 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp55.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 1 Dec 2009 05:24:36 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nB1DOZj46966; Tue, 1 Dec 2009 05:24:35 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912011324.nB1DOZj46966@magenta.juniper.net>
To: Rakesh Bhatt <RBhatt@ixiacom.com>
Subject: Re: Query on : draft-ietf-l3vpn-2547bis-mcast-07.txt 
In-Reply-To: <716209EC190CA740BA799AC4ACCBFB5D1650B6F6AA@IXCAEXCH07.ixiacom.com>
References: <716209EC190CA740BA799AC4ACCBFB5D1650B6F6AA@IXCAEXCH07.ixiacom.com>
X-MH-In-Reply-To: Rakesh Bhatt <RBhatt@ixiacom.com> message dated "Tue, 01 Dec 2009 00:20:15 -0800."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <9322.1259673875.1@juniper.net>
Date: Tue, 1 Dec 2009 05:24:35 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 01 Dec 2009 13:24:36.0476 (UTC) FILETIME=[A0A563C0:01CA7289]
Cc: "l3vpn@ietf.org" <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 01 Dec 2009 13:34:03 -0000

Rakesh,

> Hi Team
> 
> One Question on MVPN -p2mp
> 
> What should be extended community attribute ( route target ) value  for Leaf
> AD Route ( type 4 route ).

See the following sections in draft-ietf-l3vpn-2547bis-mcast-bgp-08.txt:

   9.2.3.2.1. Originating Leaf A-D route into EBGP

   9.2.3.4.1. Originating Leaf A-D route into IBGP

Yakov.

From erosen@cisco.com  Thu Dec  3 08:39:27 2009
Return-Path: <erosen@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id CB3753A676A for <l3vpn@core3.amsl.com>; Thu,  3 Dec 2009 08:39:27 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -5.639
X-Spam-Level: 
X-Spam-Status: No, score=-5.639 tagged_above=-999 required=5 tests=[AWL=-0.840, BAYES_00=-2.599, J_CHICKENPOX_13=0.6, J_CHICKENPOX_31=0.6, J_CHICKENPOX_41=0.6, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id cRJ9CpZD3AK6 for <l3vpn@core3.amsl.com>; Thu,  3 Dec 2009 08:39:25 -0800 (PST)
Received: from rtp-iport-2.cisco.com (rtp-iport-2.cisco.com [64.102.122.149]) by core3.amsl.com (Postfix) with ESMTP id 6C5DE3A66B4 for <l3vpn@ietf.org>; Thu,  3 Dec 2009 08:39:25 -0800 (PST)
Authentication-Results: rtp-iport-2.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-AV: E=Sophos;i="4.47,336,1257120000"; d="scan'208";a="71888977"
Received: from rtp-core-2.cisco.com ([64.102.124.13]) by rtp-iport-2.cisco.com with ESMTP; 03 Dec 2009 16:39:16 +0000
Received: from erosen-linux.cisco.com (erosen-linux.cisco.com [161.44.70.34]) by rtp-core-2.cisco.com (8.13.8/8.14.3) with ESMTP id nB3GdG7g008848; Thu, 3 Dec 2009 16:39:16 GMT
Received: from erosen-linux (localhost.localdomain [127.0.0.1]) by erosen-linux.cisco.com (8.13.1/8.13.1) with ESMTP id nB3GdG5g003127;  Thu, 3 Dec 2009 11:39:16 -0500
To: Thomas Morin <thomas.morin@orange-ftgroup.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 
In-reply-to: Your message of Mon, 09 Nov 2009 16:47:58 +0900. <4AF7C92E.4050904@orange-ftgroup.com> 
Date: Thu, 03 Dec 2009 11:39:16 -0500
Message-ID: <3126.1259858356@erosen-linux>
From: Eric Rosen <erosen@cisco.com>
Cc: L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: erosen@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 03 Dec 2009 16:39:27 -0000

Appendix A of the "considerations" draft purports to provide an analysis
showing that BGP C-multicast routing provides a more scalable MVPN PE-PE
control plane than does PE-PE PIM.  However, in addition to errors of fact,
the analysis has lots of methodological problems; as a result it does not
serve its intended function, which is to support the recommendations of
section 3.3.1.

Due to the large number of methodological problems, I believe the Appendix
should be removed before the draft progresses.  I also have to say that I
find it peculiar for a PIM scalability analysis to progress without having
been last called in any of the multicast WGs.  The topic itself seems out of
scope for the L3VPN WG.  Nevertheless I will try to explain as clearly as I
can why I think the analysis is defective.

A scalability analysis must show how the utilization of some resource or set
of resources on some system increases as some demand or set of demands
increases.  The appendix proposes two resources to consider:

   Though the type of processing, messages and states, may vary with the
   different approaches, we propose here a rough estimation of the load
   of PEs, in terms of number of messages processed and number of
   control plane states maintained.  A "message processed" being a
   message being parsed, a lookup being done, and some action being
   taken (such has updating a control plane or data plane state).

This is okay as one of the resources considered.  If I understand this
correctly, this omits consideration of any messages received that do not
result in any "action being taken", and thus omits consideration of any
acknowledgment messages or other messages that fall into the category of
"transport overhead".  It also omits consideration of messages transmitted,
thus does not distinguish actions that cause a new message to be sent from
actions that do not.  So it is not a complete measure of the amount of
"messaging", but it is a factor worth considering.

   A "state maintained" being a multicast state kept in the control plane
   memory of a PE, related to a interface or a PE being subscribed to a
   multicast stream.

This is also worth considering as an abstract measure of memory resources.
However, as we will see, the analysis is not very precise about what it
counts as a "state", and this leads to some mistakes.

The real methodological problems begin when the analysis attempts to be
quantitative:

   The estimation unit used is the "message.equipment" (or "m.e"): one
   "message.equipment" corresponding to "one equipment processing one
   message" (10 m.e being "10 equipments processing each one message",
   or "5 messages each processed by 2 equipments", or "1 message
   processed by 10 equipment", etc.).  Similarly, for the amount of
   control plane state, the unit used is "state.equipment" or "s.e".
   This allow to take into account the fact that a message (or a state)
   can have be processed (or maintained) by more than one node.

This "estimation unit" has no meaning whatsoever, as it does not reflect
resource utilization on any system.  Suppose there are N systems, each
running an algorithm that scales as O(N).  Contrast this with an alternative
in which a single system does the work of all N, running an algorithm that
scales as O(N^2).  Other things being equal, the former would be considered
to be the more scalable alternative.  However, if you use the "estimation
unit" of "work.equipment", you would conclude that in both schemes the work
scales as O(N^2), and hence that the schemes are equally scalable.  If this
were a valid methodology, no one would ever have heard of such notions as
"distributed routing" or "datagram networks".  Use of this peculiar
"estimation unit" will always bias the results towards a centralized system
and away from a distributed system.

Any subsequent revision of Appendix A should eliminate the notion of "s.e."
or "m.e." and focus solely on the scaling of the individual systems.

Another methodological error is the fact that the analysis focuses heavily
on the Source-Specific Multicast mode, even though the number of VPN
customer deployments of SSM is quite small when compared to the number of
VPN customer deployments of ASM.  To support the recommendations of section
3.3, the analysis needs to focus on scenarios that relevant to actual
deployments. Some of the analysis of SSM also applies to the ASM procedures
for joining/pruning source trees, but this is not made clear, and it's
difficult for a reader to know how much of the analysis is relevant to
actual deployments.

In the most recent rev of the draft, a new section devoted to ASM has been
introduced, and this section has its own set of problems, as I will discuss
shortly.

In addition to the time spent on the virtually irrelevant SSM scenario, much
time is spent on examining a "PIM with explicit tracking" scheme, even
though no such scheme has been proposed.  It's difficult to understand why
this straw man is discussed, while other actual proposals for using PIM are
not.

Another methodological error is that only the messaging needed to join a
single (S,G) flow is considered.  Thus the analysis ignores any scaling
issues that show up as the number of (S,G) flows increases.  Since the
number of (S,G) flows handled by a Route Reflector will far exceed the
number handled by any PE router, failure to consider this particular scaling
factor tends to bias the results in favor of a scheme that uses Route
Reflectors.  Also omitted from consideration are scaling issues that show up
as the rate of change of multicast states increases.

Another methodological error is that the analysis takes no account of the
increased amount of state that must be maintained when C-multicast trees are
set up using two different protocols, rather than just one.  When PIM is
used as the PE-PE control plane, the (C-S,C-G) and (C-*,C-G) states that
drive the inter-PE control plane are the same as the one as the drive the
CE-PE control plane.  When BGP is used, a given multicast state has to be
represented both as a PIM state and as a BGP state (C-multicast route).
This should really be counted as a doubling (at least) of the number of
states that need to be maintained.  This is a very conservative estimate, as
the BGP states tend to contain much more information (e.g., RTs) than the
corresponding PIM states.  It's clear that appendix A does not consider both
the BGP states and the PIM states, but it's never quite clear just what
states it is considering.

Another methodological error is that the PE-CE interactions are not taken
into account.  In many scenarios the PE-PE interactions are just "noise"
when compared to the real bottleneck, the PE-CE interactions.  The total
number of PIM messages processed by a PE includes PIM messages from CEs as
well as from PEs.  Suppose that by adding BGP, you could reduce this total
number of PIM messages by 1%, at the cost of doubling the amount of state
you need to maintain.  Would that be a good tradeoff?  In order to support
the recommendations of section 3.3, Appendix A has to show that the
tradeoff is a lot better than this.  Yet it doesn't even attempt to do so.

It's possible to "prove" any conclusion you want by carefully scoping the
set of things you look at.  For example, if one wanted to prove that BGP is
less scalable, one could say "let's compare the two schemes to see how the
total number of TCP connections in the network scales as the number of PEs
increases".  Of course, to produce a valid analysis, one has to look at more
than one issue, and one has to demonstrate that one is actually looking at
the bottlenecks.

This collection of methodological errors is more than enough to send
Appendix A  "back to the drawing board".  If the goal is to make a
choice between two protocols by considering which scales best, then all  the
scaling factors must be considered.  

Let's take a look now at some of the more detailed issues.

Consider the messaging needed by a PE to maintain an (S,G) state in the
"steady state".  By "steady state", I mean during periods (a) that are
longer than the PIM refresh period and (b) in which the PE does not join or
prune (S,G).  (I think this is similar to what Appendix A calls "baseline
processing".)  Let's also suppose that the PE has k VRF interfaces over
which there are (S,G) receivers.

Assuming a typical one-minute PIM refresh time, the number of Join(S,G) PIM
messages per minute that the PE receives from the other PEs of the same MVPN
in the steady state is either 0 or 1.  (Due to PIM Join Suppression, during
the steady state only one PE on the LAN sends a Join(S,G), the others
receive it.  A PE with (S,G) state that sees a Join(S,G) from another PE
does not send one itself.)  The number of Join(S,G) messages the PE receives
over its VRF interfaces is k per minute.

So the messages received per minute to support the (S,G) state is either k
or k+1.  Note that this has no dependence whatsoever on the number of PEs.

What if BGP is used instead of PIM for PE-PE control?  In that case, no
PE-PE message needs to be sent at all, so the number of messages received
per minute to support the (S,G) state is k per minute.  Even if k is 1, the
most one can get is a 50% reduction; if k is 10 that goes down to 10%, if k
is 100 that goes down to 1%.  Considering the extra state needed to maintain
the (S,G) in both PIM form and BGP form, these reductions seem rather
modest, to say the least.

The methodological problem here is that Appendix A does not consider the PIM
overhead on the VRF interfaces.  But when BGP C-multicast routing is added,
it does not replace the C-PIM instance for a given MVPN, it just reduces by
1 the number of PIM interfaces supported by that C-PIM interface.  By
ignoring this fact (and by ignoring the cost of adding a second multicast
protocol without removing PIM) one makes a small difference look much more
significant than it really is.  It is not valid for a scalability analysis
to focus on just one aspect of the protocols unless it has first shown that
that one aspect is the bottleneck.

(Actually, adding BGP does not even fully reduce the number of PIM
interfaces by even 1, as C-PIM on a PE still needs to keep state for the its
"interface" to the other PEs; see, e.g., section 13.2.1 of
draft-ietf-l3vpn-2547bis-mcast-bgp.)

Appendix A talks of "order of magnitude" differences; that simply doesn't
exist.  In fact, the analysis of Appendix A fails to demonstrate any real
scaling advantage, in the steady state, for one protocol over the other.

Some amount of the analysis in Appendix A is devoted to the startup and
shutdown transients, when a PE first joins an (S,G) or prunes an (S,G).  A
rather big deal is made of the "prune override" procedures.  I have focused
this critique on the steady state, because if the steady state (as I defined
it above) is actually reached, the startup/shutdown transients are not very
important.  I suppose one might disagree with that, but certainly thee has
been no analysis showing that the transients are more important than the
steady state.

It would certainly be interesting to analyze the case where there is so much
multicast churn that steady state is never reached.  However, that cannot be
done by focusing on a single (S,G) over a limited time period.

Now let's look specifically at the ASM case.

One of the features of the BGP solution is that there is no way, in BGP,
to represent the PIM operation of "prune source S off the (*,G) tree".
To compensate for this, every PE in a given VPN is forced to keep track of
every (S,G) stream for which any PE in that VPN has receivers.  This is done
by means of the Source Active A-D routes.

There are two options that can be used.  In one option, where "inter-site
shared trees are not used", each PE functions as a C-RP, and then all the PE
of a given VPN must exchange information on all the (S,G) streams that they
see locally.  This option uses BGP to do something that is very similar to
what MSDP is used for today.  Let me quote from Pekka Savola's review of
this procedure:

         Active source BGP messages.  This is a duplication of a
         similar mechanism in MSDP (RFC3618) which has caused
         much grief in Internet.  Does this meant that when a
         host does 'nmap -sU 224.0.0.0/4' at a VPN site, this
         will result in about 268 million BGP active source
         updates being sent (2^28) in the SP backbone?

The only answer to this was to beef up the Security Considerations with
some required procedures that will prevent the PE from melting down entirely
in this case.  Of course, Appendix A's methodology of only considering a
single (S,G) prevents it from seeing that this particular BGP option
uses a mechanism whose scalability is already known to be poo,r and which
the multicast WGs have decided not even to bring forward to IPv6.

In the other, more sensible option, a PE originates a Source Active A-D
route for an (S,G) only when it receives a C-multicast Join route for that
(S,G).  However, the Source Active A-D route goes to all the PEs in the VPN,
not just to those that have receivers for that (S,G).  I believe this means
that the amount of state a given PE maintains in the BGP scheme is
proportional to the number of (S,G) flows for which any PE in the VPN has
receivers.  Actually, since the SA route contains the RD of the originating
PE, the situation is a bit worse; if (S,G) traffic is sent over the backbone
by more than one PE, the other PEs in the same VPN must maintain an (S,G) SA
A-D route for each such PE.

In the BGP solution, even if a PE has only (*,G) receivers, or even if it
has no receivers for G at all, it must maintain state for all the (S,G)s in
the MVPN, multiplied by the average number of upstream PEs per source.

Strangely, this scaling issue is considered to be insignificant by Appendix
A, for the following reasons:

        when PIM-based C-multicast routing is used there are
        ... possible control plane state transitions triggered by
        the reception of (S,G) packets ; such events would induce
        processing on all PEs joined to G

Control plane state transitions triggered by the reception of data packets
have not even been mentioned to this point, and no analysis has been given
to show whether or not they cause a scaling issue, or whether any such
scaling issue is mitigated by the use of BGP.  I also really don't see what
this has to do with the issue at hand, namely the added state overhead of
the Source Active A-D routes.  Certainly there is no analysis provided that
would show the SA A-D routes trade off in a favorable manner against the
data-driven state changes.

      when PIM-based C-multicast routing is used there are ...
      possible PIM Assert messages specific to (S,G) ; this would
      induce a message processing on each PE of the VPN for each
      PIM Assert message

It is true that the Source Active A-D routes eliminate the need for Assert
messages.  The implication is that this is favorable for BGP; however, there
does not appear to be any analysis showing that.

The new material on ASM does not appear to contain any analysis at all, just
unsupported assertions.

Let me also mention a few other, less major, problems in the Appendix:

        for each PIM Prune message, all PE routers on the LAN
        work to let the upstream PE determine the answer to the
        "did the last receiver leave?" question.

I believe that only one PE router needs to send a Join in order to override
the Prune, as once the others see that Join they reset their Join timers.
Of course, actual behavior varies depending on the value of the timer and
on the random jitter.

        We can see that answering the "last receiver leaves"
        question is a significant proportion of the work that the
        C-multicast routing building block has to make, and where
        the approaches differ most.

This assertion is made without support.

        the case where the RT-Constraint mechanisms [RFC4684] is
        not used is not covered;

Since RT-constraint is not a mandatory to implement mechanism, I don't see
why this is not covered.  Note though that it doesn't really matter in the
steady state.

>From A.1.1.1.  "PIM LAN procedures, by default"

   +------------+------------+---------------+----------+--------------+
   |            | upstream   | other PEs     | RR       | total across |
   |            | PE (1)     | (total across | (none)   | all          |
   |            |            | (#mvpn_PE-1)  |          | equipments   |
   |            |            | PEs)          |          |              |
   +------------+------------+---------------+----------+--------------+
   | first PE   | 1 m.e      | #MVPN_PE-1    | /        | #MVPN_PE m.e |
   | joins      |            | m.e           |          |              |
   +------------+------------+---------------+----------+--------------+

If the above row implies that the "other PEs" must process the join, where
"process" is defined (as it is in this Appendix) as "a message being parsed,
a lookup being done, and some action being taken (such has updating a
control plane or data plane state)", then one the next row is incorrect:

   +------------+------------+---------------+----------+--------------+   
   | for *each* | 1 m.e      | #mvpn_PE-1    | /        | #mvpn_PE m.e |
   | additional |            | m.e           |          |              |
   | PE joining |            |               |          |              |
   +------------+------------+---------------+----------+--------------+

If a second PE has "processed" the Join(S,G) from the first, then it doesn't
need to send a Join if it decides to join (S,G) itself; join suppression
will be in effect.

The third and fifth columns above are examples of cases where the Appendix
misleadingly multiplies by #mvpn_PE.  In fact, the number of messages
received by a given PE does not increase in proportion to #mvpn_PE.  This
problem infects the entire set of tables in the Appendix.

Section A.1.1.3

The RR has to maintain all the multicast states of all the clients (PEs), as
long as those multicast states have a PMSI as input interface.  That's way
more multicast states than any PE has to maintain. The analysis in this
Appendix simply ignores the negative scaling effects due to the large number
of RR clients.

The claim is made that when a second PE joins (S,G), the upstream PE does
not receive a new BGP route.  That may or may not be the case.  The second
PE will always need to send a C-multicast Join route to the RRs; in PIM the
second PE might not need to send any message.  Whether the RRs will forward
the route originated by the second PE depends upon whether that route is
more preferable (according to normal BGP procedures) than the route
originated by the first PE.  A similar observation applies when C-multicast
routes are withdrawn.


















   

   





From erosen@cisco.com  Thu Dec  3 08:40:13 2009
Return-Path: <erosen@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id BF5D13A6878 for <l3vpn@core3.amsl.com>; Thu,  3 Dec 2009 08:40:13 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.399
X-Spam-Level: 
X-Spam-Status: No, score=-6.399 tagged_above=-999 required=5 tests=[AWL=0.200,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id uANvasy9MIcj for <l3vpn@core3.amsl.com>; Thu,  3 Dec 2009 08:40:13 -0800 (PST)
Received: from rtp-iport-2.cisco.com (rtp-iport-2.cisco.com [64.102.122.149]) by core3.amsl.com (Postfix) with ESMTP id D31BE3A682B for <l3vpn@ietf.org>; Thu,  3 Dec 2009 08:40:12 -0800 (PST)
Authentication-Results: rtp-iport-2.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-AV: E=Sophos;i="4.47,336,1257120000"; d="scan'208";a="71889068"
Received: from rtp-core-2.cisco.com ([64.102.124.13]) by rtp-iport-2.cisco.com with ESMTP; 03 Dec 2009 16:40:04 +0000
Received: from erosen-linux.cisco.com (erosen-linux.cisco.com [161.44.70.34]) by rtp-core-2.cisco.com (8.13.8/8.14.3) with ESMTP id nB3Ge4hN009158; Thu, 3 Dec 2009 16:40:04 GMT
Received: from erosen-linux (localhost.localdomain [127.0.0.1]) by erosen-linux.cisco.com (8.13.1/8.13.1) with ESMTP id nB3Ge3gd003158;  Thu, 3 Dec 2009 11:40:03 -0500
To: Thomas Morin <thomas.morin@orange-ftgroup.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 
In-reply-to: Your message of Mon, 09 Nov 2009 16:47:58 +0900. <4AF7C92E.4050904@orange-ftgroup.com> 
Date: Thu, 03 Dec 2009 11:40:03 -0500
Message-ID: <3157.1259858403@erosen-linux>
From: Eric Rosen <erosen@cisco.com>
Cc: L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: erosen@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 03 Dec 2009 16:40:13 -0000

I believe that comparing segmented inter-AS P-tunnels with non-segmented
inter-AS P-tunnels gives the following comparison matrix:

- Scalability for MI-PMSIs: Segmented more scalable

- Scalability for S-PMSIs: Comparable scalability

- Complexity at ASBRs: segmented more complex (requires ASBRs to have more
  VPN-specific knowledge, and some options increased configuration at ASBRs)

- Independence of tunneling technology from one AS to another: segmented
  provides this, non-segmented does not.

- Compatibility with existing implementations: non-segmented provides this,
  segmented does not.

Looking at these considerations, I would not choose "segmented" as a
required feature of the minimal implementation, because I don't think anyone
will ever use MI-PMSIs instantiated as segmented inter-AS P-tunnels, and I
think the minimal implementation should emphasize simplicity over
generality.  But whatever the document recommends, it should take all the
considerations into account, not just the ones that best support its
conclusion.


From yakov@juniper.net  Tue Dec  8 07:16:53 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 6166B3A6832 for <l3vpn@core3.amsl.com>; Tue,  8 Dec 2009 07:16:53 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.556
X-Spam-Level: 
X-Spam-Status: No, score=-6.556 tagged_above=-999 required=5 tests=[AWL=0.043,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id tFVHmHQdk3gL for <l3vpn@core3.amsl.com>; Tue,  8 Dec 2009 07:16:51 -0800 (PST)
Received: from exprod7og109.obsmtp.com (exprod7og109.obsmtp.com [64.18.2.171]) by core3.amsl.com (Postfix) with ESMTP id 3AA3228C140 for <l3vpn@ietf.org>; Tue,  8 Dec 2009 07:16:46 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob109.postini.com ([64.18.6.12]) with SMTP ID DSNKSx5t06rmaxKxLS6FyZIaiDo6G0cZVBdy@postini.com; Tue, 08 Dec 2009 07:16:38 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Tue, 8 Dec 2009 07:08:20 -0800
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 8 Dec 2009 07:08:20 -0800
Received: from emailsmtp56.jnpr.net ([172.24.60.77]) by p-emlb01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 8 Dec 2009 07:08:20 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp56.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 8 Dec 2009 07:08:20 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nB8F8Jj31633; Tue, 8 Dec 2009 07:08:20 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912081508.nB8F8Jj31633@magenta.juniper.net>
To: erosen@cisco.com
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 
In-Reply-To: <3157.1259858403@erosen-linux> 
References: <3157.1259858403@erosen-linux>
X-MH-In-Reply-To: Eric Rosen <erosen@cisco.com> message dated "Thu, 03 Dec 2009 11:40:03 -0500."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <23294.1260284899.1@juniper.net>
Date: Tue, 8 Dec 2009 07:08:19 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 08 Dec 2009 15:08:20.0341 (UTC) FILETIME=[47401250:01CA7818]
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 08 Dec 2009 15:16:53 -0000

Eric,

> I believe that comparing segmented inter-AS P-tunnels with non-segmented
> inter-AS P-tunnels gives the following comparison matrix:
> 
> - Scalability for MI-PMSIs: Segmented more scalable

This is correct.

> - Scalability for S-PMSIs: Comparable scalability

This is correct, *but only* when C-multicast routing is done with PIM.

This is *incorrect* when C-multicast routing is done with BGP, as
when BGP is used for C-multicast routing, segmented P-tunnels are
more scalable than non-segmented due to the ability to aggregate
segmented S-PMSIs with I-PMSIs.

> - Complexity at ASBRs: segmented more complex (requires ASBRs to have more
>   VPN-specific knowledge, and some options increased configuration at ASBRs)

To be more precise, depending on whether inter-AS I-PMSI auto-discovery
routes are at the granularity of <source ASBR, MVPN> or <source AS,
MVPN>, the segmented P-tunnels would require either the same or
additional configuration on ASBRs as the non-segmented P-tunnels.

(Note that the issue of configuration has been already discussed on the mailing
list - see http://www.ietf.org/mail-archive/web/l3vpn/current/msg02420.html) 
  
> - Independence of tunneling technology from one AS to another: segmented
>   provides this, non-segmented does not.

This is correct.

> - Compatibility with existing implementations: non-segmented provides this,
>   segmented does not.

>From Merriam-Webster:

   Main Entry: compatible

      1 : capable of existing together in harmony <compatible theories> 
          <compatible people>
      2 : capable of cross-fertilizing freely or uniting vegetatively
      3 : capable of forming a homogeneous mixture that neither separates 
          nor is altered by chemical interaction
      4 : capable of being used in transfusion or grafting without 
          immunological reaction (as agglutination or tissue rejection)
      5 : designed to work with another device or system without 
          modification; especially : being a computer designed to operate 
          in the same manner and use the same software as another computer

Given the above definition of "compatible", your claim is *incorrect*,
as having non-segmented P-tunnels, by itself, does not provide
compatibility with "existing implementations".

This is because an implementation of non-segmented inter-AS I-PMSI
P-tunnels with BGP-based auto-discovery, as specified in
draft-ietf-l3vpn-mcast-bgp is *not* compatible with "existing
implementations" of non-segmented inter-AS I-PMSI P-tunnels with
BGP-based auto-discovery based on MDT-SAFI, even if both of these
implementations use non-segmented P-tunnels.

Likewise, an implementation of non-segmented S-PMSI P-tunnels with
BGP-based auto-discovery, as specified in draft-ietf-l3vpn-mcast-bgp
is *not compatible* with an "existing implementation" of non-segmented
S-PMSI P-tunnels with UDP-based auto-discovery, even if both these
implementations use non-segmented P-tunnels.

As a side comment, I would also suggest to revise the following
from section 6 of draft-ietf-l3vpn-mvpn-considerations-05:

   -  that implementations support segmented inter-AS tunnels and
      consider supporting non-segmented inter-AS tunnels (in order to
      maintain backwards compatibility and for migration) ;

to better aling it with the (correct) text in section 3.5 of
draft-ietf-l3vpn-mvpn-considerations-05.

> Looking at these considerations, I would not choose "segmented" as a
> required feature of the minimal implementation, because I don't think anyone
> will ever use MI-PMSIs instantiated as segmented inter-AS P-tunnels, and I
> think the minimal implementation should emphasize simplicity over
> generality. But whatever the document recommends, it should take all the
> considerations into account, not just the ones that best support its
> conclusion.

The document should take all the *valid* considerations into account.
Moreover, these considerations should reflect the requirements,
as specified in rfc4834. The document does all this.

Yakov.

From thomas.morin@orange-ftgroup.com  Tue Dec  8 08:38:57 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id BB9D028C1FA for <l3vpn@core3.amsl.com>; Tue,  8 Dec 2009 08:38:57 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.648
X-Spam-Level: 
X-Spam-Status: No, score=-2.648 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35, HTML_MESSAGE=0.001, J_CHICKENPOX_26=0.6, RCVD_IN_DNSWL_LOW=-1]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id kzZmdvECBnnl for <l3vpn@core3.amsl.com>; Tue,  8 Dec 2009 08:38:56 -0800 (PST)
Received: from p-mail2.rd.francetelecom.com (p-mail2.rd.francetelecom.com [195.101.245.16]) by core3.amsl.com (Postfix) with ESMTP id ABA1828C1F3 for <l3vpn@ietf.org>; Tue,  8 Dec 2009 08:38:55 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp1.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Tue, 8 Dec 2009 17:38:44 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Tue, 8 Dec 2009 17:38:43 +0100
Message-ID: <4B1E8112.6070007@orange-ftgroup.com>
Date: Tue, 08 Dec 2009 17:38:42 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091130 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05
References: <3157.1259858403@erosen-linux> <200912081508.nB8F8Jj31633@magenta.juniper.net>
In-Reply-To: <200912081508.nB8F8Jj31633@magenta.juniper.net>
Content-Type: multipart/alternative; boundary="------------090706010407040409080006"
X-OriginalArrivalTime: 08 Dec 2009 16:38:43.0720 (UTC) FILETIME=[E7D60880:01CA7824]
Cc: Yakov Rekhter <yakov@juniper.net>, erosen@cisco.com
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 08 Dec 2009 16:38:57 -0000

This is a multi-part message in MIME format.
--------------090706010407040409080006
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Hello Eric, Yakov,

Given your exchange, I would propose to add to section 3.5, a modified 
version of Eric's points taking also into account Yakov's suggestions 
and the discussion in June/July (text to be added before the last 
paragraph of section 3.5, except of course text not included):

    The following is a summary comparison matrix between the "segmented
    inter-AS P-tunnels" and "non-segmented inter-AS P-tunnels" approaches:

    - Scalability for MI-PMSIs: "Segmented inter-AS P-tunnels" more scalable

    [unmodified]

    - Scalability for S-PMSIs: Comparable scalability when the "Full
    per-MVPN PIM peering across an MI-PMSI" approach is used for
    C-multicast routing; "segmented inter-AS P-tunnels" more scalable
    when BGP is used for C-multicast routing (due to the ability to
    aggregate S-PMSIs with I-PMSI when BGP is used for C-multicast routing)

    [modified to take into account S-PMSI aggregation as discussed in
    June/July and reminded below]

    - Configuration at ASBRs: depending on whether segmented P-tunnels
    have a granularity of source ASBR or source AS, the "segmented
    inter-AS P-tunnels" approach would require respectively the same or
    additional configuration on ASBRs as the "non-segmented inter-AS
    P-tunnels" approach

    [modified to take into account the discussion in June/July and
    reminded below]

    - Independence of tunneling technology from one AS to another: the
    "segmented inter-AS P-tunnels" approach provides this, the
    "non-segmented inter-AS P-tunnels" approach does not.

    [unmodified]

    - Facilitated co-existence with existing deployments, and providing
    a lighter engineering in some scenarios : the "non-segmented
    inter-AS P-tunnels" approach provides this, the "segmented inter-AS
    P-tunnels" approach does not.

    [unmodified to avoid the inexact "compatibility" term]

Moreover, the sixth item in section 6 "Summary of recommendation" can be 
rewritten to avoid the inexact term of "backward compatibility":

    that implementations support segmented inter-AS tunnels and consider
    supporting non-segmented inter-AS tunnels (in order to facilitate
    co-existence with existing deployments, and provide a lighter
    engineering in some scenarios) ;

Please send feedback on these proposed changes.

Thanks,

-Thomas






Yakov Rekhter:
> Eric,
>
>    
>> I believe that comparing segmented inter-AS P-tunnels with non-segmented
>> inter-AS P-tunnels gives the following comparison matrix:
>>
>> - Scalability for MI-PMSIs: Segmented more scalable
>>      
> This is correct.
>
>    
>> - Scalability for S-PMSIs: Comparable scalability
>>      
> This is correct, *but only* when C-multicast routing is done with PIM.
>
> This is *incorrect* when C-multicast routing is done with BGP, as
> when BGP is used for C-multicast routing, segmented P-tunnels are
> more scalable than non-segmented due to the ability to aggregate
> segmented S-PMSIs with I-PMSIs.
>
>    
>> - Complexity at ASBRs: segmented more complex (requires ASBRs to have more
>>    VPN-specific knowledge, and some options increased configuration at ASBRs)
>>      
> To be more precise, depending on whether inter-AS I-PMSI auto-discovery
> routes are at the granularity of<source ASBR, MVPN>  or<source AS,
> MVPN>, the segmented P-tunnels would require either the same or
> additional configuration on ASBRs as the non-segmented P-tunnels.
>
> (Note that the issue of configuration has been already discussed on the mailing
> list - seehttp://www.ietf.org/mail-archive/web/l3vpn/current/msg02420.html)
>
>    
>> - Independence of tunneling technology from one AS to another: segmented
>>    provides this, non-segmented does not.
>>      
> This is correct.
>
>    
>> - Compatibility with existing implementations: non-segmented provides this,
>>    segmented does not.
>>      
>  From Merriam-Webster:
>
>     Main Entry: compatible
>
>        1 : capable of existing together in harmony<compatible theories>
>            <compatible people>
>        2 : capable of cross-fertilizing freely or uniting vegetatively
>        3 : capable of forming a homogeneous mixture that neither separates
>            nor is altered by chemical interaction
>        4 : capable of being used in transfusion or grafting without
>            immunological reaction (as agglutination or tissue rejection)
>        5 : designed to work with another device or system without
>            modification; especially : being a computer designed to operate
>            in the same manner and use the same software as another computer
>
> Given the above definition of "compatible", your claim is *incorrect*,
> as having non-segmented P-tunnels, by itself, does not provide
> compatibility with "existing implementations".
>
> This is because an implementation of non-segmented inter-AS I-PMSI
> P-tunnels with BGP-based auto-discovery, as specified in
> draft-ietf-l3vpn-mcast-bgp is *not* compatible with "existing
> implementations" of non-segmented inter-AS I-PMSI P-tunnels with
> BGP-based auto-discovery based on MDT-SAFI, even if both of these
> implementations use non-segmented P-tunnels.
>
> Likewise, an implementation of non-segmented S-PMSI P-tunnels with
> BGP-based auto-discovery, as specified in draft-ietf-l3vpn-mcast-bgp
> is *not compatible* with an "existing implementation" of non-segmented
> S-PMSI P-tunnels with UDP-based auto-discovery, even if both these
> implementations use non-segmented P-tunnels.
>
> As a side comment, I would also suggest to revise the following
> from section 6 of draft-ietf-l3vpn-mvpn-considerations-05:
>
>     -  that implementations support segmented inter-AS tunnels and
>        consider supporting non-segmented inter-AS tunnels (in order to
>        maintain backwards compatibility and for migration) ;
>
> to better aling it with the (correct) text in section 3.5 of
> draft-ietf-l3vpn-mvpn-considerations-05.
>
>    
>> Looking at these considerations, I would not choose "segmented" as a
>> required feature of the minimal implementation, because I don't think anyone
>> will ever use MI-PMSIs instantiated as segmented inter-AS P-tunnels, and I
>> think the minimal implementation should emphasize simplicity over
>> generality. But whatever the document recommends, it should take all the
>> considerations into account, not just the ones that best support its
>> conclusion.
> The document should take all the *valid* considerations into account.
> Moreover, these considerations should reflect the requirements,
> as specified in rfc4834. The document does all this.
>
> Yakov.
>    

--------------090706010407040409080006
Content-Type: text/html; charset=ISO-8859-1
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">
</head>
<body text="#000000" bgcolor="#ffffff">
Hello Eric, Yakov,<br>
<br>
Given your exchange, I would propose to add to section 3.5, a modified
version of Eric's points taking also into account Yakov's suggestions
and the discussion in June/July (text
to be added before the last paragraph of section 3.5, except of course
text not included):<br>
<blockquote>The following is a summary comparison matrix between the
"segmented inter-AS P-tunnels" and "non-segmented inter-AS P-tunnels"
approaches:<br>
  <br>
- Scalability for MI-PMSIs: "Segmented inter-AS P-tunnels" more scalable<br>
  <br>
[unmodified]<br>
  <br>
- Scalability for S-PMSIs: Comparable scalability when the "Full
per-MVPN PIM peering across an MI-PMSI" approach is used for
C-multicast routing; "segmented inter-AS P-tunnels" more scalable when
BGP is used for C-multicast routing (due to the ability to aggregate
S-PMSIs with I-PMSI when BGP is used for C-multicast routing)<br>
  <br>
[modified to take into account S-PMSI aggregation as discussed in
June/July
and reminded below]<br>
  <br>
- Configuration at ASBRs: depending on whether segmented P-tunnels have
a granularity of source ASBR or source AS, the "segmented inter-AS
P-tunnels" approach would require respectively the same or additional
configuration on ASBRs as the "non-segmented inter-AS P-tunnels"
approach<br>
  <br>
[modified to take into account the discussion in June/July and reminded
below]<br>
  <br>
- Independence of tunneling technology from one AS to another: the
"segmented inter-AS P-tunnels" approach provides this, the
"non-segmented inter-AS P-tunnels" approach does not.<br>
  <br>
[unmodified]<br>
  <br>
- Facilitated co-existence with existing deployments, and providing a
lighter engineering in some scenarios : the "non-segmented inter-AS
P-tunnels" approach
provides this, the "segmented inter-AS P-tunnels" approach does not.<br>
  <br>
[unmodified to avoid the inexact "compatibility" term]<br>
  <br>
</blockquote>
Moreover, the sixth item in section 6 "Summary of recommendation" can
be rewritten to avoid the inexact term of "backward compatibility":<br>
<blockquote>that implementations support segmented inter-AS tunnels
and consider supporting non-segmented inter-AS tunnels (in order to
facilitate co-existence with existing deployments, and provide a
lighter engineering in some scenarios) ;<br>
</blockquote>
Please send feedback on these proposed changes.<br>
<br>
Thanks,<br>
<br>
-Thomas<br>
<br>
<br>
<br>
<br>
<br>
<br>
Yakov Rekhter:
<blockquote cite="mid:200912081508.nB8F8Jj31633@magenta.juniper.net"
 type="cite">
  <pre wrap="">Eric,

  </pre>
  <blockquote type="cite">
    <pre wrap="">I believe that comparing segmented inter-AS P-tunnels with non-segmented
inter-AS P-tunnels gives the following comparison matrix:

- Scalability for MI-PMSIs: Segmented more scalable
    </pre>
  </blockquote>
  <pre wrap="">This is correct.

  </pre>
  <blockquote type="cite">
    <pre wrap="">- Scalability for S-PMSIs: Comparable scalability
    </pre>
  </blockquote>
  <pre wrap="">This is correct, *but only* when C-multicast routing is done with PIM.

This is *incorrect* when C-multicast routing is done with BGP, as
when BGP is used for C-multicast routing, segmented P-tunnels are
more scalable than non-segmented due to the ability to aggregate
segmented S-PMSIs with I-PMSIs.

  </pre>
  <blockquote type="cite">
    <pre wrap="">- Complexity at ASBRs: segmented more complex (requires ASBRs to have more
  VPN-specific knowledge, and some options increased configuration at ASBRs)
    </pre>
  </blockquote>
  <pre wrap="">To be more precise, depending on whether inter-AS I-PMSI auto-discovery
routes are at the granularity of &lt;source ASBR, MVPN&gt; or &lt;source AS,
MVPN&gt;, the segmented P-tunnels would require either the same or
additional configuration on ASBRs as the non-segmented P-tunnels.

(Note that the issue of configuration has been already discussed on the mailing
list - see <a class="moz-txt-link-freetext"
 href="http://www.ietf.org/mail-archive/web/l3vpn/current/msg02420.html">http://www.ietf.org/mail-archive/web/l3vpn/current/msg02420.html</a>) 
  
  </pre>
  <blockquote type="cite">
    <pre wrap="">- Independence of tunneling technology from one AS to another: segmented
  provides this, non-segmented does not.
    </pre>
  </blockquote>
  <pre wrap="">This is correct.

  </pre>
  <blockquote type="cite">
    <pre wrap="">- Compatibility with existing implementations: non-segmented provides this,
  segmented does not.
    </pre>
  </blockquote>
  <pre wrap="">From Merriam-Webster:

   Main Entry: compatible

      1 : capable of existing together in harmony &lt;compatible theories&gt; 
          &lt;compatible people&gt;
      2 : capable of cross-fertilizing freely or uniting vegetatively
      3 : capable of forming a homogeneous mixture that neither separates 
          nor is altered by chemical interaction
      4 : capable of being used in transfusion or grafting without 
          immunological reaction (as agglutination or tissue rejection)
      5 : designed to work with another device or system without 
          modification; especially : being a computer designed to operate 
          in the same manner and use the same software as another computer

Given the above definition of "compatible", your claim is *incorrect*,
as having non-segmented P-tunnels, by itself, does not provide
compatibility with "existing implementations".

This is because an implementation of non-segmented inter-AS I-PMSI
P-tunnels with BGP-based auto-discovery, as specified in
draft-ietf-l3vpn-mcast-bgp is *not* compatible with "existing
implementations" of non-segmented inter-AS I-PMSI P-tunnels with
BGP-based auto-discovery based on MDT-SAFI, even if both of these
implementations use non-segmented P-tunnels.

Likewise, an implementation of non-segmented S-PMSI P-tunnels with
BGP-based auto-discovery, as specified in draft-ietf-l3vpn-mcast-bgp
is *not compatible* with an "existing implementation" of non-segmented
S-PMSI P-tunnels with UDP-based auto-discovery, even if both these
implementations use non-segmented P-tunnels.

As a side comment, I would also suggest to revise the following
from section 6 of draft-ietf-l3vpn-mvpn-considerations-05:

   -  that implementations support segmented inter-AS tunnels and
      consider supporting non-segmented inter-AS tunnels (in order to
      maintain backwards compatibility and for migration) ;

to better aling it with the (correct) text in section 3.5 of
draft-ietf-l3vpn-mvpn-considerations-05.

  </pre>
  <blockquote type="cite">
    <pre wrap="">Looking at these considerations, I would not choose "segmented" as a
required feature of the minimal implementation, because I don't think anyone
will ever use MI-PMSIs instantiated as segmented inter-AS P-tunnels, and I
think the minimal implementation should emphasize simplicity over
generality. But whatever the document recommends, it should take all the
considerations into account, not just the ones that best support its
conclusion.    </pre>
  </blockquote>
  <pre wrap="">The document should take all the *valid* considerations into account.
Moreover, these considerations should reflect the requirements,
as specified in rfc4834. The document does all this.

Yakov.
  </pre>
</blockquote>
</body>
</html>

--------------090706010407040409080006--

From yakov@juniper.net  Thu Dec 10 07:42:27 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id B2C7B3A672F for <l3vpn@core3.amsl.com>; Thu, 10 Dec 2009 07:42:27 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.532
X-Spam-Level: 
X-Spam-Status: No, score=-6.532 tagged_above=-999 required=5 tests=[AWL=0.067,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ZswRJYqxtYtY for <l3vpn@core3.amsl.com>; Thu, 10 Dec 2009 07:42:27 -0800 (PST)
Received: from exprod7og127.obsmtp.com (exprod7og127.obsmtp.com [64.18.2.210]) by core3.amsl.com (Postfix) with ESMTP id D94D53A6A81 for <l3vpn@ietf.org>; Thu, 10 Dec 2009 07:42:25 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob127.postini.com ([64.18.6.12]) with SMTP ID DSNKSyEW1n0YfXTH+7Io5Y1+pkQoGM6VGlbX@postini.com; Thu, 10 Dec 2009 07:42:16 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Thu, 10 Dec 2009 07:35:48 -0800
Received: from p-emlb02-sac.jnpr.net ([66.129.254.47]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Thu, 10 Dec 2009 07:35:48 -0800
Received: from emailsmtp56.jnpr.net ([172.24.60.77]) by p-emlb02-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Thu, 10 Dec 2009 07:35:48 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp56.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Thu, 10 Dec 2009 07:35:47 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBAFZlj04113; Thu, 10 Dec 2009 07:35:47 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912101535.nBAFZlj04113@magenta.juniper.net>
To: Thomas Morin <thomas.morin@orange-ftgroup.com>
Subject: "Other PEs" in draft-ietf-l3vpn-mvpn-considerations-05 
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <21893.1260459347.1@juniper.net>
Date: Thu, 10 Dec 2009 07:35:47 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 10 Dec 2009 15:35:47.0636 (UTC) FILETIME=[71F0D740:01CA79AE]
Cc: L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 10 Dec 2009 15:42:28 -0000

Thomas,

In A.1.1.3 why does the "Other PEs" column show 2 m.e. on first join ?
Ditto for each additional PE joining.

Yakov.

From tnadeau@lucidvision.com  Thu Dec 10 12:28:12 2009
Return-Path: <tnadeau@lucidvision.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 8E2843A6A5F for <l3vpn@core3.amsl.com>; Thu, 10 Dec 2009 12:28:12 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.796
X-Spam-Level: 
X-Spam-Status: No, score=-3.796 tagged_above=-999 required=5 tests=[AWL=0.801,  BAYES_00=-2.599, GB_I_LETTER=-2, HTML_MESSAGE=0.001]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id S2OEsNmWpcJb for <l3vpn@core3.amsl.com>; Thu, 10 Dec 2009 12:28:11 -0800 (PST)
Received: from lucidvision.com (lucidvision.com [72.71.250.34]) by core3.amsl.com (Postfix) with ESMTP id 5644A3A6A4E for <l3vpn@ietf.org>; Thu, 10 Dec 2009 12:28:11 -0800 (PST)
Received: from [192.168.1.120] (unknown [72.71.250.36]) by lucidvision.com (Postfix) with ESMTP id E0EF0422D1E4 for <l3vpn@ietf.org>; Thu, 10 Dec 2009 15:27:59 -0500 (EST)
From: Thomas Nadeau <tnadeau@lucidvision.com>
Content-Type: multipart/alternative; boundary=Apple-Mail-13-849322099
Subject: ONTC-PRISM Newsletter: Call for Papers - Special Issue on Cloud Computing and Data Center Networks
Date: Thu, 10 Dec 2009 15:27:59 -0500
Message-Id: <27404495-7069-4E80-A92D-59D88143A63B@lucidvision.com>
To: l3vpn@ietf.org
Mime-Version: 1.0 (Apple Message framework v1077)
X-Mailer: Apple Mail (2.1077)
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 10 Dec 2009 20:28:12 -0000

--Apple-Mail-13-849322099
Content-Transfer-Encoding: 7bit
Content-Type: text/plain;
	charset=us-ascii

FYI:


ONTC-PRISM Newsletter: Call for Papers - Special Issue on Cloud Computing
and Data Center Networks

Scope: IT services have proliferated significantly through most
enterprises and even to end-users. This explosion of IT services has led
to the emergence, growth and consolidation of computationally intensive
data centers. Data centers are principle network wide entities that are
responsible for the successful provisioning and scaling of IT services
across disparate users over a heterogenous network. We seek to understand
the network-wide design, implementation, architecture and technology
aspects of data centers and of emerging applications such as cloud
computing. We seek original review articles on technologies,
architectures, implementation excercises associated with data centers
including cloud computing, distributed and virtual data centers and
applications using such infrastructures. Articles must be written as per
the format prescribed below.

Keywords: Cloud computing, data centers, Carrier Ethernet, MPLS-TP,
PBB-TE, Virtualization, Consolidation.

Format: up to 3 pages with single column format, Times new roman 10 point
font for the text and 8 point font for the figure captions.

Submission information: Please send manuscripts to
submissions@ontc-prism.org. All articles will be reviewed by the TAB
and/or external reviewers.

Timeline:
Submission Deadline: Dec 31, 2009
Notification: Feb 1, 2010
Issue date: Feb 15, 2010

TAB in-charge: Thomas Nadeau BT Group LLC.

About ONTC Prism (www.ontc-prism.org): Prism is a quarterly newsletter
aimed at the dissemination of relevant content pertaining to optical and
high-speed networks and related growth areas across industry and academia.
Prism seeks to promote the growth of optical networking activity by
creation of a unified knowledge base and aims to create a communication
bridge between industry and academia in terms of research frontiers and
complementary strategies for future growth.

--Apple-Mail-13-849322099
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=us-ascii

<html><head></head><body style=3D"word-wrap: break-word; =
-webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><span =
class=3D"Apple-style-span" style=3D"font-family: monospace; =
"><div>FYI:</div><div><br></div><div><br></div>ONTC-PRISM Newsletter: =
Call for Papers - Special Issue on Cloud Computing<br>and Data Center =
Networks<br><br>Scope: IT services have proliferated significantly =
through most<br>enterprises and even to end-users. This explosion of IT =
services has led<br>to the emergence, growth and consolidation of =
computationally intensive<br>data centers. Data centers are principle =
network wide entities that are<br>responsible for the successful =
provisioning and scaling of IT services<br>across disparate users over a =
heterogenous network. We seek to understand<br>the network-wide design, =
implementation, architecture and technology<br>aspects of data centers =
and of emerging applications such as cloud<br>computing. We seek =
original review articles on technologies,<br>architectures, =
implementation excercises associated with data centers<br>including =
cloud computing, distributed and virtual data centers =
and<br>applications using such infrastructures. Articles must be written =
as per<br>the format prescribed below.<br><br>Keywords: Cloud computing, =
data centers, Carrier Ethernet, MPLS-TP,<br>PBB-TE, Virtualization, =
Consolidation.<br><br>Format: up to 3 pages with single column format, =
Times new roman 10 point<br>font for the text and 8 point font for the =
figure captions.<br><br>Submission information: Please send manuscripts =
to<br><a =
href=3D"mailto:submissions@ontc-prism.org">submissions@ontc-prism.org</a>.=
 All articles will be reviewed by the TAB<br>and/or external =
reviewers.<br><br>Timeline:<br>Submission Deadline: Dec 31, =
2009<br>Notification: Feb 1, 2010<br>Issue date: Feb 15, 2010<br><br>TAB =
in-charge: Thomas Nadeau BT Group LLC.<br><br>About ONTC Prism (<a =
href=3D"http://www.ontc-prism.org/">www.ontc-prism.org</a>): Prism is a =
quarterly newsletter<br>aimed at the dissemination of relevant content =
pertaining to optical and<br>high-speed networks and related growth =
areas across industry and academia.<br>Prism seeks to promote the growth =
of optical networking activity by<br>creation of a unified knowledge =
base and aims to create a communication<br>bridge between industry and =
academia in terms of research frontiers and<br>complementary strategies =
for future growth.<br></span></body></html>=

--Apple-Mail-13-849322099--

From thomas.morin@orange-ftgroup.com  Fri Dec 11 09:25:38 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id F14443A687B for <l3vpn@core3.amsl.com>; Fri, 11 Dec 2009 09:25:37 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.649
X-Spam-Level: 
X-Spam-Status: No, score=-1.649 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35, J_CHICKENPOX_36=0.6]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id qkunKX2d56Dj for <l3vpn@core3.amsl.com>; Fri, 11 Dec 2009 09:25:37 -0800 (PST)
Received: from r-mail1.rd.francetelecom.com (r-mail1.rd.francetelecom.com [217.108.152.41]) by core3.amsl.com (Postfix) with ESMTP id CD31E3A67D3 for <l3vpn@ietf.org>; Fri, 11 Dec 2009 09:25:36 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp2.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Fri, 11 Dec 2009 18:25:21 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Fri, 11 Dec 2009 18:25:22 +0100
Message-ID: <4B228081.3060409@orange-ftgroup.com>
Date: Fri, 11 Dec 2009 18:25:21 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: Yakov Rekhter <yakov@juniper.net>
Subject: Re: "Other PEs" in draft-ietf-l3vpn-mvpn-considerations-05
References: <200912101535.nBAFZlj04113@magenta.juniper.net>
In-Reply-To: <200912101535.nBAFZlj04113@magenta.juniper.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 11 Dec 2009 17:25:22.0334 (UTC) FILETIME=[EB2DE3E0:01CA7A86]
Cc: L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 11 Dec 2009 17:25:38 -0000

Yakov,

Yakov Rekhter:
> In A.1.1.3 why does the "Other PEs" column show 2 m.e. on first join ?
> Ditto for each additional PE joining.
>    
The "Other PEs" column counts the m.e for all PEs,except for the upstream PE, which is in a separate column.

In the BGP case you mention (A.1.1.3) the only PE that processes a 
message is the upstream PE. The number should thus be 0 m.e and not 2 
m.e.  The mistake is a left-over from previous versions of this analyses 
where the recommendation on using RT-constraint in 
draft-ietf-2547bis-mcast-bgp was not taken into account.

In fact, section A.1.1.1 has a similar "off by one" unit mistake : the 
joiner PE does not itself see and process its own joins, and where the 
table read "#mVPN_PEs - 1", it should in fact read "#mVPN_PEs - 2".

I will update the two tables with correct numbers.

-Thomas

From mn1921@att.com  Fri Dec 11 09:08:53 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id EC9433A69F4 for <l3vpn@core3.amsl.com>; Fri, 11 Dec 2009 09:08:53 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -105.999
X-Spam-Level: 
X-Spam-Status: No, score=-105.999 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, J_CHICKENPOX_51=0.6, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id HRRNDnErKOlP for <l3vpn@core3.amsl.com>; Fri, 11 Dec 2009 09:08:52 -0800 (PST)
Received: from mail161.messagelabs.com (mail161.messagelabs.com [216.82.253.115]) by core3.amsl.com (Postfix) with ESMTP id A61463A69A0 for <l3vpn@ietf.org>; Fri, 11 Dec 2009 09:08:52 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-5.tower-161.messagelabs.com!1260551319!13210554!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 10028 invoked from network); 11 Dec 2009 17:08:39 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-5.tower-161.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 11 Dec 2009 17:08:39 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBBH8ZBp000994 for <l3vpn@ietf.org>; Fri, 11 Dec 2009 12:08:35 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBBH8Wgd000987 for <l3vpn@ietf.org>; Fri, 11 Dec 2009 12:08:32 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
x-cr-puzzleid: {A4892589-B0D9-459A-833C-9F22C29EB1AA}
x-cr-hashedpuzzle: KDE= AxXk Bbb6 CnDj C8pH DkUU FEhv FRyP GP6d HWJk IKzH MKgz P2iZ P8zM QTcz VQZ1; 2; bAAzAHYAcABuAEAAaQBlAHQAZgAuAG8AcgBnADsAdABoAG8AbQBhAHMALgBtAG8AcgBpAG4AQABvAHIAYQBuAGcAZQAtAGYAdABnAHIAbwB1AHAALgBjAG8AbQA=; Sosha1_v1; 7; {A4892589-B0D9-459A-833C-9F22C29EB1AA}; bQBuADEAOQAyADEAQABhAHQAdAAuAGMAbwBtAA==; Fri, 11 Dec 2009 17:08:06 GMT; ZAByAGEAZgB0AC0AaQBlAHQAZgAtAGwAMwB2AHAAbgAtAG0AdgBwAG4ALQBjAG8AbgBzAGkAZABlAHIAYQB0AGkAbwBuAHMALQAwADUA
Content-class: urn:content-classes:message
Subject: draft-ietf-l3vpn-mvpn-considerations-05 
Date: Fri, 11 Dec 2009 12:08:06 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A200446CE07@misout7msgusr7e.ugd.att.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 
Thread-Index: Acp1O0ao6zG7WAmHTMqAD8KxvPngiQAAFm5QAF/E+IAAJwVhYAABU4kQAACY0BAABnS14AAsu16wAAKm7yAAA3cIcAAHkxqwAAbz/qAAAsdU4AAB/pDQAAA/YPAAG9YXMAAAccVwAABSn8AAAI7dAAA0HKXgAAI6KmAAAYeeMAAIpTRgAB6tGoA=
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: "Thomas Morin" <thomas.morin@orange-ftgroup.com>
X-Mailman-Approved-At: Fri, 11 Dec 2009 09:33:09 -0800
Cc: l3vpn@ietf.org
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 11 Dec 2009 17:12:10 -0000

Hi Thomas,

I have several comments on the latest version of the draft:

1. I believe that the draft should contain some analysis on how PIM-SM
differs when handled by BGP signaling.  The difference is significant
enough to warrant at least a theoretical consideration if no significant
field experience is available.  Currently, vast majority of multicast
applications used in the enterprise networks are PIM-SM based. It is
important to analyze the BGP translation of pruning source traffic off
the shared tree because this is the most significant change from PIM-SM.

When BGP is used for PE-PE C-multicast signaling the data-driven PIM-SM
state changes are replaced by additional BGP (Source Active) routes and
timer-driven PIM messages. Let me point out just few aspects related to
those changes that should be considered, starting from timer-driven
Prune(S,G,rpt).

When a single forwarder selection is the procedure used to avoid sending
duplicate packets to MVPN customers during switching from RPT to SPT,
the correct setting of the above mentioned timer to prune source traffic
off the shared tree is important. It is not going to be trivial to tune
such timer in large networks, because it can cause either duplicate
traffic (if timer is set too high) or loss of traffic (if timer is set
too low). This approach to PIM-SM changes the way the PIM-SM multicast
trees are constructed, and it is not known to what extend this can
impact existing multicast applications.
If the solution to this is to make the timer for sending Prune(S,G,rpt)
sufficiently high (to assure that SPT is created) and avoid duplicates
sent to customers by having egress PEs discard data packets received
from the "wrong" upstream PE, this should be included in MVPN
considerations draft. It should be also considered that it may not
always be possible for the egress PE to determine the upstream PE, in
which case the single forwarder selection has to be used.=20

Regarding Source Active A-D routes, it was pointed out recently and in
the past that the amount of BGP state they generate should be analyzed.
The Appendix A ignores this important scalability aspect of
BGP-signaling.=20
The field experience shows that there are multicast applications used in
enterprise VPNs such that only control messages, but no multicast data,
are sent across MVPN backbone. When BGP-multicast signaling is used, the
Source Active A-D route is originated for every C-(S,G) flow that
registers with C-RP and such that the C-RP joins "S" across MVPN
backbone, even if no inter-site data is sent for this flow. Hence, the
BGP scheme requires not only that for any given C-(S,G) flow that enters
MVPN backbone an SA message is announced and maintained by all PEs in a
given MVPN, but that SA messages are also announced for flows that never
enter the MVPN backbone.

2. The considerations draft should also point out the difference how BGP
vs. PIM (used as PE-PE signaling protocol) handles the upstream PE
selection in case of multi-homed sources/RPs.=20

Typically, in a layer 3 VPN deployment it is not known whether or not
all VRFs of a given VPN have the same best route to a given source.  In
other words, VRFs routing policies are provisioned independently of each
other. If all VRFs of a given MVPN have the same best/ installed route
to a multi-homed source/RP then this VPN customer *expects* that the
source/RP will be joined via the installed route.=20

When PIM is used as PE-PE C-signaling, the installed route towards the
source is always chosen as UHM. If different VRFs have different best
routes to C-S or C-RP, the Assert process is a tie breaker. No steady
state duplicate traffic is sent across MVPN backbone.

Now let's consider BGP as the PE-PE signaling protocol and let's assume
the UMH policy is to choose the installed/best route towards C-source or
C-RP. If different VRFs have different installed routes to C-S/C-RP then
there will be a steady state duplication of traffic across the MVPN
backbone (to the same set of PEs if MI-PMSI is used). This also requires
that the egress PEs be able to discard data traffic that arrives from
the "wrong" upstream PE. Such duplication of traffic might not be
operationally acceptable to a service provider because of too much
network bandwidth being consumed. The other UMH policy that can be used
with BGP signaling where duplicates are not generated, is a single or
default forwarder selection. However, single forwarder selection might
not choose the installed/best route (to a multi-homed C-S/C-RP) *even*
if all VRFs have the same best route to C-S/C-RP. Hence, in a MVPN where
all VRFs have the same best path to C-S or C-RP, its multicast traffic
might not flow via the best path as is expected by the VPN customer.

3. In section A.1.1.3, why the "each additional PE joining" row for the
"upstream PE" column shows "0"? The same question for "upstream PEs"
column and the "each PE leaving" row.

4. Why Appendix A assumes that the RT-constraint is deployed when BGP
signaling is used?  This is not a mandatory implementation.

5. Why Appendix A analyzes PIM LAN procedure with explicit tracking but
not the explicit tracking using BGP-signaling?=20

6. The appendix states that "we propose here a rough estimation of the
load of PEs". And it is in fact a very *rough* estimation and not even a
complete *rough* estimation as it almost dismisses ASM analysis. The
appendix treats different and, in fact, not comparable kinds of
"processing" as *equal* and uses the same unit "m.e" to measure them.=20
For example, the first row in the table in section A.1.1.1, says that to
process the first join on the upstream PE takes "1 m.e" and that the
same amount of processing, "1 m.e", is required on every other PE in
MVPN. Clearly, the amount of processing on the upstream PE to which the
joined is destined is not the same as the "processing" on other PEs. Any
other PE than the upstream PE will simply discard this Join because the
neighbor address in the PIM Join is not this PE address. While the
upstream PE has to create the (s,g) state in the control plane, program
the data plane for forwarding, and send Join (s,g) message upstream.
This is by any measure much more processing than discarding the Join
based on the address. Yet, it is assumed that the resources used in both
cases are the same.=20
Another example is the row "each PE leaving" (but last) in the same
table in section A.1.1.1. The only processing the upstream PE does is
setting a timer for receiving prune-override PIM Join message, and when
such PIM join is received it cancels the timer. This is estimated as
"2 m.e", yet, this is much less "work" to be done by the upstream PE
than in the previous "first PE joins" example. In addition, the column
"other PEs" in the row "each PE leaving" states that every other PE in
MVPN will also have to do an action worth "2 m.e". Yet, only the PEs
that have interested receivers in (s,g) have to perform a task (i.e.,
randomly generate a join message delay timer and cancel it when they
hear a join from another PE; only one of those PEs will send a PIM Join
message). The PEs that do need to receive (s,g) traffic will simply drop
the (s,g) prune message. Hence, the application of the multiplier
#MVPN_PE in this row/column is also not correct.
This "problem" of treating not equal actions as equal repeats itself
throughout the scaling analysis in the appendix. Such analysis is not
only *rough* but misleading, and because of this may lead to wrong
conclusions. I believe that unless the appendix A is corrected by
removing all not comparable, or even wrong estimations, it should be
removed from the draft.=20

Maria


From wwwrun@core3.amsl.com  Fri Dec 11 10:58:15 2009
Return-Path: <wwwrun@core3.amsl.com>
X-Original-To: l3vpn@ietf.org
Delivered-To: l3vpn@core3.amsl.com
Received: by core3.amsl.com (Postfix, from userid 30) id 767F83A68DA; Fri, 11 Dec 2009 10:58:15 -0800 (PST)
X-idtracker: yes
From: The IESG <iesg-secretary@ietf.org>
To: IETF-Announce <ietf-announce@ietf.org>
Subject: Protocol Action: 'BGP Encodings and Procedures for Multicast in MPLS/BGP IP VPNs' to Proposed Standard
Message-Id: <20091211185815.767F83A68DA@core3.amsl.com>
Date: Fri, 11 Dec 2009 10:58:15 -0800 (PST)
Cc: l3vpn chair <l3vpn-chairs@tools.ietf.org>, Internet Architecture Board <iab@iab.org>, l3vpn mailing list <l3vpn@ietf.org>, RFC Editor <rfc-editor@rfc-editor.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 11 Dec 2009 18:58:15 -0000

The IESG has approved the following document:

- 'BGP Encodings and Procedures for Multicast in MPLS/BGP IP VPNs '
   <draft-ietf-l3vpn-2547bis-mcast-bgp-08.txt> as a Proposed Standard


This document is the product of the Layer 3 Virtual Private Networks Working Group. 

The IESG contact persons are Ross Callon and Adrian Farrel.

A URL of this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-l3vpn-2547bis-mcast-bgp-08.txt

Technical Summary

   This document describes the BGP encodings and procedures for
   exchanging the information elements required by Multicast in 
   MPLS/BGP IP VPNs, as specified in the related document
   draft-ietf-l3vpn-2547bis-mcast. 

Working Group Summary

   The L3VPN WG has been working on multicast for many years, with 
   significant difficulty coming to consensus. This document is put 
   forth with draft-ietf-l3vpn-2547bis-mcast. Interoperability and
   default deployment modes are a concern because of the large number
   of options and features provided in each of these specifications.
   These concerns are being addressed by draft-ietf-l3vpn-mvpn-
   considerations, which will be submitted very shortly as well.  
   After much deliberation and consideration in the WG and among 
   relevant stake holders this is the most feasible path forward.

Document Quality

   Many of the options described in this document are implemented and
   widely deployed. The related document draft-ietf-l3vpn-mvpn-
   considerations is co-authored by experts from multiple service 
   providers based largely on deployement experience. 

Personnel

   Danny McPherson is the Document Shepherd for this document. Ross
   Callon is is the Responsible Area Director.

RFC Editor Note

  In section 2 ("Introduction"), second paragraph:

  OLD
    This document defines a new NLRI, MCAST-VPN NLRI. The MCAST-VPN NLRI
    is used for MVPN auto-discovery, advertising MVPN to I-PMSI tunnel
    binding, advertising (C-S, C-G) to S-PMSI tunnel binding, VPN
    customer multicast routing information exchange among PEs, choosing a
    single forwarder PE, and for procedures in support of co-locating a
    C-RP on a PE.

  NEW
    This document defines a new NLRI, MCAST-VPN NLRI. The MCAST-VPN NLRI
    is used for MVPN auto-discovery, advertising MVPN to I-PMSI tunnel
    binding, advertising (C-S, C-G) to S-PMSI tunnel binding, VPN
    customer multicast routing information exchange among PEs, choosing a
    single forwarder PE, and for procedures in support of co-locating a
    C-RP on a PE.  This NLRI does not apply to communications between CE
    and PE equipment.

  In section 8 ("PE Distinguisher Labels Attribute"), third paragraph:

  OLD
    The attribute is also considered to be malformed if the PE Address 
    field is expected to be an IPv4 address, and the length of the
    attribute minus 4 is not a multiple of 3, or the PE Address field
    is expected to be an IPv6 address, and the length of the attribute
    minus 16 is not a multiple of 3.

  NEW
    The attribute is also considered to be malformed if the PE Address
    field is expected to be an IPv4 address, and the length of the
    attribute is not a multiple of 7, or the PE Address field is
    expected to be an IPv6 address, and the length of the attribute is
    not a multiple of 19.

  In section 17 ("Security Considerations"), replace first paragraph:

  OLD  
    The mechanisms described in this document could re-use the existing
    BGP security mechanisms.
  
  NEW  
    The mechanisms described in this document could re-use the existing
    BGP security mechanisms [RFC4271][RFC4272]. The security model and
    threats specific to Provider Provisioned VPNs, including L3VPNs, are
    discussed in [RFC4111]. [MVPN] discusses 
    additional threats specific to the use of multicast in L3VPNs. There
    is currently work in progress to improve the security of TCP
    authentication. When the document is finalized as an RFC, the method
    defined in [draft-ietf-tcpm-tcp-auth-opt] SHOULD be used where
    authentication of BGP control packets is needed.

  In Section 18 ("IANA Considerations"), first paragraph:

  OLD
    This document defines a new BGP Extended Community called Source AS.
    This community is 2-octet AS specific, of an extended type, and is
    transitive.

  NEW
    This document defines a new BGP Extended Community called Source
    AS. This community is of an extended type, and is transitive.
    This document requests allocation of the Type value for this
    community from both the Two-octet AS Specific Extended Community
    and the Four-octet AS Specific Extended Community registries.
    Moreover, the document requests allocating the same Type value
    from both of these registries.

  Please add an informative reference (section 21.2) to RFC4111, 
  RFC4272, and draft-ietf-tcpm-tcp-auth-opt.


From erosen@cisco.com  Mon Dec 14 07:24:34 2009
Return-Path: <erosen@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id D9C9028C129 for <l3vpn@core3.amsl.com>; Mon, 14 Dec 2009 07:24:34 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.428
X-Spam-Level: 
X-Spam-Status: No, score=-6.428 tagged_above=-999 required=5 tests=[AWL=0.171,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id XmyMfYZuOE1b for <l3vpn@core3.amsl.com>; Mon, 14 Dec 2009 07:24:34 -0800 (PST)
Received: from rtp-iport-1.cisco.com (rtp-iport-1.cisco.com [64.102.122.148]) by core3.amsl.com (Postfix) with ESMTP id C1D8128C124 for <l3vpn@ietf.org>; Mon, 14 Dec 2009 07:24:33 -0800 (PST)
Authentication-Results: rtp-iport-1.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-AV: E=Sophos;i="4.47,395,1257120000"; d="scan'208";a="74131942"
Received: from rtp-core-1.cisco.com ([64.102.124.12]) by rtp-iport-1.cisco.com with ESMTP; 14 Dec 2009 15:24:20 +0000
Received: from erosen-linux.cisco.com (erosen-linux.cisco.com [161.44.70.34]) by rtp-core-1.cisco.com (8.13.8/8.14.3) with ESMTP id nBEFOKLK028195; Mon, 14 Dec 2009 15:24:20 GMT
Received: from erosen-linux (localhost.localdomain [127.0.0.1]) by erosen-linux.cisco.com (8.13.1/8.13.1) with ESMTP id nBEFOJ9G004299;  Mon, 14 Dec 2009 10:24:19 -0500
To: Yakov Rekhter <yakov@juniper.net>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 
In-reply-to: Your message of Tue, 08 Dec 2009 07:08:19 -0800. <200912081508.nB8F8Jj31633@magenta.juniper.net> 
Date: Mon, 14 Dec 2009 10:24:19 -0500
Message-ID: <4298.1260804259@erosen-linux>
From: Eric Rosen <erosen@cisco.com>
Cc: erosen@cisco.com, Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: erosen@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 14 Dec 2009 15:24:35 -0000

Eric> Scalability for S-PMSIs: Comparable scalability

Yakov> when C-multicast routing is done with BGP, as when BGP is used for
Yakov> C-multicast routing, segmented P-tunnels are more scalable than
Yakov> non-segmented due to the ability to aggregate segmented S-PMSIs with
Yakov> I-PMSIs.

I don't see the relevance of this point.  What does "the ability to
aggregate segmented S-PMSIs with I-PMSIs" have to do with the scalability of
S-PMSIs.

Eric> Complexity at ASBRs: segmented more complex (requires ASBRs to have
Eric> more VPN-specific knowledge, and some options increased configuration
Eric> at ASBRs)

Yakov> To be more precise, depending on whether inter-AS I-PMSI auto-discovery
Yakov> routes are at the granularity of <source ASBR, MVPN> or <source AS,
Yakov> MVPN>, the segmented P-tunnels would require either the same or
Yakov> additional configuration on ASBRs as the non-segmented P-tunnels.

To be even more precise, the more scalable option requires per-MVPN
configuration on each ASBR, and even the less scalable option requires more
complexity on the ASBRs.

Eric> Compatibility with existing implementations: non-segmented provides
Eric> this, segmented does not.

Yakov> having non-segmented P-tunnels, by itself, does not provide
Yakov> compatibility with "existing implementations".

It would indeed be more precise to say that non-segmented P-tunnels is a
necessary but not always sufficient feature for interworking new deployments
with old.

Yakov> an implementation of non-segmented inter-AS I-PMSI P-tunnels with
Yakov> BGP-based auto-discovery, as specified in draft-ietf-l3vpn-mcast-bgp
Yakov> is *not* compatible with "existing implementations" of non-segmented
Yakov> inter-AS I-PMSI P-tunnels with BGP-based auto-discovery based on
Yakov> MDT-SAFI, even if both of these implementations use non-segmented
Yakov> P-tunnels.

Actually, I believe these two implementations will interwork correctly, as
is, when there is an "option C" interconnect, PIM/MI-PMSI is used as the
PE-PE control protocol, S-PMSI Joins are used, and the MI-PMSI is set up
using a sparse mode group address.

In an "option B" interconnect, new implementations would have to include and
parse the "connector attribute" as well as the "RT VRF Import Attribute" in
order to interwork with old implementations; a very simple and
straightforward interworking technique that is not available with segmented
inter-AS P-tunnels.

Thomas proposes to change the phrase "in order to maintain backwards
compatibility and for migration" to "in order to facilitate co-existence
with existing deployments, and provide a lighter engineering in some
scenarios".  I think it would be better to say "in order to facilitate
interworking with and migration from existing deployments, and to provide a
simpler solution when the complexity of segmented inter-AS P-tunnels is not
needed."










From yakov@juniper.net  Mon Dec 14 07:56:46 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id C04FE3A686E for <l3vpn@core3.amsl.com>; Mon, 14 Dec 2009 07:56:46 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.539
X-Spam-Level: 
X-Spam-Status: No, score=-6.539 tagged_above=-999 required=5 tests=[AWL=0.060,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id EdDsqvi0POET for <l3vpn@core3.amsl.com>; Mon, 14 Dec 2009 07:56:46 -0800 (PST)
Received: from exprod7og119.obsmtp.com (exprod7og119.obsmtp.com [64.18.2.16]) by core3.amsl.com (Postfix) with ESMTP id BC7F83A68D4 for <l3vpn@ietf.org>; Mon, 14 Dec 2009 07:56:40 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob119.postini.com ([64.18.6.12]) with SMTP ID DSNKSyZgK/0cvthoktrjNuP56LSQ/URBv7Pd@postini.com; Mon, 14 Dec 2009 07:56:29 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Mon, 14 Dec 2009 07:51:25 -0800
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 14 Dec 2009 07:51:25 -0800
Received: from emailsmtp56.jnpr.net ([172.24.60.77]) by p-emlb01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 14 Dec 2009 07:51:24 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp56.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 14 Dec 2009 07:51:24 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBEFpOj67017; Mon, 14 Dec 2009 07:51:24 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912141551.nBEFpOj67017@magenta.juniper.net>
To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
Subject: Re: Fwd: draft-ietf-l3vpn-mvpn-considerations-05 
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <97811.1260805884.1@juniper.net>
Date: Mon, 14 Dec 2009 07:51:24 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 14 Dec 2009 15:51:24.0968 (UTC) FILETIME=[4A495A80:01CA7CD5]
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, l3vpn@ietf.org
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 14 Dec 2009 15:56:46 -0000

Maria,

> 6. The appendix states that "we propose here a rough estimation of the
> load of PEs". And it is in fact a very *rough* estimation and not even a
> complete *rough* estimation as it almost dismisses ASM analysis. The
> appendix treats different and, in fact, not comparable kinds of
> "processing" as *equal* and uses the same unit "m.e" to measure them.
> For example, the first row in the table in section A.1.1.1, says that to
> process the first join on the upstream PE takes "1 m.e" and that the
> same amount of processing, "1 m.e", is required on every other PE in
> MVPN. Clearly, the amount of processing on the upstream PE to which the
> joined is destined is not the same as the "processing" on other PEs. Any
> other PE than the upstream PE will simply discard this Join because the
> neighbor address in the PIM Join is not this PE address. While the
> upstream PE has to create the (s,g) state in the control plane, program
> the data plane for forwarding, and send Join (s,g) message upstream.
> This is by any measure much more processing than discarding the Join
> based on the address. Yet, it is assumed that the resources used in both
> cases are the same.

To address the above I would propose to replace: 

   Though the type of processing, messages and states, may vary with the
   different approaches, we propose here a rough estimation of the load
   of PEs, in terms of number of messages processed and number of
   control plane states maintained. 

with the following:

   Though the overhead of message processing and the state overhead
   may vary between different approaches, and even within the same
   approach between different equipments (PEs or RRs), we propose
   here a rough estimation of the load of PEs, in terms of number
   of messages processed and number of control plane states maintained.

Yakov.

From yakov@juniper.net  Mon Dec 14 08:09:10 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 1E8503A6898 for <l3vpn@core3.amsl.com>; Mon, 14 Dec 2009 08:09:10 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.544
X-Spam-Level: 
X-Spam-Status: No, score=-6.544 tagged_above=-999 required=5 tests=[AWL=0.055,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id XJj5A9c7tO+p for <l3vpn@core3.amsl.com>; Mon, 14 Dec 2009 08:09:09 -0800 (PST)
Received: from exprod7og104.obsmtp.com (exprod7og104.obsmtp.com [64.18.2.161]) by core3.amsl.com (Postfix) with ESMTP id 4AB1C3A67E2 for <l3vpn@ietf.org>; Mon, 14 Dec 2009 08:09:08 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob104.postini.com ([64.18.6.12]) with SMTP ID DSNKSyZjFpvjD5NR1OIEpZOfMaEu89fMdyYO@postini.com; Mon, 14 Dec 2009 08:08:56 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Mon, 14 Dec 2009 08:05:16 -0800
Received: from p-emlb02-sac.jnpr.net ([66.129.254.47]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 14 Dec 2009 08:05:16 -0800
Received: from emailsmtp56.jnpr.net ([172.24.60.77]) by p-emlb02-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 14 Dec 2009 08:05:16 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp56.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 14 Dec 2009 08:05:16 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBEG5Fj73715; Mon, 14 Dec 2009 08:05:15 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912141605.nBEG5Fj73715@magenta.juniper.net>
To: erosen@cisco.com
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 
In-Reply-To: <4298.1260804259@erosen-linux> 
References: <4298.1260804259@erosen-linux>
X-MH-In-Reply-To: Eric Rosen <erosen@cisco.com> message dated "Mon, 14 Dec 2009 10:24:19 -0500."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <98972.1260806715.1@juniper.net>
Date: Mon, 14 Dec 2009 08:05:15 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 14 Dec 2009 16:05:16.0239 (UTC) FILETIME=[39C33DF0:01CA7CD7]
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 14 Dec 2009 16:09:10 -0000

Eric,

> Eric> Scalability for S-PMSIs: Comparable scalability
> 
> Yakov> when C-multicast routing is done with BGP, as when BGP is used for
> Yakov> C-multicast routing, segmented P-tunnels are more scalable than
> Yakov> non-segmented due to the ability to aggregate segmented S-PMSIs with
> Yakov> I-PMSIs.
> 
> I don't see the relevance of this point.  What does "the ability to
> aggregate segmented S-PMSIs with I-PMSIs" have to do with the scalability of
> S-PMSIs.

For one thing, in the multi-AS scenario it allows (downstream) ASes
to reduce the overhead due to S-PMSIs by aggregating them into
I-PMSIs.

In addition, in the multi-AS scenario it allows (downstream) ASes
to aggregate P-tunnels of the intra-AS segments of S-PMSIs, thus
reducing the overhead due to S-PMSIs.

Yakov. 

From yakov@juniper.net  Mon Dec 14 08:27:01 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 2127A3A68F1 for <l3vpn@core3.amsl.com>; Mon, 14 Dec 2009 08:27:01 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.549
X-Spam-Level: 
X-Spam-Status: No, score=-6.549 tagged_above=-999 required=5 tests=[AWL=0.050,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Vwm+3ISTjyOK for <l3vpn@core3.amsl.com>; Mon, 14 Dec 2009 08:27:00 -0800 (PST)
Received: from exprod7og104.obsmtp.com (exprod7og104.obsmtp.com [64.18.2.161]) by core3.amsl.com (Postfix) with ESMTP id 65D483A67D1 for <l3vpn@ietf.org>; Mon, 14 Dec 2009 08:26:53 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob104.postini.com ([64.18.6.12]) with SMTP ID DSNKSyZnP+QSYkjisZjXLn9NCL2T+MCcHn9E@postini.com; Mon, 14 Dec 2009 08:26:45 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.71) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Mon, 14 Dec 2009 08:23:40 -0800
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 14 Dec 2009 08:23:40 -0800
Received: from emailsmtp56.jnpr.net ([172.24.60.77]) by p-emlb01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 14 Dec 2009 08:23:40 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp56.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 14 Dec 2009 08:23:40 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBEGNej84217; Mon, 14 Dec 2009 08:23:40 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912141623.nBEGNej84217@magenta.juniper.net>
To: erosen@cisco.com
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 
In-Reply-To: <4298.1260804259@erosen-linux> 
References: <4298.1260804259@erosen-linux>
X-MH-In-Reply-To: Eric Rosen <erosen@cisco.com> message dated "Mon, 14 Dec 2009 10:24:19 -0500."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <615.1260807819.1@juniper.net>
Date: Mon, 14 Dec 2009 08:23:40 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 14 Dec 2009 16:23:40.0412 (UTC) FILETIME=[CBE6ABC0:01CA7CD9]
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 14 Dec 2009 16:27:01 -0000

Eric,

> Eric> Compatibility with existing implementations: non-segmented provides
> Eric> this, segmented does not.
> 
> Yakov> having non-segmented P-tunnels, by itself, does not provide
> Yakov> compatibility with "existing implementations".
> 
> It would indeed be more precise to say that non-segmented P-tunnels is a
> necessary but not always sufficient feature for interworking new deployments
> with old.
> 
> Yakov> an implementation of non-segmented inter-AS I-PMSI P-tunnels with
> Yakov> BGP-based auto-discovery, as specified in draft-ietf-l3vpn-mcast-bgp
> Yakov> is *not* compatible with "existing implementations" of non-segmented
> Yakov> inter-AS I-PMSI P-tunnels with BGP-based auto-discovery based on
> Yakov> MDT-SAFI, even if both of these implementations use non-segmented
> Yakov> P-tunnels.
> 
> Actually, I believe these two implementations will interwork correctly, as
> is, when there is an "option C" interconnect, PIM/MI-PMSI is used as the
> PE-PE control protocol, S-PMSI Joins are used, and the MI-PMSI is set up
> using a sparse mode group address.

Incorrect.

Even if "PIM/MI-PMSI is used as the PE-PE control protocol, S-PMSI
Joins are used and the MI-PMSI is set up using a sparse mode group
address", an implementation that follows draft-ietf-l3vpn-* will
*not* interwork with "existing implementations" as according to
*your own draft* (draft-rosen-vpn-mcast) the existing implementation
"REQUIRES the use of the MDT-SAFI NLRI (even if PIM-SSM is not used
to set up the default MDT)" (while an implemendation that follows
draft-ietf-l3vpn-* is not required to use MDT-SAFI NLRI).

Yakov.

From erosen@cisco.com  Tue Dec 15 08:05:32 2009
Return-Path: <erosen@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id B5F8A3A6A99 for <l3vpn@core3.amsl.com>; Tue, 15 Dec 2009 08:05:32 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.449
X-Spam-Level: 
X-Spam-Status: No, score=-6.449 tagged_above=-999 required=5 tests=[AWL=0.150,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id wx1J58-GdTEj for <l3vpn@core3.amsl.com>; Tue, 15 Dec 2009 08:05:31 -0800 (PST)
Received: from rtp-iport-2.cisco.com (rtp-iport-2.cisco.com [64.102.122.149]) by core3.amsl.com (Postfix) with ESMTP id 9B60F3A6A8D for <l3vpn@ietf.org>; Tue, 15 Dec 2009 08:05:31 -0800 (PST)
Authentication-Results: rtp-iport-2.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-AV: E=Sophos;i="4.47,400,1257120000"; d="scan'208";a="74494934"
Received: from rtp-core-2.cisco.com ([64.102.124.13]) by rtp-iport-2.cisco.com with ESMTP; 15 Dec 2009 16:05:15 +0000
Received: from erosen-linux.cisco.com (erosen-linux.cisco.com [161.44.70.34]) by rtp-core-2.cisco.com (8.13.8/8.14.3) with ESMTP id nBFG5EKg021368; Tue, 15 Dec 2009 16:05:15 GMT
Received: from erosen-linux (localhost.localdomain [127.0.0.1]) by erosen-linux.cisco.com (8.13.1/8.13.1) with ESMTP id nBFG5EOx015667;  Tue, 15 Dec 2009 11:05:14 -0500
To: Yakov Rekhter <yakov@juniper.net>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 
In-reply-to: Your message of Mon, 14 Dec 2009 08:05:15 -0800. <200912141605.nBEG5Fj73715@magenta.juniper.net> 
Date: Tue, 15 Dec 2009 11:05:14 -0500
Message-ID: <15666.1260893114@erosen-linux>
From: Eric Rosen <erosen@cisco.com>
Cc: erosen@cisco.com, Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: erosen@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 15 Dec 2009 16:05:32 -0000

Eric> What does "the ability to aggregate segmented S-PMSIs with I-PMSIs"
Eric> have to do with the scalability of S-PMSIs.

Yakov> For one thing, in the multi-AS scenario it allows (downstream) ASes
Yakov> to reduce the overhead due to S-PMSIs by aggregating them into
Yakov> I-PMSIs.

This means that a downstream provider can use I-PMSIs even if an upstream
provider uses S-PMSIs.  That may or may not be a useful feature, but it
doesn't make the use of S-PMSIs more scalable.

Furthermore, this cannot be done using only the purported mandatory features
of the considerations draft, it requires optional features.  

Yakov> In addition, in the multi-AS scenario it allows (downstream) ASes
Yakov> to aggregate P-tunnels of the intra-AS segments of S-PMSIs, thus
Yakov> reducing the overhead due to S-PMSIs.

I don't think the ability to aggregate multiple S-PMSIs into a single
P-tunnel has anything to do with I-PMSIs.  In addition, non-segmented
S-PMSIs can also be aggregated into a single P-tunnel.

In a transit AS, the aggregation of multiple S-PMSI intra-AS segments into a
single P-tunnel across that transit AS cannot be done using only the
purported mandatory features of the considerations draft.

I think it is somewhat misleading to say that a certain feature should be
mandatory because it has certain properties, if those properties cannot be
realized without the implementation of optional features.

Yakov> according to *your own draft* (draft-rosen-vpn-mcast) the existing
Yakov> implementation "REQUIRES the use of the MDT-SAFI NLRI

That just means that an implementation without MDT-SAFI is not compliant
with that draft.  I don't think I ever said otherwise.





From yakov@juniper.net  Tue Dec 15 09:06:21 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id DBE833A6AA7 for <l3vpn@core3.amsl.com>; Tue, 15 Dec 2009 09:06:21 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.553
X-Spam-Level: 
X-Spam-Status: No, score=-6.553 tagged_above=-999 required=5 tests=[AWL=0.046,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id zYwW+i2gVw7t for <l3vpn@core3.amsl.com>; Tue, 15 Dec 2009 09:06:20 -0800 (PST)
Received: from exprod7og117.obsmtp.com (exprod7og117.obsmtp.com [64.18.2.6]) by core3.amsl.com (Postfix) with SMTP id 5EF103A688F for <l3vpn@ietf.org>; Tue, 15 Dec 2009 09:06:19 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob117.postini.com ([64.18.6.12]) with SMTP ID DSNKSyfB/c2pIaauXVEu/5Oa+T79qJcf4r03@postini.com; Tue, 15 Dec 2009 09:06:07 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Tue, 15 Dec 2009 08:55:41 -0800
Received: from p-emlb02-sac.jnpr.net ([66.129.254.47]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 15 Dec 2009 08:55:40 -0800
Received: from emailsmtp55.jnpr.net ([172.24.18.132]) by p-emlb02-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 15 Dec 2009 08:55:40 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp55.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 15 Dec 2009 08:55:40 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBFGtdj27964; Tue, 15 Dec 2009 08:55:39 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912151655.nBFGtdj27964@magenta.juniper.net>
To: erosen@cisco.com
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 
In-Reply-To: <15666.1260893114@erosen-linux> 
References: <15666.1260893114@erosen-linux>
X-MH-In-Reply-To: Eric Rosen <erosen@cisco.com> message dated "Tue, 15 Dec 2009 11:05:14 -0500."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <37970.1260896139.1@juniper.net>
Date: Tue, 15 Dec 2009 08:55:39 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 15 Dec 2009 16:55:40.0183 (UTC) FILETIME=[6E963A70:01CA7DA7]
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 15 Dec 2009 17:06:22 -0000

Eric,

> Eric> What does "the ability to aggregate segmented S-PMSIs with I-PMSIs"
> Eric> have to do with the scalability of S-PMSIs.
> 
> Yakov> For one thing, in the multi-AS scenario it allows (downstream) ASes
> Yakov> to reduce the overhead due to S-PMSIs by aggregating them into
> Yakov> I-PMSIs.
> 
> This means that a downstream provider can use I-PMSIs even if an upstream
> provider uses S-PMSIs.  That may or may not be a useful feature, but it
> doesn't make the use of S-PMSIs more scalable.

Sure it does.

> Furthermore, this cannot be done using only the purported mandatory features
> of the considerations draft, it requires optional features.  
> 
> Yakov> In addition, in the multi-AS scenario it allows (downstream) ASes
> Yakov> to aggregate P-tunnels of the intra-AS segments of S-PMSIs, thus
> Yakov> reducing the overhead due to S-PMSIs.
> 
> I don't think the ability to aggregate multiple S-PMSIs into a single
> P-tunnel has anything to do with I-PMSIs.  

This has to do with scalability of segmented vs non-segmented S-PMSIs.

> In addition, non-segmented
> S-PMSIs can also be aggregated into a single P-tunnel.

By virtue of being non-segmented, non-segmentd S-PMSIs do *not*
provide the ability to aggregate P-tunnels of the intra-AS *segments*
of S-PMSIs (for one thing, non-segmented S-PMSIs do not have "intra-AS 
segments").

Yakov.

From thomas.morin@orange-ftgroup.com  Wed Dec 16 00:30:52 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id E072F3A6971 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:30:52 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.949
X-Spam-Level: 
X-Spam-Status: No, score=-1.949 tagged_above=-999 required=5 tests=[AWL=0.300,  BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id z143+VSR0679 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:30:52 -0800 (PST)
Received: from r-mail1.rd.francetelecom.com (r-mail1.rd.francetelecom.com [217.108.152.41]) by core3.amsl.com (Postfix) with ESMTP id A6D763A696F for <l3vpn@ietf.org>; Wed, 16 Dec 2009 00:30:51 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp1.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:30:36 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:30:36 +0100
Message-ID: <4B289AAB.9060107@orange-ftgroup.com>
Date: Wed, 16 Dec 2009 09:30:35 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Eric Rosen <erosen@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / counting states in both protocols
References: <3126.1259858356@erosen-linux>
In-Reply-To: <3126.1259858356@erosen-linux>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Dec 2009 08:30:36.0084 (UTC) FILETIME=[0A596F40:01CA7E2A]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 08:30:53 -0000

Eric, working group,

Eric Rosen:
[...]
 > [...] as we will see, the analysis is not very precise about what it
 > counts as a "state", and this leads to some mistakes.
 >
[...]
 > Another methodological error is that the analysis takes no account of the
 > increased amount of state that must be maintained when C-multicast 
trees are
 > set up using two different protocols, rather than just one.  When PIM is
 > used as the PE-PE control plane, the (C-S,C-G) and (C-*,C-G) states that
 > drive the inter-PE control plane are the same as the one as the drive the
 > CE-PE control plane.  When BGP is used, a given multicast state has to be
 > represented both as a PIM state and as a BGP state (C-multicast route).
 > This should really be counted as a doubling (at least) of the number of
 > states that need to be maintained.  This is a very conservative 
estimate, as
 > the BGP states tend to contain much more information (e.g., RTs) than the
 > corresponding PIM states.  It's clear that appendix A does not 
consider both
 > the BGP states and the PIM states, but it's never quite clear just what
 > states it is considering.


In revision -02 of the draft, section A.3 (which is now A.1.1.3) 
included a change to properly take
into account what you describe above, after a similar comment by Maria 
Napierala we received February 23th 2009.

http://tools.ietf.org/rfcdiff?url2=draft-ietf-l3vpn-mvpn-considerations-02.txt#part-l12

I notice now that the correction was lost when A.1.1.3 was revised in -05.

Given this:
- the table in A.1.1.3 will be corrected, so that a factor 2 is added in 
columns 2 and 3, and the "total" updated accordingly.
- I will also add a note in the introduction to indicate that a state is 
counted twice when both present in PIM and BGP, and a note  to A.1.1.3 
indicating that both PIM and BGP states have been accounted for on the 
receiver PEs and upstream PE.

(So, this is in fact not a "methodological error", but a simple counting 
error.)

-Thomas


From thomas.morin@orange-ftgroup.com  Wed Dec 16 00:33:24 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 445983A696F for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:33:24 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.499
X-Spam-Level: 
X-Spam-Status: No, score=-1.499 tagged_above=-999 required=5 tests=[AWL=-0.450, BAYES_00=-2.599, HELO_EQ_FR=0.35, J_CHICKENPOX_41=0.6, J_CHICKENPOX_51=0.6]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id HKhikAQETDf0 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:33:23 -0800 (PST)
Received: from r-mail1.rd.francetelecom.com (r-mail1.rd.francetelecom.com [217.108.152.41]) by core3.amsl.com (Postfix) with ESMTP id 01E3E3A6A16 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 00:33:22 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp1.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:33:08 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:33:07 +0100
Message-ID: <4B289B43.9020200@orange-ftgroup.com>
Date: Wed, 16 Dec 2009 09:33:07 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Eric Rosen <erosen@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / last receiver leaves
References: <3126.1259858356@erosen-linux>
In-Reply-To: <3126.1259858356@erosen-linux>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Dec 2009 08:33:07.0630 (UTC) FILETIME=[64AD84E0:01CA7E2A]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 08:33:24 -0000

Eric, working group,

Eric Rosen:
 >
 >         for each PIM Prune message, all PE routers on the LAN
 >         work to let the upstream PE determine the answer to the
 >         "did the last receiver leave?" question.
 >
 > I believe that only one PE router needs to send a Join in order to 
override
 > the Prune, as once the others see that Join they reset their Join timers.
 > Of course, actual behavior varies depending on the value of the timer and
 > on the random jitter.

The text was already clarified in version -04, after your comments from 
2009-06-11, and I provided an explanation on 2009-06-19.

  Previous text:
      The "did the last receiver leave?" question is thus implicitly 
replied to by all PE
      routers, for each PIM Prune message.

You had claimed in June that the text "lead one to believe that a 
Prune(S,G) immediately causes lots of Join(S,G) messages to be sent", 
which could be argued sense because the phrase "replied to" could be 
understood a "a message is sent", even though it was not the intent of 
the text.

The text now reads as follows:
      Thus, for each PIM Prune message, all PE routers work to let the 
upstream PE
      determine the answer to the "did the last receiver leave?" question.

Indeed, when a PE sends a Prune(S,G):
   - all PEs parse this message, and try to find a matching state in 
their TIB ==> reflected by "all PE routers on the LAN work..."
   (- if a PE finds one, it setup a short timer to send a Prune to 
override the Join)
   - if no PE find any, after some delay, the upstream PE can determine 
that the last receiver has left ==> reflected by "... to let the 
upstream PE determine the answer to the ``did the last receiver 
leave?``question"

I don't think that the new text carries any wrong or misleading term, 
since "work" properly designate what the PEs of the MVPN do, without 
carrying a misleading idea of multiple Join message being sent.

If you can explain why you think the revised text above does not 
properly reflect the PIM Prune override procedures, feel free to suggest 
a better text (just repeating your comment from June is not helpful).


 >        We can see that answering the "last receiver leaves"
 >        question is a significant proportion of the work that the
 >        C-multicast routing building block has to make, and where
 >        the approaches differ most.
 >
 > This assertion is made without support.

Ok, the introduction of appendix A is not supposed to carry quantitative 
conclusions, so let's just say "is part of the work" instead of "is a 
significant proportion of the work" in this introduction.

Thanks,

-Thomas


From thomas.morin@orange-ftgroup.com  Wed Dec 16 00:39:56 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 28C203A694E for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:39:56 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.949
X-Spam-Level: 
X-Spam-Status: No, score=-1.949 tagged_above=-999 required=5 tests=[AWL=0.300,  BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id p16sRcVyzaJl for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:39:55 -0800 (PST)
Received: from r-mail1.rd.francetelecom.com (r-mail1.rd.francetelecom.com [217.108.152.41]) by core3.amsl.com (Postfix) with ESMTP id DCBCB3A6861 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 00:39:54 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp2.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:39:40 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:39:39 +0100
Message-ID: <4B289CCB.7090702@orange-ftgroup.com>
Date: Wed, 16 Dec 2009 09:39:39 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Eric Rosen <erosen@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / PIM with explicit tracking
References: <3126.1259858356@erosen-linux>
In-Reply-To: <3126.1259858356@erosen-linux>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Dec 2009 08:39:39.0878 (UTC) FILETIME=[4E79D060:01CA7E2B]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 08:39:56 -0000

Eric, working group,

Eric Rosen:
 > [...] much time is spent on examining a "PIM with explicit tracking" 
scheme, even
 > though no such scheme has been proposed. It's difficult to understand 
why this straw
 > man is discussed, while other actual proposals for using PIM are not.

"PIM-SM" specifications (RFC4601, section 4.3.3, page 36):

    Although the mechanisms are not specified in this document, it is
    possible for upstream routers to explicitly track the join membership
    of individual downstream routers if Join suppression is disabled.  A
    router can advertise its willingness to disable Join suppression by
    using the T bit in the LAN Prune Delay Hello option.  Unless all PIM
    routers on a link negotiate this capability, explicit tracking and
    the disabling of the Join suppression mechanism are not possible.

Nothing, as far as I know, exclude this modality of PIM from the "Full 
Per-MVPN PIM Peering Across a MI-PMSI" procedures described in 
draft-ietf-l3vpn-2547bis-mcast.  So we are definitely not "discussing a 
straw-man".  I wonder if you are possibly confusing "PIM-SM 
specifications" and "a PIM implementation".

However, if nobody thinks it is useful to look at this use of PIM with 
Explicit tracking, then we can certainly remove it.

As far as "other actual proposal for PIM", I guess you may be referring 
to proposals in draft-rosen-l3vpn-mvpn-mspmsi: of course, they are not 
in the scope of this document, which is restricted to the solutions 
described in draft-ietf-l3vpn-2547bis-mcast and 
draft-ietf-l3vpn-2547bis-mcast-bgp.

-Thomas

PS: Let me note that you did not bring up this up during the LC in June, 
and that this comment is not related to any new content in the revisions 
made since this LC.


From thomas.morin@orange-ftgroup.com  Wed Dec 16 00:40:12 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 018CB3A6AA0 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:40:12 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.024
X-Spam-Level: 
X-Spam-Status: No, score=-2.024 tagged_above=-999 required=5 tests=[AWL=0.225,  BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id a7WUyiwR7uRg for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:40:11 -0800 (PST)
Received: from r-mail2.rd.francetelecom.com (r-mail2.rd.francetelecom.com [217.108.152.42]) by core3.amsl.com (Postfix) with ESMTP id AAFD13A6A84 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 00:40:10 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp2.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:39:56 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:39:55 +0100
Message-ID: <4B289CDB.8090404@orange-ftgroup.com>
Date: Wed, 16 Dec 2009 09:39:55 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Eric Rosen <erosen@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / RT Constraint recommendation
References: <3126.1259858356@erosen-linux>
In-Reply-To: <3126.1259858356@erosen-linux>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Dec 2009 08:39:55.0862 (UTC) FILETIME=[5800C760:01CA7E2B]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 08:40:12 -0000

Eric, working group,

Eric Rosen:
 >        the case where the RT-Constraint mechanisms [RFC4684] is
 >        not used is not covered;
 >
 > Since RT-constraint is not a mandatory to implement mechanism, I 
don't see
 > why this is not covered.

You omitted the first part of the sentence you quoted, which actually 
provides the answer:

       following the recommendation of [I-D.ietf-l3vpn-2547bis-mcast-bgp]
       the case where the RT-Constraint mechanisms [RFC4684] is not used 
is not
       covered;

Here is the corresponding quote in draft-ieft-2547bis-mcast-bgp (which 
you happen co-author) :

    16. Scalability Considerations

    A PE should use Route Target Constrain [RT-CONSTRAIN] to advertise
    the Route Targets that the PE uses for the VRF Route Imports extended
    community (note that doing this requires just a single Route Target
    Constraint advertisement by the PE). This allows each C-multicast
    route to reach only the relevant PE, rather than all the PEs
    participating the an MVPN.

Hence the considerations document does not cover the case where the 
RT-Constraint mechanism is not used.

-Thomas

PS: Let me note that you did not bring up this up during the LC in June, 
and that this comment is not related to any new content in the revisions 
made since this LC.



From thomas.morin@orange-ftgroup.com  Wed Dec 16 00:42:39 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 937953A67D8 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:42:39 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.069
X-Spam-Level: 
X-Spam-Status: No, score=-2.069 tagged_above=-999 required=5 tests=[AWL=0.180,  BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id AHjwkFKUwJy7 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:42:38 -0800 (PST)
Received: from r-mail2.rd.francetelecom.com (r-mail2.rd.francetelecom.com [217.108.152.42]) by core3.amsl.com (Postfix) with ESMTP id 2A1AB3A6782 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 00:42:38 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp1.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:42:24 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:42:23 +0100
Message-ID: <4B289D6F.7080507@orange-ftgroup.com>
Date: Wed, 16 Dec 2009 09:42:23 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Eric Rosen <erosen@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / totals across equipments
References: <3126.1259858356@erosen-linux>
In-Reply-To: <3126.1259858356@erosen-linux>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Dec 2009 08:42:23.0439 (UTC) FILETIME=[AFF73DF0:01CA7E2B]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 08:42:39 -0000

Eric, working group,

Eric Rosen:
 >
 > A scalability analysis must show how the utilization of some resource 
or set
 > of resources on some system increases as some demand or set of demands
 > increases.  The appendix proposes two resources to consider:
 >
 >    Though the type of processing, messages and states, may vary with the
 >    different approaches, we propose here a rough estimation of the load
 >    of PEs, in terms of number of messages processed and number of
 >    control plane states maintained.  A "message processed" being a
 >    message being parsed, a lookup being done, and some action being
 >    taken (such has updating a control plane or data plane state).
 >
 > This is okay as one of the resources considered.  If I understand this
 > correctly, this omits consideration of any messages received that do not
 > result in any "action being taken", and thus omits consideration of any
 > acknowledgment messages or other messages that fall into the category of
 > "transport overhead".  It also omits consideration of messages 
transmitted,
 > thus does not distinguish actions that cause a new message to be sent 
from
 > actions that do not.   So it is not a complete measure of the amount of
 > "messaging", but it is a factor worth considering.

Yes, you could always think about refining and distinguishing more, but 
with a goal of looking at the orders of magnitude, I think the document 
already balances this aspect finely.



 >    A "state maintained" being a multicast state kept in the control plane
 >    memory of a PE, related to a interface or a PE being subscribed to a
 >    multicast stream.
 >
 > This is also worth considering as an abstract measure of memory 
resources.


Ok.


 > The real methodological problems begin when the analysis attempts to be
 > quantitative:
 >
 >    The estimation unit used is the "message.equipment" (or "m.e"): one
 >    "message.equipment" corresponding to "one equipment processing one
 >    message" (10 m.e being "10 equipments processing each one message",
 >    or "5 messages each processed by 2 equipments", or "1 message
 >    processed by 10 equipment", etc.).  Similarly, for the amount of
 >    control plane state, the unit used is "state.equipment" or "s.e".
 >    This allow to take into account the fact that a message (or a state)
 >    can have be processed (or maintained) by more than one node.
 >
 > This "estimation unit" has no meaning whatsoever, as it does not reflect
 > resource utilization on any system.  Suppose there are N systems, each
 > running an algorithm that scales as O(N).  Contrast this with an 
alternative
 > in which a single system does the work of all N, running an algorithm 
that
 > scales as O(N^2).  Other things being equal, the former would be 
considered
 > to be the more scalable alternative.  However, if you use the "estimation
 > unit" of "work.equipment", you would conclude that in both schemes 
the work
 > scales as O(N^2), and hence that the schemes are equally scalable.  
If this
 > were a valid methodology, no one would ever have heard of such notions as
 > "distributed routing" or "datagram networks".  Use of this peculiar
 > "estimation unit" will always bias the results towards a centralized 
system
 > and away from a distributed system.
 >
 > Any subsequent revision of Appendix A should eliminate the notion of 
"s.e."
 > or "m.e." and focus solely on the scaling of the individual systems.


You already brought up a slightly different form of this argument six 
month ago during WG LC (June 11th).

You could certainly claim that your argument "has not been addressed" 
and thus deserves being repeated, but the issue is that you did not care 
about answering me in the thread that followed on this particular point 
(my last email from July 10th), thus artificially delaying the 
resolution of this point.

But, for the sake of completeness, here are elements that justify how 
the study is currently done:
- as explained in revision -04, the choice of using these units is 
motivated; we need some efficient way to compare BGP and PIM on a common 
ground, even though the type and amount of "invididual systems" 
concerned by the processing of one message are not the same (sometimes 
only the upstream PE, sometimes only RRs, sometimes all PEs, sometimes a 
mix)
- your argument is abstract and does not prove that the comparison done 
in our specific case are irrelevant
- contrary to the imaginary example you take, we are looking at two 
distributed systems, and we observe orders of magnitude of difference 
(in that sense, you are beating a strawman with your example)
- your argument of "no meaning whatsoever" would hold if we were looking 
at individual numbers like O(N**2) and saying "this is big" in the 
absolute, but we don't do and only do comparisons based on these 
estimation units
- in last revisions, we reviewed the text and tables of Appendix A, and 
other sections, so that when a counting is mentioned and refers to a 
total across multiple equipement, we make sure that the reader will not 
be mislead into believing that we talk about a count for a given PE

I can concede to you that there may be more than one way to analyze the 
scalability of C-multicast routing options (and Appendix A.1 proposes 
one of them).  But the opportunity was missed to explore what an 
alternative way of doing the analysis would bring, as after the document 
went for the LC this summer you did not answer me in the thread that 
followed on this particular point (my last email from July 10th) ; you 
did not suggest a different approach until today (~6 month after).  
Moreover, other than from you, there were no objections to the way the 
analysis is done. Given all that, it is too late to request that the 
whole analysis be redone.

-Thomas


From thomas.morin@orange-ftgroup.com  Wed Dec 16 00:47:54 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 45BF43A6782 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:47:54 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.799
X-Spam-Level: 
X-Spam-Status: No, score=-2.799 tagged_above=-999 required=5 tests=[AWL=-0.150, BAYES_00=-2.599, HELO_EQ_FR=0.35, J_CHICKENPOX_41=0.6, RCVD_IN_DNSWL_LOW=-1]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Wize+0FfQcpG for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:47:53 -0800 (PST)
Received: from p-mail1.rd.francetelecom.com (p-mail1.rd.francetelecom.com [195.101.245.15]) by core3.amsl.com (Postfix) with ESMTP id A0FEB3A692C for <l3vpn@ietf.org>; Wed, 16 Dec 2009 00:47:52 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp1.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:47:38 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:47:37 +0100
Message-ID: <4B289EA9.2090500@orange-ftgroup.com>
Date: Wed, 16 Dec 2009 09:47:37 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Eric Rosen <erosen@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / PIM LAN procedures
References: <3126.1259858356@erosen-linux>
In-Reply-To: <3126.1259858356@erosen-linux>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Dec 2009 08:47:37.0719 (UTC) FILETIME=[6B4A9070:01CA7E2C]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 08:47:54 -0000

Eric, working group,

Eric Rosen:
 >
 > From A.1.1.1.  "PIM LAN procedures, by default"
 >
 >    +------------+------------+---------------+----------+--------------+
 >    |            | upstream   | other PEs     | RR       | total across |
 >    |            | PE (1)     | (total across | (none)   | all          |
 >    |            |            | (#mvpn_PE-1)  |          | equipments   |
 >    |            |            | PEs)          |          |              |
 >    +------------+------------+---------------+----------+--------------+
 >    | first PE   | 1 m.e      | #MVPN_PE-1    | /        | #MVPN_PE m.e |
 >    | joins      |            | m.e           |          |              |
 >    +------------+------------+---------------+----------+--------------+
 >[...]

In the sequence above the first PE joining results in all PEs processing 
a message (except the PE joining):
  - the upstream PE processes the Join and ends up propagating the state 
upstream and setting up forwarding state
  - the other PEs parse the message, and lookup their TIB to possibly 
find an (S,G) state; they find none and discard the information in this 
Join) ; note that they don't have any (S,G) state at this point

 >[...]
 >
 >    +------------+------------+---------------+----------+--------------+
 >    | for *each* | 1 m.e      | #mvpn_PE-1    | /        | #mvpn_PE m.e |
 >    | additional |            | m.e           |          |              |
 >    | PE joining |            |               |          |              |
 >    +------------+------------+---------------+----------+--------------+
 >

For each subsequent PE joining (second line in the table, quoted above):
   - the joining PE is transitioning its Upstream (S,G) state machine 
from NotJoined to Joined, and thus sends a Join (RFC 4601, section 
4.5.7, p.73)
   - for other PEs that are already joined to (S,G) : parse the message, 
and lookup their TIB to find an (S,G) state, find one, and reset their 
timer to further delay the moment at which they would send their Join
   - for the other PEs that are not joined to (S,G) : parse the message, 
and lookup their TIB to try find an (S,G) state, find none, and discard 
the information in this Join)


[...]
 > If a second PE has "processed" the Join(S,G) from the first, then it 
doesn't
 > need to send a Join if it decides to join (S,G) itself;

As PIM procedures are specified (RFC 4601, section 4.5.7, p.73), even if 
the second PE had initially processed a message (parsed, done a lookup, 
discarded) from the first joining PE, this PE will still send a Join 
when it starts to need to send (S,G) traffic towards a CE.



 >
 > If the [first] row implies that the "other PEs" must process the 
join, where
 > "process" is defined (as it is in this Appendix) as "a message being 
parsed,
 > a lookup being done, and some action being taken (such has updating a
 > control plane or data plane state)", then one the [second] row is 
incorrect:

Let's look at your reasoning, and how it could hold:
- if you want to ignore that the definition of "process" includes 
"discarding a message" as "an action being taken", then yes, your 
reasoning would hold ("if first [row] implies, [..] then second row is 
incorrect"), but your reasoning would actually be of low interest, 
because in that case the first row would be invalid too...
- if we count as one m.e when a PE parses a Join or Prune for which it 
is not the upstream, does a lookup for the corresponding (S,G) to 
finally discard the message, then the *two* rows are both valid

The "(such has updating a control plane or data plane state)" phrase is, 
obviously, a non-exhaustive list of examples. When assessing the 
processing done by PEs, we have to reflect the work done by a PE to 
identify it can discard a message because it does not have any (S,G) 
state, after parsing the Join or Prune and doing a lookup in its TIB.

We can certainly add "or discard the information in the message" in the 
list of examples.
I see this sentence has a typo ("has"' instead of "as"), which possibly 
was the source of confusion : it will be fixed too.

So, to avoid propagating any confusion any further, we can update the 
document as follows:

    A "message processed" being a message being parsed, a lookup being done, and some action being
    taken (such as, for instance, updating a control plane or data plane state or discarding the
    information in the message).


 >
 > The third and fifth columns above are examples of cases where the 
Appendix
 > misleadingly multiplies by #mvpn_PE.
 >


As we see above, the #mVPN_PEs order of magnitude is definitely correct 
for each PE joining.

Now, if we let's look at when PE leave:
- each PE leaving causes a Prune to be sent (RFC 4601, section 4.5.7, 
end of p.73)
- PEs that are joined to (S,G), parse the message, try to find an (S,G) 
state in their TIB, find one, and then setup a short timer to send a 
Prune to override the Join
- PEs that are not joined to (S,G), parse the message, try to find an 
(S,G) state in their TIB, find none, and decide they can discard the 
information
- the upstream PE parses the message, matches it with its (S,G) state, 
setups a timer and :
   - in the "last PE leaving" case, after a delay stop forwarding traffic
   - else, will not stop forwarding because of the Join sent by another 
PE to override the Prune


The #mVPN_PEs factor here also correctly reflects the amount of work 
done by PEs.


 >
 > In fact, the number of messages received by a given PE does not 
increase in proportion to #mvpn_PE.
 >

The content of Appendix A certainly does not imply that "the number of 
messages
received by a given PE increases in proportion to #mvpn_PE".

The #mVPN_PE factor is for the total across all PEs, not for a given PE.

( this exact same comment was in your email from June 11th, and I 
already replied to you June 19th : you deliberately ignored my reply, I 
don't think it is fair to again further propagate such irrelevant 
implicit claims on what the document says...)


 >
 > This problem infects the entire set of tables in the Appendix.
 >

See above, your point is only right if you tweak what "processing a 
message" means up to the point of being irrelevant and not accounting 
for all the processing done by PEs on a LAN.

Let me now highlight that the points about "prune override" and "join 
suppression" in your recent email where forming a large part of your 
comments during the LC in June, and where fueling your claims of 
"incorrectness" of the analysis (such as the claim that the "join 
suppression" was not taken into account).  By not providing a precise 
quote of what exactly would have been "incorrect" until last week -- 
despite multiple requests to you to provide them, and despite the 
multiple clarifications I provided on the fact that "join suppression" 
and "prune override" where relevantly accounted for -- you have 
artificially delayed the clarification of this issue for roughly six 
months.

-Thomas

From thomas.morin@orange-ftgroup.com  Wed Dec 16 00:51:11 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 556563A6782 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:51:11 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.099
X-Spam-Level: 
X-Spam-Status: No, score=-2.099 tagged_above=-999 required=5 tests=[AWL=0.150,  BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 3jDOKaQhWAgJ for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 00:51:10 -0800 (PST)
Received: from r-mail1.rd.francetelecom.com (r-mail1.rd.francetelecom.com [217.108.152.41]) by core3.amsl.com (Postfix) with ESMTP id 2D3423A692C for <l3vpn@ietf.org>; Wed, 16 Dec 2009 00:51:09 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp2.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:50:55 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 09:50:54 +0100
Message-ID: <4B289F6E.3020509@orange-ftgroup.com>
Date: Wed, 16 Dec 2009 09:50:54 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Eric Rosen <erosen@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / ASM
References: <3126.1259858356@erosen-linux>
In-Reply-To: <3126.1259858356@erosen-linux>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Dec 2009 08:50:54.0358 (UTC) FILETIME=[E07F4B60:01CA7E2C]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 08:51:11 -0000

Eric, working group,

Eric Rosen:
 >
 > Another methodological error is the fact that the analysis focuses 
heavily
 > on the Source-Specific Multicast mode,
[...]

You made this comment in June, and the document has since precisely been 
revised to cover ASM (section A.1.2).
Still claiming that "the analysis focuses heavily on the Source-Specific 
Multicast mode" is just misleading.

The only reason why A.1.2 (on ASM) it is smaller than A.1.1 (on SSM) is 
because it avoids being redundant, for the portion of the procedures for 
the setup and maintenance of one (S,G) SPT state or one (*,G), that are 
common with one (S,G) SSM state.


 > [...] To support the recommendations of section
 > 3.3, the analysis needs to focus on scenarios that relevant to actual
 > deployments. Some of the analysis of SSM also applies to the ASM 
procedures
 > for joining/pruning source trees, but this is not made clear, and it's
 > difficult for a reader to know how much of the analysis is relevant to
 > actual deployments.

I think that the introduction of A.1.2 is clear on this:

   PEs will generally have to maintain one shared tree, plus one source
   tree for each source sending toward G; each tree resulting in an
   amount of processing and state maintenance similar to what is
   described in the scenario in Appendix A.1.1,

We can make it even clearer by adding :
- in the start of section A.1.2 (ASM):

      The conclusion in A.1.1 are reused in this section, for the parts 
that are
      common to the setup and maintenance of states related to a source 
tree
      or a shared tree.

- in the introduction of section A.1.1 (on SSM):

       For all options provided for C-multicast routing, the procedures 
to setup
       and maintain a shortest path tree toward the source of an SSM 
group are
       the same than the procedures used to setup and maintain a 
shortest path
       tree toward an RP or a non-SSM source ; the results of this 
section are
       thus re-used in section A.1.2.



 >
 > [...] In addition to the time spent on the virtually irrelevant SSM 
scenario, [...]
 >

This statement looks to me as totally abusive:
- SSM by itself is certainly not an "irrelevant scenario" ; 
one-to-many-receivers applications certainly are part the applications 
of enterprise VPN customers (look at section 4.1 and 4.1.1 of RFC4834) ; 
at the very least you'd need to support such a claim with actual data...
- as explained above, many results from the SSM case are reused in the 
ASM case, the behavior related to the maintenance of a tree on the the 
shortest path to a source or RP being common

So, this "time" is not spent unwisely...


 >
 > Now let's look specifically at the ASM case.
 >
 > One of the features of the BGP solution is that there is no way, in BGP,
 > to represent the PIM operation of "prune source S off the (*,G) tree".
 > To compensate for this, every PE in a given VPN is forced to keep 
track of
 > every (S,G) stream for which any PE in that VPN has receivers.  This 
is done
 > by means of the Source Active A-D routes.
 >
 > There are two options that can be used.  In one option, where "inter-site
 > shared trees are not used", each PE functions as a C-RP, and then all 
the PE
 > of a given VPN must exchange information on all the (S,G) streams 
that they
 > see locally.  This option uses BGP to do something that is very 
similar to
 > what MSDP is used for today.  Let me quote from Pekka Savola's review of
 > this procedure:
 >
 >          Active source BGP messages.  This is a duplication of a
 >          similar mechanism in MSDP (RFC3618) which has caused
 >          much grief in Internet.  Does this meant that when a
 >          host does 'nmap -sU 224.0.0.0/4' at a VPN site, this
 >          will result in about 268 million BGP active source
 >          updates being sent (2^28) in the SP backbone?
 >
 > The only answer to this was to beef up the Security Considerations with
 > some required procedures that will prevent the PE from melting down 
entirely
 > in this case.  Of course, Appendix A's methodology of only considering a
 > single (S,G) prevents it from seeing that this particular BGP option
 > uses a mechanism whose scalability is already known to be poor and which
 > the multicast WGs have decided not even to bring forward to IPv6.

Let me note that the potential security issues of this way of supporting 
ASM with BGP-based C-multicast routing, is one of the reasons identified 
in section 4 of the document that lead us to make the recommendation 
that this approach should not be the only one implemented.

Do you want us to remind this in Appendix A.1.2 ?

( The other claim you have that the scaling analysis in Appendix A 
should be extended to look at more than a single (S,G) : I will address 
this in a separate email. )



 > In the other, more sensible option, a PE originates a Source Active A-D
 > route for an (S,G) only when it receives a C-multicast Join route for 
that
 > (S,G).  However, the Source Active A-D route goes to all the PEs in 
the VPN,
 > not just to those that have receivers for that (S,G).  I believe this 
means
 > that the amount of state a given PE maintains in the BGP scheme is
 > proportional to the number of (S,G) flows for which any PE in the VPN has
 > receivers.  Actually, since the SA route contains the RD of the 
originating
 > PE, the situation is a bit worse; if (S,G) traffic is sent over the 
backbone
 > by more than one PE, the other PEs in the same VPN must maintain an 
(S,G) SA
 > A-D route for each such PE.
 >
 > In the BGP solution, even if a PE has only (*,G) receivers, or even if it
 > has no receivers for G at all, it must maintain state for all the 
(S,G)s in
 > the MVPN, multiplied by the average number of upstream PEs per source.


You are correct that for each source sending to a group, there may be 
one S-A A-D route per PE upstream for this source.

I will update the text accordingly.

Let's also note that this won't change the orders of magnitude.

(What you say in the rest of your comment is already mentioned in A.1.2.)




 > Strangely, this scaling issue is considered to be insignificant by 
Appendix
 > A, for the following reasons:


Your claim that this scaling issue would be "considered insignificant" 
is not supported.  Your "following reasons" is only followed by stuff 
related to PIM, and I find nothing in A.1.2 indicating we would downplay 
the dependency of BGP of the number of sources  (and not either for PIM 
actually).



 >
 >         when PIM-based C-multicast routing is used there are
 >         ... possible control plane state transitions triggered by
 >         the reception of (S,G) packets ; such events would induce
 >         processing on all PEs joined to G
 >
 > Control plane state transitions triggered by the reception of data 
packets
 > have not even been mentioned to this point, and no analysis has been 
given
 > to show whether or not they cause a scaling issue, or whether any such
 > scaling issue is mitigated by the use of BGP.  I also really don't 
see what
 > this has to do with the issue at hand, namely the added state overhead of
 > the Source Active A-D routes.


When saying "I also really don't see what this has to do with the issue 
at hand" and mentioning Source Active A-D routes you are attempting to 
redefine the scope of the section you comment.

To avoid confusion, I will address your claims on the overhead of 
per-source routes or messages in a separate message.


 > Certainly there is no analysis provided that
 > would show the SA A-D routes trade off in a favorable manner against the
 > data-driven state changes.
 >
 >       when PIM-based C-multicast routing is used there are ...
 >       possible PIM Assert messages specific to (S,G) ; this would
 >       induce a message processing on each PE of the VPN for each
 >       PIM Assert message
 >
 > It is true that the Source Active A-D routes eliminate the need for 
Assert
 > messages.  The implication is that this is favorable for BGP; 
however, there
 > does not appear to be any analysis showing that.


I don't share with you that what we state implies that the comparison 
with the corresponding mechanisms in PIM would be "favorable for BGP".

The goal for this section is to thoroughly assess the load put on BGP or 
PIM as C-multicast routing protocols in the ASM case.

The hard part is to find a way to compare very distinct mechanisms in 
PIM and BGP (as you said, one relies on dataplane driven events where 
the other does not, and one relies on an Assert mechanisms where the 
other does not, and one does not rely on BGP S-A A-D messages where the 
other does not). And the key is to get something meaningful in the 
context of A.1 (scalability wrt. number of PEs) taking into account the 
mechanisms in play for both PIM and BGP protocols, with hypothesis that 
are relevant for an operator trying to choose between the two.

I think that the section properly explains why the orders of magnitude 
for a increased number of PEs are essentially the same for ASM than the 
one found for SSM.

That said, precise suggestions to amend this subsection are still welcome.

( However, note that we would be wrong to give too much weight to the 
scenario your focus on right above ("if a PE has only (*,G) receivers, 
or if it has no receivers for G at all") as it is not that useful for an 
operator willing chose a C-multicast routing option for a deployment, 
and looking at the PE-PE scalability and dimensioning aspect of the 
issue, because you cannot count on some PEs on the MVPN not having  
(*,G) state or having (*,G) state but no (S,G) state, since this depends 
on customer multicast policy and activity.  )


 > The new material on ASM does not appear to contain any analysis at 
all, just
 > unsupported assertions.

I don't think this is true, and your comments certainly don't support 
this claim.
(But without anything more explicit from you, your claim is just not 
refutable...)

That said, we can possibly make A.1.2 better, and all suggestions are 
welcome.
(easier if we don't get an artificial 6 month delay before we get some 
usable feedback)

-Thomas


From thomas.morin@orange-ftgroup.com  Wed Dec 16 06:02:12 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 2923A3A6881 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 06:02:12 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.049
X-Spam-Level: 
X-Spam-Status: No, score=-3.049 tagged_above=-999 required=5 tests=[AWL=0.200,  BAYES_00=-2.599, HELO_EQ_FR=0.35, RCVD_IN_DNSWL_LOW=-1]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id bNQJql7OlEZJ for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 06:02:11 -0800 (PST)
Received: from p-mail1.rd.francetelecom.com (p-mail1.rd.francetelecom.com [195.101.245.15]) by core3.amsl.com (Postfix) with ESMTP id 8CF323A681C for <l3vpn@ietf.org>; Wed, 16 Dec 2009 06:02:10 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp2.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 15:01:54 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 15:01:52 +0100
Message-ID: <4B28E84F.4000604@orange-ftgroup.com>
Date: Wed, 16 Dec 2009 15:01:51 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Eric Rosen <erosen@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / section A.1.1.3
References: <3126.1259858356@erosen-linux>
In-Reply-To: <3126.1259858356@erosen-linux>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Dec 2009 14:01:52.0192 (UTC) FILETIME=[516ED400:01CA7E58]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 14:02:12 -0000

Eric, working group,

Eric Rosen:
> Section A.1.1.3
>
> The RR has to maintain all the multicast states of all the clients (PEs), as
> long as those multicast states have a PMSI as input interface.  That's way more multicast states than any PE has to maintain.

More exactly, each RR, has to keep the states of all the PEs in its cluster.
Note that this is not specific to multicast, but already true with unicast.


> The analysis in this Appendix simply ignores the negative scaling effects due to the large number
> of RR clients.
>    

I think the analysis in A.1.1.3 properly takes into account the "number 
of RR clients": the "total state maintained" for RRs (last line of the 
table, fourth column) includes a 2x#R_PE factor, corresponding to each 
PE being a client of two RRs.

As far as "negative scaling effects" are concerned, this is your own 
"conclusion". Based on the analysis,  section A.1.1 concludes that the 
different approaches for C-multicast all have to overall maintain an 
amount of state comparable in order of magnitude (linear with the number 
of PEs), but spread differently on the different types of equipments.

(let me also note that this specific comment above could have been done 
during the LC in June, but wasn't)



> The claim is made that when a second PE joins (S,G), the upstream PE does
> not receive a new BGP route.  That may or may not be the case.  The second
> PE will always need to send a C-multicast Join route to the RRs;

> in PIM the second PE might not need to send any message.

This assertion on PIM is incorrect:
RFC4601, section 4.5.7, end of p.73 : when JoinDesired transitions to 
true a Join is always sent.



> Whether the RRs will forward
> the route originated by the second PE depends upon whether that route is
> more preferable (according to normal BGP procedures) than the route
> originated by the first PE.

This is incomplete :
- for a re-advertisement to happen, the route need to be more preferable 
*and* have different attributes
- we have to take into account the aggregation done by RRs (described in 
section 11.4 of draft-ietf-l3vpn-2547bis-mcast-bgp)
- for each RR, whether or not he might have to re-advertise will depend 
on which clusters the "second PE" and "upstream PE" are in


(More on this below...)


> A similar observation applies when C-multicast
> routes are withdrawn.

(yes, similar, and especially with the same correction to make on the 
PIM side : a Prune is always sent when JoinDesired transitions to False 
for a said (S,G) because the last CE has left this stream, RFC4601 
section 4.5.7, end of p.73)


Now, to go back on the computation in  section A.1.1.3 on BGP, you are 
correct that the current text does not yet properly reflects the cases 
where an RR may re-advertises a route they were already advertising. 
Let's look practically at the attributes that could vary :

   - next-hop : updates of the value of this attribute do not "follow" 
PE advertisements, being rewritten by the RR when aggregating 
C-multicast routes having same NLRI (section 11.4 of 
draft-ietf-l3vpn-2547bis-mcast-bgp) ; the advertisements made by 
different RRs may lead to updates of the value of this attribute

   - ORIGINATOR_ID : updates of the value of this attributes do not "follow" PE advertisements, as the RR rewrite it when aggregating C-multicast routes having same NLRI; the advertisements made by different RRs may lead possibly to updates of the value of this attribute, given that each RR would set ORIGINATOR_ID to a value of its own

   - CLUSTER_LIST : the value of this attribute only varies between clusters, changes of the value of this attributes does not "follow" PE advertisements, and only advertisements made by different RRs may lead possibly to updates of the value of this attribute

- LOCAL_PREF : the value of this attribute is determined locally, this 
is true both for the routes advertized by each PE (which could all be 
configured to use the same value) and for a route that results from the 
aggregation by an RR of the route with the same NLRI advertised by the 
PEs of his cluster (the RRs could also be configured to use a local pref 
independent from the local_pref of the routes advertised to him) ; 
updates of the value of this attribute don't "follow" PE or RR 
advertisements

Let's keep in mind that these attributes don't play any active role in 
C-multicast routing (contrary to the role they play in unicast).  Other 
BGP attributes do not have, AFAIK, any particular reason to be set for 
C-multicast routes, and if they were, an RR (or ASBR for attributes 
relevant for inter-AS) could overwrite these values when aggregating 
these routes.

Given the above, for a said C-multicast Source Tree Join (S,G) NLRI, 
what may force an RR to re-advertise the route with different attributes 
to the upstream PE would be the case of an RR of another cluster 
advertising a route better than its current best route.

If we consider our (#R_PE -1 ) joining a said (S,G), one after the other 
after the first PE joining, some of these events may thus lead to a 
re-advertisement to the upstream PE, but the number of times this can 
happen is at worse the number of RRs in clusters having receivers ; plus 
one because of the same route for the local cluster.  Given that we look 
at scalability with an increased number of PEs, we need to consider the 
possibility where all clusters have a receiver PE.

With these elements in mind, it seems to me that the section A.1.1.3 
need to be updated, along with an explanation in the lines of the one 
provided above, so that the second column reads like this:

+------------+----------
|            | upstream
|            | PE (1)
|            |
|            |
|            |
+------------+----------
| first PE   | 2 m.e
| joins      |            (unchanged)
+------------+----------
| for *each* | between 0
| additional | and 2 m.e         (was zero)
| PE joining |
|            |
+------------+----------
| baseline   | 0
| processing | (unchanged)
| over a     |
| period T   |
+------------+----------
| for *each* | between 0
| PE leaving | and 2 m.e       (was zero)
|            |
+------------+----------
| the last   | 2 m.e (unchanged)
| PE leaves  |
+------------+----------
| total for  | at most
| #R_PE PEs  | #RRs m.e    (was 4 m.e)
|            |
|            |
|            |
|            |
+------------+----------

(for clarity, the cells in lines 2,4 and 6 would also need to point to 
the explanation)

Note that between "4 m.e" and "at most #RRs", we do not change the order 
of magnitude when the number of PE increases. This is not a "surprise" 
given the aggregation made by RRs.

-Thomas

PS: Let me also note, again, that your comment could have been done 
during the LC in June, but wasn't.

From thomas.morin@orange-ftgroup.com  Wed Dec 16 07:12:52 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 8E3893A691B for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 07:12:52 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.12
X-Spam-Level: 
X-Spam-Status: No, score=-2.12 tagged_above=-999 required=5 tests=[AWL=0.129,  BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id B3cmrCWX2oUh for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 07:12:47 -0800 (PST)
Received: from r-mail1.rd.francetelecom.com (r-mail1.rd.francetelecom.com [217.108.152.41]) by core3.amsl.com (Postfix) with ESMTP id 22AFC3A680F for <l3vpn@ietf.org>; Wed, 16 Dec 2009 07:12:45 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp1.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 16:12:30 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 16:12:30 +0100
Message-ID: <4B28F8DD.1020704@orange-ftgroup.com>
Date: Wed, 16 Dec 2009 16:12:29 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Eric Rosen <erosen@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / PE-CE
References: <3126.1259858356@erosen-linux>
In-Reply-To: <3126.1259858356@erosen-linux>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Dec 2009 15:12:30.0389 (UTC) FILETIME=[2F985E50:01CA7E62]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 15:12:52 -0000

Eric, working group,


Eric Rosen:
> Another methodological error is that the PE-CE interactions are not taken
> into account.  In many scenarios the PE-PE interactions are just "noise"
> when compared to the real bottleneck, the PE-CE interactions.  The total
> number of PIM messages processed by a PE includes PIM messages from CEs as
> well as from PEs.

This is a comment you had already made during the LC in June, a short 
thread followed on this point, and I provided reasons explaining why 
covering PE-CE aspects would not be a relevant extension to Appendix A.  
Unfortunately, you left my email from July 10th without a reply.


> Suppose that by adding BGP, you could reduce this total
> number of PIM messages by 1%, at the cost of doubling the amount of state
> you need to maintain.  Would that be a good tradeoff?

Come on, Eric, who do you think will buy such a cheap tradesman trick 
?...  ;-)


> In order to support
> the recommendations of section 3.3, Appendix A has to show that the
> tradeoff is a lot better than this.  Yet it doesn't even attempt to do so.

The scope of Appendix A is to provide insight to the arguments in 
section 3.3 which aim at comparing the C-multicast PE-PE multicast 
routing options, by looking at scalability with an increased number of 
PEs, and the cost of a PE joining or leaving.

What you only did in your recent comments is build a specific scenario 
where you show that PE-CE is "the bottleneck". It certainly does not 
mean that this is always the case.   And the concluding recommendations 
do address the case when PE-CE would be the bottleneck (as explained below).

Thus, I stand that what is said in section 3.3 or Appendix A for these 
matters does not mandate extending the scope of the PE-PE scaling 
analysis to the PE-CE overhead.



>
> The methodological problem here is that Appendix A does not consider 
> the PIM
> overhead on the VRF interfaces.  But when BGP C-multicast routing is 
> added,
> it does not replace the C-PIM instance for a given MVPN, it just 
> reduces by
> 1 the number of PIM interfaces supported by that C-PIM interface.  By
> ignoring this fact (and by ignoring the cost of adding a second multicast
> protocol without removing PIM) one makes a small difference look much more
> significant than it really is.  It is not valid for a scalability analysis
> to focus on just one aspect of the protocols unless it has first shown 
> that
> that one aspect is the bottleneck.

I have already explained in the past, and again in the beginning of this 
email, why PE-CE is not part of the analysis in Appendix A.

What *you* are doing is focusing on "steady state" (with a definition of 
steady state that is probably not relevant by the way) to convince us 
that PE-CE is the bottleneck.

While the analysis in A.1 is precisely *not* focusing on steady state or 
non-steady state, but looking at both, so I don't think why we would 
have to show that any is the bottleneck.


> In fact, the analysis of Appendix A fails to demonstrate any real
> scaling advantage, in the steady state, for one protocol over the other.

Appendix A does not "fail" to do what you suggest, since it does not 
need to do so and not claim doing so.

Now, when you say that PE-CE may sometimes be a significant factor of 
the overall load due to multicast routing, I think this is reasonable, 
which is why the following text was added to a new sub -section of 
section 3.3 (section 3.3.2):

    The overhead associated with the PE-CE exchange of C-multicast
    routing is independent of the choice of the mechanism used for the
    PE-PE C-multicast routing.  Therefore, the impact of the PE-CE
    C-multicast routing overhead on the overall system scalability is
    independent of the protocol used for PE-PE signaling, and therefore
    is not relevant when comparing the different approaches proposed for
    the PE-PE C-multicast routing.  This is true even if in some
    operational contexts the PE-CE C-multicast routing overhead is a
    significant factor in the overall system overhead.

Furthermore, to reflect your point that PE-CE may in some cases be an important factor, and to also reflect the fact that actually a provider will have to pounder each argument in its own deployment context, I propose to replace the following from 3.3.7:

    In the meantime, this recommendation would enable service providers
    to choose between the first and the fourth mechanism, without
    this choice being constrained by vendors implementation choices.

  with:

    In the meantime, this recommendation would enable a service
    provider to choose between the first and the fourth mechanism,
    without this choice being constrained by vendors implementation
    choices, and taking into account the peculiarities of its own
    deployment context by pondering the weight of the different
    factors into account.

-Thomas



From thomas.morin@orange-ftgroup.com  Wed Dec 16 07:20:27 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id B45453A6994 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 07:20:27 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.099
X-Spam-Level: 
X-Spam-Status: No, score=-3.099 tagged_above=-999 required=5 tests=[AWL=0.150,  BAYES_00=-2.599, HELO_EQ_FR=0.35, RCVD_IN_DNSWL_LOW=-1]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id rSqpiHNFcqr9 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 07:20:25 -0800 (PST)
Received: from p-mail1.rd.francetelecom.com (p-mail1.rd.francetelecom.com [195.101.245.15]) by core3.amsl.com (Postfix) with ESMTP id 32B843A6991 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 07:20:25 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp1.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 16:20:10 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 16:20:10 +0100
Message-ID: <4B28FAA9.1090304@orange-ftgroup.com>
Date: Wed, 16 Dec 2009 16:20:09 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Eric Rosen <erosen@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / C-multicast routing dynamics
References: <3126.1259858356@erosen-linux>
In-Reply-To: <3126.1259858356@erosen-linux>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Dec 2009 15:20:10.0182 (UTC) FILETIME=[41A73660:01CA7E63]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 15:20:28 -0000

Eric, working group,

Let me first note that these comments quoted below are not specific to 
new content added in releases -04/-05 and should have been done during 
the LC in June.

Eric Rosen:
[...]
> Also omitted from consideration are scaling issues that show up
> as the rate of change of multicast states increases.


No, section A.2 describes the cost of one individual PE joining and 
leaving, which provide the required insight to compare the behavior of 
the different C-multicast routing protocols for different rates of 
change of multicast subscription by PEs.

(This section details what was briefly mentioned in the end of section 
A.4 since version -00 of the WG draft)


> Some amount of the analysis in Appendix A is devoted to the startup and
> shutdown transients, when a PE first joins an (S,G) or prunes an (S,G). 


Indeed...


> A rather big deal is made of the "prune override" procedures.


( I don't think that the analysis focuses more on the "prune override" 
procedures than on any other PIM or BGP procedures.  Please be more 
explicit. )


> I have focused this critique on the steady state, because if the 
> steady state (as I defined
> it above) is actually reached, the startup/shutdown transients are not 
> very
> important.  I suppose one might disagree with that, but certainly thee has
> been no analysis showing that the transients are more important than the
> steady state.


The statements above are "surprising":
- first, you imply that "a steady state" may be "reached" : this 
certainly does not take into account that PEs joining and leaving is 
something that, unlike unicast routing changes, happens following 
customer multicast application use patterns and customer routing 
policy   (by the way, "transient" is probably not the best term, since 
it is most often used for unicast routing which has specific dynamics)
- second, you state that "thee has been no analysis showing that the 
transients are more important than the steady state" :
     * this is certainly true, but at the same time, there is nothing is 
the analysis that claim that
     * and the reverse (steady state more important than dynamic events) 
is the basis for your claim that we do not take into account the PE-CE 
side, but certainly has not been shown either

Given the above, let me remind you of the following : Appendix A looks both at the transient scenario with some join/leave ("for each additional PE joining/leaving")*and*  the steady state scenario ("baseline processing over a period T"), where for some period no PE joins or leaves.

If we were to constrain the analysis to*only*  the scenario "where no PE joins or leaves" (that is purely imaginary by the way, as such a period must at least start with a period where some PE join a streams, and would have an end too), we would precisely loose generality and irrelevantly narrow the scope.


> It would certainly be interesting to analyze the case where there is 
> so much
> multicast churn that steady state is never reached.  However, that 
> cannot be
> done by focusing on a single (S,G) over a limited time period.


(The above statement is unsupported.)

-Thomas


From yakov@juniper.net  Wed Dec 16 07:21:31 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id AAE093A6984 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 07:21:30 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.256
X-Spam-Level: 
X-Spam-Status: No, score=-6.256 tagged_above=-999 required=5 tests=[AWL=-0.257, BAYES_00=-2.599, J_CHICKENPOX_41=0.6, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id BDpFPdBhQ0Cj for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 07:21:29 -0800 (PST)
Received: from exprod7og112.obsmtp.com (exprod7og112.obsmtp.com [64.18.2.177]) by core3.amsl.com (Postfix) with SMTP id 201493A682C for <l3vpn@ietf.org>; Wed, 16 Dec 2009 07:21:26 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob112.postini.com ([64.18.6.12]) with SMTP ID DSNKSyj659kjwuNhlAOPP8YOleCvwzc0Nzhp@postini.com; Wed, 16 Dec 2009 07:21:15 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Wed, 16 Dec 2009 07:13:10 -0800
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Wed, 16 Dec 2009 07:13:09 -0800
Received: from emailsmtp55.jnpr.net ([172.24.18.132]) by p-emlb01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Wed, 16 Dec 2009 07:13:09 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp55.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Wed, 16 Dec 2009 07:13:09 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBGFD8j71754; Wed, 16 Dec 2009 07:13:09 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912161513.nBGFD8j71754@magenta.juniper.net>
To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
Subject: Re: Fwd: draft-ietf-l3vpn-mvpn-considerations-05 
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <53280.1260976388.1@juniper.net>
Date: Wed, 16 Dec 2009 07:13:08 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 16 Dec 2009 15:13:09.0419 (UTC) FILETIME=[46DBDFB0:01CA7E62]
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, l3vpn@ietf.org
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 15:21:31 -0000

Maria,

> Clearly, the amount of processing on the upstream PE to which the
> joined is destined is not the same as the "processing" on other PEs. Any
> other PE than the upstream PE will simply discard this Join because the
> neighbor address in the PIM Join is not this PE address. 

The last sentence above is incorrect. Specifically, some other than
the upstream PE can *not* "simply discard" the Join(C-S, C-G)/Join(C-*, C-G)
just "because the neighbor address in the Join is not this PE
address", as that other PE may wish to propagate a
Join(C-S, C-G)/Join(C-*, C-G) upstream.

>From 4.5.7 of rfc4601:

    If a router wishes to propagate a Join(S,G) upstream, it must also
    watch for messages on its upstream interface from other routers on
    that subnet, and these may modify its behavior. 

Yakov.

From mn1921@att.com  Wed Dec 16 08:39:10 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 598FA3A67AB for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 08:39:10 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -105.988
X-Spam-Level: 
X-Spam-Status: No, score=-105.988 tagged_above=-999 required=5 tests=[AWL=0.011, BAYES_00=-2.599, J_CHICKENPOX_41=0.6, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ZonkYEoAdY-3 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 08:39:09 -0800 (PST)
Received: from mail167.messagelabs.com (mail167.messagelabs.com [216.82.253.179]) by core3.amsl.com (Postfix) with ESMTP id 77B963A67FA for <l3vpn@ietf.org>; Wed, 16 Dec 2009 08:39:09 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-12.tower-167.messagelabs.com!1260981532!26860742!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 18544 invoked from network); 16 Dec 2009 16:38:53 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-12.tower-167.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 16 Dec 2009 16:38:53 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBGGcmas031301; Wed, 16 Dec 2009 11:38:48 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBGGchMX031244; Wed, 16 Dec 2009 11:38:44 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: Fwd: draft-ietf-l3vpn-mvpn-considerations-05 
Date: Wed, 16 Dec 2009 11:38:45 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A200451C7A0@misout7msgusr7e.ugd.att.com>
In-Reply-To: <200912161513.nBGFD8j71754@magenta.juniper.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Fwd: draft-ietf-l3vpn-mvpn-considerations-05 
Thread-Index: Acp+Y3tc0isAOBkaQz6gTjpISVsJDQACoExA
References: <200912161513.nBGFD8j71754@magenta.juniper.net>
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: "Yakov Rekhter" <yakov@juniper.net>
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, l3vpn@ietf.org
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 16:39:10 -0000

Yakov,

as Thomas correctly stated in one the e-mails he sent today:

"the other PEs parse the message, and lookup their TIB to possibly=20
find an (S,G) state; they find none and discard the information in this=20
Join) ; note that they don't have any (S,G) state at this point"

This is what I meant.

Maria

> -----Original Message-----
> From: l3vpn-bounces@ietf.org [mailto:l3vpn-bounces@ietf.org] On Behalf
> Of Yakov Rekhter
> Sent: Wednesday, December 16, 2009 10:13 AM
> To: NAPIERALA, MARIA H (ATTLABS)
> Cc: Thomas Morin; l3vpn@ietf.org
> Subject: Re: Fwd: draft-ietf-l3vpn-mvpn-considerations-05
>=20
> Maria,
>=20
> > Clearly, the amount of processing on the upstream PE to which the
> > joined is destined is not the same as the "processing" on other PEs.
> Any
> > other PE than the upstream PE will simply discard this Join because
> the
> > neighbor address in the PIM Join is not this PE address.
>=20
> The last sentence above is incorrect. Specifically, some other than
> the upstream PE can *not* "simply discard" the Join(C-S,
C-G)/Join(C-*,
> C-G)
> just "because the neighbor address in the Join is not this PE
> address", as that other PE may wish to propagate a
> Join(C-S, C-G)/Join(C-*, C-G) upstream.
>=20
> From 4.5.7 of rfc4601:
>=20
>     If a router wishes to propagate a Join(S,G) upstream, it must also
>     watch for messages on its upstream interface from other routers on
>     that subnet, and these may modify its behavior.
>=20
> Yakov.

From thomas.morin@orange-ftgroup.com  Wed Dec 16 09:11:12 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 687633A67A4 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 09:11:12 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.829
X-Spam-Level: 
X-Spam-Status: No, score=-2.829 tagged_above=-999 required=5 tests=[AWL=-0.180, BAYES_00=-2.599, HELO_EQ_FR=0.35, J_CHICKENPOX_51=0.6, RCVD_IN_DNSWL_LOW=-1]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id BXAG2nk60jIn for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 09:11:08 -0800 (PST)
Received: from p-mail1.rd.francetelecom.com (p-mail1.rd.francetelecom.com [195.101.245.15]) by core3.amsl.com (Postfix) with ESMTP id 535613A69EE for <l3vpn@ietf.org>; Wed, 16 Dec 2009 09:11:06 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp1.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 18:10:50 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Wed, 16 Dec 2009 18:10:49 +0100
Message-ID: <4B291498.6060105@orange-ftgroup.com>
Date: Wed, 16 Dec 2009 18:10:48 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Maria Napierala <mnapierala@att.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / M.Napierala's comments
References: <2F1DE4DFCFF32144B771BD2C246E6A200446CE07@misout7msgusr7e.ugd.att.com>
In-Reply-To: <2F1DE4DFCFF32144B771BD2C246E6A200446CE07@misout7msgusr7e.ugd.att.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 16 Dec 2009 17:10:49.0468 (UTC) FILETIME=[B6F9DBC0:01CA7E72]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 17:11:12 -0000

Maria, working group,

Find below some replies to some of your comments.

But first of all : correct me if I'm wrong, but it does not seem to me 
that any of them is specific to changes made to the document since the 
WG Last Call in June.... ??


NAPIERALA, MARIA H (ATTLABS):
> I have several comments on the latest version of the draft:
>
> 1. I believe that the draft should contain some analysis on how PIM-SM
> differs when handled by BGP signaling.  The difference is significant
> enough to warrant at least a theoretical consideration if no significant
> field experience is available.  Currently, vast majority of multicast
> applications used in the enterprise networks are PIM-SM based. It is
> important to analyze the BGP translation of pruning source traffic off
> the shared tree because this is the most significant change from PIM-SM.
>    

"significant enough to warrant" ... but not worth making the suggestion 
six month after the LC ?  ;-)


> When BGP is used for PE-PE C-multicast signaling the data-driven PIM-SM
> state changes are replaced by additional BGP (Source Active) routes and
> timer-driven PIM messages. Let me point out just few aspects related to
> those changes that should be considered, starting from timer-driven
> Prune(S,G,rpt).
>
> When a single forwarder selection is the procedure used to avoid sending
> duplicate packets to MVPN customers during switching from RPT to SPT,
> the correct setting of the above mentioned timer to prune source traffic
> off the shared tree is important. It is not going to be trivial to tune
> such timer in large networks, because it can cause either duplicate
> traffic (if timer is set too high) or loss of traffic (if timer is set
> too low). This approach to PIM-SM changes the way the PIM-SM multicast
> trees are constructed, and it is not known to what extend this can
> impact existing multicast applications.
> If the solution to this is to make the timer for sending Prune(S,G,rpt)
> sufficiently high (to assure that SPT is created) and avoid duplicates
> sent to customers by having egress PEs discard data packets received
> from the "wrong" upstream PE, this should be included in MVPN
> considerations draft.


I personnally agree with you it would be reasonable that the draft 
recommends implementing these procedures ( 
http://tools.ietf.org/html/draft-ietf-l3vpn-2547bis-mcast-09#section-9.1.1 )

Let's propose this to the working group (keeping in mind the delay after 
WG LC...).


> It should be also considered that it may not
> always be possible for the egress PE to determine the upstream PE, in
> which case the single forwarder selection has to be used.
>    


> Regarding Source Active A-D routes, it was pointed out recently and in
> the past that the amount of BGP state they generate should be analyzed.
> The Appendix A ignores this important scalability aspect of
> BGP-signaling.
>    

( I'll comment on this one in a later email, in reply to Eric Rosen 
comments)


> The field experience shows that there are multicast applications used in
> enterprise VPNs such that only control messages, but no multicast data,
> are sent across MVPN backbone. When BGP-multicast signaling is used, the
> Source Active A-D route is originated for every C-(S,G) flow that
> registers with C-RP and such that the C-RP joins "S" across MVPN
> backbone, even if no inter-site data is sent for this flow. Hence, the
> BGP scheme requires not only that for any given C-(S,G) flow that enters
> MVPN backbone an SA message is announced and maintained by all PEs in a
> given MVPN, but that SA messages are also announced for flows that never
> enter the MVPN backbone.
>    

What you state above it true only with one way of using the BGP approach 
to support ASM (the one in section 14 of 
draft-ietf-l3vpn-2547bis-mcast-bgp)?

The other way (section 13) does not have this effect, and happens to 
indirectly be the one recommended by 
draft-ietf-l3vpn-mvpn-considerations (because it is the one of the two 
not requiring an RP function co-located in the PE).

So well, I'm not sure that this point would be very important to cover.


> 2. The considerations draft should also point out the difference how BGP
> vs. PIM (used as PE-PE signaling protocol) handles the upstream PE
> selection in case of multi-homed sources/RPs.
>
> Typically, in a layer 3 VPN deployment it is not known whether or not
> all VRFs of a given VPN have the same best route to a given source.  In
> other words, VRFs routing policies are provisioned independently of each
> other. If all VRFs of a given MVPN have the same best/ installed route
> to a multi-homed source/RP then this VPN customer *expects* that the
> source/RP will be joined via the installed route.
>
> When PIM is used as PE-PE C-signaling, the installed route towards the
> source is always chosen as UHM.

This is not my understanding, to quote 
draft-ietf-l3vpn-2547bis-mcast-09, section-5.1.3:

    The default procedure, which MUST be implemented, is to select the
    route whose corresponding upstream PE address is numerically highest,
    where a 32-bit IP address is treated as a 32 bit unsigned integer.
    Call this the "default upstream PE selection".

    [...]

    Another procedure, which SHOULD be implemented, is to use the
    Installed UMH Route as the Selected UMH Route.


>   If different VRFs have different best
> routes to C-S or C-RP, the Assert process is a tie breaker. No steady
> state duplicate traffic is sent across MVPN backbone.
>
> Now let's consider BGP as the PE-PE signaling protocol and let's assume
> the UMH policy is to choose the installed/best route towards C-source or
> C-RP.

Given the above, why is this assumption relevant ?

>   If different VRFs have different installed routes to C-S/C-RP then
> there will be a steady state duplication of traffic across the MVPN
> backbone (to the same set of PEs if MI-PMSI is used). This also requires
> that the egress PEs be able to discard data traffic that arrives from
> the "wrong" upstream PE. Such duplication of traffic might not be
> operationally acceptable to a service provider because of too much
> network bandwidth being consumed.

I don't think this is a relevant drawback :
- if a provider chooses the non-default UMH selection you mention,  it 
is likely that he is fine with the bandwidth aspect, and has made sure 
the vendors implement the procedures to discard data traffic that 
arrives from the "wrong" upstream PE .
- see above for the suggestion to recommend implementing the procedures 
required to discard data traffic that arrives from the "wrong" upstream 
PE (making it easier for the provider to make sure that it is ok for him 
to use the non-default UMH selection you mention ).


> The other UMH policy that can be used
> with BGP signaling where duplicates are not generated, is a single or
> default forwarder selection. However, single forwarder selection might
> not choose the installed/best route (to a multi-homed C-S/C-RP) *even*
> if all VRFs have the same best route to C-S/C-RP. Hence, in a MVPN where
> all VRFs have the same best path to C-S or C-RP, its multicast traffic
> might not flow via the best path as is expected by the VPN customer.
>    

See above.


> 3. In section A.1.1.3, why the "each additional PE joining" row for the
> "upstream PE" column shows "0"? The same question for "upstream PEs"
> column and the "each PE leaving" row.
>    

See reply to Eric Rosen "draft-ietf-l3vpn-mvpn-considerations-05 / 
Appendix A / section A.1.1.3".

> 4. Why Appendix A assumes that the RT-constraint is deployed when BGP
> signaling is used?  This is not a mandatory implementation.
>
>    

See reply to Eric Rosen "draft-ietf-l3vpn-mvpn-considerations-05 / 
Appendix A / RT Constraint recommendation".

> 5. Why Appendix A analyzes PIM LAN procedure with explicit tracking but
> not the explicit tracking using BGP-signaling?
>    

I'm not sure as what you call "explicit tracking using BGP-signaling".

There is, AFAIK, no such thing as "explicit tracking" in the BGP 
procedures for C-multicast routing, and Appendix A is only about 
C-multicast routing options.

( Possibly you refer to the case where BGP auto-discovery procedures are 
used to allow a PE to know about which PEs would be leaves of S-PMSI 
tunnels ?)


> 6. The appendix states that "we propose here a rough estimation of the
> load of PEs". And it is in fact a very *rough* estimation and not even a
> complete *rough* estimation as it almost dismisses ASM analysis.

On "almost dismisses ASM analysis", see reply to Eric Rosen 
"draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / ASM".


> The appendix treats different and, in fact, not comparable kinds of
> "processing" as *equal* and uses the same unit "m.e" to measure them.
>    

On "non comparable" I disagree; moreover this is not something having 
been brought up by anyone until now, AFAICT.


> For example, the first row in the table in section A.1.1.1, says that to
> process the first join on the upstream PE takes "1 m.e" and that the
> same amount of processing, "1 m.e", is required on every other PE in
> MVPN. Clearly, the amount of processing on the upstream PE to which the
> joined is destined is not the same as the "processing" on other PEs.
>    
>   Any other PE than the upstream PE will simply discard this Join because the
> neighbor address in the PIM Join is not this PE address.

See correction by Yakov, which you acknowledged, and let me point out 
that, if you write "simply discard this Join because the neighbor 
address in the PIM Join is not this PE address" while you mean "parse + 
lookup in TIB + discard", then I disagree with you because that "parse + 
lookup in TIB" does not seem to me as insignificant.


>   While the
> upstream PE has to create the (s,g) state in the control plane, program
> the data plane for forwarding, and send Join (s,g) message upstream.
> This is by any measure much more processing than discarding the Join
> based on the address.

(see above, PE that are not the upstream have more to do than merely 
discarding the Join based on the upstream address ).


That said it is true that among the upstream PE for the first join, the 
upstream PE for later joins or  the PEs that have (S,G) state, and the 
other PEs that have no (S,G) state, not all of them have the exact 
amount of processing to do.

Where I'm not sure is about, when some equipment has more to do than 
another one, "how much more" it is ; it seems to me that it is in fact 
likely to be very much vendor specific.

>   Yet, it is assumed that the resources used in both
> cases are the same.
>    

Well the idea is to accept this approximation, and document it.

The introduction of Appendix A already provides such an explanation, and 
your comment would lead to think of improving it.  I see that this is 
what Yakov Rekhter as suggested, and I think that this is a good idea.

> Another example is the row "each PE leaving" (but last) in the same
> table in section A.1.1.1. The only processing the upstream PE does is
> setting a timer for receiving prune-override PIM Join message, and when
> such PIM join is received it cancels the timer. This is estimated as
> "2 m.e"

Not exactly, the "2 m.e" includes the processing of the Prune *and* the 
processing of the Join send by one PE to override the Prune.

> , yet, this is much less "work" to be done by the upstream PE
> than in the previous "first PE joins" example.

(see above with reference to how we can reasonably take into account 
"how much less" work it is)


>   In addition, the column
> "other PEs" in the row "each PE leaving" states that every other PE in
> MVPN will also have to do an action worth "2 m.e". Yet, only the PEs
> that have interested receivers in (s,g) have to perform a task (i.e.,
> randomly generate a join message delay timer and cancel it when they
> hear a join from another PE; only one of those PEs will send a PIM Join
> message). The PEs that do need to receive (s,g) traffic will simply drop
> the (s,g) prune message.
>    

This is not exact.
(see the correction brought up by Yakov, for the details)


>   Hence, the application of the multiplier
> #MVPN_PE in this row/column is also not correct.
>    

(it is correct: see the reply to Eric Rosen 
"draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / PIM LAN procedures")


> This "problem" of treating not equal actions as equal repeats itself
> throughout the scaling analysis in the appendix. Such analysis is not
> only *rough* but misleading, and because of this may lead to wrong
> conclusions. I believe that unless the appendix A is corrected by
> removing all not comparable, or even wrong estimations, it should be
> removed from the draft.

...but you only find this suggestion worth enough bring brought forward 
only six month or so after the WG LC... ;-)

I think that the proposed modification to the introduction of Appendix A 
made by Yakov is much more reasonable (copy-pasted below for easier 
follow up) :

    Though the overhead of message processing and the state overhead
    may vary between different approaches, and even within the same
    approach between different equipments (PEs or RRs), we propose
    here a rough estimation of the load of PEs, in terms of number
    of messages processed and number of control plane states maintained.


-Thomas

From mn1921@att.com  Wed Dec 16 10:58:40 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id EAB4E3A6A19 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 10:58:40 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -105.991
X-Spam-Level: 
X-Spam-Status: No, score=-105.991 tagged_above=-999 required=5 tests=[AWL=0.008, BAYES_00=-2.599, J_CHICKENPOX_51=0.6, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 05HOABEsp5PK for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 10:58:39 -0800 (PST)
Received: from mail129.messagelabs.com (mail129.messagelabs.com [216.82.250.147]) by core3.amsl.com (Postfix) with ESMTP id 098583A69C6 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 10:58:39 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-8.tower-129.messagelabs.com!1260989903!37204200!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 7762 invoked from network); 16 Dec 2009 18:58:24 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-8.tower-129.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 16 Dec 2009 18:58:24 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBGIwK8B002740 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 13:58:20 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBGIwFc0002705 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 13:58:16 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
x-cr-puzzleid: {3BCFFDA4-C652-47C4-9299-649772EBEDB9}
x-cr-hashedpuzzle: B4Rv Cr5j ETwk EduZ FO46 FhGN GGhH Ghme H5HB Kb23 NnA1 OY6W PNND Qgj9 TSVL UQes; 2; bAAzAHYAcABuAEAAaQBlAHQAZgAuAG8AcgBnADsAdABoAG8AbQBhAHMALgBtAG8AcgBpAG4AQABvAHIAYQBuAGcAZQAtAGYAdABnAHIAbwB1AHAALgBjAG8AbQA=; Sosha1_v1; 7; {3BCFFDA4-C652-47C4-9299-649772EBEDB9}; bQBuADEAOQAyADEAQABhAHQAdAAuAGMAbwBtAA==; Wed, 16 Dec 2009 18:57:59 GMT; UgBFADoAIABkAHIAYQBmAHQALQBpAGUAdABmAC0AbAAzAHYAcABuAC0AbQB2AHAAbgAtAGMAbwBuAHMAaQBkAGUAcgBhAHQAaQBvAG4AcwAtADAANQAgAC8AIABNAC4ATgBhAHAAaQBlAHIAYQBsAGEAJwBzACAAYwBvAG0AbQBlAG4AdABzAA==
Content-class: urn:content-classes:message
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / M.Napierala's comments
Date: Wed, 16 Dec 2009 13:57:59 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A200451C8AE@misout7msgusr7e.ugd.att.com>
In-Reply-To: <4B291498.6060105@orange-ftgroup.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 / M.Napierala's comments
Thread-Index: Acp+csFhDUTG/gPqQBWv8WrbLM948gAAoxzg
References: <2F1DE4DFCFF32144B771BD2C246E6A200446CE07@misout7msgusr7e.ugd.att.com> <4B291498.6060105@orange-ftgroup.com>
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: "Thomas Morin" <thomas.morin@orange-ftgroup.com>, <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 18:58:41 -0000

Thomas,

> > 1. I believe that the draft should contain some analysis on how PIM-
> SM
> > differs when handled by BGP signaling.  The difference is
significant
> > enough to warrant at least a theoretical consideration if no
> significant
> > field experience is available.  Currently, vast majority of
multicast
> > applications used in the enterprise networks are PIM-SM based. It is
> > important to analyze the BGP translation of pruning source traffic
> off
> > the shared tree because this is the most significant change from
PIM-
> SM.
> >
>=20
> "significant enough to warrant" ... but not worth making the
suggestion
> six month after the LC ?  ;-)

You must recall that the request to analyze the BGP translation of
pruning source traffic off the shared tree in MVPN-consideration draft
was made several times by Eric Rosen (as well as by me) in the past. I
don't think that this significant change from PIM-SM procedures can be
ignored.

> > When BGP is used for PE-PE C-multicast signaling the data-driven
PIM-
> SM
> > state changes are replaced by additional BGP (Source Active) routes
> and
> > timer-driven PIM messages. Let me point out just few aspects related
> to
> > those changes that should be considered, starting from timer-driven
> > Prune(S,G,rpt).
> >
> > When a single forwarder selection is the procedure used to avoid
> sending
> > duplicate packets to MVPN customers during switching from RPT to
SPT,
> > the correct setting of the above mentioned timer to prune source
> traffic
> > off the shared tree is important. It is not going to be trivial to
> tune
> > such timer in large networks, because it can cause either duplicate
> > traffic (if timer is set too high) or loss of traffic (if timer is
> set
> > too low). This approach to PIM-SM changes the way the PIM-SM
> multicast
> > trees are constructed, and it is not known to what extend this can
> > impact existing multicast applications.
> > If the solution to this is to make the timer for sending
> Prune(S,G,rpt)
> > sufficiently high (to assure that SPT is created) and avoid
> duplicates
> > sent to customers by having egress PEs discard data packets received
> > from the "wrong" upstream PE, this should be included in MVPN
> > considerations draft.
>=20
>=20
> I personnally agree with you it would be reasonable that the draft
> recommends implementing these procedures (
> http://tools.ietf.org/html/draft-ietf-l3vpn-2547bis-mcast-09#section-
> 9.1.1 )

Good.


> > It should be also considered that it may not
> > always be possible for the egress PE to determine the upstream PE,
in
> > which case the single forwarder selection has to be used.
> >
>=20
>=20
> > Regarding Source Active A-D routes, it was pointed out recently and
> in
> > the past that the amount of BGP state they generate should be
> analyzed.
> > The Appendix A ignores this important scalability aspect of
> > BGP-signaling.
> >
>=20
> ( I'll comment on this one in a later email, in reply to Eric Rosen
> comments)

OK

> > The field experience shows that there are multicast applications
used
> in
> > enterprise VPNs such that only control messages, but no multicast
> data,
> > are sent across MVPN backbone. When BGP-multicast signaling is used,
> the
> > Source Active A-D route is originated for every C-(S,G) flow that
> > registers with C-RP and such that the C-RP joins "S" across MVPN
> > backbone, even if no inter-site data is sent for this flow. Hence,
> the
> > BGP scheme requires not only that for any given C-(S,G) flow that
> enters
> > MVPN backbone an SA message is announced and maintained by all PEs
in
> a
> > given MVPN, but that SA messages are also announced for flows that
> never
> > enter the MVPN backbone.
> >
>=20
> What you state above it true only with one way of using the BGP
> approach
> to support ASM (the one in section 14 of
> draft-ietf-l3vpn-2547bis-mcast-bgp)?

> The other way (section 13) does not have this effect, and happens to
> indirectly be the one recommended by
> draft-ietf-l3vpn-mvpn-considerations (because it is the one of the two
> not requiring an RP function co-located in the PE).
>=20

No, what I stated above applies to section 13 (I did not even considered
the implications of section 14..). The inter-site shared trees are used.
When a PE receives C-multicast Join route (which triggers Source Active
A-D route) it does not necessarily mean that data was actually sent
across MVPN backbone. We see a lot of applications behaving that way in
the field.
=20
>=20
> > 2. The considerations draft should also point out the difference how
> BGP
> > vs. PIM (used as PE-PE signaling protocol) handles the upstream PE
> > selection in case of multi-homed sources/RPs.
> >
> > Typically, in a layer 3 VPN deployment it is not known whether or
not
> > all VRFs of a given VPN have the same best route to a given source.
> In
> > other words, VRFs routing policies are provisioned independently of
> each
> > other. If all VRFs of a given MVPN have the same best/ installed
> route
> > to a multi-homed source/RP then this VPN customer *expects* that the
> > source/RP will be joined via the installed route.
> >
> > When PIM is used as PE-PE C-signaling, the installed route towards
> the
> > source is always chosen as UHM.
>=20
> This is not my understanding, to quote
> draft-ietf-l3vpn-2547bis-mcast-09, section-5.1.3:
>=20
>     The default procedure, which MUST be implemented, is to select the
>     route whose corresponding upstream PE address is numerically
> highest,
>     where a 32-bit IP address is treated as a 32 bit unsigned integer.
>     Call this the "default upstream PE selection".
>=20
>     [...]
>=20
>     Another procedure, which SHOULD be implemented, is to use the
>     Installed UMH Route as the Selected UMH Route.
>=20
>=20
> >   If different VRFs have different best
> > routes to C-S or C-RP, the Assert process is a tie breaker. No
steady
> > state duplicate traffic is sent across MVPN backbone.
> >
> > Now let's consider BGP as the PE-PE signaling protocol and let's
> assume
> > the UMH policy is to choose the installed/best route towards
C-source
> or
> > C-RP.
>=20
> Given the above, why is this assumption relevant ?
>=20
> >   If different VRFs have different installed routes to C-S/C-RP then
> > there will be a steady state duplication of traffic across the MVPN
> > backbone (to the same set of PEs if MI-PMSI is used). This also
> requires
> > that the egress PEs be able to discard data traffic that arrives
from
> > the "wrong" upstream PE. Such duplication of traffic might not be
> > operationally acceptable to a service provider because of too much
> > network bandwidth being consumed.
>=20
> I don't think this is a relevant drawback :
> - if a provider chooses the non-default UMH selection you mention,  it
> is likely that he is fine with the bandwidth aspect, and has made sure
> the vendors implement the procedures to discard data traffic that
> arrives from the "wrong" upstream PE.

Why not to state it? This does not come for free and requires "strong"
RPF check i.e., checking the upstream PE. Are you aware of any such
implementations?=20

> - see above for the suggestion to recommend implementing the
procedures
> required to discard data traffic that arrives from the "wrong"
upstream
> PE (making it easier for the provider to make sure that it is ok for
> him
> to use the non-default UMH selection you mention ).
>=20
>=20
> > The other UMH policy that can be used
> > with BGP signaling where duplicates are not generated, is a single
or
> > default forwarder selection. However, single forwarder selection
> might
> > not choose the installed/best route (to a multi-homed C-S/C-RP)
> *even*
> > if all VRFs have the same best route to C-S/C-RP. Hence, in a MVPN
> where
> > all VRFs have the same best path to C-S or C-RP, its multicast
> traffic
> > might not flow via the best path as is expected by the VPN customer.
> >
>=20
> See above.

Same as above for me also.
=20
> > 3. In section A.1.1.3, why the "each additional PE joining" row for
> the
> > "upstream PE" column shows "0"? The same question for "upstream PEs"
> > column and the "each PE leaving" row.
> >
>=20
> See reply to Eric Rosen "draft-ietf-l3vpn-mvpn-considerations-05 /
> Appendix A / section A.1.1.3".

OK. I will read and reply there.
=20
> > 4. Why Appendix A assumes that the RT-constraint is deployed when
BGP
> > signaling is used?  This is not a mandatory implementation.
> >
> >
>=20
> See reply to Eric Rosen "draft-ietf-l3vpn-mvpn-considerations-05 /
> Appendix A / RT Constraint recommendation".

I will read it.

> > 5. Why Appendix A analyzes PIM LAN procedure with explicit tracking
> but
> > not the explicit tracking using BGP-signaling?
> >
>=20
> I'm not sure as what you call "explicit tracking using BGP-signaling".
>=20
> There is, AFAIK, no such thing as "explicit tracking" in the BGP
> procedures for C-multicast routing, and Appendix A is only about
> C-multicast routing options.

Well, in that case, PIM LAN procedure with explicit tracking should be
removed also. Why would you do explicit tracking of receivers when using
PIM?
=20
>=20
> > 6. The appendix states that "we propose here a rough estimation of
> the
> > load of PEs". And it is in fact a very *rough* estimation and not
> even a
> > complete *rough* estimation as it almost dismisses ASM analysis.
>=20
> On "almost dismisses ASM analysis", see reply to Eric Rosen
> "draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / ASM".

OK. I will read and reply there.
=20
> > The appendix treats different and, in fact, not comparable kinds of
> > "processing" as *equal* and uses the same unit "m.e" to measure
them.
> >
>=20
> On "non comparable" I disagree; moreover this is not something having
> been brought up by anyone until now, AFAICT.
>=20
>=20
> > For example, the first row in the table in section A.1.1.1, says
that
> to
> > process the first join on the upstream PE takes "1 m.e" and that the
> > same amount of processing, "1 m.e", is required on every other PE in
> > MVPN. Clearly, the amount of processing on the upstream PE to which
> the
> > joined is destined is not the same as the "processing" on other PEs.
> >
> >   Any other PE than the upstream PE will simply discard this Join
> because the
> > neighbor address in the PIM Join is not this PE address.
>=20
> See correction by Yakov, which you acknowledged, and let me point out
> that, if you write "simply discard this Join because the neighbor
> address in the PIM Join is not this PE address" while you mean "parse
+
> lookup in TIB + discard", then I disagree with you because that "parse
> +
> lookup in TIB" does not seem to me as insignificant.

First of all, this is still not the same amount of work that the
upstream PE has to perform, i.e., to create the (S,G) state, propagate
it upstream, and set up the forwarding state.

Second, this is still just *discarding* the massage based on two things:
"I am not the upstream router" and "I don't have (S,G) state". This is
all the work the "other PE" has to do in this case. You cannot claim
that this is equal with what the upstream PE has to do, right?

Third, regarding the workload on the upstream PE when BGP signaling is
used, how is BGP processing accounted for? All the work that upstream PE
needs to do in PIM signaling case has also to be done in BGP signaling
case. But with BGP signaling, upstream PE has also to perform the BGP
processing (and *parsing*), including the best path selection. I don't
see this being accounted for at all on PEs.
If you worry so much about the work a PE has to do to *discard* a
message, why would you ignore a significant amount of BGP related work a
PE?
=20
> >   While the
> > upstream PE has to create the (s,g) state in the control plane,
> program
> > the data plane for forwarding, and send Join (s,g) message upstream.
> > This is by any measure much more processing than discarding the Join
> > based on the address.
>=20
> (see above, PE that are not the upstream have more to do than merely
> discarding the Join based on the upstream address ).

See my comment above also.

> That said it is true that among the upstream PE for the first join,
the
> upstream PE for later joins or  the PEs that have (S,G) state, and the
> other PEs that have no (S,G) state, not all of them have the exact
> amount of processing to do.
>=20
> Where I'm not sure is about, when some equipment has more to do than
> another one, "how much more" it is ; it seems to me that it is in fact
> likely to be very much vendor specific.

That is why I believe that attempting such an analysis as in Appendix A
is not a right way to go. And this is why I recommend removing it from
the draft.=20

>=20
> >   Yet, it is assumed that the resources used in both
> > cases are the same.
> >
>=20
> Well the idea is to accept this approximation, and document it.

This is precisely the problem. Such approximations are quite arbitrary
and the danger is that an analysis made of such approximations might
lead to wrong conclusions.=20

>=20
> The introduction of Appendix A already provides such an explanation,
> and
> your comment would lead to think of improving it.  I see that this is
> what Yakov Rekhter as suggested, and I think that this is a good idea.

But improving the *description* of the issue will not solve the issue
itself.

>=20
> > Another example is the row "each PE leaving" (but last) in the same
> > table in section A.1.1.1. The only processing the upstream PE does
is
> > setting a timer for receiving prune-override PIM Join message, and
> when
> > such PIM join is received it cancels the timer. This is estimated as
> > "2 m.e"
>=20
> Not exactly, the "2 m.e" includes the processing of the Prune *and*
the
> processing of the Join send by one PE to override the Prune.
>=20
> > , yet, this is much less "work" to be done by the upstream PE
> > than in the previous "first PE joins" example.
>=20
> (see above with reference to how we can reasonably take into account
> "how much less" work it is)
>=20
>=20
> >   In addition, the column
> > "other PEs" in the row "each PE leaving" states that every other PE
> in
> > MVPN will also have to do an action worth "2 m.e". Yet, only the PEs
> > that have interested receivers in (s,g) have to perform a task
(i.e.,
> > randomly generate a join message delay timer and cancel it when they
> > hear a join from another PE; only one of those PEs will send a PIM
> Join
> > message). The PEs that do need to receive (s,g) traffic will simply
> drop
> > the (s,g) prune message.
> >
>=20
> This is not exact.
> (see the correction brought up by Yakov, for the details)

See my reply above.

> >   Hence, the application of the multiplier
> > #MVPN_PE in this row/column is also not correct.
> >
>=20
> (it is correct: see the reply to Eric Rosen
> "draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / PIM LAN
> procedures")
>=20
>=20
> > This "problem" of treating not equal actions as equal repeats itself
> > throughout the scaling analysis in the appendix. Such analysis is
not
> > only *rough* but misleading, and because of this may lead to wrong
> > conclusions. I believe that unless the appendix A is corrected by
> > removing all not comparable, or even wrong estimations, it should be
> > removed from the draft.
>=20
> ...but you only find this suggestion worth enough bring brought
forward
> only six month or so after the WG LC... ;-)

The analysis kept changing, so I waited until it is finalized (at least
by the author). I believe that it is more important to make the draft as
accurate as possible than worry about the timing. I don't think anybody
would want that important scaling conclusions in an IETF document be
based on *rough approximations*.

>=20
> I think that the proposed modification to the introduction of Appendix
> A
> made by Yakov is much more reasonable (copy-pasted below for easier
> follow up) :
>=20
>     Though the overhead of message processing and the state overhead
>     may vary between different approaches, and even within the same
>     approach between different equipments (PEs or RRs), we propose
>     here a rough estimation of the load of PEs, in terms of number
>     of messages processed and number of control plane states
> maintained.
>=20
See my comment above on this.
=20
> -Thomas

From yakov@juniper.net  Wed Dec 16 11:27:01 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id C2AC93A6842 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 11:27:01 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.239
X-Spam-Level: 
X-Spam-Status: No, score=-6.239 tagged_above=-999 required=5 tests=[AWL=-0.240, BAYES_00=-2.599, J_CHICKENPOX_41=0.6, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id pT5SpUoornkk for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 11:27:00 -0800 (PST)
Received: from exprod7og101.obsmtp.com (exprod7og101.obsmtp.com [64.18.2.155]) by core3.amsl.com (Postfix) with SMTP id B84323A69FC for <l3vpn@ietf.org>; Wed, 16 Dec 2009 11:26:58 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob101.postini.com ([64.18.6.12]) with SMTP ID DSNKSyk0dA89P7545ViTzooNeRAVltzrorbY@postini.com; Wed, 16 Dec 2009 11:26:46 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Wed, 16 Dec 2009 11:23:11 -0800
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Wed, 16 Dec 2009 11:23:10 -0800
Received: from emailsmtp55.jnpr.net ([172.24.18.132]) by p-emlb01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Wed, 16 Dec 2009 11:23:10 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp55.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Wed, 16 Dec 2009 11:23:10 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBGJN9j30307; Wed, 16 Dec 2009 11:23:10 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912161923.nBGJN9j30307@magenta.juniper.net>
To: Thomas Morin <thomas.morin@orange-ftgroup.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / PIM LAN procedures 
In-Reply-To: <4B289EA9.2090500@orange-ftgroup.com> 
References: <3126.1259858356@erosen-linux> <4B289EA9.2090500@orange-ftgroup.com>
X-MH-In-Reply-To: Thomas Morin <thomas.morin@orange-ftgroup.com> message dated "Wed, 16 Dec 2009 09:47:37 +0100."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <78167.1260991389.1@juniper.net>
Date: Wed, 16 Dec 2009 11:23:09 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 16 Dec 2009 19:23:10.0273 (UTC) FILETIME=[34108B10:01CA7E85]
Cc: Eric Rosen <erosen@cisco.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 19:27:01 -0000

Thomas,

> Eric, working group,
> 
> Eric Rosen:
>  >
>  > From A.1.1.1.  "PIM LAN procedures, by default"
>  >
>  >    +------------+------------+---------------+----------+--------------+
>  >    |            | upstream   | other PEs     | RR       | total across |
>  >    |            | PE (1)     | (total across | (none)   | all          |
>  >    |            |            | (#mvpn_PE-1)  |          | equipments   |
>  >    |            |            | PEs)          |          |              |
>  >    +------------+------------+---------------+----------+--------------+
>  >    | first PE   | 1 m.e      | #MVPN_PE-1    | /        | #MVPN_PE m.e |
>  >    | joins      |            | m.e           |          |              |
>  >    +------------+------------+---------------+----------+--------------+
>  >[...]
> 
> In the sequence above the first PE joining results in all PEs processing 
> a message (except the PE joining):
>   - the upstream PE processes the Join and ends up propagating the state 
> upstream and setting up forwarding state
>   - the other PEs parse the message, and lookup their TIB to possibly 
> find an (S,G) state; they find none and discard the information in this 
> Join) ; note that they don't have any (S,G) state at this point
> 
>  >[...]
>  >
>  >    +------------+------------+---------------+----------+--------------+
>  >    | for *each* | 1 m.e      | #mvpn_PE-1    | /        | #mvpn_PE m.e |
>  >    | additional |            | m.e           |          |              |
>  >    | PE joining |            |               |          |              |
>  >    +------------+------------+---------------+----------+--------------+
>  >
> 
> For each subsequent PE joining (second line in the table, quoted above):
>    - the joining PE is transitioning its Upstream (S,G) state machine 
> from NotJoined to Joined, and thus sends a Join (RFC 4601, section 
> 4.5.7, p.73)
>    - for other PEs that are already joined to (S,G) : parse the message, 
> and lookup their TIB to find an (S,G) state, find one, and reset their 
> timer to further delay the moment at which they would send their Join
>    - for the other PEs that are not joined to (S,G) : parse the message, 
> and lookup their TIB to try find an (S,G) state, find none, and discard 
> the information in this Join)
> 
> 
> [...]
>  > If a second PE has "processed" the Join(S,G) from the first, then it 
> doesn't
>  > need to send a Join if it decides to join (S,G) itself;
> 
> As PIM procedures are specified (RFC 4601, section 4.5.7, p.73), even if 
> the second PE had initially processed a message (parsed, done a lookup, 
> discarded) from the first joining PE, this PE will still send a Join 
> when it starts to need to send (S,G) traffic towards a CE.
  
Just to add, from 4.5.7 of rfc4601:

   If a router wishes to propagate a Join(S,G) upstream, it must also
   watch for messages on its upstream interface from other routers on
   that subnet, and these may modify its behavior.  If it sees a
   Join(S,G) to the correct upstream neighbor, it should suppress its
   own Join(S,G).

which means that in order to suppress its own Join, the router
should be in the state where it "wishes to propagate a Join (S,G)
upstream". Clearly, that until the second PE receives Join
from downstream, the second PE is *not* in the state where it
"wishes" to propagate a Join upstream, and thus will *not*
suppress its first Join.

Yakov.

From yakov@juniper.net  Wed Dec 16 11:41:04 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id DDADB3A6800 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 11:41:04 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.524
X-Spam-Level: 
X-Spam-Status: No, score=-6.524 tagged_above=-999 required=5 tests=[AWL=0.075,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 5ND3VDs8efvZ for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 11:41:04 -0800 (PST)
Received: from exprod7og125.obsmtp.com (exprod7og125.obsmtp.com [64.18.2.28]) by core3.amsl.com (Postfix) with SMTP id 366C83A6822 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 11:41:02 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob125.postini.com ([64.18.6.12]) with SMTP ID DSNKSyk3wKS8e0+AfxMAjR/HWrH/7z93HIXA@postini.com; Wed, 16 Dec 2009 11:40:50 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Wed, 16 Dec 2009 11:35:41 -0800
Received: from p-emlb02-sac.jnpr.net ([66.129.254.47]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Wed, 16 Dec 2009 11:35:40 -0800
Received: from emailsmtp55.jnpr.net ([172.24.18.132]) by p-emlb02-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Wed, 16 Dec 2009 11:35:40 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp55.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Wed, 16 Dec 2009 11:35:40 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBGJZej37195; Wed, 16 Dec 2009 11:35:40 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912161935.nBGJZej37195@magenta.juniper.net>
To: Thomas Morin <thomas.morin@orange-ftgroup.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / C-multicast routing dynamics 
In-Reply-To: <4B28FAA9.1090304@orange-ftgroup.com> 
References: <3126.1259858356@erosen-linux> <4B28FAA9.1090304@orange-ftgroup.com>
X-MH-In-Reply-To: Thomas Morin <thomas.morin@orange-ftgroup.com> message dated "Wed, 16 Dec 2009 16:20:09 +0100."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <79639.1260992139.1@juniper.net>
Date: Wed, 16 Dec 2009 11:35:39 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 16 Dec 2009 19:35:40.0508 (UTC) FILETIME=[F33D51C0:01CA7E86]
Cc: Eric Rosen <erosen@cisco.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 19:41:05 -0000

Thomas,

> Eric, working group,
> 
> Let me first note that these comments quoted below are not specific to 
> new content added in releases -04/-05 and should have been done during 
> the LC in June.
> 
> Eric Rosen:
> [...]
> > Also omitted from consideration are scaling issues that show up
> > as the rate of change of multicast states increases.
> 
> No, section A.2 describes the cost of one individual PE joining and 
> leaving, which provide the required insight to compare the behavior of 
> the different C-multicast routing protocols for different rates of 
> change of multicast subscription by PEs.

On the topic of scaling "as the rate of change of multicast states
increases" that Eric brough in his e-mail, let's keep in mind that
the rate of change of multicast state is of practical significance
only when it starts to impose a non-neglidgeable load on the control
plane. You covered this in section 3.3.5 already. So, Eric's
claim that this is "omitted from consideration" is incorrect.

Yakov.

From mn1921@att.com  Wed Dec 16 11:52:34 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id B66B53A68BC for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 11:52:34 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -106.293
X-Spam-Level: 
X-Spam-Status: No, score=-106.293 tagged_above=-999 required=5 tests=[AWL=0.307, BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id cQ0SgfgrTm+q for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 11:52:34 -0800 (PST)
Received: from mail161.messagelabs.com (mail161.messagelabs.com [216.82.253.115]) by core3.amsl.com (Postfix) with ESMTP id DA0773A6894 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 11:52:33 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-14.tower-161.messagelabs.com!1260993138!19822041!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 4223 invoked from network); 16 Dec 2009 19:52:19 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-14.tower-161.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 16 Dec 2009 19:52:19 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBGJqEu6007344 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 14:52:14 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBGJqAqN007285 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 14:52:10 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / RTConstraint recommendation
Date: Wed, 16 Dec 2009 14:52:12 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A200451C91C@misout7msgusr7e.ugd.att.com>
In-Reply-To: <4B289CDB.8090404@orange-ftgroup.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / RTConstraint recommendation
Thread-Index: Acp+K2AWGqAIks0nReqFRbIHqe87xQAXX1Fg
References: <3126.1259858356@erosen-linux> <4B289CDB.8090404@orange-ftgroup.com>
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: "Thomas Morin" <thomas.morin@orange-ftgroup.com>, "L3VPN" <l3vpn@ietf.org>, "Eric Rosen" <erosen@cisco.com>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 19:52:34 -0000
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From mn1921@att.com  Wed Dec 16 13:16:59 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id A6C0E3A6A98 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 13:16:59 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -106.344
X-Spam-Level: 
X-Spam-Status: No, score=-106.344 tagged_above=-999 required=5 tests=[AWL=0.255, BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id gtYgOaQWV1-V for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 13:16:58 -0800 (PST)
Received: from mail167.messagelabs.com (mail167.messagelabs.com [216.82.253.179]) by core3.amsl.com (Postfix) with ESMTP id ADA483A6A97 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 13:16:58 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-12.tower-167.messagelabs.com!1260998203!26877233!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 10615 invoked from network); 16 Dec 2009 21:16:43 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-12.tower-167.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 16 Dec 2009 21:16:43 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBGLGdi4032215 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 16:16:39 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBGLGYux032151 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 16:16:34 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / sectionA.1.1.3
Date: Wed, 16 Dec 2009 16:16:36 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com>
In-Reply-To: <4B28E84F.4000604@orange-ftgroup.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / sectionA.1.1.3
Thread-Index: Acp+WF6X19mBAQCxSGuqq7bOpMWjTQANavlw
References: <3126.1259858356@erosen-linux> <4B28E84F.4000604@orange-ftgroup.com>
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: "Thomas Morin" <thomas.morin@orange-ftgroup.com>, "L3VPN" <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 21:16:59 -0000

VGhvbWFzLA0KDQpyZWdhcmRpbmcgbXkgcXVlc3Rpb246ICJJbiBzZWN0aW9uIEEuMS4xLjMsIHdo
eSB0aGUgJ2VhY2ggYWRkaXRpb25hbCBQRSBqb2luaW5nJyByb3cgZm9yIHRoZSAndXBzdHJlYW0g
UEUnIGNvbHVtbiBzaG93cyAiMCI/IiwgeW91IHBvaW50ZWQgbWUgdG8gdGhpcyBlLW1haWwuDQpQ
bGVhc2Ugc2VlIG15IHJlcGxpZXMgaW4tbGluZS4NCg0KTWFyaWENCg0KDQo+IE5vdywgdG8gZ28g
YmFjayBvbiB0aGUgY29tcHV0YXRpb24gaW4gIHNlY3Rpb24gQS4xLjEuMyBvbiBCR1AsIHlvdSBh
cmUNCj4gY29ycmVjdCB0aGF0IHRoZSBjdXJyZW50IHRleHQgZG9lcyBub3QgeWV0IHByb3Blcmx5
IHJlZmxlY3RzIHRoZSBjYXNlcw0KPiB3aGVyZSBhbiBSUiBtYXkgcmUtYWR2ZXJ0aXNlcyBhIHJv
dXRlIHRoZXkgd2VyZSBhbHJlYWR5IGFkdmVydGlzaW5nLg0KPiBMZXQncyBsb29rIHByYWN0aWNh
bGx5IGF0IHRoZSBhdHRyaWJ1dGVzIHRoYXQgY291bGQgdmFyeSA6DQo+IA0KPiAgICAtIG5leHQt
aG9wIDogdXBkYXRlcyBvZiB0aGUgdmFsdWUgb2YgdGhpcyBhdHRyaWJ1dGUgZG8gbm90ICJmb2xs
b3ciDQo+IFBFIGFkdmVydGlzZW1lbnRzLCBiZWluZyByZXdyaXR0ZW4gYnkgdGhlIFJSIHdoZW4g
YWdncmVnYXRpbmcNCj4gQy1tdWx0aWNhc3Qgcm91dGVzIGhhdmluZyBzYW1lIE5MUkkgKHNlY3Rp
b24gMTEuNCBvZg0KPiBkcmFmdC1pZXRmLWwzdnBuLTI1NDdiaXMtbWNhc3QtYmdwKSA7IHRoZSBh
ZHZlcnRpc2VtZW50cyBtYWRlIGJ5DQo+IGRpZmZlcmVudCBSUnMgbWF5IGxlYWQgdG8gdXBkYXRl
cyBvZiB0aGUgdmFsdWUgb2YgdGhpcyBhdHRyaWJ1dGUNCg0KVGhpcyBkb2VzIG5vdCBoZWxwIGlu
IHJvdXRlIGFnZ3JlZ2F0aW9uIGJlY2F1c2UgUlIgYXBwbGllcyB0aGUgT1JJR0lOQVRPUl9JRCB0
byBlYWNoIGNsaWVudCdzL2VncmVzcyBQRSByb3V0ZS4NCg0KPiAgICAtIE9SSUdJTkFUT1JfSUQg
OiB1cGRhdGVzIG9mIHRoZSB2YWx1ZSBvZiB0aGlzIGF0dHJpYnV0ZXMgZG8gbm90DQo+ICJmb2xs
b3ciIFBFIGFkdmVydGlzZW1lbnRzLCBhcyB0aGUgUlIgcmV3cml0ZSBpdCB3aGVuIGFnZ3JlZ2F0
aW5nIEMtDQo+IG11bHRpY2FzdCByb3V0ZXMgaGF2aW5nIHNhbWUgTkxSSTsgdGhlIGFkdmVydGlz
ZW1lbnRzIG1hZGUgYnkgZGlmZmVyZW50DQo+IFJScyBtYXkgbGVhZCBwb3NzaWJseSB0byB1cGRh
dGVzIG9mIHRoZSB2YWx1ZSBvZiB0aGlzIGF0dHJpYnV0ZSwgZ2l2ZW4NCj4gdGhhdCBlYWNoIFJS
IHdvdWxkIHNldCBPUklHSU5BVE9SX0lEIHRvIGEgdmFsdWUgb2YgaXRzIG93bg0KPiANCg0KUmVn
YXJkaW5nIHJvdXRlICphZ2dyZWdhdGlvbiogdGhlIGRyYWZ0LWlldGYtbDN2cG4tMjU0N2Jpcy1t
Y2FzdC1iZ3AtMDggb25seSBzdGF0ZXMgdGhlIGZvbGxvd2luZzoNCg0KICAgIlRvIGZ1cnRoZXIg
cmVkdWNlIHRoZSByb3V0aW5nIGNodXJuIGR1ZSB0byBDLW11bHRpY2FzdCByb3V0ZXMgY2hhbmdl
cw0KICAgYSBSb3V0ZSBSZWZsZWN0b3IgdGhhdCByZS1hZHZlcnRpc2VzIGEgQy1tdWx0aWNhc3Qg
cm91dGUgU0hPVUxEIHNldA0KICAgdGhlIE5leHQgSG9wIGZpZWxkIG9mIHRoZSBNUF9SRUFDSF9O
TFJJIGF0dHJpYnV0ZSBvZiB0aGUgcm91dGUgdG8gYW4NCiAgIElQIGFkZHJlc3Mgb2YgdGhlIFJv
dXRlIFJlZmxlY3Rvci4gTGlrZXdpc2UsIGFuIEFTQlIgdGhhdCByZS0NCiAgIGFkdmVydGlzZXMg
YSBDLW11bHRpY2FzdCByb3V0ZSBTSE9VTEQgc2V0IHRoZSBOZXh0IEhvcCBmaWVsZCBvZiB0aGUN
CiAgIE1QX1JFQUNIX05MUkkgYXR0cmlidXRlIG9mIHRoZSByb3V0ZSB0byBhbiBJUCBhZGRyZXNz
IG9mIHRoZSBBU0JSLiINCg0KSXQgc3RhdGVzIG5vdGhpbmcgYWJvdXQgc2V0dGluZyBPUklHSU5B
VE9SX0lEIHRvIHNlbGYgb24gUlIuIA0KDQpNYXJpYQ0K

From raszuk@cisco.com  Wed Dec 16 15:28:45 2009
Return-Path: <raszuk@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 57A2F3A6AA5 for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 15:28:45 -0800 (PST)
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 ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id OjKCKrMCMdUZ for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 15:28:44 -0800 (PST)
Received: from sj-iport-4.cisco.com (sj-iport-4.cisco.com [171.68.10.86]) by core3.amsl.com (Postfix) with ESMTP id 7C6283A68C8 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 15:28:44 -0800 (PST)
Authentication-Results: sj-iport-4.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AikHAJ77KEurRN+J/2dsb2JhbACEC5R2pR6BHAgBhVuQB4EvgipSBA
X-IronPort-AV: E=Sophos;i="4.47,409,1257120000"; d="scan'208";a="63824789"
Received: from sj-core-3.cisco.com ([171.68.223.137]) by sj-iport-4.cisco.com with ESMTP; 16 Dec 2009 23:28:30 +0000
Received: from [192.168.1.57] (sjc-raszuk-87113.cisco.com [10.20.147.254]) by sj-core-3.cisco.com (8.13.8/8.14.3) with ESMTP id nBGNSTgk015353; Wed, 16 Dec 2009 23:28:29 GMT
Message-ID: <4B296D1C.6000001@cisco.com>
Date: Thu, 17 Dec 2009 00:28:28 +0100
From: Robert Raszuk <raszuk@cisco.com>
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /	sectionA.1.1.3
References: <3126.1259858356@erosen-linux>	<4B28E84F.4000604@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com>
In-Reply-To: <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: raszuk@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 16 Dec 2009 23:28:45 -0000

Hi Maria,

Excellent point. Looking back at the archive of this WG I have been 
pointing out this issue in the past, but it was always never addressed.

It seems that for this mvpn solution to work classic "bgp route 
reflectors" would not do the right job.

They should be replaced with "multicast control aggregation" devices 
which would have to support it's own set of requirements and it's own 
advertisement rules. In fact draft already recommends to use such 
devices as a standalone control plane boxes.

Therefor before proceeding in this WG I would like to either see all 
references to BGP route reflectors use removed. Alternatively they could 
be replaced with the new term describing their newly defined and 
apparently required functionality for any form of sensible operation of 
the proposed herein solution.

Cheers,
R.


> Thomas,
> 
> regarding my question: "In section A.1.1.3, why the 'each additional PE joining' row for the 'upstream PE' column shows "0"?", you pointed me to this e-mail.
> Please see my replies in-line.
> 
> Maria
> 
> 
>> Now, to go back on the computation in  section A.1.1.3 on BGP, you are
>> correct that the current text does not yet properly reflects the cases
>> where an RR may re-advertises a route they were already advertising.
>> Let's look practically at the attributes that could vary :
>>
>>    - next-hop : updates of the value of this attribute do not "follow"
>> PE advertisements, being rewritten by the RR when aggregating
>> C-multicast routes having same NLRI (section 11.4 of
>> draft-ietf-l3vpn-2547bis-mcast-bgp) ; the advertisements made by
>> different RRs may lead to updates of the value of this attribute
> 
> This does not help in route aggregation because RR applies the ORIGINATOR_ID to each client's/egress PE route.
> 
>>    - ORIGINATOR_ID : updates of the value of this attributes do not
>> "follow" PE advertisements, as the RR rewrite it when aggregating C-
>> multicast routes having same NLRI; the advertisements made by different
>> RRs may lead possibly to updates of the value of this attribute, given
>> that each RR would set ORIGINATOR_ID to a value of its own
>>
> 
> Regarding route *aggregation* the draft-ietf-l3vpn-2547bis-mcast-bgp-08 only states the following:
> 
>    "To further reduce the routing churn due to C-multicast routes changes
>    a Route Reflector that re-advertises a C-multicast route SHOULD set
>    the Next Hop field of the MP_REACH_NLRI attribute of the route to an
>    IP address of the Route Reflector. Likewise, an ASBR that re-
>    advertises a C-multicast route SHOULD set the Next Hop field of the
>    MP_REACH_NLRI attribute of the route to an IP address of the ASBR."
> 
> It states nothing about setting ORIGINATOR_ID to self on RR. 
> 
> Maria


From mn1921@att.com  Wed Dec 16 16:39:29 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 122113A6ABF for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 16:39:29 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -106.08
X-Spam-Level: 
X-Spam-Status: No, score=-106.08 tagged_above=-999 required=5 tests=[AWL=-0.081, BAYES_00=-2.599, J_CHICKENPOX_41=0.6, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id FnUi3pm1nlba for <l3vpn@core3.amsl.com>; Wed, 16 Dec 2009 16:39:27 -0800 (PST)
Received: from mail121.messagelabs.com (mail121.messagelabs.com [216.82.242.3]) by core3.amsl.com (Postfix) with ESMTP id 5CA253A6AB8 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 16:39:27 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-9.tower-121.messagelabs.com!1261010351!29782588!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 1891 invoked from network); 17 Dec 2009 00:39:12 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-9.tower-121.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 17 Dec 2009 00:39:12 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBH0d8lv015773 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 19:39:08 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBH0d00N015757 for <l3vpn@ietf.org>; Wed, 16 Dec 2009 19:39:00 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / PIM LANprocedures
Date: Wed, 16 Dec 2009 19:39:02 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A200457F855@misout7msgusr7e.ugd.att.com>
In-Reply-To: <4B289EA9.2090500@orange-ftgroup.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / PIM LANprocedures
Thread-Index: Acp+LHTyPrGgXiRMQp6n/wExTN/ooAAhCiww
References: <3126.1259858356@erosen-linux> <4B289EA9.2090500@orange-ftgroup.com>
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: "Thomas Morin" <thomas.morin@orange-ftgroup.com>, "L3VPN" <l3vpn@ietf.org>, "Eric Rosen" <erosen@cisco.com>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 17 Dec 2009 00:39:29 -0000

VGhvbWFzLA0KDQpMZXQgbWUgcmVzdGF0ZSBzb21lIG9mIHRoZSBjb21tZW50cyBmcm9tIG15IG9y
aWdpbmFsIGUtbWFpbCBzaW5jZSB5b3Ugc2FpZCB0aGF0IHNvbWUgb2YgdGhlbSB5b3UgYWRkcmVz
c2VkIGluIHRoZSByZXBsaWVzIHRvIEVyaWMuIFRoaXMgaXMgKGZyYWdtZW50cykgb2Ygb25lIG9m
IHRob3NlIGUtbWFpbHMuDQoNCk1hcmlhDQoNCg0KPiAgPg0KPiAgPiAgICArLS0tLS0tLS0tLS0t
Ky0tLS0tLS0tLS0tLSstLS0tLS0tLS0tLS0tLS0rLS0tLS0tLS0tLSstLS0tLS0tLS0tLQ0KPiAt
LS0rDQo+ICA+ICAgIHwgICAgICAgICAgICB8IHVwc3RyZWFtICAgfCBvdGhlciBQRXMgICAgIHwg
UlIgICAgICAgfCB0b3RhbA0KPiBhY3Jvc3MgfA0KPiAgPiAgICB8ICAgICAgICAgICAgfCBQRSAo
MSkgICAgIHwgKHRvdGFsIGFjcm9zcyB8IChub25lKSAgIHwgYWxsDQo+IHwNCj4gID4gICAgfCAg
ICAgICAgICAgIHwgICAgICAgICAgICB8ICgjbXZwbl9QRS0xKSAgfCAgICAgICAgICB8IGVxdWlw
bWVudHMNCj4gfA0KPiAgPiAgICB8ICAgICAgICAgICAgfCAgICAgICAgICAgIHwgUEVzKSAgICAg
ICAgICB8ICAgICAgICAgIHwNCj4gfA0KPiAgPiAgICArLS0tLS0tLS0tLS0tKy0tLS0tLS0tLS0t
LSstLS0tLS0tLS0tLS0tLS0rLS0tLS0tLS0tLSstLS0tLS0tLS0tLQ0KPiAtLS0rDQo+ICA+ICAg
IHwgZmlyc3QgUEUgICB8IDEgbS5lICAgICAgfCAjTVZQTl9QRS0xICAgIHwgLyAgICAgICAgfCAj
TVZQTl9QRQ0KPiBtLmUgfA0KPiAgPiAgICB8IGpvaW5zICAgICAgfCAgICAgICAgICAgIHwgbS5l
ICAgICAgICAgICB8ICAgICAgICAgIHwNCj4gfA0KPiAgPiAgICArLS0tLS0tLS0tLS0tKy0tLS0t
LS0tLS0tLSstLS0tLS0tLS0tLS0tLS0rLS0tLS0tLS0tLSstLS0tLS0tLS0tLQ0KPiAtLS0rDQo+
ICA+Wy4uLl0NCj4gDQo+IEluIHRoZSBzZXF1ZW5jZSBhYm92ZSB0aGUgZmlyc3QgUEUgam9pbmlu
ZyByZXN1bHRzIGluIGFsbCBQRXMNCj4gcHJvY2Vzc2luZw0KPiBhIG1lc3NhZ2UgKGV4Y2VwdCB0
aGUgUEUgam9pbmluZyk6DQo+ICAgLSB0aGUgdXBzdHJlYW0gUEUgcHJvY2Vzc2VzIHRoZSBKb2lu
IGFuZCBlbmRzIHVwIHByb3BhZ2F0aW5nIHRoZQ0KPiAgICAgc3RhdGUgdXBzdHJlYW0gYW5kIHNl
dHRpbmcgdXAgZm9yd2FyZGluZyBzdGF0ZQ0KPiAgIC0gdGhlIG90aGVyIFBFcyBwYXJzZSB0aGUg
bWVzc2FnZSwgYW5kIGxvb2t1cCB0aGVpciBUSUIgdG8gcG9zc2libHkNCj4gZmluZCBhbiAoUyxH
KSBzdGF0ZTsgdGhleSBmaW5kIG5vbmUgYW5kIGRpc2NhcmQgdGhlIGluZm9ybWF0aW9uIGluIHRo
aXMNCj4gSm9pbikgOyBub3RlIHRoYXQgdGhleSBkb24ndCBoYXZlIGFueSAoUyxHKSBzdGF0ZSBh
dCB0aGlzIHBvaW50DQoNClRoZSBwcm9jZXNzaW5nIGRlc2NyaWJlZCB1bmRlciB0aGUgZmlyc3Qg
YnVsbGV0IGFib3ZlIGFsc28gbWVhbnMgc2V0dGluZyB1cCAocyxnKSBzdGF0ZSBpbiB0aGUgY29u
dHJvbCBwbGFuZS4gIEkgYXNzdW1lIHRoaXMgaXMgd2hhdCB5b3UgbWVhbiBieSAicHJvY2Vzc2lu
ZyIgdnMuICJwYXJzaW5nIiB1bmRlciB0aGUgc2Vjb25kIGJ1bGxldC4gKEJUVywgaWYgdGhpcyBu
ZWVkcyB0byBiZSBhbmFseXplZCBkb3duIHRvIHRoZSBsZXZlbCBvZiAqcGFyc2luZyogb2YgdGhl
IG1lc3NhZ2VzLCBsZXQncyBiZSBpdC4uKS4NCkNsZWFybHksIHdvcmtsb2FkIHVuZGVyIHRoZSBm
aXJzdCBidWxsZXQgaXMgbm90IHRoZSBzYW1lIGFtb3VudCBvZiB3b3JrIGFzIHRoYXQgZGVzY3Jp
YmVkIHVuZGVyIHRoZSBzZWNvbmQgYnVsbGV0LiBUaGUgc2Vjb25kIGJ1bGxldCBpcyBqdXN0ICpk
aXNjYXJkaW5nKiB0aGUgbWFzc2FnZSBiYXNlZCBvbiB0d28gdGhpbmdzOiAiSSBhbSBub3QgdGhl
IHVwc3RyZWFtIHJvdXRlciIgYW5kICJJIGRvbid0IGhhdmUgKFMsRykgc3RhdGUiLiBJdCBjYW5u
b3QgYmUgY2xhaW1lZCB0aGF0IHRoaXMgaXMgZXF1YWwgd2l0aCB3aGF0IHRoZSB1cHN0cmVhbSBQ
RSBoYXMgdG8gZG8uDQoNCkxldCdzIGNvbXBhcmUgaXQgd2l0aCB0aGUgZmlyc3Qgcm93ICJmaXJz
dCBQRSBqb2lucyIgb2YgdGFibGUgaW4gc2VjdGlvbiBBLjEuMS4zLiBpLmUuLCB3aXRoIHRoZSB3
b3JrbG9hZCBvbiB0aGUgdXBzdHJlYW0gUEUgd2hlbiBCR1Agc2lnbmFsaW5nIGlzIHVzZWQuIEFs
bCB0aGUgd29yayB0aGF0IHVwc3RyZWFtIFBFIG5lZWRzIHRvIGRvIGluIFBJTSBzaWduYWxpbmcg
Y2FzZSBhbHNvIG5lZWRzIHRvIGJlIGRvbmUgaW4gQkdQIHNpZ25hbGluZyBjYXNlLiBCdXQgd2l0
aCBCR1Agc2lnbmFsaW5nLCB1cHN0cmVhbSBQRSBoYXMgYWxzbyB0byBwZXJmb3JtIHRoZSBCR1Ag
cHJvY2Vzc2luZyAoYW5kICpwYXJzaW5nKiksIGluY2x1ZGluZyB0aGUgYmVzdCBwYXRoIHNlbGVj
dGlvbi4gSSBkb24ndCBzZWUgdGhpcyBiZWluZyBhY2NvdW50ZWQgZm9yIGF0IGFsbCBvbiBQRXMu
DQpJZiB5b3Ugd29ycnkgc28gbXVjaCBhYm91dCB0aGUgd29yayBhIFBFIGhhcyB0byBkbyB0byAq
ZGlzY2FyZCogYSBtZXNzYWdlLCB3aHkgd291bGQgeW91IGlnbm9yZSBhIHNpZ25pZmljYW50IGFt
b3VudCBvZiBCR1AgcmVsYXRlZCB3b3JrIG9uIGEgUEU/DQoNCg0KDQo+ICA+ICAgICstLS0tLS0t
LS0tLS0rLS0tLS0tLS0tLS0tKy0tLS0tLS0tLS0tLS0tLSstLS0tLS0tLS0tKy0tLS0tLS0tLS0t
DQo+IC0tLSsNCj4gID4gICAgfCBmb3IgKmVhY2gqIHwgMSBtLmUgICAgICB8ICNtdnBuX1BFLTEg
ICAgfCAvICAgICAgICB8ICNtdnBuX1BFDQo+IG0uZSB8DQo+ICA+ICAgIHwgYWRkaXRpb25hbCB8
ICAgICAgICAgICAgfCBtLmUgICAgICAgICAgIHwgICAgICAgICAgfA0KPiB8DQo+ICA+ICAgIHwg
UEUgam9pbmluZyB8ICAgICAgICAgICAgfCAgICAgICAgICAgICAgIHwgICAgICAgICAgfA0KPiB8
DQo+ICA+ICAgICstLS0tLS0tLS0tLS0rLS0tLS0tLS0tLS0tKy0tLS0tLS0tLS0tLS0tLSstLS0t
LS0tLS0tKy0tLS0tLS0tLS0tDQo+IC0tLSsNCj4gID4NCj4gDQo+IEZvciBlYWNoIHN1YnNlcXVl
bnQgUEUgam9pbmluZyAoc2Vjb25kIGxpbmUgaW4gdGhlIHRhYmxlLCBxdW90ZWQNCj4gYWJvdmUp
Og0KPiAgICAtIHRoZSBqb2luaW5nIFBFIGlzIHRyYW5zaXRpb25pbmcgaXRzIFVwc3RyZWFtIChT
LEcpIHN0YXRlIG1hY2hpbmUNCj4gZnJvbSBOb3RKb2luZWQgdG8gSm9pbmVkLCBhbmQgdGh1cyBz
ZW5kcyBhIEpvaW4gKFJGQyA0NjAxLCBzZWN0aW9uDQo+IDQuNS43LCBwLjczKQ0KPiAgICAtIGZv
ciBvdGhlciBQRXMgdGhhdCBhcmUgYWxyZWFkeSBqb2luZWQgdG8gKFMsRykgOiBwYXJzZSB0aGUN
Cj4gbWVzc2FnZSwNCj4gYW5kIGxvb2t1cCB0aGVpciBUSUIgdG8gZmluZCBhbiAoUyxHKSBzdGF0
ZSwgZmluZCBvbmUsIGFuZCByZXNldCB0aGVpcg0KPiB0aW1lciB0byBmdXJ0aGVyIGRlbGF5IHRo
ZSBtb21lbnQgYXQgd2hpY2ggdGhleSB3b3VsZCBzZW5kIHRoZWlyIEpvaW4NCj4gICAgLSBmb3Ig
dGhlIG90aGVyIFBFcyB0aGF0IGFyZSBub3Qgam9pbmVkIHRvIChTLEcpIDogcGFyc2UgdGhlDQo+
IG1lc3NhZ2UsDQo+IGFuZCBsb29rdXAgdGhlaXIgVElCIHRvIHRyeSBmaW5kIGFuIChTLEcpIHN0
YXRlLCBmaW5kIG5vbmUsIGFuZCBkaXNjYXJkDQo+IHRoZSBpbmZvcm1hdGlvbiBpbiB0aGlzIEpv
aW4pDQoNClRoZSB3b3JrbG9hZCBvbiB0aGUgdXBzdHJlYW0gUEUgKGFuZCBvdGhlciBQRXMgdGhh
dCBhbHJlYWR5IGhhdmUgKFMsRykgc3RhdGUpIGluIHRoaXMgY2FzZSBpcyBzaWduaWZpY2FudGx5
IGxlc3MgdGhhbiBpbiB0aGUgY2FzZSBvZiAidGhlIGZpcnN0IFBFIGpvaW5pbmciIHNpbmNlIHRo
ZSBQRXMgYWxyZWFkeSBoYXZlIHRoZSAoUyxHKSBzdGF0ZS4gQnV0IGl0IGlzIGVzdGltYXRlZCBh
cyB0aGUgc2FtZS4gT24gdGhvc2UgUEVzIHRoYXQgZG8gbm90IGhhdmUgKFMsRykgdGhlcmUgaXMg
ZXZlbiBsZXNzIHdvcmsuIFlldCwgYWxsIGlzIGVzdGltYXRlZCBhcyB0aGUgc2FtZSBhbW91bnQg
b2Ygd29yay4gDQoNCg0KPiAtLS0tLU9yaWdpbmFsIE1lc3NhZ2UtLS0tLQ0KPiBGcm9tOiBsM3Zw
bi1ib3VuY2VzQGlldGYub3JnIFttYWlsdG86bDN2cG4tYm91bmNlc0BpZXRmLm9yZ10gT24gQmVo
YWxmDQo+IE9mIFRob21hcyBNb3Jpbg0KPiBTZW50OiBXZWRuZXNkYXksIERlY2VtYmVyIDE2LCAy
MDA5IDM6NDggQU0NCj4gVG86IEwzVlBOOyBFcmljIFJvc2VuDQo+IFN1YmplY3Q6IFJlOiBkcmFm
dC1pZXRmLWwzdnBuLW12cG4tY29uc2lkZXJhdGlvbnMtMDUgLyBBcHBlbmRpeCBBIC8gUElNDQo+
IExBTnByb2NlZHVyZXMNCj4gDQo+IEVyaWMsIHdvcmtpbmcgZ3JvdXAsDQo+IA0KPiBFcmljIFJv
c2VuOg0KPiAgPg0KPiAgPiBGcm9tIEEuMS4xLjEuICAiUElNIExBTiBwcm9jZWR1cmVzLCBieSBk
ZWZhdWx0Ig0KPiAgPg0KPiAgPiAgICArLS0tLS0tLS0tLS0tKy0tLS0tLS0tLS0tLSstLS0tLS0t
LS0tLS0tLS0rLS0tLS0tLS0tLSstLS0tLS0tLS0tLQ0KPiAtLS0rDQo+ICA+ICAgIHwgICAgICAg
ICAgICB8IHVwc3RyZWFtICAgfCBvdGhlciBQRXMgICAgIHwgUlIgICAgICAgfCB0b3RhbA0KPiBh
Y3Jvc3MgfA0KPiAgPiAgICB8ICAgICAgICAgICAgfCBQRSAoMSkgICAgIHwgKHRvdGFsIGFjcm9z
cyB8IChub25lKSAgIHwgYWxsDQo+IHwNCj4gID4gICAgfCAgICAgICAgICAgIHwgICAgICAgICAg
ICB8ICgjbXZwbl9QRS0xKSAgfCAgICAgICAgICB8IGVxdWlwbWVudHMNCj4gfA0KPiAgPiAgICB8
ICAgICAgICAgICAgfCAgICAgICAgICAgIHwgUEVzKSAgICAgICAgICB8ICAgICAgICAgIHwNCj4g
fA0KPiAgPiAgICArLS0tLS0tLS0tLS0tKy0tLS0tLS0tLS0tLSstLS0tLS0tLS0tLS0tLS0rLS0t
LS0tLS0tLSstLS0tLS0tLS0tLQ0KPiAtLS0rDQo+ICA+ICAgIHwgZmlyc3QgUEUgICB8IDEgbS5l
ICAgICAgfCAjTVZQTl9QRS0xICAgIHwgLyAgICAgICAgfCAjTVZQTl9QRQ0KPiBtLmUgfA0KPiAg
PiAgICB8IGpvaW5zICAgICAgfCAgICAgICAgICAgIHwgbS5lICAgICAgICAgICB8ICAgICAgICAg
IHwNCj4gfA0KPiAgPiAgICArLS0tLS0tLS0tLS0tKy0tLS0tLS0tLS0tLSstLS0tLS0tLS0tLS0t
LS0rLS0tLS0tLS0tLSstLS0tLS0tLS0tLQ0KPiAtLS0rDQo+ICA+Wy4uLl0NCj4gDQo+IEluIHRo
ZSBzZXF1ZW5jZSBhYm92ZSB0aGUgZmlyc3QgUEUgam9pbmluZyByZXN1bHRzIGluIGFsbCBQRXMN
Cj4gcHJvY2Vzc2luZw0KPiBhIG1lc3NhZ2UgKGV4Y2VwdCB0aGUgUEUgam9pbmluZyk6DQo+ICAg
LSB0aGUgdXBzdHJlYW0gUEUgcHJvY2Vzc2VzIHRoZSBKb2luIGFuZCBlbmRzIHVwIHByb3BhZ2F0
aW5nIHRoZQ0KPiBzdGF0ZQ0KPiB1cHN0cmVhbSBhbmQgc2V0dGluZyB1cCBmb3J3YXJkaW5nIHN0
YXRlDQo+ICAgLSB0aGUgb3RoZXIgUEVzIHBhcnNlIHRoZSBtZXNzYWdlLCBhbmQgbG9va3VwIHRo
ZWlyIFRJQiB0byBwb3NzaWJseQ0KPiBmaW5kIGFuIChTLEcpIHN0YXRlOyB0aGV5IGZpbmQgbm9u
ZSBhbmQgZGlzY2FyZCB0aGUgaW5mb3JtYXRpb24gaW4gdGhpcw0KPiBKb2luKSA7IG5vdGUgdGhh
dCB0aGV5IGRvbid0IGhhdmUgYW55IChTLEcpIHN0YXRlIGF0IHRoaXMgcG9pbnQNCj4gDQo+ICA+
Wy4uLl0NCj4gID4NCj4gID4gICAgKy0tLS0tLS0tLS0tLSstLS0tLS0tLS0tLS0rLS0tLS0tLS0t
LS0tLS0tKy0tLS0tLS0tLS0rLS0tLS0tLS0tLS0NCj4gLS0tKw0KPiAgPiAgICB8IGZvciAqZWFj
aCogfCAxIG0uZSAgICAgIHwgI212cG5fUEUtMSAgICB8IC8gICAgICAgIHwgI212cG5fUEUNCj4g
bS5lIHwNCj4gID4gICAgfCBhZGRpdGlvbmFsIHwgICAgICAgICAgICB8IG0uZSAgICAgICAgICAg
fCAgICAgICAgICB8DQo+IHwNCj4gID4gICAgfCBQRSBqb2luaW5nIHwgICAgICAgICAgICB8ICAg
ICAgICAgICAgICAgfCAgICAgICAgICB8DQo+IHwNCj4gID4gICAgKy0tLS0tLS0tLS0tLSstLS0t
LS0tLS0tLS0rLS0tLS0tLS0tLS0tLS0tKy0tLS0tLS0tLS0rLS0tLS0tLS0tLS0NCj4gLS0tKw0K
PiAgPg0KPiANCj4gRm9yIGVhY2ggc3Vic2VxdWVudCBQRSBqb2luaW5nIChzZWNvbmQgbGluZSBp
biB0aGUgdGFibGUsIHF1b3RlZA0KPiBhYm92ZSk6DQo+ICAgIC0gdGhlIGpvaW5pbmcgUEUgaXMg
dHJhbnNpdGlvbmluZyBpdHMgVXBzdHJlYW0gKFMsRykgc3RhdGUgbWFjaGluZQ0KPiBmcm9tIE5v
dEpvaW5lZCB0byBKb2luZWQsIGFuZCB0aHVzIHNlbmRzIGEgSm9pbiAoUkZDIDQ2MDEsIHNlY3Rp
b24NCj4gNC41LjcsIHAuNzMpDQo+ICAgIC0gZm9yIG90aGVyIFBFcyB0aGF0IGFyZSBhbHJlYWR5
IGpvaW5lZCB0byAoUyxHKSA6IHBhcnNlIHRoZQ0KPiBtZXNzYWdlLA0KPiBhbmQgbG9va3VwIHRo
ZWlyIFRJQiB0byBmaW5kIGFuIChTLEcpIHN0YXRlLCBmaW5kIG9uZSwgYW5kIHJlc2V0IHRoZWly
DQo+IHRpbWVyIHRvIGZ1cnRoZXIgZGVsYXkgdGhlIG1vbWVudCBhdCB3aGljaCB0aGV5IHdvdWxk
IHNlbmQgdGhlaXIgSm9pbg0KPiAgICAtIGZvciB0aGUgb3RoZXIgUEVzIHRoYXQgYXJlIG5vdCBq
b2luZWQgdG8gKFMsRykgOiBwYXJzZSB0aGUNCj4gbWVzc2FnZSwNCj4gYW5kIGxvb2t1cCB0aGVp
ciBUSUIgdG8gdHJ5IGZpbmQgYW4gKFMsRykgc3RhdGUsIGZpbmQgbm9uZSwgYW5kIGRpc2NhcmQN
Cj4gdGhlIGluZm9ybWF0aW9uIGluIHRoaXMgSm9pbikNCj4gDQo+IA0KPiBbLi4uXQ0KPiAgPiBJ
ZiBhIHNlY29uZCBQRSBoYXMgInByb2Nlc3NlZCIgdGhlIEpvaW4oUyxHKSBmcm9tIHRoZSBmaXJz
dCwgdGhlbiBpdA0KPiBkb2Vzbid0DQo+ICA+IG5lZWQgdG8gc2VuZCBhIEpvaW4gaWYgaXQgZGVj
aWRlcyB0byBqb2luIChTLEcpIGl0c2VsZjsNCj4gDQo+IEFzIFBJTSBwcm9jZWR1cmVzIGFyZSBz
cGVjaWZpZWQgKFJGQyA0NjAxLCBzZWN0aW9uIDQuNS43LCBwLjczKSwgZXZlbg0KPiBpZg0KPiB0
aGUgc2Vjb25kIFBFIGhhZCBpbml0aWFsbHkgcHJvY2Vzc2VkIGEgbWVzc2FnZSAocGFyc2VkLCBk
b25lIGEgbG9va3VwLA0KPiBkaXNjYXJkZWQpIGZyb20gdGhlIGZpcnN0IGpvaW5pbmcgUEUsIHRo
aXMgUEUgd2lsbCBzdGlsbCBzZW5kIGEgSm9pbg0KPiB3aGVuIGl0IHN0YXJ0cyB0byBuZWVkIHRv
IHNlbmQgKFMsRykgdHJhZmZpYyB0b3dhcmRzIGEgQ0UuDQo+IA0KPiANCj4gDQo+ICA+DQo+ICA+
IElmIHRoZSBbZmlyc3RdIHJvdyBpbXBsaWVzIHRoYXQgdGhlICJvdGhlciBQRXMiIG11c3QgcHJv
Y2VzcyB0aGUNCj4gam9pbiwgd2hlcmUNCj4gID4gInByb2Nlc3MiIGlzIGRlZmluZWQgKGFzIGl0
IGlzIGluIHRoaXMgQXBwZW5kaXgpIGFzICJhIG1lc3NhZ2UgYmVpbmcNCj4gcGFyc2VkLA0KPiAg
PiBhIGxvb2t1cCBiZWluZyBkb25lLCBhbmQgc29tZSBhY3Rpb24gYmVpbmcgdGFrZW4gKHN1Y2gg
aGFzIHVwZGF0aW5nDQo+IGENCj4gID4gY29udHJvbCBwbGFuZSBvciBkYXRhIHBsYW5lIHN0YXRl
KSIsIHRoZW4gb25lIHRoZSBbc2Vjb25kXSByb3cgaXMNCj4gaW5jb3JyZWN0Og0KPiANCj4gTGV0
J3MgbG9vayBhdCB5b3VyIHJlYXNvbmluZywgYW5kIGhvdyBpdCBjb3VsZCBob2xkOg0KPiAtIGlm
IHlvdSB3YW50IHRvIGlnbm9yZSB0aGF0IHRoZSBkZWZpbml0aW9uIG9mICJwcm9jZXNzIiBpbmNs
dWRlcw0KPiAiZGlzY2FyZGluZyBhIG1lc3NhZ2UiIGFzICJhbiBhY3Rpb24gYmVpbmcgdGFrZW4i
LCB0aGVuIHllcywgeW91cg0KPiByZWFzb25pbmcgd291bGQgaG9sZCAoImlmIGZpcnN0IFtyb3dd
IGltcGxpZXMsIFsuLl0gdGhlbiBzZWNvbmQgcm93IGlzDQo+IGluY29ycmVjdCIpLCBidXQgeW91
ciByZWFzb25pbmcgd291bGQgYWN0dWFsbHkgYmUgb2YgbG93IGludGVyZXN0LA0KPiBiZWNhdXNl
IGluIHRoYXQgY2FzZSB0aGUgZmlyc3Qgcm93IHdvdWxkIGJlIGludmFsaWQgdG9vLi4uDQo+IC0g
aWYgd2UgY291bnQgYXMgb25lIG0uZSB3aGVuIGEgUEUgcGFyc2VzIGEgSm9pbiBvciBQcnVuZSBm
b3Igd2hpY2ggaXQNCj4gaXMgbm90IHRoZSB1cHN0cmVhbSwgZG9lcyBhIGxvb2t1cCBmb3IgdGhl
IGNvcnJlc3BvbmRpbmcgKFMsRykgdG8NCj4gZmluYWxseSBkaXNjYXJkIHRoZSBtZXNzYWdlLCB0
aGVuIHRoZSAqdHdvKiByb3dzIGFyZSBib3RoIHZhbGlkDQo+IA0KPiBUaGUgIihzdWNoIGhhcyB1
cGRhdGluZyBhIGNvbnRyb2wgcGxhbmUgb3IgZGF0YSBwbGFuZSBzdGF0ZSkiIHBocmFzZQ0KPiBp
cywNCj4gb2J2aW91c2x5LCBhIG5vbi1leGhhdXN0aXZlIGxpc3Qgb2YgZXhhbXBsZXMuIFdoZW4g
YXNzZXNzaW5nIHRoZQ0KPiBwcm9jZXNzaW5nIGRvbmUgYnkgUEVzLCB3ZSBoYXZlIHRvIHJlZmxl
Y3QgdGhlIHdvcmsgZG9uZSBieSBhIFBFIHRvDQo+IGlkZW50aWZ5IGl0IGNhbiBkaXNjYXJkIGEg
bWVzc2FnZSBiZWNhdXNlIGl0IGRvZXMgbm90IGhhdmUgYW55IChTLEcpDQo+IHN0YXRlLCBhZnRl
ciBwYXJzaW5nIHRoZSBKb2luIG9yIFBydW5lIGFuZCBkb2luZyBhIGxvb2t1cCBpbiBpdHMgVElC
Lg0KPiANCj4gV2UgY2FuIGNlcnRhaW5seSBhZGQgIm9yIGRpc2NhcmQgdGhlIGluZm9ybWF0aW9u
IGluIHRoZSBtZXNzYWdlIiBpbiB0aGUNCj4gbGlzdCBvZiBleGFtcGxlcy4NCj4gSSBzZWUgdGhp
cyBzZW50ZW5jZSBoYXMgYSB0eXBvICgiaGFzIicgaW5zdGVhZCBvZiAiYXMiKSwgd2hpY2ggcG9z
c2libHkNCj4gd2FzIHRoZSBzb3VyY2Ugb2YgY29uZnVzaW9uIDogaXQgd2lsbCBiZSBmaXhlZCB0
b28uDQo+IA0KPiBTbywgdG8gYXZvaWQgcHJvcGFnYXRpbmcgYW55IGNvbmZ1c2lvbiBhbnkgZnVy
dGhlciwgd2UgY2FuIHVwZGF0ZSB0aGUNCj4gZG9jdW1lbnQgYXMgZm9sbG93czoNCj4gDQo+ICAg
ICBBICJtZXNzYWdlIHByb2Nlc3NlZCIgYmVpbmcgYSBtZXNzYWdlIGJlaW5nIHBhcnNlZCwgYSBs
b29rdXAgYmVpbmcNCj4gZG9uZSwgYW5kIHNvbWUgYWN0aW9uIGJlaW5nDQo+ICAgICB0YWtlbiAo
c3VjaCBhcywgZm9yIGluc3RhbmNlLCB1cGRhdGluZyBhIGNvbnRyb2wgcGxhbmUgb3IgZGF0YQ0K
PiBwbGFuZSBzdGF0ZSBvciBkaXNjYXJkaW5nIHRoZQ0KPiAgICAgaW5mb3JtYXRpb24gaW4gdGhl
IG1lc3NhZ2UpLg0KPiANCj4gDQo+ICA+DQo+ICA+IFRoZSB0aGlyZCBhbmQgZmlmdGggY29sdW1u
cyBhYm92ZSBhcmUgZXhhbXBsZXMgb2YgY2FzZXMgd2hlcmUgdGhlDQo+IEFwcGVuZGl4DQo+ICA+
IG1pc2xlYWRpbmdseSBtdWx0aXBsaWVzIGJ5ICNtdnBuX1BFLg0KPiAgPg0KPiANCj4gDQo+IEFz
IHdlIHNlZSBhYm92ZSwgdGhlICNtVlBOX1BFcyBvcmRlciBvZiBtYWduaXR1ZGUgaXMgZGVmaW5p
dGVseSBjb3JyZWN0DQo+IGZvciBlYWNoIFBFIGpvaW5pbmcuDQo+IA0KPiBOb3csIGlmIHdlIGxl
dCdzIGxvb2sgYXQgd2hlbiBQRSBsZWF2ZToNCj4gLSBlYWNoIFBFIGxlYXZpbmcgY2F1c2VzIGEg
UHJ1bmUgdG8gYmUgc2VudCAoUkZDIDQ2MDEsIHNlY3Rpb24gNC41LjcsDQo+IGVuZCBvZiBwLjcz
KQ0KPiAtIFBFcyB0aGF0IGFyZSBqb2luZWQgdG8gKFMsRyksIHBhcnNlIHRoZSBtZXNzYWdlLCB0
cnkgdG8gZmluZCBhbiAoUyxHKQ0KPiBzdGF0ZSBpbiB0aGVpciBUSUIsIGZpbmQgb25lLCBhbmQg
dGhlbiBzZXR1cCBhIHNob3J0IHRpbWVyIHRvIHNlbmQgYQ0KPiBQcnVuZSB0byBvdmVycmlkZSB0
aGUgSm9pbg0KPiAtIFBFcyB0aGF0IGFyZSBub3Qgam9pbmVkIHRvIChTLEcpLCBwYXJzZSB0aGUg
bWVzc2FnZSwgdHJ5IHRvIGZpbmQgYW4NCj4gKFMsRykgc3RhdGUgaW4gdGhlaXIgVElCLCBmaW5k
IG5vbmUsIGFuZCBkZWNpZGUgdGhleSBjYW4gZGlzY2FyZCB0aGUNCj4gaW5mb3JtYXRpb24NCj4g
LSB0aGUgdXBzdHJlYW0gUEUgcGFyc2VzIHRoZSBtZXNzYWdlLCBtYXRjaGVzIGl0IHdpdGggaXRz
IChTLEcpIHN0YXRlLA0KPiBzZXR1cHMgYSB0aW1lciBhbmQgOg0KPiAgICAtIGluIHRoZSAibGFz
dCBQRSBsZWF2aW5nIiBjYXNlLCBhZnRlciBhIGRlbGF5IHN0b3AgZm9yd2FyZGluZw0KPiB0cmFm
ZmljDQo+ICAgIC0gZWxzZSwgd2lsbCBub3Qgc3RvcCBmb3J3YXJkaW5nIGJlY2F1c2Ugb2YgdGhl
IEpvaW4gc2VudCBieSBhbm90aGVyDQo+IFBFIHRvIG92ZXJyaWRlIHRoZSBQcnVuZQ0KPiANCj4g
DQo+IFRoZSAjbVZQTl9QRXMgZmFjdG9yIGhlcmUgYWxzbyBjb3JyZWN0bHkgcmVmbGVjdHMgdGhl
IGFtb3VudCBvZiB3b3JrDQo+IGRvbmUgYnkgUEVzLg0KPiANCj4gDQo+ICA+DQo+ICA+IEluIGZh
Y3QsIHRoZSBudW1iZXIgb2YgbWVzc2FnZXMgcmVjZWl2ZWQgYnkgYSBnaXZlbiBQRSBkb2VzIG5v
dA0KPiBpbmNyZWFzZSBpbiBwcm9wb3J0aW9uIHRvICNtdnBuX1BFLg0KPiAgPg0KPiANCj4gVGhl
IGNvbnRlbnQgb2YgQXBwZW5kaXggQSBjZXJ0YWlubHkgZG9lcyBub3QgaW1wbHkgdGhhdCAidGhl
IG51bWJlciBvZg0KPiBtZXNzYWdlcw0KPiByZWNlaXZlZCBieSBhIGdpdmVuIFBFIGluY3JlYXNl
cyBpbiBwcm9wb3J0aW9uIHRvICNtdnBuX1BFIi4NCj4gDQo+IFRoZSAjbVZQTl9QRSBmYWN0b3Ig
aXMgZm9yIHRoZSB0b3RhbCBhY3Jvc3MgYWxsIFBFcywgbm90IGZvciBhIGdpdmVuDQo+IFBFLg0K
PiANCj4gKCB0aGlzIGV4YWN0IHNhbWUgY29tbWVudCB3YXMgaW4geW91ciBlbWFpbCBmcm9tIEp1
bmUgMTF0aCwgYW5kIEkNCj4gYWxyZWFkeSByZXBsaWVkIHRvIHlvdSBKdW5lIDE5dGggOiB5b3Ug
ZGVsaWJlcmF0ZWx5IGlnbm9yZWQgbXkgcmVwbHksIEkNCj4gZG9uJ3QgdGhpbmsgaXQgaXMgZmFp
ciB0byBhZ2FpbiBmdXJ0aGVyIHByb3BhZ2F0ZSBzdWNoIGlycmVsZXZhbnQNCj4gaW1wbGljaXQg
Y2xhaW1zIG9uIHdoYXQgdGhlIGRvY3VtZW50IHNheXMuLi4pDQo+IA0KPiANCj4gID4NCj4gID4g
VGhpcyBwcm9ibGVtIGluZmVjdHMgdGhlIGVudGlyZSBzZXQgb2YgdGFibGVzIGluIHRoZSBBcHBl
bmRpeC4NCj4gID4NCj4gDQo+IFNlZSBhYm92ZSwgeW91ciBwb2ludCBpcyBvbmx5IHJpZ2h0IGlm
IHlvdSB0d2VhayB3aGF0ICJwcm9jZXNzaW5nIGENCj4gbWVzc2FnZSIgbWVhbnMgdXAgdG8gdGhl
IHBvaW50IG9mIGJlaW5nIGlycmVsZXZhbnQgYW5kIG5vdCBhY2NvdW50aW5nDQo+IGZvciBhbGwg
dGhlIHByb2Nlc3NpbmcgZG9uZSBieSBQRXMgb24gYSBMQU4uDQo+IA0KPiBMZXQgbWUgbm93IGhp
Z2hsaWdodCB0aGF0IHRoZSBwb2ludHMgYWJvdXQgInBydW5lIG92ZXJyaWRlIiBhbmQgImpvaW4N
Cj4gc3VwcHJlc3Npb24iIGluIHlvdXIgcmVjZW50IGVtYWlsIHdoZXJlIGZvcm1pbmcgYSBsYXJn
ZSBwYXJ0IG9mIHlvdXINCj4gY29tbWVudHMgZHVyaW5nIHRoZSBMQyBpbiBKdW5lLCBhbmQgd2hl
cmUgZnVlbGluZyB5b3VyIGNsYWltcyBvZg0KPiAiaW5jb3JyZWN0bmVzcyIgb2YgdGhlIGFuYWx5
c2lzIChzdWNoIGFzIHRoZSBjbGFpbSB0aGF0IHRoZSAiam9pbg0KPiBzdXBwcmVzc2lvbiIgd2Fz
IG5vdCB0YWtlbiBpbnRvIGFjY291bnQpLiAgQnkgbm90IHByb3ZpZGluZyBhIHByZWNpc2UNCj4g
cXVvdGUgb2Ygd2hhdCBleGFjdGx5IHdvdWxkIGhhdmUgYmVlbiAiaW5jb3JyZWN0IiB1bnRpbCBs
YXN0IHdlZWsgLS0NCj4gZGVzcGl0ZSBtdWx0aXBsZSByZXF1ZXN0cyB0byB5b3UgdG8gcHJvdmlk
ZSB0aGVtLCBhbmQgZGVzcGl0ZSB0aGUNCj4gbXVsdGlwbGUgY2xhcmlmaWNhdGlvbnMgSSBwcm92
aWRlZCBvbiB0aGUgZmFjdCB0aGF0ICJqb2luIHN1cHByZXNzaW9uIg0KPiBhbmQgInBydW5lIG92
ZXJyaWRlIiB3aGVyZSByZWxldmFudGx5IGFjY291bnRlZCBmb3IgLS0geW91IGhhdmUNCj4gYXJ0
aWZpY2lhbGx5IGRlbGF5ZWQgdGhlIGNsYXJpZmljYXRpb24gb2YgdGhpcyBpc3N1ZSBmb3Igcm91
Z2hseSBzaXgNCj4gbW9udGhzLg0KPiANCj4gLVRob21hcw0K

From thomas.morin@orange-ftgroup.com  Thu Dec 17 05:25:06 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 98B4C3A67EA for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 05:25:06 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.099
X-Spam-Level: 
X-Spam-Status: No, score=-3.099 tagged_above=-999 required=5 tests=[AWL=0.150,  BAYES_00=-2.599, HELO_EQ_FR=0.35, RCVD_IN_DNSWL_LOW=-1]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Puqs-EhUi9LO for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 05:25:05 -0800 (PST)
Received: from p-mail2.rd.francetelecom.com (p-mail2.rd.francetelecom.com [195.101.245.16]) by core3.amsl.com (Postfix) with ESMTP id 6FF9A3A68F1 for <l3vpn@ietf.org>; Thu, 17 Dec 2009 05:25:05 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp2.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Thu, 17 Dec 2009 14:24:47 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Thu, 17 Dec 2009 14:24:47 +0100
Message-ID: <4B2A311E.8060002@orange-ftgroup.com>
Date: Thu, 17 Dec 2009 14:24:46 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>,  L3VPN <l3vpn@ietf.org>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / M.Napierala's comments
References: <2F1DE4DFCFF32144B771BD2C246E6A200446CE07@misout7msgusr7e.ugd.att.com>	<4B291498.6060105@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C8AE@misout7msgusr7e.ugd.att.com>
In-Reply-To: <2F1DE4DFCFF32144B771BD2C246E6A200451C8AE@misout7msgusr7e.ugd.att.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 17 Dec 2009 13:24:47.0205 (UTC) FILETIME=[4DA68150:01CA7F1C]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 17 Dec 2009 13:25:06 -0000

Maria,

NAPIERALA, MARIA H (ATTLABS):
>>> This "problem" of treating not equal actions as equal repeats itself
>>> throughout the scaling analysis in the appendix. Such analysis is not
>>>        
>>> only *rough* but misleading, and because of this may lead to wrong
>>> conclusions. I believe that unless the appendix A is corrected by
>>> removing all not comparable, or even wrong estimations, it should be
>>> removed from the draft.
>>>        
>> ...but you only find this suggestion worth enough being brought forward
>>      
>> only six month or so after the WG LC... ;-)
>>      
> The analysis kept changing, so I waited until it is finalized (at least
> by the author). I believe that it is more important to make the draft as
> accurate as possible than worry about the timing.

This is not the usual understanding of "Last-Call", to say the least !
I've got no more comment on this except suggest reading the BCPs 
describing the IETF process...

( Oh, in fact, let me also add for the sake of precision that it is just 
totally false to state that
"the analysis kept changing", since it had mostly only editorial changes 
between draft-morin-l3vpn-mvpn-considerations-03, in July 2008, and the 
time the WG LC was made. The particular part you care about here has 
been _strictly_ unchanged in the draft since July 2008. )

-Thomas

From yakov@juniper.net  Thu Dec 17 05:59:18 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id A502028C12C for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 05:59:18 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.528
X-Spam-Level: 
X-Spam-Status: No, score=-6.528 tagged_above=-999 required=5 tests=[AWL=0.071,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id EflPk8cL5RPn for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 05:59:17 -0800 (PST)
Received: from exprod7og122.obsmtp.com (exprod7og122.obsmtp.com [64.18.2.22]) by core3.amsl.com (Postfix) with SMTP id 28A9828C124 for <l3vpn@ietf.org>; Thu, 17 Dec 2009 05:59:16 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob122.postini.com ([64.18.6.12]) with SMTP ID DSNKSyo5JVpOoSxqQ4C9zidu06E1QvXgQmj0@postini.com; Thu, 17 Dec 2009 05:59:03 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Thu, 17 Dec 2009 05:52:38 -0800
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Thu, 17 Dec 2009 05:52:37 -0800
Received: from emailsmtp56.jnpr.net ([172.24.60.77]) by p-emlb01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Thu, 17 Dec 2009 05:52:37 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp56.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Thu, 17 Dec 2009 05:52:37 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBHDqbj62561; Thu, 17 Dec 2009 05:52:37 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912171352.nBHDqbj62561@magenta.juniper.net>
To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / M.Napierala's comments 
In-Reply-To: <2F1DE4DFCFF32144B771BD2C246E6A200451C8AE@misout7msgusr7e.ugd.att.com>
References: <2F1DE4DFCFF32144B771BD2C246E6A200446CE07@misout7msgusr7e.ugd.att.com> <4B291498.6060105@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C8AE@misout7msgusr7e.ugd.att.com>
X-MH-In-Reply-To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com> message dated "Wed, 16 Dec 2009 13:57:59 -0500."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <81507.1261057957.1@juniper.net>
Date: Thu, 17 Dec 2009 05:52:37 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 17 Dec 2009 13:52:37.0643 (UTC) FILETIME=[314F1DB0:01CA7F20]
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, l3vpn@ietf.org
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 17 Dec 2009 13:59:18 -0000

Maria,

Maria> The field experience shows that there are multicast applications used in
Maria> enterprise VPNs such that only control messages, but no multicast data,
Maria> are sent across MVPN backbone. When BGP-multicast signaling is used, the
Maria> Source Active A-D route is originated for every C-(S,G) flow that
Maria> registers with C-RP and such that the C-RP joins "S" across MVPN
Maria> backbone, even if no inter-site data is sent for this flow. Hence, the
Maria> BGP scheme requires not only that for any given C-(S,G) flow that enters
Maria> MVPN backbone an SA message is announced and maintained by all PEs in a
Maria> given MVPN, but that SA messages are also announced for flows that never
Maria> enter the MVPN backbone.

By definition the MVPN service is a Layer 3 service, and as such it
does not distinguish whether customer's IP multicast packets carry
some application specific control information, or whether they are
the data traffic of a particular application. Thus from the MVPN
service point of view there is no different between a customer
multicast flow that carries control traffic of a particular application
and a customer multicast flow that carries data traffic of that
application. 

Given the above, if C-RP joins "S" across the MVPN backbone for a
particular (C-S, C-G), that means that *contrary to your claim above*,
the (C-S, C-G) traffic/flow is going to flow from "S" towards C-RP,
and it is going to flow *across the MVPN backbone*. And as I said
above, from the MVPN service point of view it is just a multicast
flow, irrespective of whether this flow carries some application
control information or a data of this application.

Thomas> What you state above it true only with one way of using the BGP 
Thomas> approach to support ASM (the one in section 14 of
Thomas> draft-ietf-l3vpn-2547bis-mcast-bgp)?
Thomas>
Thomas> The other way (section 13) does not have this effect, and happens to
Thomas> indirectly be the one recommended by
Thomas> draft-ietf-l3vpn-mvpn-considerations (because it is the one of the two
Thomas> not requiring an RP function co-located in the PE).
  
Maria> No, what I stated above applies to section 13 (I did not
Maria> even considered the implications of section 14..). The
Maria> inter-site shared trees are used.  When a PE receives
Maria> C-multicast Join route (which triggers Source Active A-D
Maria> route) it does not necessarily mean that data was actually
Maria> sent across MVPN backbone. 

Incorrect.

This is because for a PE to receive a Source Tree Join C-multicast
route for (C-S, C-G), it would require the PE that originates this
route to first receive PIM Join (C-S, C-G) from one of its CEs, and
that, in turn, would first require (C-S, C-G) to be sent across the
MVPN backbone. Therefore, this (C-S, C-G) has to be sent across the
MVPN backbone prior to the PE receiving the Source Tree Join
C-multicast route.

Yakov.

From yakov@juniper.net  Thu Dec 17 06:19:34 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 3E3A33A6AD2 for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 06:19:34 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.532
X-Spam-Level: 
X-Spam-Status: No, score=-6.532 tagged_above=-999 required=5 tests=[AWL=0.067,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id X2sCA1nP7gWt for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 06:19:33 -0800 (PST)
Received: from exprod7og113.obsmtp.com (exprod7og113.obsmtp.com [64.18.2.179]) by core3.amsl.com (Postfix) with SMTP id 7E40C3A6892 for <l3vpn@ietf.org>; Thu, 17 Dec 2009 06:19:31 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob113.postini.com ([64.18.6.12]) with SMTP ID DSNKSyo95C8xxNXBKfRuUKMAI5TOh+BasEbt@postini.com; Thu, 17 Dec 2009 06:19:19 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.71) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Thu, 17 Dec 2009 06:14:44 -0800
Received: from p-emlb02-sac.jnpr.net ([66.129.254.47]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Thu, 17 Dec 2009 06:14:43 -0800
Received: from emailsmtp56.jnpr.net ([172.24.60.77]) by p-emlb02-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Thu, 17 Dec 2009 06:14:43 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp56.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Thu, 17 Dec 2009 06:14:43 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBHEEhj72179; Thu, 17 Dec 2009 06:14:43 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912171414.nBHEEhj72179@magenta.juniper.net>
To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / RTConstraint recommendation 
In-Reply-To: <2F1DE4DFCFF32144B771BD2C246E6A200451C91C@misout7msgusr7e.ugd.att.com>
References: <3126.1259858356@erosen-linux> <4B289CDB.8090404@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C91C@misout7msgusr7e.ugd.att.com>
X-MH-In-Reply-To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com> message dated "Wed, 16 Dec 2009 14:52:12 -0500."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <83439.1261059282.1@juniper.net>
Date: Thu, 17 Dec 2009 06:14:42 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 17 Dec 2009 14:14:43.0423 (UTC) FILETIME=[478916F0:01CA7F23]
Cc: Eric Rosen <erosen@cisco.com>, Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 17 Dec 2009 14:19:34 -0000

Maria,

> Does this mean that BGP C-multicast signaling should never be deployed 
> without RT-Constraint because otherwise it would not work correctly/would 
> not scale?

BGP C-multicast signaling works correctly without RT-Constraint.
 
The scale at which a service provider can start wanting RT constraint
is very much dependent on the scale of deployment and performance
of the equipments used.

Yakov.


> Maria
> 
> > -----Original Message-----
> > From: l3vpn-bounces@ietf.org [mailto:l3vpn-bounces@ietf.org] On Behalf
> > Of Thomas Morin
> > Sent: Wednesday, December 16, 2009 3:40 AM
> > To: L3VPN; Eric Rosen
> > Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /
> > RTConstraint recommendation
> > 
> > Eric, working group,
> > 
> > Eric Rosen:
> >  >        the case where the RT-Constraint mechanisms [RFC4684] is
> >  >        not used is not covered;
> >  >
> >  > Since RT-constraint is not a mandatory to implement mechanism, I
> > don't see
> >  > why this is not covered.
> > 
> > You omitted the first part of the sentence you quoted, which actually
> > provides the answer:
> > 
> >        following the recommendation of [I-D.ietf-l3vpn-2547bis-mcast-
> > bgp]
> >        the case where the RT-Constraint mechanisms [RFC4684] is not
> > used
> > is not
> >        covered;
> > 
> > Here is the corresponding quote in draft-ieft-2547bis-mcast-bgp (which
> > you happen co-author) :
> > 
> >     16. Scalability Considerations
> > 
> >     A PE should use Route Target Constrain [RT-CONSTRAIN] to advertise
> >     the Route Targets that the PE uses for the VRF Route Imports
> > extended
> >     community (note that doing this requires just a single Route Target
> >     Constraint advertisement by the PE). This allows each C-multicast
> >     route to reach only the relevant PE, rather than all the PEs
> >     participating the an MVPN.
> > 
> > Hence the considerations document does not cover the case where the
> > RT-Constraint mechanism is not used.
> > 
> > -Thomas
> > 
> > PS: Let me note that you did not bring up this up during the LC in
> > June,
> > and that this comment is not related to any new content in the
> > revisions
> > made since this LC.
> > 
> 
> 

From mn1921@att.com  Thu Dec 17 08:14:11 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id AF4BE3A6AD2 for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 08:14:10 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -106.37
X-Spam-Level: 
X-Spam-Status: No, score=-106.37 tagged_above=-999 required=5 tests=[AWL=0.229, BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id JTkUxRCK+6o0 for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 08:14:04 -0800 (PST)
Received: from mail161.messagelabs.com (mail161.messagelabs.com [216.82.253.115]) by core3.amsl.com (Postfix) with ESMTP id 4DB153A681E for <l3vpn@ietf.org>; Thu, 17 Dec 2009 08:14:03 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-13.tower-161.messagelabs.com!1261066427!23032977!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 27887 invoked from network); 17 Dec 2009 16:13:48 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-13.tower-161.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 17 Dec 2009 16:13:48 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBHGDhl2009402; Thu, 17 Dec 2009 11:13:43 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBHGDd4w009364; Thu, 17 Dec 2009 11:13:40 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / RTConstraint recommendation 
Date: Thu, 17 Dec 2009 11:13:41 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A200457FA64@misout7msgusr7e.ugd.att.com>
In-Reply-To: <200912171414.nBHEEhj72179@magenta.juniper.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / RTConstraint recommendation 
Thread-Index: Acp/I/Hj9qVrredfRCmX6dBuAb8MdgAD7D9Q
References: <3126.1259858356@erosen-linux> <4B289CDB.8090404@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C91C@misout7msgusr7e.ugd.att.com> <200912171414.nBHEEhj72179@magenta.juniper.net>
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: "Yakov Rekhter" <yakov@juniper.net>
Cc: Eric Rosen <erosen@cisco.com>, Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 17 Dec 2009 16:14:11 -0000

That is what I thought, in which case the consideration draft should
consider both BGP C-multicast signaling scenarios: with and without
using RT-Constraint.

Maria

> -----Original Message-----
> From: Yakov Rekhter [mailto:yakov@juniper.net]
> Sent: Thursday, December 17, 2009 9:15 AM
> To: NAPIERALA, MARIA H (ATTLABS)
> Cc: Thomas Morin; L3VPN; Eric Rosen
> Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /
> RTConstraint recommendation
>=20
> Maria,
>=20
> > Does this mean that BGP C-multicast signaling should never be
> deployed
> > without RT-Constraint because otherwise it would not work
> correctly/would
> > not scale?
>=20
> BGP C-multicast signaling works correctly without RT-Constraint.
>=20
> The scale at which a service provider can start wanting RT constraint
> is very much dependent on the scale of deployment and performance
> of the equipments used.
>=20
> Yakov.
>=20
>=20
> > Maria
> >
> > > -----Original Message-----
> > > From: l3vpn-bounces@ietf.org [mailto:l3vpn-bounces@ietf.org] On
> Behalf
> > > Of Thomas Morin
> > > Sent: Wednesday, December 16, 2009 3:40 AM
> > > To: L3VPN; Eric Rosen
> > > Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A
/
> > > RTConstraint recommendation
> > >
> > > Eric, working group,
> > >
> > > Eric Rosen:
> > >  >        the case where the RT-Constraint mechanisms [RFC4684] is
> > >  >        not used is not covered;
> > >  >
> > >  > Since RT-constraint is not a mandatory to implement mechanism,
I
> > > don't see
> > >  > why this is not covered.
> > >
> > > You omitted the first part of the sentence you quoted, which
> actually
> > > provides the answer:
> > >
> > >        following the recommendation of [I-D.ietf-l3vpn-2547bis-
> mcast-
> > > bgp]
> > >        the case where the RT-Constraint mechanisms [RFC4684] is
not
> > > used
> > > is not
> > >        covered;
> > >
> > > Here is the corresponding quote in draft-ieft-2547bis-mcast-bgp
> (which
> > > you happen co-author) :
> > >
> > >     16. Scalability Considerations
> > >
> > >     A PE should use Route Target Constrain [RT-CONSTRAIN] to
> advertise
> > >     the Route Targets that the PE uses for the VRF Route Imports
> > > extended
> > >     community (note that doing this requires just a single Route
> Target
> > >     Constraint advertisement by the PE). This allows each C-
> multicast
> > >     route to reach only the relevant PE, rather than all the PEs
> > >     participating the an MVPN.
> > >
> > > Hence the considerations document does not cover the case where
the
> > > RT-Constraint mechanism is not used.
> > >
> > > -Thomas
> > >
> > > PS: Let me note that you did not bring up this up during the LC in
> > > June,
> > > and that this comment is not related to any new content in the
> > > revisions
> > > made since this LC.
> > >
> >
> >

From John.E.Drake2@boeing.com  Thu Dec 17 08:26:17 2009
Return-Path: <John.E.Drake2@boeing.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 408663A69BC for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 08:26:17 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.299
X-Spam-Level: 
X-Spam-Status: No, score=-6.299 tagged_above=-999 required=5 tests=[AWL=-0.300, BAYES_00=-2.599, J_CHICKENPOX_13=0.6, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id TRw+dIkImAgY for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 08:26:16 -0800 (PST)
Received: from slb-smtpout-01.boeing.com (slb-smtpout-01.boeing.com [130.76.64.48]) by core3.amsl.com (Postfix) with ESMTP id 371A63A69BD for <l3vpn@ietf.org>; Thu, 17 Dec 2009 08:26:16 -0800 (PST)
Received: from slb-av-01.boeing.com (slb-av-01.boeing.com [129.172.13.4]) by slb-smtpout-01.ns.cs.boeing.com (8.14.0/8.14.0/8.14.0/SMTPOUT) with ESMTP id nBHGPvvf016665 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 17 Dec 2009 08:25:57 -0800 (PST)
Received: from slb-av-01.boeing.com (localhost [127.0.0.1]) by slb-av-01.boeing.com (8.14.0/8.14.0/DOWNSTREAM_RELAY) with ESMTP id nBHGPve9026882; Thu, 17 Dec 2009 08:25:57 -0800 (PST)
Received: from xch-swht-02.sw.nos.boeing.com (xch-swht-02.sw.nos.boeing.com [129.172.142.5]) by slb-av-01.boeing.com (8.14.0/8.14.0/UPSTREAM_RELAY) with ESMTP id nBHGPvcZ026878 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=OK); Thu, 17 Dec 2009 08:25:57 -0800 (PST)
Received: from XCH-SW-03V.sw.nos.boeing.com ([129.172.197.23]) by xch-swht-02.sw.nos.boeing.com ([129.172.142.5]) with mapi; Thu, 17 Dec 2009 08:25:57 -0800
From: "Drake, John E" <John.E.Drake2@boeing.com>
To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>, Yakov Rekhter <yakov@juniper.net>
Date: Thu, 17 Dec 2009 08:25:55 -0800
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /RTConstraint recommendation 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /RTConstraint recommendation 
Thread-Index: Acp/I/Hj9qVrredfRCmX6dBuAb8MdgAD7D9QAAB5yNA=
Message-ID: <83C701404B004C4094B604670B3F5FFA395C58AB35@XCH-SW-03V.sw.nos.boeing.com>
References: <3126.1259858356@erosen-linux><4B289CDB.8090404@orange-ftgroup.c om><2F1DE4DFCFF32144B771BD2C246E6A200451C91C@misout7msgusr7e.ugd.att.com><200912171414.nBHEEhj72179@magenta.juniper.net> <2F1DE4DFCFF32144B771BD2C246E6A200457FA64@misout7msgusr7e.ugd.att.com>
In-Reply-To: <2F1DE4DFCFF32144B771BD2C246E6A200457FA64@misout7msgusr7e.ugd.att.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Cc: Eric Rosen <erosen@cisco.com>, Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 17 Dec 2009 16:26:17 -0000

Would it be possible to call the question?=20

> -----Original Message-----
> From: l3vpn-bounces@ietf.org [mailto:l3vpn-bounces@ietf.org]=20
> On Behalf Of NAPIERALA, MARIA H (ATTLABS)
> Sent: Thursday, December 17, 2009 8:14 AM
> To: Yakov Rekhter
> Cc: Eric Rosen; Thomas Morin; L3VPN
> Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 /=20
> Appendix A /RTConstraint recommendation=20
>=20
> That is what I thought, in which case the consideration draft=20
> should consider both BGP C-multicast signaling scenarios:=20
> with and without using RT-Constraint.
>=20
> Maria
>=20
> > -----Original Message-----
> > From: Yakov Rekhter [mailto:yakov@juniper.net]
> > Sent: Thursday, December 17, 2009 9:15 AM
> > To: NAPIERALA, MARIA H (ATTLABS)
> > Cc: Thomas Morin; L3VPN; Eric Rosen
> > Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /=20
> > RTConstraint recommendation
> >=20
> > Maria,
> >=20
> > > Does this mean that BGP C-multicast signaling should never be
> > deployed
> > > without RT-Constraint because otherwise it would not work
> > correctly/would
> > > not scale?
> >=20
> > BGP C-multicast signaling works correctly without RT-Constraint.
> >=20
> > The scale at which a service provider can start wanting RT=20
> constraint=20
> > is very much dependent on the scale of deployment and=20
> performance of=20
> > the equipments used.
> >=20
> > Yakov.
> >=20
> >=20
> > > Maria
> > >
> > > > -----Original Message-----
> > > > From: l3vpn-bounces@ietf.org [mailto:l3vpn-bounces@ietf.org] On
> > Behalf
> > > > Of Thomas Morin
> > > > Sent: Wednesday, December 16, 2009 3:40 AM
> > > > To: L3VPN; Eric Rosen
> > > > Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 /=20
> Appendix A
> /
> > > > RTConstraint recommendation
> > > >
> > > > Eric, working group,
> > > >
> > > > Eric Rosen:
> > > >  >        the case where the RT-Constraint mechanisms=20
> [RFC4684] is
> > > >  >        not used is not covered;
> > > >  >
> > > >  > Since RT-constraint is not a mandatory to implement=20
> mechanism,
> I
> > > > don't see
> > > >  > why this is not covered.
> > > >
> > > > You omitted the first part of the sentence you quoted, which
> > actually
> > > > provides the answer:
> > > >
> > > >        following the recommendation of [I-D.ietf-l3vpn-2547bis-
> > mcast-
> > > > bgp]
> > > >        the case where the RT-Constraint mechanisms [RFC4684] is
> not
> > > > used
> > > > is not
> > > >        covered;
> > > >
> > > > Here is the corresponding quote in draft-ieft-2547bis-mcast-bgp
> > (which
> > > > you happen co-author) :
> > > >
> > > >     16. Scalability Considerations
> > > >
> > > >     A PE should use Route Target Constrain [RT-CONSTRAIN] to
> > advertise
> > > >     the Route Targets that the PE uses for the VRF=20
> Route Imports=20
> > > > extended
> > > >     community (note that doing this requires just a single Route
> > Target
> > > >     Constraint advertisement by the PE). This allows each C-
> > multicast
> > > >     route to reach only the relevant PE, rather than all the PEs
> > > >     participating the an MVPN.
> > > >
> > > > Hence the considerations document does not cover the case where
> the
> > > > RT-Constraint mechanism is not used.
> > > >
> > > > -Thomas
> > > >
> > > > PS: Let me note that you did not bring up this up=20
> during the LC in=20
> > > > June, and that this comment is not related to any new=20
> content in=20
> > > > the revisions made since this LC.
> > > >
> > >
> > >
> =

From thomas.morin@orange-ftgroup.com  Thu Dec 17 09:56:58 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 0F4C83A67DB for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 09:56:58 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.837
X-Spam-Level: 
X-Spam-Status: No, score=-1.837 tagged_above=-999 required=5 tests=[AWL=-0.188, BAYES_00=-2.599, HELO_EQ_FR=0.35, J_CHICKENPOX_51=0.6]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id GogkygiFzxEc for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 09:56:57 -0800 (PST)
Received: from r-mail1.rd.francetelecom.com (r-mail1.rd.francetelecom.com [217.108.152.41]) by core3.amsl.com (Postfix) with ESMTP id 9969E3A68EA for <l3vpn@ietf.org>; Thu, 17 Dec 2009 09:56:56 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp1.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Thu, 17 Dec 2009 18:56:38 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Thu, 17 Dec 2009 18:56:37 +0100
Message-ID: <4B2A70D5.1030608@orange-ftgroup.com>
Date: Thu, 17 Dec 2009 18:56:37 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>,  L3VPN <l3vpn@ietf.org>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / M.Napierala's comments
References: <2F1DE4DFCFF32144B771BD2C246E6A200446CE07@misout7msgusr7e.ugd.att.com>	<4B291498.6060105@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C8AE@misout7msgusr7e.ugd.att.com>
In-Reply-To: <2F1DE4DFCFF32144B771BD2C246E6A200451C8AE@misout7msgusr7e.ugd.att.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 17 Dec 2009 17:56:37.0821 (UTC) FILETIME=[4788FED0:01CA7F42]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 17 Dec 2009 17:56:58 -0000

Maria, working group,

Maria:
> >  >  If the solution to this is to make the timer for sending
>    
> >  >  Prune(S,G,rpt)  sufficiently high (to assure that SPT is created) and avoid
>    
> >  >  duplicates  sent to customers by having egress PEs discard data packets received
> >  >  from the "wrong" upstream PE, this should be included in MVPN
> >  >  considerations draft.
>    
> >  
> >  
> >  I personnally agree with you it would be reasonable that the draft
> >  recommends implementing these procedures (
> >  http://tools.ietf.org/html/draft-ietf-l3vpn-2547bis-mcast-09#section-
> >  9.1.1 )
>    
 >
> Good

Ok.

We can propose this to the working group, supposing noone disagrees 
despite this coming far after WG LC.


>>> 2. The considerations draft should also point out the difference how BGP
>>>        
>>> vs. PIM (used as PE-PE signaling protocol) handles the upstream PE
>>> selection in case of multi-homed sources/RPs.
>>>
>>> Typically, in a layer 3 VPN deployment it is not known whether or
>>> not
>>>        
>>> all VRFs of a given VPN have the same best route to a given source. In
>>> other words, VRFs routing policies are provisioned independently of each
>>>        
>>> other. If all VRFs of a given MVPN have the same best/ installed route
>>>        
>>> to a multi-homed source/RP then this VPN customer *expects* that the
>>> source/RP will be joined via the installed route.
>>>
>>> When PIM is used as PE-PE C-signaling, the installed route towards the
>>>        
>>> source is always chosen as UHM.
>>>        
>> This is not my understanding, to quote
>> draft-ietf-l3vpn-2547bis-mcast-09, section-5.1.3:
>>
>>      The default procedure, which MUST be implemented, is to select the
>>      route whose corresponding upstream PE address is numerically highest,
>>      where a 32-bit IP address is treated as a 32 bit unsigned integer.
>>      Call this the "default upstream PE selection".
>>
>>      [...]
>>
>>      Another procedure, which SHOULD be implemented, is to use the
>>      Installed UMH Route as the Selected UMH Route.
>>
>>
>>      
>>>    If different VRFs have different best
>>> routes to C-S or C-RP, the Assert process is a tie breaker. No steady
>>>        
>>> state duplicate traffic is sent across MVPN backbone.
>>>
>>> Now let's consider BGP as the PE-PE signaling protocol and let's assume
>>>        
>>> the UMH policy is to choose the installed/best route towards C-source or
>>>        
>>> C-RP.
>>>        
>> Given the above, why is this assumption relevant ?
>>
>>      
>>>    If different VRFs have different installed routes to C-S/C-RP then
>>> there will be a steady state duplication of traffic across the MVPN
>>> backbone (to the same set of PEs if MI-PMSI is used). This also requires
>>>        
>>> that the egress PEs be able to discard data traffic that arrives from
>>>        
>>> the "wrong" upstream PE. Such duplication of traffic might not be
>>> operationally acceptable to a service provider because of too much
>>> network bandwidth being consumed.
>>>        
>> I don't think this is a relevant drawback :
>> - if a provider chooses the non-default UMH selection you mention,  it
>> is likely that he is fine with the bandwidth aspect, and has made sure
>> the vendors implement the procedures to discard data traffic that
>> arrives from the "wrong" upstream PE.
>>      
> Why not to state it?

At this point in the process the question is "why state it ?" rather 
than "why not" (supposing we have a question to ask)


> This does not come for free and requires "strong" RPF check i.e., checking the upstream PE. Are you aware of any such implementations?

(ask your vendors)



>> - see above for the suggestion to recommend implementing the procedures
>>      
>> required to discard data traffic that arrives from the "wrong" upstream
>>      
>> PE (making it easier for the provider to make sure that it is ok for
>> him to use the non-default UMH selection you mention ).
>>      

Isn't it sufficient ?



>>> 5. Why Appendix A analyzes PIM LAN procedure with explicit tracking but
>>>        
>>> not the explicit tracking using BGP-signaling?
>>>
>>>        
>> I'm not sure as what you call "explicit tracking using BGP-signaling".
>>
>> There is, AFAIK, no such thing as "explicit tracking" in the BGP
>> procedures for C-multicast routing, and Appendix A is only about
>> C-multicast routing options.
>>      
> Well, in that case, PIM LAN procedure with explicit tracking should be
> removed also. Why would you do explicit tracking of receivers when using
> PIM?

Well, see my reply to Eric Rosen 
"draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / PIM with 
explicit tracking":
- there is perfect rationale to cover it (part of RFC4601 => part of the 
PIM-based C-multicast routing options)
- at the same time, if there is consensus that it is not useful I don't 
mind removing it

-Thomas


From mn1921@att.com  Thu Dec 17 11:59:23 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id AD79D3A69F8 for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 11:59:23 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -106.395
X-Spam-Level: 
X-Spam-Status: No, score=-106.395 tagged_above=-999 required=5 tests=[AWL=0.204, BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id SKYjyEtVJnIp for <l3vpn@core3.amsl.com>; Thu, 17 Dec 2009 11:59:23 -0800 (PST)
Received: from mail167.messagelabs.com (mail167.messagelabs.com [216.82.253.179]) by core3.amsl.com (Postfix) with ESMTP id F057F3A690E for <l3vpn@ietf.org>; Thu, 17 Dec 2009 11:59:22 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-7.tower-167.messagelabs.com!1261079946!21680510!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 12364 invoked from network); 17 Dec 2009 19:59:07 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-7.tower-167.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 17 Dec 2009 19:59:07 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBHJx2wP000424; Thu, 17 Dec 2009 14:59:03 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBHJx0iD000403; Thu, 17 Dec 2009 14:59:00 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / M.Napierala's comments 
Date: Thu, 17 Dec 2009 14:59:02 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A200457FC84@misout7msgusr7e.ugd.att.com>
In-Reply-To: <200912171352.nBHDqbj62561@magenta.juniper.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 / M.Napierala's comments
Thread-Index: Acp/IRx9BW0jxCgwS9S+gYWYLTB70QAMjAiw
References: <2F1DE4DFCFF32144B771BD2C246E6A200446CE07@misout7msgusr7e.ugd.att.com> <4B291498.6060105@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C8AE@misout7msgusr7e.ugd.att.com> <200912171352.nBHDqbj62561@magenta.juniper.net>
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: "Yakov Rekhter" <yakov@juniper.net>
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, l3vpn@ietf.org
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 17 Dec 2009 19:59:23 -0000

Yakov,

> Given the above, if C-RP joins "S" across the MVPN backbone for a
> particular (C-S, C-G), that means that *contrary to your claim above*,
> the (C-S, C-G) traffic/flow is going to flow from "S" towards C-RP,
> and it is going to flow *across the MVPN backbone*. And as I said

Actually, the multicast traffic does not necessarily flow across the
MVPN backbone. Customers use techniques (like multicast ttl-threshold)
to keep multicast data traffic local to a site. We see this happening
frequently.=20

> above, from the MVPN service point of view it is just a multicast
> flow, irrespective of whether this flow carries some application
> control information or a data of this application.=20

This is not what I meant, i.e., that the flow itself carries some
"control" information.
I meant that the traffic does no flow at all, as I explained above.
=20
Maria


From thomas.morin@orange-ftgroup.com  Fri Dec 18 07:02:36 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 588AE3A6A89 for <l3vpn@core3.amsl.com>; Fri, 18 Dec 2009 07:02:36 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.816
X-Spam-Level: 
X-Spam-Status: No, score=-1.816 tagged_above=-999 required=5 tests=[AWL=-0.167, BAYES_00=-2.599, HELO_EQ_FR=0.35, J_CHICKENPOX_31=0.6]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id FrVh-aFL0rH2 for <l3vpn@core3.amsl.com>; Fri, 18 Dec 2009 07:02:35 -0800 (PST)
Received: from r-mail2.rd.francetelecom.com (r-mail2.rd.francetelecom.com [217.108.152.42]) by core3.amsl.com (Postfix) with ESMTP id D156D3A67FB for <l3vpn@ietf.org>; Fri, 18 Dec 2009 07:02:34 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp2.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Fri, 18 Dec 2009 16:02:19 +0100
Received: from [10.193.15.39] ([10.193.15.39]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Fri, 18 Dec 2009 16:02:18 +0100
Message-ID: <4B2B9979.4070708@orange-ftgroup.com>
Date: Fri, 18 Dec 2009 16:02:17 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: L3VPN <l3vpn@ietf.org>, Eric Rosen <erosen@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / looking at multiple (S,G)
References: <3126.1259858356@erosen-linux>
In-Reply-To: <3126.1259858356@erosen-linux>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 18 Dec 2009 15:02:18.0072 (UTC) FILETIME=[1773B580:01CA7FF3]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 18 Dec 2009 15:02:36 -0000

Eric, working group,

In different parts of the document (quoted below) you claim that the 
scaling analysis should consider scalability with an increased number of 
sources sending to a group.

Before looking at the relevance question, let's see how it relates to 
the current state of the document.
Section 3.3 of the document relates to scalability with an increased 
number of PEs, and refers to the scaling analysis in Appendix A.1 
"Scalability with an increased number of PEs". Given this, claiming that 
not doing a scaling analysis with an increased number of (S,G) is "a 
methodological error" in this scaling analysis is not correct.

That "scaling with an increased number of (S,G)" should be studied is 
something that noone has proposed to do yet, as far as I can tell.

Now, that said, let's try to see how interesting it would be to consider 
looking more at this aspect:
- in your comments, you insist on the linear dependency of BGP 
C-multicast routing with the number of (S,G), but scaling linearly with 
the number of sources sending to a group is not a peculiarity of the BGP 
Source Active auto-discovery mechanism, but is also common for PIM-SM 
SPTs and SSM trees (whether they are carried across the provider core 
through PIM and BGP, by the way)
- "the number of sources sending to a group", or the number of SPTs or 
SSM trees, is not something typically under the control of the provider, 
as it completely depends on the multicast routing policy and 
subscriptions patterns of each MVPN customer; so it should be obvious to 
anyone that the provider will have to prepare its network to scale with 
these unknowns in mind and put bounds to the amount of per-source state 
on his PEs (and RRs) ; this is true independently of whether PIM-SM or 
BGP is used for C-multicast routing (and true whether or not MSDP is 
used in that deployment)
- note also also that "scaling with reference to the number of sources" 
is trivially extrapolated from the single (S,G) case
- as far as RR are concerned (a point on which you focus below): having 
an amount of state to maintain on RRs that scales linearly with the 
scale of deployments and customer routes is not unusual, especially when 
it offers scaling advantages

Given the above, I don't think it is obvious that looking at scalability 
with the number of sources sending to a group, will be helpful for a 
provider to choose the protocol he will use for PE-PE C-multicast routing.

Speaking for myself as a contributor, these elements explain why I did 
not contribute a scaling study for an increased number of sources.   
"//Your mileage may vary", but we are six month after WG LC....

-Thomas

PS: a few more comments below

Eric Rosen:
> Another methodological error is that only the messaging needed to join a
> single (S,G) flow is considered.  Thus the analysis ignores any scaling
> issues that show up as the number of (S,G) flows increases.  Since the
> number of (S,G) flows handled by a Route Reflector will far exceed the
> number handled by any PE router,

(This is not exact, as a said PE could use different RRs for different 
VPNs.)

>   failure to consider this particular scaling
> factor tends to bias the results in favor of a scheme that uses Route
> Reflectors.  [...]

I disagree: as said, above, PIM approaches are subject to the same type 
of linear dependency with the number of (S,G).


[...]
> Let's take a look now at some of the more detailed issues.
>
> [...] let's look specifically at the ASM case.
>
> One of the features of the BGP solution is that there is no way, in BGP,
> to represent the PIM operation of "prune source S off the (*,G) tree".
> To compensate for this, every PE in a given VPN is forced to keep track of
> every (S,G) stream for which any PE in that VPN has receivers.  This is done
> by means of the Source Active A-D routes.
>
>    


> There are two options that can be used.  In one option, where "inter-site
> shared trees are not used", each PE functions as a C-RP, and then all the PE
> of a given VPN must exchange information on all the (S,G) streams that they
> see locally.  This option uses BGP to do something that is very similar to
> what MSDP is used for today.  Let me quote from Pekka Savola's review of
> this procedure:
>
>           Active source BGP messages.  This is a duplication of a
>           similar mechanism in MSDP (RFC3618) which has caused
>           much grief in Internet.  Does this meant that when a
>           host does 'nmap -sU 224.0.0.0/4' at a VPN site, this
>           will result in about 268 million BGP active source
>           updates being sent (228) in the SP backbone?
>
> The only answer to this was to beef up the Security Considerations with
> some required procedures that will prevent the PE from melting down entirely
> in this case.  Of course, Appendix A's methodology of only considering a
> single (S,G) prevents it from seeing that this particular BGP option
> uses a mechanism whose scalability is already known to be poo,r and which
> the multicast WGs have decided not even to bring forward to IPv6.[...]
>    

Your are generalizing Pekka's comment on a *security* issue MSDP *in the 
Internet*, to the issue of *scaling* in a multicast VPN context.  I 
think this is an abusive generalization of his quote, and that it does 
not help much seeing why draft-ietf-l3vpn-mvpn-considerations should 
look at scaling with reference to the number of sources.   Nobody 
objects that security issues related to the amount of "per source" state 
is a legitimate concern, but that does not mechanically make it an 
interesting scalability question in our context.

Second, let me note that you replied yourself to Pekka Savola to his 
comment, by referring to security mechanisms common to PIM and BGP, that 
allows to control the amount of multicast state in PEs.


> In the other, more sensible option, a PE originates a Source Active A-D
> route for an (S,G) only when it receives a C-multicast Join route for that
> (S,G).  However, the Source Active A-D route goes to all the PEs in the VPN,
> not just to those that have receivers for that (S,G).  I believe this means
> that the amount of state a given PE maintains in the BGP scheme is
> proportional to the number of (S,G) flows for which any PE in the VPN has
> receivers.  Actually, since the SA route contains the RD of the originating
> PE, the situation is a bit worse; if (S,G) traffic is sent over the backbone
> by more than one PE, the other PEs in the same VPN must maintain an (S,G) SA
> A-D route for each such PE.
>
> In the BGP solution, even if a PE has only (*,G) receivers, or even if it
> has no receivers for G at all, it must maintain state for all the (S,G)s in
> the MVPN, multiplied by the average number of upstream PEs per source.
>    
[...]
> Strangely, this scaling issue is considered to be insignificant by Appendix
> A [...]

(See start of this email)


From thomas.morin@orange-ftgroup.com  Fri Dec 18 12:50:31 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 133A83A6ABC for <l3vpn@core3.amsl.com>; Fri, 18 Dec 2009 12:50:31 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.099
X-Spam-Level: 
X-Spam-Status: No, score=-2.099 tagged_above=-999 required=5 tests=[AWL=0.150,  BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 0ZQhqUSDCDuL for <l3vpn@core3.amsl.com>; Fri, 18 Dec 2009 12:50:30 -0800 (PST)
Received: from r-mail1.rd.francetelecom.com (r-mail1.rd.francetelecom.com [217.108.152.41]) by core3.amsl.com (Postfix) with ESMTP id DD6CB3A69BF for <l3vpn@ietf.org>; Fri, 18 Dec 2009 12:50:29 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp1.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Fri, 18 Dec 2009 21:49:58 +0100
Received: from [10.193.116.45] ([10.193.116.45]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Fri, 18 Dec 2009 21:49:57 +0100
Message-ID: <4B2BEAF5.4020609@orange-ftgroup.com>
Date: Fri, 18 Dec 2009 21:49:57 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: Maria Napierala <mnapierala@att.com>, L3VPN <l3vpn@ietf.org>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /	sectionA.1.1.3
References: <3126.1259858356@erosen-linux>	<4B28E84F.4000604@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com>
In-Reply-To: <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 18 Dec 2009 20:49:57.0954 (UTC) FILETIME=[A8E91A20:01CA8023]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 18 Dec 2009 20:50:31 -0000

Maria,

NAPIERALA, MARIA H (ATTLABS):
>>     - ORIGINATOR_ID : updates of the value of this attributes do not
>> "follow" PE advertisements, as the RR rewrite it when aggregating C-
>> multicast routes having same NLRI; the advertisements made by different
>> RRs may lead possibly to updates of the value of this attribute, given
>> that each RR would set ORIGINATOR_ID to a value of its own
>>      
 >
> Regarding route *aggregation* the draft-ietf-l3vpn-2547bis-mcast-bgp-08 only states the following:
>
>     "To further reduce the routing churn due to C-multicast routes changes
>     a Route Reflector that re-advertises a C-multicast route SHOULD set
>     the Next Hop field of the MP_REACH_NLRI attribute of the route to an
>     IP address of the Route Reflector. Likewise, an ASBR that re-
>     advertises a C-multicast route SHOULD set the Next Hop field of the
>     MP_REACH_NLRI attribute of the route to an IP address of the ASBR."
>
> It states nothing about setting ORIGINATOR_ID to self on RR.

It also says:

      [...] C-multicast routes for a given (S,G) of a given MVPN 
originated by PEs that are clients of
      a given Route Reflector are aggregated by the Route Reflector. [...]

A BGP aggregated route is a route originated by the BGP peer doing the 
aggregation, why would it carry another ORIGINATOR_ID than the one of 
the BGP peer doing the aggregation (here the RR) ?

-Thomas

From thomas.morin@orange-ftgroup.com  Fri Dec 18 13:02:46 2009
Return-Path: <thomas.morin@orange-ftgroup.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id F2D883A6ACC for <l3vpn@core3.amsl.com>; Fri, 18 Dec 2009 13:02:45 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.113
X-Spam-Level: 
X-Spam-Status: No, score=-2.113 tagged_above=-999 required=5 tests=[AWL=0.136,  BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id cPEzP7FxWtXi for <l3vpn@core3.amsl.com>; Fri, 18 Dec 2009 13:02:44 -0800 (PST)
Received: from r-mail1.rd.francetelecom.com (r-mail1.rd.francetelecom.com [217.108.152.41]) by core3.amsl.com (Postfix) with ESMTP id DE0283A6AC3 for <l3vpn@ietf.org>; Fri, 18 Dec 2009 13:02:42 -0800 (PST)
Received: from ftrdmel10.rd.francetelecom.fr ([10.192.128.44]) by ftrdsmtp1.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Fri, 18 Dec 2009 22:02:24 +0100
Received: from [10.193.116.45] ([10.193.116.45]) by ftrdmel10.rd.francetelecom.fr with Microsoft SMTPSVC(6.0.3790.3959);  Fri, 18 Dec 2009 22:02:24 +0100
Message-ID: <4B2BEDDF.3090308@orange-ftgroup.com>
Date: Fri, 18 Dec 2009 22:02:23 +0100
From: Thomas Morin <thomas.morin@orange-ftgroup.com>
Organization: France Telecom Orange
User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.9.1.5) Gecko/20091204 Lightning/1.0b1pre Thunderbird/3.0
MIME-Version: 1.0
To: Robert Raszuk <raszuk@cisco.com>, L3VPN <l3vpn@ietf.org>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /	sectionA.1.1.3
References: <3126.1259858356@erosen-linux>	<4B28E84F.4000604@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com> <4B296D1C.6000001@cisco.com>
In-Reply-To: <4B296D1C.6000001@cisco.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 18 Dec 2009 21:02:24.0183 (UTC) FILETIME=[65B29C70:01CA8025]
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 18 Dec 2009 21:02:46 -0000

Hello Robert,

Robert Raszuk:
 >
> It seems that for this mvpn solution to work classic "bgp route 
> reflectors" would not do the right job.

You challenge the fact that BGP MVPN would not work with "classic" bgp 
route reflectors: why do it after the specs are long gone out of the 
working group ?  how does it relate to the current discussion ?

(supposing it is even relevant to claim that, and pending some 
definition of "classic"...)


> They should be replaced with "multicast control aggregation" devices 
> which would have to support it's own set of requirements and it's own 
> advertisement rules.


(Your claim is vague enough to not be even refutable, but:)
What is true is that the aggregation done by RR for mVPN following 
section 11.4 of draft-ietf-l3vpn-2547bis-mcast-bgp relies on mechanisms 
already defined in other specs.


> In fact draft already recommends to use such devices as a standalone 
> control plane boxes.


What in fact is written in the draft is this:

    In the higher scale scenarios, it may be
    required to adapt the route reflector infrastructure to the MVPN
    routing load by using, for example:

    o  a separation of resources for unicast and multicast VPN routing:
       using dedicated MVPN Route Reflector(s) (or using dedicated MVPN
       BGP sessions or dedicated MVPN BGP instances) ;

    o  the deployment of additional route reflector resources, for
       example increasing the processing resources on existing route
       reflectors or deployment of additional route reflectors.

    Among the above, the most straightforward approach is to consider the
    introduction of route reflectors dedicated to the MVPN service and
    dimension them accordingly to the need of that service (but doing so
    is not required and is left as an operator engineering decision).


So :
- this text refers to the use of RRs following 
draft-ietf-l3vpn-2547bis-mcast-bgp, but *not* to anything like 
"multicast control aggregation devices" which, as you say, would "have 
to support it's own set of requirements and it's own advertisement rules"
- the reasons for suggesting the use of RRs or dedicated RRs are purely 
*quantitative* (scale) and *not qualitative* as you suggest.

Cheers,

-Thomas




>> Thomas,
>>
>> regarding my question: "In section A.1.1.3, why the 'each additional 
>> PE joining' row for the 'upstream PE' column shows "0"?", you pointed 
>> me to this e-mail.
>> Please see my replies in-line.
>>
>> Maria
>>
>>
>>> Now, to go back on the computation in  section A.1.1.3 on BGP, you are
>>> correct that the current text does not yet properly reflects the cases
>>> where an RR may re-advertises a route they were already advertising.
>>> Let's look practically at the attributes that could vary :
>>>
>>>    - next-hop : updates of the value of this attribute do not "follow"
>>> PE advertisements, being rewritten by the RR when aggregating
>>> C-multicast routes having same NLRI (section 11.4 of
>>> draft-ietf-l3vpn-2547bis-mcast-bgp) ; the advertisements made by
>>> different RRs may lead to updates of the value of this attribute
>>
>> This does not help in route aggregation because RR applies the 
>> ORIGINATOR_ID to each client's/egress PE route.
>>
>>>    - ORIGINATOR_ID : updates of the value of this attributes do not
>>> "follow" PE advertisements, as the RR rewrite it when aggregating C-
>>> multicast routes having same NLRI; the advertisements made by different
>>> RRs may lead possibly to updates of the value of this attribute, given
>>> that each RR would set ORIGINATOR_ID to a value of its own
>>>
>>
>> Regarding route *aggregation* the 
>> draft-ietf-l3vpn-2547bis-mcast-bgp-08 only states the following:
>>
>>    "To further reduce the routing churn due to C-multicast routes 
>> changes
>>    a Route Reflector that re-advertises a C-multicast route SHOULD set
>>    the Next Hop field of the MP_REACH_NLRI attribute of the route to an
>>    IP address of the Route Reflector. Likewise, an ASBR that re-
>>    advertises a C-multicast route SHOULD set the Next Hop field of the
>>    MP_REACH_NLRI attribute of the route to an IP address of the ASBR."
>>
>> It states nothing about setting ORIGINATOR_ID to self on RR.
>> Maria
>

From raszuk@cisco.com  Fri Dec 18 13:12:36 2009
Return-Path: <raszuk@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 4E04C3A67F8 for <l3vpn@core3.amsl.com>; Fri, 18 Dec 2009 13:12:36 -0800 (PST)
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 ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id SWEYhEyy4XmZ for <l3vpn@core3.amsl.com>; Fri, 18 Dec 2009 13:12:35 -0800 (PST)
Received: from sj-iport-5.cisco.com (sj-iport-5.cisco.com [171.68.10.87]) by core3.amsl.com (Postfix) with ESMTP id 3483C3A6AB7 for <l3vpn@ietf.org>; Fri, 18 Dec 2009 13:12:24 -0800 (PST)
Authentication-Results: sj-iport-5.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AvsFALJ+K0urRN+K/2dsb2JhbACECpR/pWuBHAgBhVuQHYEvgi1SBA
X-IronPort-AV: E=Sophos;i="4.47,420,1257120000"; d="scan'208";a="122191844"
Received: from sj-core-4.cisco.com ([171.68.223.138]) by sj-iport-5.cisco.com with ESMTP; 18 Dec 2009 21:12:09 +0000
Received: from [192.168.1.57] (sjc-raszuk-87113.cisco.com [10.20.147.254]) by sj-core-4.cisco.com (8.13.8/8.14.3) with ESMTP id nBILC8UQ006431; Fri, 18 Dec 2009 21:12:08 GMT
Message-ID: <4B2BF022.8020505@cisco.com>
Date: Fri, 18 Dec 2009 22:12:02 +0100
From: Robert Raszuk <raszuk@cisco.com>
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
To: Thomas Morin <thomas.morin@orange-ftgroup.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A	/	sectionA.1.1.3
References: <3126.1259858356@erosen-linux>	<4B28E84F.4000604@orange-ftgroup.com>	<2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com> <4B2BEAF5.4020609@orange-ftgroup.com>
In-Reply-To: <4B2BEAF5.4020609@orange-ftgroup.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Cc: L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: raszuk@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 18 Dec 2009 21:12:36 -0000

Hi Thomas,

>      [...] C-multicast routes for a given (S,G) of a given MVPN 
> originated by PEs that are clients of
>      a given Route Reflector are aggregated by the Route Reflector. [...]
> 
> A BGP aggregated route is a route originated by the BGP peer doing the 
> aggregation, why would it carry another ORIGINATOR_ID than the one of 
> the BGP peer doing the aggregation (here the RR) ?

Because RR behavior is very well already defined in RFC 4456  which says:

    ORIGINATOR_ID is a new optional, non-transitive BGP attribute of Type
    code 9.  This attribute is 4 bytes long and it will be created by an
    RR in reflecting a route.  This attribute will carry the BGP
    Identifier of the originator of the route in the local AS.

Here RR is doing _aggregation_ and not _origination_ of the route. Sorry 
but this is not the same.

C-multicast routes are originated by the PEs and to be compliant to RFC 
4456 each time best path shifts from one PE to the other the RR should 
reflect it to all of it's clients with new corresponding to the new best 
path ORIGINATOR_ID.

Cheers,
R.


From raszuk@cisco.com  Fri Dec 18 13:19:55 2009
Return-Path: <raszuk@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 088973A67A8 for <l3vpn@core3.amsl.com>; Fri, 18 Dec 2009 13:19:55 -0800 (PST)
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 ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id pU5G11CLsNBA for <l3vpn@core3.amsl.com>; Fri, 18 Dec 2009 13:19:53 -0800 (PST)
Received: from sj-iport-5.cisco.com (sj-iport-5.cisco.com [171.68.10.87]) by core3.amsl.com (Postfix) with ESMTP id B78AC3A68C2 for <l3vpn@ietf.org>; Fri, 18 Dec 2009 13:19:52 -0800 (PST)
Authentication-Results: sj-iport-5.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AvsFAAuBK0urRN+J/2dsb2JhbACECpR/pXaBHAgBhVuQIYEvgi1SBA
X-IronPort-AV: E=Sophos;i="4.47,420,1257120000"; d="scan'208";a="122193926"
Received: from sj-core-3.cisco.com ([171.68.223.137]) by sj-iport-5.cisco.com with ESMTP; 18 Dec 2009 21:19:38 +0000
Received: from [192.168.1.57] (sjc-raszuk-87113.cisco.com [10.20.147.254]) by sj-core-3.cisco.com (8.13.8/8.14.3) with ESMTP id nBILJaAA007484; Fri, 18 Dec 2009 21:19:37 GMT
Message-ID: <4B2BF1E2.3070102@cisco.com>
Date: Fri, 18 Dec 2009 22:19:30 +0100
From: Robert Raszuk <raszuk@cisco.com>
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
To: Thomas Morin <thomas.morin@orange-ftgroup.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /	sectionA.1.1.3
References: <3126.1259858356@erosen-linux>	<4B28E84F.4000604@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com> <4B296D1C.6000001@cisco.com> <4B2BEDDF.3090308@orange-ftgroup.com>
In-Reply-To: <4B2BEDDF.3090308@orange-ftgroup.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Cc: L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: raszuk@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 18 Dec 2009 21:19:55 -0000

Thomas,

My comments were concerning scaling the solution to be able to support 
real network's demands. Not some theoretical ones.

And the proposed solution of using RRs as aggregators of c-multicast 
routes is not what RR RFC 4456 specifies for reflection rules.

Cheers,
R.


> Hello Robert,
> 
> Robert Raszuk:
>  >
>> It seems that for this mvpn solution to work classic "bgp route 
>> reflectors" would not do the right job.
> 
> You challenge the fact that BGP MVPN would not work with "classic" bgp 
> route reflectors: why do it after the specs are long gone out of the 
> working group ?  how does it relate to the current discussion ?
> 
> (supposing it is even relevant to claim that, and pending some 
> definition of "classic"...)
> 
> 
>> They should be replaced with "multicast control aggregation" devices 
>> which would have to support it's own set of requirements and it's own 
>> advertisement rules.
> 
> 
> (Your claim is vague enough to not be even refutable, but:)
> What is true is that the aggregation done by RR for mVPN following 
> section 11.4 of draft-ietf-l3vpn-2547bis-mcast-bgp relies on mechanisms 
> already defined in other specs.
> 
> 
>> In fact draft already recommends to use such devices as a standalone 
>> control plane boxes.
> 
> 
> What in fact is written in the draft is this:
> 
>    In the higher scale scenarios, it may be
>    required to adapt the route reflector infrastructure to the MVPN
>    routing load by using, for example:
> 
>    o  a separation of resources for unicast and multicast VPN routing:
>       using dedicated MVPN Route Reflector(s) (or using dedicated MVPN
>       BGP sessions or dedicated MVPN BGP instances) ;
> 
>    o  the deployment of additional route reflector resources, for
>       example increasing the processing resources on existing route
>       reflectors or deployment of additional route reflectors.
> 
>    Among the above, the most straightforward approach is to consider the
>    introduction of route reflectors dedicated to the MVPN service and
>    dimension them accordingly to the need of that service (but doing so
>    is not required and is left as an operator engineering decision).
> 
> 
> So :
> - this text refers to the use of RRs following 
> draft-ietf-l3vpn-2547bis-mcast-bgp, but *not* to anything like 
> "multicast control aggregation devices" which, as you say, would "have 
> to support it's own set of requirements and it's own advertisement rules"
> - the reasons for suggesting the use of RRs or dedicated RRs are purely 
> *quantitative* (scale) and *not qualitative* as you suggest.
> 
> Cheers,
> 
> -Thomas
> 
> 
> 
> 
>>> Thomas,
>>>
>>> regarding my question: "In section A.1.1.3, why the 'each additional 
>>> PE joining' row for the 'upstream PE' column shows "0"?", you pointed 
>>> me to this e-mail.
>>> Please see my replies in-line.
>>>
>>> Maria
>>>
>>>
>>>> Now, to go back on the computation in  section A.1.1.3 on BGP, you are
>>>> correct that the current text does not yet properly reflects the cases
>>>> where an RR may re-advertises a route they were already advertising.
>>>> Let's look practically at the attributes that could vary :
>>>>
>>>>    - next-hop : updates of the value of this attribute do not "follow"
>>>> PE advertisements, being rewritten by the RR when aggregating
>>>> C-multicast routes having same NLRI (section 11.4 of
>>>> draft-ietf-l3vpn-2547bis-mcast-bgp) ; the advertisements made by
>>>> different RRs may lead to updates of the value of this attribute
>>>
>>> This does not help in route aggregation because RR applies the 
>>> ORIGINATOR_ID to each client's/egress PE route.
>>>
>>>>    - ORIGINATOR_ID : updates of the value of this attributes do not
>>>> "follow" PE advertisements, as the RR rewrite it when aggregating C-
>>>> multicast routes having same NLRI; the advertisements made by different
>>>> RRs may lead possibly to updates of the value of this attribute, given
>>>> that each RR would set ORIGINATOR_ID to a value of its own
>>>>
>>>
>>> Regarding route *aggregation* the 
>>> draft-ietf-l3vpn-2547bis-mcast-bgp-08 only states the following:
>>>
>>>    "To further reduce the routing churn due to C-multicast routes 
>>> changes
>>>    a Route Reflector that re-advertises a C-multicast route SHOULD set
>>>    the Next Hop field of the MP_REACH_NLRI attribute of the route to an
>>>    IP address of the Route Reflector. Likewise, an ASBR that re-
>>>    advertises a C-multicast route SHOULD set the Next Hop field of the
>>>    MP_REACH_NLRI attribute of the route to an IP address of the ASBR."
>>>
>>> It states nothing about setting ORIGINATOR_ID to self on RR.
>>> Maria
>>
> 


From yakov@juniper.net  Mon Dec 21 06:17:10 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 9423E3A6A0B for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 06:17:10 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.539
X-Spam-Level: 
X-Spam-Status: No, score=-6.539 tagged_above=-999 required=5 tests=[AWL=0.060,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id XQACSMzsWrcC for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 06:17:09 -0800 (PST)
Received: from exprod7og110.obsmtp.com (exprod7og110.obsmtp.com [64.18.2.173]) by core3.amsl.com (Postfix) with SMTP id 37E4B3A67D9 for <l3vpn@ietf.org>; Mon, 21 Dec 2009 06:17:08 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob110.postini.com ([64.18.6.12]) with SMTP ID DSNKSy+DUpxkqLzTIcFAlPM0OHJC2124g6PD@postini.com; Mon, 21 Dec 2009 06:16:53 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.71) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Mon, 21 Dec 2009 06:13:24 -0800
Received: from p-emlb02-sac.jnpr.net ([66.129.254.47]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 06:13:24 -0800
Received: from emailsmtp56.jnpr.net ([172.24.60.77]) by p-emlb02-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 06:13:24 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp56.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 06:13:24 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBLEDNj02373; Mon, 21 Dec 2009 06:13:23 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912211413.nBLEDNj02373@magenta.juniper.net>
To: Thomas Morin <thomas.morin@orange-ftgroup.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / looking at multiple (S, G) 
In-Reply-To: <4B2B9979.4070708@orange-ftgroup.com> 
References: <3126.1259858356@erosen-linux> <4B2B9979.4070708@orange-ftgroup.com>
X-MH-In-Reply-To: Thomas Morin <thomas.morin@orange-ftgroup.com> message dated "Fri, 18 Dec 2009 16:02:17 +0100."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <47821.1261404803.1@juniper.net>
Date: Mon, 21 Dec 2009 06:13:23 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 21 Dec 2009 14:13:24.0067 (UTC) FILETIME=[C1E35330:01CA8247]
Cc: Eric Rosen <erosen@cisco.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 14:17:10 -0000

Thomas,

> Eric, working group,
> 
> In different parts of the document (quoted below) you claim that the 
> scaling analysis should consider scalability with an increased number of 
> sources sending to a group.
> 
> Before looking at the relevance question, let's see how it relates to 
> the current state of the document.
> Section 3.3 of the document relates to scalability with an increased 
> number of PEs, and refers to the scaling analysis in Appendix A.1 
> "Scalability with an increased number of PEs". Given this, claiming that 
> not doing a scaling analysis with an increased number of (S,G) is "a 
> methodological error" in this scaling analysis is not correct.
> 
> That "scaling with an increased number of (S,G)" should be studied is 
> something that noone has proposed to do yet, as far as I can tell.
> 
> Now, that said, let's try to see how interesting it would be to consider 
> looking more at this aspect:
> - in your comments, you insist on the linear dependency of BGP 
> C-multicast routing with the number of (S,G), but scaling linearly with 
> the number of sources sending to a group is not a peculiarity of the BGP 
> Source Active auto-discovery mechanism, but is also common for PIM-SM 
> SPTs and SSM trees (whether they are carried across the provider core 
> through PIM and BGP, by the way)
> - "the number of sources sending to a group", or the number of SPTs or 
> SSM trees, is not something typically under the control of the provider, 
> as it completely depends on the multicast routing policy and 
> subscriptions patterns of each MVPN customer; so it should be obvious to 
> anyone that the provider will have to prepare its network to scale with 
> these unknowns in mind and put bounds to the amount of per-source state 
> on his PEs (and RRs) ; this is true independently of whether PIM-SM or 
> BGP is used for C-multicast routing (and true whether or not MSDP is 
> used in that deployment)

Just to add to the point above, the following from Security
Considerations section of draft-ietf-l3vpn-2547bis-mcast-09.txt:

   In particular, an implementation SHOULD provide mechanisms that allow
   a SP to place limitations on the following:

     - total number of (C-*,C-G) and/or (C-S,C-G) states per VRF

And as you mentioned above, "this is true independently of whether
PIM-SM or BGP is used for C-multicast routing (and true whether or
not MSDP is used in that deployment)"

Yakov.

From yakov@juniper.net  Mon Dec 21 06:20:47 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id BACDA3A6A0F for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 06:20:47 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.542
X-Spam-Level: 
X-Spam-Status: No, score=-6.542 tagged_above=-999 required=5 tests=[AWL=0.057,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id uVz14whfGhps for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 06:20:46 -0800 (PST)
Received: from exprod7og101.obsmtp.com (exprod7og101.obsmtp.com [64.18.2.155]) by core3.amsl.com (Postfix) with SMTP id 78D1B3A6A0E for <l3vpn@ietf.org>; Mon, 21 Dec 2009 06:20:45 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob101.postini.com ([64.18.6.12]) with SMTP ID DSNKSy+EJ5o+ImSV499taXKkWj5SUlvrNCvX@postini.com; Mon, 21 Dec 2009 06:20:31 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Mon, 21 Dec 2009 06:16:26 -0800
Received: from p-emlb02-sac.jnpr.net ([66.129.254.47]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 06:16:26 -0800
Received: from emailsmtp56.jnpr.net ([172.24.60.77]) by p-emlb02-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 06:16:25 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp56.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 06:16:25 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBLEGPj03527; Mon, 21 Dec 2009 06:16:25 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912211416.nBLEGPj03527@magenta.juniper.net>
To: raszuk@cisco.com
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / sectionA.1.1.3 
In-Reply-To: <4B2BF022.8020505@cisco.com> 
References: <3126.1259858356@erosen-linux> <4B28E84F.4000604@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com> <4B2BEAF5.4020609@orange-ftgroup.com> <4B2BF022.8020505@cisco.com>
X-MH-In-Reply-To: Robert Raszuk <raszuk@cisco.com> message dated "Fri, 18 Dec 2009 22:12:02 +0100."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <48208.1261404985.1@juniper.net>
Date: Mon, 21 Dec 2009 06:16:25 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 21 Dec 2009 14:16:25.0270 (UTC) FILETIME=[2DE4B560:01CA8248]
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 14:20:47 -0000

Robert,

> Hi Thomas,
> 
> >      [...] C-multicast routes for a given (S,G) of a given MVPN 
> > originated by PEs that are clients of
> >      a given Route Reflector are aggregated by the Route Reflector. [...]
> > 
> > A BGP aggregated route is a route originated by the BGP peer doing the 
> > aggregation, why would it carry another ORIGINATOR_ID than the one of 
> > the BGP peer doing the aggregation (here the RR) ?
> 
> Because RR behavior is very well already defined in RFC 4456  which says:
> 
>     ORIGINATOR_ID is a new optional, non-transitive BGP attribute of Type
>     code 9.  This attribute is 4 bytes long and it will be created by an
>     RR in reflecting a route.  This attribute will carry the BGP
>     Identifier of the originator of the route in the local AS.
> 
> Here RR is doing _aggregation_ and not _origination_ of the route. Sorry 
> but this is not the same.

As the RR performs the aggregation, it is *the RR* that originates
the *aggregated route*.

Yakov.

From raszuk@cisco.com  Mon Dec 21 06:55:17 2009
Return-Path: <raszuk@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 95FD33A69E7 for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 06:55:17 -0800 (PST)
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 ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id XG47isAPAVsi for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 06:55:16 -0800 (PST)
Received: from sj-iport-4.cisco.com (sj-iport-4.cisco.com [171.68.10.86]) by core3.amsl.com (Postfix) with ESMTP id CB9643A6825 for <l3vpn@ietf.org>; Mon, 21 Dec 2009 06:55:16 -0800 (PST)
Authentication-Results: sj-iport-4.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AmUGAJsaL0urRN+K/2dsb2JhbACZAKdOgRwIAZN/hC4E
X-IronPort-AV: E=Sophos;i="4.47,431,1257120000"; d="scan'208";a="65817791"
Received: from sj-core-4.cisco.com ([171.68.223.138]) by sj-iport-4.cisco.com with ESMTP; 21 Dec 2009 14:55:01 +0000
Received: from [192.168.1.57] (sjc-raszuk-87113.cisco.com [10.20.147.254]) by sj-core-4.cisco.com (8.13.8/8.14.3) with ESMTP id nBLEsxWg015190; Mon, 21 Dec 2009 14:55:00 GMT
Message-ID: <4B2F8C43.2000500@cisco.com>
Date: Mon, 21 Dec 2009 15:54:59 +0100
From: Robert Raszuk <raszuk@cisco.com>
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
To: Yakov Rekhter <yakov@juniper.net>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / sectionA.1.1.3
References: <3126.1259858356@erosen-linux> <4B28E84F.4000604@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com> <4B2BEAF5.4020609@orange-ftgroup.com> <4B2BF022.8020505@cisco.com> <200912211416.nBLEGPj03527@magenta.juniper.net>
In-Reply-To: <200912211416.nBLEGPj03527@magenta.juniper.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: raszuk@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 14:55:17 -0000

Yakov,

The spec started with just talking about setting the next hop to self on 
the RR, then is was discovered that this is not sufficient and 
ORIGINATOR_ID was added.

Now we have a case where this is not really a route reflector as 
definition of reflection means to receive a path and *reflect* such path 
to the clients/non-clients.

Now this "RR" is an originator of the C-multicast routes ... I think it 
would be much better to call it correctly ... as already proposed to 
call it "C-multicast route aggregator" or perhaps just "C-multicast 
route proxy" - in any case it is clear that IBGP aggregation and route 
origination by a control plane only RR is a new network function - not 
that common in IP or even in VPN networks.

Now regardless of the name where can one find a definition of such IBGP 
aggregation ?

As an example: should such "IBGP aggregation device" remove local 
preference if it is send by the PEs along with c-multicast routes ? If 
not having two such IBGP aggregation devices which is likely to be the 
practical case may still generate unnecessary churn on all PEs involved.

Robert.



> Robert,
> 
>> Hi Thomas,
>>
>>>      [...] C-multicast routes for a given (S,G) of a given MVPN 
>>> originated by PEs that are clients of
>>>      a given Route Reflector are aggregated by the Route Reflector. [...]
>>>
>>> A BGP aggregated route is a route originated by the BGP peer doing the 
>>> aggregation, why would it carry another ORIGINATOR_ID than the one of 
>>> the BGP peer doing the aggregation (here the RR) ?
>> Because RR behavior is very well already defined in RFC 4456  which says:
>>
>>     ORIGINATOR_ID is a new optional, non-transitive BGP attribute of Type
>>     code 9.  This attribute is 4 bytes long and it will be created by an
>>     RR in reflecting a route.  This attribute will carry the BGP
>>     Identifier of the originator of the route in the local AS.
>>
>> Here RR is doing _aggregation_ and not _origination_ of the route. Sorry 
>> but this is not the same.
> 
> As the RR performs the aggregation, it is *the RR* that originates
> the *aggregated route*.
> 
> Yakov.
> 


From yakov@juniper.net  Mon Dec 21 07:08:42 2009
Return-Path: <yakov@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 699563A693E for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 07:08:42 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.544
X-Spam-Level: 
X-Spam-Status: No, score=-6.544 tagged_above=-999 required=5 tests=[AWL=0.055,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id PmRgSTz3-35e for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 07:08:41 -0800 (PST)
Received: from exprod7og105.obsmtp.com (exprod7og105.obsmtp.com [64.18.2.163]) by core3.amsl.com (Postfix) with SMTP id C70C93A68A4 for <l3vpn@ietf.org>; Mon, 21 Dec 2009 07:08:39 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob105.postini.com ([64.18.6.12]) with SMTP ID DSNKSy+PZ5aw5Z+toCvwwd6ckwP7TuI+v3O3@postini.com; Mon, 21 Dec 2009 07:08:25 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Mon, 21 Dec 2009 07:05:39 -0800
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 07:05:39 -0800
Received: from emailsmtp56.jnpr.net ([172.24.60.77]) by p-emlb01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 07:05:39 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp56.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 07:05:39 -0800
Received: from juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBLF5dj20847; Mon, 21 Dec 2009 07:05:39 -0800 (PST)	(envelope-from yakov@juniper.net)
Message-ID: <200912211505.nBLF5dj20847@magenta.juniper.net>
To: raszuk@cisco.com
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / sectionA.1.1.3 
In-Reply-To: <4B296D1C.6000001@cisco.com> 
References: <3126.1259858356@erosen-linux> <4B28E84F.4000604@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com> <4B296D1C.6000001@cisco.com>
X-MH-In-Reply-To: Robert Raszuk <raszuk@cisco.com> message dated "Thu, 17 Dec 2009 00:28:28 +0100."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <52973.1261407939.1@juniper.net>
Date: Mon, 21 Dec 2009 07:05:39 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-OriginalArrivalTime: 21 Dec 2009 15:05:39.0201 (UTC) FILETIME=[0E92CF10:01CA824F]
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 15:08:42 -0000

Robert,

> Therefor before proceeding in this WG I would like to either see all 
> references to BGP route reflectors use removed. Alternatively they could 
> be replaced with the new term describing their newly defined and 
> apparently required functionality for any form of sensible operation of 
> the proposed herein solution.

Let me remind you that modifying the ORIGINATOR_ID and Next_hop
attributes by Route Reflectors (by setting up the ORIGINATOR_ID
attribute to self) is by no means new to C-multicast routing. From
rfc4684, section 3.2:

  i.   When advertising RT membership NLRI to a route-reflector client,
       the Originator attribute shall be set to the router-id of the
       advertiser, and the Next-hop attribute shall be set of the local
       address for that session.

Interestingly enough at the time this rfc was published, you did
*not* raise any objections/concerns to Route Reflectors modifying
the ORIGINATOR_ID and the Next_Hop attributes (by setting both of
them to self).

Moreover, at that time you did *not* request that in rfc4684 "all
references to BGP route reflectors use removed". You did *not*
request to replace all references to BGP route reflectors in that
rfc "with the new term".

All this is especially interesting since you are one of the co-authors
of rfc4684.

Yakov.

From raszuk@cisco.com  Mon Dec 21 07:20:37 2009
Return-Path: <raszuk@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id B53CF3A6A1C for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 07:20:37 -0800 (PST)
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 ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 5xO9lkf27Yvg for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 07:20:36 -0800 (PST)
Received: from sj-iport-4.cisco.com (sj-iport-4.cisco.com [171.68.10.86]) by core3.amsl.com (Postfix) with ESMTP id 8B5363A6969 for <l3vpn@ietf.org>; Mon, 21 Dec 2009 07:20:36 -0800 (PST)
Authentication-Results: sj-iport-4.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AmUGAHcgL0urRN+J/2dsb2JhbACZAKdcgRwIAZQBhC4E
X-IronPort-AV: E=Sophos;i="4.47,431,1257120000"; d="scan'208";a="65834567"
Received: from sj-core-3.cisco.com ([171.68.223.137]) by sj-iport-4.cisco.com with ESMTP; 21 Dec 2009 15:20:13 +0000
Received: from [192.168.1.57] (sjc-raszuk-87113.cisco.com [10.20.147.254]) by sj-core-3.cisco.com (8.13.8/8.14.3) with ESMTP id nBLFKBJN007386; Mon, 21 Dec 2009 15:20:12 GMT
Message-ID: <4B2F922A.9070307@cisco.com>
Date: Mon, 21 Dec 2009 16:20:10 +0100
From: Robert Raszuk <raszuk@cisco.com>
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
To: Yakov Rekhter <yakov@juniper.net>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / sectionA.1.1.3
References: <3126.1259858356@erosen-linux> <4B28E84F.4000604@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com> <4B296D1C.6000001@cisco.com> <200912211505.nBLF5dj20847@magenta.juniper.net>
In-Reply-To: <200912211505.nBLF5dj20847@magenta.juniper.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: raszuk@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 15:20:37 -0000

Yakov,

The goals of overwriting Originator_ID by the RR in RFC4684 are totally 
different then those described in mvpn spec.

To quote:

    The first of these route advertisement rules is designed such that
    the originator of an RT membership NLRI does not drop an RT
    membership NLRI that is reflected back to it, thus allowing the route
    reflector to use this RT membership NLRI in order to signal the
    client that it should distribute VPN routes with the specific target
    towards the reflector.

So it has no bearing on limiting the churn of RTs ... contrary it makes 
sure that reflected NLRI is just not dropped.

On the other hand in current mvpn draft the goal is to stabilize the 
control plane behavior to make it comparable with alternative solutions 
to the same application space.

Cheers,
R.


> Robert,
> 
>> Therefor before proceeding in this WG I would like to either see all 
>> references to BGP route reflectors use removed. Alternatively they could 
>> be replaced with the new term describing their newly defined and 
>> apparently required functionality for any form of sensible operation of 
>> the proposed herein solution.
> 
> Let me remind you that modifying the ORIGINATOR_ID and Next_hop
> attributes by Route Reflectors (by setting up the ORIGINATOR_ID
> attribute to self) is by no means new to C-multicast routing. From
> rfc4684, section 3.2:
> 
>   i.   When advertising RT membership NLRI to a route-reflector client,
>        the Originator attribute shall be set to the router-id of the
>        advertiser, and the Next-hop attribute shall be set of the local
>        address for that session.
> 
> Interestingly enough at the time this rfc was published, you did
> *not* raise any objections/concerns to Route Reflectors modifying
> the ORIGINATOR_ID and the Next_Hop attributes (by setting both of
> them to self).
> 
> Moreover, at that time you did *not* request that in rfc4684 "all
> references to BGP route reflectors use removed". You did *not*
> request to replace all references to BGP route reflectors in that
> rfc "with the new term".
> 
> All this is especially interesting since you are one of the co-authors
> of rfc4684.
> 
> Yakov.


From wwwrun@core3.amsl.com  Mon Dec 21 09:28:49 2009
Return-Path: <wwwrun@core3.amsl.com>
X-Original-To: l3vpn@ietf.org
Delivered-To: l3vpn@core3.amsl.com
Received: by core3.amsl.com (Postfix, from userid 30) id 625393A6A64; Mon, 21 Dec 2009 09:28:49 -0800 (PST)
X-idtracker: yes
From: The IESG <iesg-secretary@ietf.org>
To: IETF-Announce <ietf-announce@ietf.org>
Subject: Protocol Action: 'OSPFv3 as a PE-CE routing protocol' to Proposed Standard
Message-Id: <20091221172849.625393A6A64@core3.amsl.com>
Date: Mon, 21 Dec 2009 09:28:49 -0800 (PST)
Cc: l3vpn chair <l3vpn-chairs@tools.ietf.org>, Internet Architecture Board <iab@iab.org>, l3vpn mailing list <l3vpn@ietf.org>, RFC Editor <rfc-editor@rfc-editor.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 17:28:49 -0000

The IESG has approved the following document:

- 'OSPFv3 as a PE-CE routing protocol '
   <draft-ietf-l3vpn-ospfv3-pece-04.txt> as a Proposed Standard


This document is the product of the Layer 3 Virtual Private Networks Working Group. 

The IESG contact persons are Ross Callon and Adrian Farrel.

A URL of this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-l3vpn-ospfv3-pece-04.txt

Technical Summary

   Many Service Providers (SPs) offer Virtual Private Network (VPN)
   services to their customers using a technique in which Customer Edge
   (CE) routers are routing peers of Provider Edge (PE) routers.  The
   Border Gateway Protocol (BGP) is used to distribute the customer's
   routes across the provider's IP backbone network, and Multiprotocol
   Label Switching (MPLS) is used to tunnel customer packets across the
   provider's backbone.  This is known as a "BGP/MPLS IP VPN".
   Originally only IPv4 was supported and it was later extended to
   support IPv6 VPNs as well.  Extensions were later added for the
   support of the Open Shortest Path First protocol version 2 (OSPFv2)
   as a PE-CE routing protocol for the IPv4 VPNs.  This document extends
   those specifications to support OSPF version 3 (OSPFv3) as a PE-CE
   routing protocol.  The OSPFv3 PE-CE functionality is very similar to
   the previously-specified approached for OSPFv2.

Working Group Summary

   No controversy reported (see PROTO writeup by Danny McPherson in the 
   ID Tracker). There were no objections during the WG Last Call on the
   document, and the utility of this specification is straight-forward,
   as well as both obvious and intuitive.

Document Quality

   No known implementations. 

Personnel

   Danny McPherson is the Document Shepherd for this document. Ross
   Callon is the Responsible Area Director.

RFC Editor Note

   All references to [BGP-EXTCOMM-IPV6] should be replaced by a 
   references to [RFC5701].

   In the informative references section (11.2), please remove the
   reference to [BGP-EXTCOMM-IPV6].

   Please add a reference to RFC5701 in the normative references 
   section (11.1).


From mn1921@att.com  Mon Dec 21 10:17:08 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id AAE6C3A6A9F for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 10:17:08 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -106.416
X-Spam-Level: 
X-Spam-Status: No, score=-106.416 tagged_above=-999 required=5 tests=[AWL=0.183, BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id UkrV0iKzG-Zv for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 10:17:07 -0800 (PST)
Received: from mail121.messagelabs.com (mail121.messagelabs.com [216.82.242.3]) by core3.amsl.com (Postfix) with ESMTP id AF39028C121 for <l3vpn@ietf.org>; Mon, 21 Dec 2009 10:16:57 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-3.tower-121.messagelabs.com!1261419400!35833223!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 29172 invoked from network); 21 Dec 2009 18:16:40 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-3.tower-121.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 21 Dec 2009 18:16:40 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBLIGbbD003648 for <l3vpn@ietf.org>; Mon, 21 Dec 2009 13:16:37 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBLIGaUR003641 for <l3vpn@ietf.org>; Mon, 21 Dec 2009 13:16:36 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /	sectionA.1.1.3
Date: Mon, 21 Dec 2009 13:16:35 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A2004580452@misout7msgusr7e.ugd.att.com>
In-Reply-To: <4B2BEAF5.4020609@orange-ftgroup.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /	sectionA.1.1.3
Thread-Index: AcqAI7iCPqzZ5H9AQ6ulRjiQRusvtACRXDqw
References: <3126.1259858356@erosen-linux>	<4B28E84F.4000604@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com> <4B2BEAF5.4020609@orange-ftgroup.com>
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: "Thomas Morin" <thomas.morin@orange-ftgroup.com>, "L3VPN" <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 18:17:08 -0000
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==

From rahul@juniper.net  Mon Dec 21 10:59:24 2009
Return-Path: <rahul@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id BB7C628C129 for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 10:59:24 -0800 (PST)
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 ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id T42gY2m42Pec for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 10:59:23 -0800 (PST)
Received: from exprod7og101.obsmtp.com (exprod7og101.obsmtp.com [64.18.2.155]) by core3.amsl.com (Postfix) with SMTP id B3EF43A68F4 for <l3vpn@ietf.org>; Mon, 21 Dec 2009 10:59:21 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob101.postini.com ([64.18.6.12]) with SMTP ID DSNKSy/FeftJvQ8qQuxEkQEYeLgBoSbr+VzK@postini.com; Mon, 21 Dec 2009 10:59:07 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Mon, 21 Dec 2009 10:54:29 -0800
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 10:54:29 -0800
Received: from emailsmtp55.jnpr.net ([172.24.18.132]) by p-emlb01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 10:54:28 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp55.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 10:54:28 -0800
Received: from sapphire.juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBLIsSj07165; Mon, 21 Dec 2009 10:54:28 -0800 (PST)	(envelope-from rahul@juniper.net)
Date: Mon, 21 Dec 2009 10:54:28 -0800
From: Rahul Aggarwal <rahul@juniper.net>
To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / sectionA.1.1.3
In-Reply-To: <2F1DE4DFCFF32144B771BD2C246E6A2004580452@misout7msgusr7e.ugd.att.com>
Message-ID: <20091221104548.N69632@sapphire.juniper.net>
References: <3126.1259858356@erosen-linux> <4B28E84F.4000604@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com> <4B2BEAF5.4020609@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A2004580452@misout7msgusr7e.ugd.att.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: QUOTED-PRINTABLE
X-OriginalArrivalTime: 21 Dec 2009 18:54:28.0708 (UTC) FILETIME=[05FF5240:01CA826F]
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 18:59:24 -0000

Maria,

<snipped>

On Mon, 21 Dec 2009, NAPIERALA, MARIA H (ATTLABS) wrote:

> Thomas,
>
> >
> > Maria,
> >
> > NAPIERALA, MARIA H (ATTLABS):
> > >>     - ORIGINATOR_ID : updates of the value of this attributes do not
> > >> "follow" PE advertisements, as the RR rewrite it when aggregating C-
> > >> multicast routes having same NLRI; the advertisements made by
> > different
> > >> RRs may lead possibly to updates of the value of this attribute,
> > given
> > >> that each RR would set ORIGINATOR_ID to a value of its own
> > >>
> >  >
> > > Regarding route *aggregation* the draft-ietf-l3vpn-2547bis-mcast-bgp-
> > 08 only states the following:
> > >
> > >     "To further reduce the routing churn due to C-multicast routes
> > changes
> > >     a Route Reflector that re-advertises a C-multicast route SHOULD
> > set
> > >     the Next Hop field of the MP_REACH_NLRI attribute of the route to
> > an
> > >     IP address of the Route Reflector. Likewise, an ASBR that re-
> > >     advertises a C-multicast route SHOULD set the Next Hop field of
> > the
> > >     MP_REACH_NLRI attribute of the route to an IP address of the
> > ASBR."
> > >
> > > It states nothing about setting ORIGINATOR_ID to self on RR.
> >
> > It also says:
> >
> >       [...] C-multicast routes for a given (S,G) of a given MVPN
> > originated by PEs that are clients of
> >       a given Route Reflector are aggregated by the Route Reflector.
> > [...]
> >
>
> You omitted very important text from the same section 11.4 that precedes =
the one you quoted, namely:
>
>   "Note that C-multicast routes are "de facto" aggregated by BGP. This
>    is because the MCAST-VPN NLRIs advertised by multiple PEs, for a C-
>    multicast route for a particular C-S and C-G (or a particular C-* and
>    C-G) of a given MVPN are identical."
>
> This is the route aggregation the text you quoted refers to.
>
> Regarding the routing *churn*, section 11.4 says further down that:
>
>   =E2=80=9DTo further reduce the routing churn due to C-multicast routes =
changes
>    a Route Reflector that re-advertises a C-multicast route SHOULD set
>    the Next Hop field of the MP_REACH_NLRI attribute of the route to an
>    IP address of the Route Reflector."
>
> Which, in fact, doesn=E2=80=99t reduce routing churn.
>

Incorrect. The text you quoted *does* reduce routing churn due to
C-multicast route changes. This is because it ensures that a Route
Reflector advertises a C-multicast route, for a given (C-S/*, C-G) only
when it receives the first route for the (C-S/*. C-G) from a client. It
also ensures that a Route Reflector doesn't withdraw a C-multicast route
for a given (C-S/*, C-G) as long as it has at least one route from a
client for the (C-S/*, C-G).

rahul

From raszuk@cisco.com  Mon Dec 21 11:12:47 2009
Return-Path: <raszuk@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 1EAA53A67FA for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 11:12:47 -0800 (PST)
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=[AWL=0.000,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 4I8jdekLNED7 for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 11:12:36 -0800 (PST)
Received: from sj-iport-5.cisco.com (sj-iport-5.cisco.com [171.68.10.87]) by core3.amsl.com (Postfix) with ESMTP id 8B2323A6893 for <l3vpn@ietf.org>; Mon, 21 Dec 2009 11:12:35 -0800 (PST)
Authentication-Results: sj-iport-5.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AoQGACJXL0urRN+J/2dsb2JhbACECpRyqEyBHAgBhVuOTYEvgi1SBA
X-IronPort-AV: E=Sophos;i="4.47,432,1257120000"; d="scan'208";a="123355798"
Received: from sj-core-3.cisco.com ([171.68.223.137]) by sj-iport-5.cisco.com with ESMTP; 21 Dec 2009 19:11:17 +0000
Received: from [192.168.1.57] (sjc-raszuk-87113.cisco.com [10.20.147.254]) by sj-core-3.cisco.com (8.13.8/8.14.3) with ESMTP id nBLJBGpX017053; Mon, 21 Dec 2009 19:11:16 GMT
Message-ID: <4B2FC853.4050208@cisco.com>
Date: Mon, 21 Dec 2009 20:11:15 +0100
From: Robert Raszuk <raszuk@cisco.com>
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
To: Rahul Aggarwal <rahul@juniper.net>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /	sectionA.1.1.3
References: <3126.1259858356@erosen-linux>	<4B28E84F.4000604@orange-ftgroup.com>	<2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com>	<4B2BEAF5.4020609@orange-ftgroup.com>	<2F1DE4DFCFF32144B771BD2C246E6A2004580452@misout7msgusr7e.ugd.att.com> <20091221104548.N69632@sapphire.juniper.net>
In-Reply-To: <20091221104548.N69632@sapphire.juniper.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: raszuk@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 19:12:47 -0000

Rahul,

>> Regarding the routing *churn*, section 11.4 says further down that:
>>
>>   ”To further reduce the routing churn due to C-multicast routes changes
>>    a Route Reflector that re-advertises a C-multicast route SHOULD set
>>    the Next Hop field of the MP_REACH_NLRI attribute of the route to an
>>    IP address of the Route Reflector."
>>
>> Which, in fact, doesn’t reduce routing churn.
>>
> 
> Incorrect. The text you quoted *does* reduce routing churn due to
> C-multicast route changes. This is because it ensures that a Route
> Reflector advertises a C-multicast route, for a given (C-S/*, C-G) only
> when it receives the first route for the (C-S/*. C-G) from a client. It
> also ensures that a Route Reflector doesn't withdraw a C-multicast route
> for a given (C-S/*, C-G) as long as it has at least one route from a
> client for the (C-S/*, C-G).

Incorrect.

As Maria have said correctly *the* quoted text does not reduce routing 
churn.

The procedure you are spelling out are not reflected in the above 
paragraph. At least not in version 8 of the 
draft-ietf-l3vpn-2547bis-mcast-bgp.

Thx,
R.

From rahul@juniper.net  Mon Dec 21 11:23:58 2009
Return-Path: <rahul@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id D03823A692A for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 11:23:58 -0800 (PST)
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 ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id QXplLHAC3MCG for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 11:23:58 -0800 (PST)
Received: from exprod7og113.obsmtp.com (exprod7og113.obsmtp.com [64.18.2.179]) by core3.amsl.com (Postfix) with SMTP id B6BE83A6892 for <l3vpn@ietf.org>; Mon, 21 Dec 2009 11:23:56 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob113.postini.com ([64.18.6.12]) with SMTP ID DSNKSy/LPHoAptk9J6pbvQa3eDc6pOvtm/Ni@postini.com; Mon, 21 Dec 2009 11:23:42 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Mon, 21 Dec 2009 11:20:37 -0800
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 11:20:37 -0800
Received: from emailsmtp55.jnpr.net ([172.24.18.132]) by p-emlb01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 11:20:36 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp55.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Mon, 21 Dec 2009 11:20:36 -0800
Received: from sapphire.juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBLJKaj18380; Mon, 21 Dec 2009 11:20:36 -0800 (PST)	(envelope-from rahul@juniper.net)
Date: Mon, 21 Dec 2009 11:20:36 -0800
From: Rahul Aggarwal <rahul@juniper.net>
To: Robert Raszuk <raszuk@cisco.com>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A / sectionA.1.1.3
In-Reply-To: <4B2FC853.4050208@cisco.com>
Message-ID: <20091221111529.G69632@sapphire.juniper.net>
References: <3126.1259858356@erosen-linux> <4B28E84F.4000604@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com> <4B2BEAF5.4020609@orange-ftgroup.com> <2F1DE4DFCFF32144B771BD2C246E6A2004580452@misout7msgusr7e.ugd.att.com> <20091221104548.N69632@sapphire.juniper.net> <4B2FC853.4050208@cisco.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: QUOTED-PRINTABLE
X-OriginalArrivalTime: 21 Dec 2009 19:20:36.0895 (UTC) FILETIME=[ACB5AAF0:01CA8272]
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 19:23:58 -0000

Robert,

On Mon, 21 Dec 2009, Robert Raszuk wrote:

> Rahul,
>
> >> Regarding the routing *churn*, section 11.4 says further down that:
> >>
> >>   =E2=80=9DTo further reduce the routing churn due to C-multicast rout=
es changes
> >>    a Route Reflector that re-advertises a C-multicast route SHOULD set
> >>    the Next Hop field of the MP_REACH_NLRI attribute of the route to a=
n
> >>    IP address of the Route Reflector."
> >>
> >> Which, in fact, doesn=E2=80=99t reduce routing churn.
> >>
> >
> > Incorrect. The text you quoted *does* reduce routing churn due to
> > C-multicast route changes. This is because it ensures that a Route
> > Reflector advertises a C-multicast route, for a given (C-S/*, C-G) only
> > when it receives the first route for the (C-S/*. C-G) from a client. It
> > also ensures that a Route Reflector doesn't withdraw a C-multicast rout=
e
> > for a given (C-S/*, C-G) as long as it has at least one route from a
> > client for the (C-S/*, C-G).
>
> Incorrect.
>
> As Maria have said correctly *the* quoted text does not reduce routing
> churn.
>
> The procedure you are spelling out are not reflected in the above
> paragraph. At least not in version 8 of the
> draft-ietf-l3vpn-2547bis-mcast-bgp.
>

That is an incorrect interpretation of the text in section 11.4. The
procedure I spelled out follows from the text in section 11.4.


rahul

From mn1921@att.com  Mon Dec 21 11:52:24 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id E7AD33A681C for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 11:52:24 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -106.132
X-Spam-Level: 
X-Spam-Status: No, score=-106.132 tagged_above=-999 required=5 tests=[AWL=-0.133, BAYES_00=-2.599, J_CHICKENPOX_13=0.6, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id jGLSNrPRHtiE for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 11:52:24 -0800 (PST)
Received: from mail146.messagelabs.com (mail146.messagelabs.com [216.82.241.147]) by core3.amsl.com (Postfix) with ESMTP id B824F3A67FA for <l3vpn@ietf.org>; Mon, 21 Dec 2009 11:52:23 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-7.tower-146.messagelabs.com!1261425125!8773683!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 16333 invoked from network); 21 Dec 2009 19:52:05 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-7.tower-146.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 21 Dec 2009 19:52:05 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBLJq331008165; Mon, 21 Dec 2009 14:52:03 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBLJpx88008086; Mon, 21 Dec 2009 14:51:59 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /sectionA.1.1.3
Date: Mon, 21 Dec 2009 14:52:00 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A20045804CC@misout7msgusr7e.ugd.att.com>
In-Reply-To: <20091221111529.G69632@sapphire.juniper.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /sectionA.1.1.3
Thread-Index: AcqCcywZmgKKtOXCSM2nDxPARZbC5gAA3NDQ
References: <3126.1259858356@erosen-linux><4B28E84F.4000604@orange-ftgroup.com><2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com><4B2BEAF5.4020609@orange-ftgroup.com><2F1DE4DFCFF32144B771BD2C246E6A2004580452@misout7msgusr7e.ugd.att.com><20091221104548.N69632@sapphire.juniper.net><4B2FC853.4050208@cisco.com> <20091221111529.G69632@sapphire.juniper.net>
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: "Rahul Aggarwal" <rahul@juniper.net>, "Robert Raszuk" <raszuk@cisco.com>
Cc: Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 19:52:25 -0000
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From robert@raszuk.net  Mon Dec 21 11:56:21 2009
Return-Path: <robert@raszuk.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 76B943A6AA9 for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 11:56:21 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.149
X-Spam-Level: 
X-Spam-Status: No, score=-6.149 tagged_above=-999 required=5 tests=[AWL=-0.150, BAYES_00=-2.599, J_CHICKENPOX_13=0.6, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id t91sxES44lz5 for <l3vpn@core3.amsl.com>; Mon, 21 Dec 2009 11:56:20 -0800 (PST)
Received: from sj-iport-4.cisco.com (sj-iport-4.cisco.com [171.68.10.86]) by core3.amsl.com (Postfix) with ESMTP id A09303A6AA6 for <l3vpn@ietf.org>; Mon, 21 Dec 2009 11:56:20 -0800 (PST)
Authentication-Results: sj-iport-4.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-AV: E=Sophos;i="4.47,432,1257120000"; d="scan'208";a="65929288"
Received: from sj-core-4.cisco.com ([171.68.223.138]) by sj-iport-4.cisco.com with ESMTP; 21 Dec 2009 19:56:04 +0000
Received: from [192.168.1.57] (sjc-raszuk-87113.cisco.com [10.20.147.254]) by sj-core-4.cisco.com (8.13.8/8.14.3) with ESMTP id nBLJu3B1021787; Mon, 21 Dec 2009 19:56:03 GMT
Message-ID: <4B2FD2D2.9030206@raszuk.net>
Date: Mon, 21 Dec 2009 20:56:02 +0100
From: Robert Raszuk <robert@raszuk.net>
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
To: Rahul Aggarwal <rahul@juniper.net>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /	sectionA.1.1.3
References: <3126.1259858356@erosen-linux>	<4B28E84F.4000604@orange-ftgroup.com>	<2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com>	<4B2BEAF5.4020609@orange-ftgroup.com>	<2F1DE4DFCFF32144B771BD2C246E6A2004580452@misout7msgusr7e.ugd.att.com>	<20091221104548.N69632@sapphire.juniper.net>	<4B2FC853.4050208@cisco.com> <20091221111529.G69632@sapphire.juniper.net>
In-Reply-To: <20091221111529.G69632@sapphire.juniper.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Cc: Robert Raszuk <raszuk@cisco.com>, Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: robert@raszuk.net
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2009 19:56:21 -0000

Rahul,

>>>> Regarding the routing *churn*, section 11.4 says further down that:
>>>>
>>>>   ”To further reduce the routing churn due to C-multicast routes changes
>>>>    a Route Reflector that re-advertises a C-multicast route SHOULD set
>>>>    the Next Hop field of the MP_REACH_NLRI attribute of the route to an
>>>>    IP address of the Route Reflector."
>>>>
>>>> Which, in fact, doesn’t reduce routing churn.
>>>>
>>> Incorrect. The text you quoted *does* reduce routing churn due to
>>> C-multicast route changes. This is because it ensures that a Route
>>> Reflector advertises a C-multicast route, for a given (C-S/*, C-G) only
>>> when it receives the first route for the (C-S/*. C-G) from a client. It
>>> also ensures that a Route Reflector doesn't withdraw a C-multicast route
>>> for a given (C-S/*, C-G) as long as it has at least one route from a
>>> client for the (C-S/*, C-G).
>> Incorrect.
>>
>> As Maria have said correctly *the* quoted text does not reduce routing
>> churn.
>>
>> The procedure you are spelling out are not reflected in the above
>> paragraph. At least not in version 8 of the
>> draft-ietf-l3vpn-2547bis-mcast-bgp.
>>
> 
> That is an incorrect interpretation of the text in section 11.4. The
> procedure I spelled out follows from the text in section 11.4.
> 

Unfortunately not. Route reflector advertises *best* path not the 
*first* path. The text as stated today does not accomplish C-multicast 
routes aggregation. Also the concept of path virtualization you are 
essentially recommending at the withdraw is new.

Moreover it makes the use of route reflector "aggregators" a mandatory 
device in the solution.

Btw how service provider interested in the service and having full mesh 
of IBGP provide any c-multicast aggregation ?

How about service provider which uses confederations ?

I think this thread has reached it's limits. The discussion about route 
reflection as described in section 11.4 should take place on IDR list.

I bet vast majority of the members of L3VPN WG do not even read emails 
when the subject contains "mvpn" part :).

Cheers and Happy Holidays,
R.



From rahul@juniper.net  Tue Dec 22 15:06:28 2009
Return-Path: <rahul@juniper.net>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id AE62C3A68EA for <l3vpn@core3.amsl.com>; Tue, 22 Dec 2009 15:06:28 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.299
X-Spam-Level: 
X-Spam-Status: No, score=-6.299 tagged_above=-999 required=5 tests=[AWL=-0.300, BAYES_00=-2.599, J_CHICKENPOX_13=0.6, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 6n0MTpJjy3fV for <l3vpn@core3.amsl.com>; Tue, 22 Dec 2009 15:06:27 -0800 (PST)
Received: from exprod7og113.obsmtp.com (exprod7og113.obsmtp.com [64.18.2.179]) by core3.amsl.com (Postfix) with SMTP id C5B1E3A67E7 for <l3vpn@ietf.org>; Tue, 22 Dec 2009 15:05:57 -0800 (PST)
Received: from source ([66.129.224.36]) (using TLSv1) by exprod7ob113.postini.com ([64.18.6.12]) with SMTP ID DSNKSzFQxccc7WFul5CEzsU/g3QwEUxA7vPI@postini.com; Tue, 22 Dec 2009 15:06:11 PST
Received: from p-emfe01-sac.jnpr.net (66.129.254.72) by P-EMHUB01-HQ.jnpr.net (172.24.192.35) with Microsoft SMTP Server id 8.1.393.1; Tue, 22 Dec 2009 15:03:32 -0800
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emfe01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 22 Dec 2009 15:03:31 -0800
Received: from emailsmtp56.jnpr.net ([172.24.60.77]) by p-emlb01-sac.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 22 Dec 2009 15:03:31 -0800
Received: from magenta.juniper.net ([172.17.27.123]) by emailsmtp56.jnpr.net with Microsoft SMTPSVC(6.0.3790.3959); Tue, 22 Dec 2009 15:03:31 -0800
Received: from sapphire.juniper.net (sapphire.juniper.net [172.17.28.108])	by magenta.juniper.net (8.11.3/8.11.3) with ESMTP id nBMN3Vj37525; Tue, 22 Dec 2009 15:03:31 -0800 (PST)	(envelope-from rahul@juniper.net)
Date: Tue, 22 Dec 2009 15:03:31 -0800
From: Rahul Aggarwal <rahul@juniper.net>
To: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /sectionA.1.1.3
In-Reply-To: <2F1DE4DFCFF32144B771BD2C246E6A20045804CC@misout7msgusr7e.ugd.att.com>
Message-ID: <20091222144315.T2504@sapphire.juniper.net>
References: <3126.1259858356@erosen-linux><4B28E84F.4000604@orange-ftgroup.com><2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com><4B2BEAF5.4020609@orange-ftgroup.com><2F1DE4DFCFF32144B771BD2C246E6A2004580452@misout7msgusr7e.ugd.att.com><20091221104548.N69632@sapphire.juniper.net><4B2FC853.4050208@cisco.com> <20091221111529.G69632@sapphire.juniper.net> <2F1DE4DFCFF32144B771BD2C246E6A20045804CC@misout7msgusr7e.ugd.att.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="X-UNKNOWN"
Content-Transfer-Encoding: QUOTED-PRINTABLE
X-OriginalArrivalTime: 22 Dec 2009 23:03:31.0413 (UTC) FILETIME=[FAF4E850:01CA835A]
Cc: Robert Raszuk <raszuk@cisco.com>, Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 22 Dec 2009 23:06:28 -0000

Maria,

On Mon, 21 Dec 2009, NAPIERALA, MARIA H (ATTLABS) wrote:

> Rahul,
>
> the Route Reflector advertizes the first C-multicast (S/*,G) Join route a=
nd any
> subsequent C-multicast (S/*,G) Join route if it is better than previously
> selected,

No. As specified in section 11.4 the Route Reflector advertises an
aggregated C-multicast route.

> with the corresponding to the new best route ORIGINATOR_ID.
>
> Which text in section 11.4 states otherwise?
>

The entire text in section 11.4 describes how a Route Reflector
aggregates C-multicast routes.

At least two vendors with an inter-operable implementation have had no
trouble interpreting section 11.4. So despite your mis-interpretations
the spec is serving its purpose.

And finally, I am done repeating myself.

rahul




> Maria
>
>
> > -----Original Message-----
> > From: l3vpn-bounces@ietf.org [mailto:l3vpn-bounces@ietf.org] On Behalf
> > Of Rahul Aggarwal
> > Sent: Monday, December 21, 2009 2:21 PM
> > To: Robert Raszuk
> > Cc: Thomas Morin; L3VPN
> > Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A
> > /sectionA.1.1.3
> >
> >
> > Robert,
> >
> > On Mon, 21 Dec 2009, Robert Raszuk wrote:
> >
> > > Rahul,
> > >
> > > >> Regarding the routing *churn*, section 11.4 says further down
> > that:
> > > >>
> > > >>   =E2=80=9DTo further reduce the routing churn due to C-multicast =
routes
> > changes
> > > >>    a Route Reflector that re-advertises a C-multicast route SHOULD
> > set
> > > >>    the Next Hop field of the MP_REACH_NLRI attribute of the route
> > to an
> > > >>    IP address of the Route Reflector."
> > > >>
> > > >> Which, in fact, doesn=E2=80=99t reduce routing churn.
> > > >>
> > > >
> > > > Incorrect. The text you quoted *does* reduce routing churn due to
> > > > C-multicast route changes. This is because it ensures that a Route
> > > > Reflector advertises a C-multicast route, for a given (C-S/*, C-G)
> > only
> > > > when it receives the first route for the (C-S/*. C-G) from a
> > client. It
> > > > also ensures that a Route Reflector doesn't withdraw a C-multicast
> > route
> > > > for a given (C-S/*, C-G) as long as it has at least one route from
> > a
> > > > client for the (C-S/*, C-G).
> > >
> > > Incorrect.
> > >
> > > As Maria have said correctly *the* quoted text does not reduce
> > routing
> > > churn.
> > >
> > > The procedure you are spelling out are not reflected in the above
> > > paragraph. At least not in version 8 of the
> > > draft-ietf-l3vpn-2547bis-mcast-bgp.
> > >
> >
> > That is an incorrect interpretation of the text in section 11.4. The
> > procedure I spelled out follows from the text in section 11.4.
> >
> >
> > rahul
>
>

From mn1921@att.com  Tue Dec 22 15:50:46 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 65CB13A68C6 for <l3vpn@core3.amsl.com>; Tue, 22 Dec 2009 15:50:46 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -106.121
X-Spam-Level: 
X-Spam-Status: No, score=-106.121 tagged_above=-999 required=5 tests=[AWL=-0.122, BAYES_00=-2.599, J_CHICKENPOX_13=0.6, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id qfaOCO8TlrIg for <l3vpn@core3.amsl.com>; Tue, 22 Dec 2009 15:50:45 -0800 (PST)
Received: from mail129.messagelabs.com (mail129.messagelabs.com [216.82.250.147]) by core3.amsl.com (Postfix) with ESMTP id 732183A6867 for <l3vpn@ietf.org>; Tue, 22 Dec 2009 15:50:45 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-5.tower-129.messagelabs.com!1261525827!23537293!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 26373 invoked from network); 22 Dec 2009 23:50:28 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-5.tower-129.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 22 Dec 2009 23:50:28 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBMNoOqV017588; Tue, 22 Dec 2009 18:50:24 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBMNoJv9017561; Tue, 22 Dec 2009 18:50:19 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /sectionA.1.1.3
Date: Tue, 22 Dec 2009 18:49:56 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A200463340D@misout7msgusr7e.ugd.att.com>
In-Reply-To: <20091222144315.T2504@sapphire.juniper.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /sectionA.1.1.3
Thread-Index: AcqDW19tQvx+DrQXRaSTGymvJNXF/wABU/aw
References: <3126.1259858356@erosen-linux><4B28E84F.4000604@orange-ftgroup.com><2F1DE4DFCFF32144B771BD2C246E6A200451C9E0@misout7msgusr7e.ugd.att.com><4B2BEAF5.4020609@orange-ftgroup.com><2F1DE4DFCFF32144B771BD2C246E6A2004580452@misout7msgusr7e.ugd.att.com><20091221104548.N69632@sapphire.juniper.net><4B2FC853.4050208@cisco.com> <20091221111529.G69632@sapphire.juniper.net> <2F1DE4DFCFF32144B771BD2C246E6A20045804CC@misout7msgusr7e.ugd.att.com> <20091222144315.T2504@sapphire.juniper.net>
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: "Rahul Aggarwal" <rahul@juniper.net>
Cc: Robert Raszuk <raszuk@cisco.com>, Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 22 Dec 2009 23:50:46 -0000
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From erosen@cisco.com  Wed Dec 23 08:03:55 2009
Return-Path: <erosen@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 182E53A699A for <l3vpn@core3.amsl.com>; Wed, 23 Dec 2009 08:03:55 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.466
X-Spam-Level: 
X-Spam-Status: No, score=-6.466 tagged_above=-999 required=5 tests=[AWL=0.133,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id bxuOx1xlTqcK for <l3vpn@core3.amsl.com>; Wed, 23 Dec 2009 08:03:54 -0800 (PST)
Received: from rtp-iport-1.cisco.com (rtp-iport-1.cisco.com [64.102.122.148]) by core3.amsl.com (Postfix) with ESMTP id E6DB53A6897 for <l3vpn@ietf.org>; Wed, 23 Dec 2009 08:03:53 -0800 (PST)
Authentication-Results: rtp-iport-1.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-AV: E=Sophos;i="4.47,442,1257120000"; d="scan'208";a="76259713"
Received: from rtp-core-2.cisco.com ([64.102.124.13]) by rtp-iport-1.cisco.com with ESMTP; 23 Dec 2009 16:03:34 +0000
Received: from erosen-linux.cisco.com (erosen-linux.cisco.com [161.44.70.34]) by rtp-core-2.cisco.com (8.13.8/8.14.3) with ESMTP id nBNG3Ykx015983; Wed, 23 Dec 2009 16:03:34 GMT
Received: from erosen-linux (localhost.localdomain [127.0.0.1]) by erosen-linux.cisco.com (8.13.1/8.13.1) with ESMTP id nBNG3VNV001989;  Wed, 23 Dec 2009 11:03:31 -0500
To: Yakov Rekhter <yakov@juniper.net>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 
In-reply-to: Your message of Tue, 15 Dec 2009 08:55:39 -0800. <200912151655.nBFGtdj27964@magenta.juniper.net> 
Date: Wed, 23 Dec 2009 11:03:31 -0500
Message-ID: <1988.1261584211@erosen-linux>
From: Eric Rosen <erosen@cisco.com>
Cc: erosen@cisco.com, Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: erosen@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 23 Dec 2009 16:03:55 -0000

Yakov> By virtue of being non-segmented, non-segmented S-PMSIs do *not*
Yakov> provide the ability to aggregate P-tunnels of the intra-AS *segments*
Yakov> of S-PMSIs (for one thing, non-segmented S-PMSIs do not have "intra-AS 
Yakov> segments").

Well, by that logic one could say that non-segmented S-PMSIs are more
scalable on the grounds that they require fewer segments ;-)  This would fit
right in with the logic of Appendix A ;-)

Note that there are Carrier's Carrier techniques that can be used to
aggregate S-PMSIs through a transit AS when the S-PMSIs are not segmented.  

Eric> a downstream provider can use I-PMSIs even if an upstream provider
Eric> uses S-PMSIs.  That may or may not be a useful feature, but it doesn't
Eric> make the use of S-PMSIs more scalable.

Yakov> Sure it does.

I can't wait to see the multi-provider deployment in which one provider
decides to improve scale by using I-PMSIs and the other provider decides to
improve scale by using S-PMSIs ;-)








From erosen@cisco.com  Wed Dec 23 11:31:34 2009
Return-Path: <erosen@cisco.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id A0CA63A6A06 for <l3vpn@core3.amsl.com>; Wed, 23 Dec 2009 11:31:34 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.479
X-Spam-Level: 
X-Spam-Status: No, score=-6.479 tagged_above=-999 required=5 tests=[AWL=0.120,  BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Z5Eiby0FHOOD for <l3vpn@core3.amsl.com>; Wed, 23 Dec 2009 11:31:33 -0800 (PST)
Received: from rtp-iport-1.cisco.com (rtp-iport-1.cisco.com [64.102.122.148]) by core3.amsl.com (Postfix) with ESMTP id A333D3A685E for <l3vpn@ietf.org>; Wed, 23 Dec 2009 11:31:33 -0800 (PST)
Authentication-Results: rtp-iport-1.cisco.com; dkim=neutral (message not signed) header.i=none
X-IronPort-AV: E=Sophos;i="4.47,444,1257120000"; d="scan'208";a="76302624"
Received: from rtp-core-2.cisco.com ([64.102.124.13]) by rtp-iport-1.cisco.com with ESMTP; 23 Dec 2009 19:31:13 +0000
Received: from erosen-linux.cisco.com (erosen-linux.cisco.com [161.44.70.34]) by rtp-core-2.cisco.com (8.13.8/8.14.3) with ESMTP id nBNJVDsg004750; Wed, 23 Dec 2009 19:31:13 GMT
Received: from erosen-linux (localhost.localdomain [127.0.0.1]) by erosen-linux.cisco.com (8.13.1/8.13.1) with ESMTP id nBNJVBBx014292;  Wed, 23 Dec 2009 14:31:11 -0500
To: Rahul Aggarwal <rahul@juniper.net>
Subject: Re: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /sectionA.1.1.3 
In-reply-to: Your message of Tue, 22 Dec 2009 15:03:31 -0800. <20091222144315.T2504@sapphire.juniper.net> 
Date: Wed, 23 Dec 2009 14:31:11 -0500
Message-ID: <14291.1261596671@erosen-linux>
From: Eric Rosen <erosen@cisco.com>
Cc: Robert Raszuk <raszuk@cisco.com>, Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
Reply-To: erosen@cisco.com
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 23 Dec 2009 19:48:37 -0000

> As specified in section 11.4 the Route Reflector advertises an aggregated
> C-multicast route.

Well, section 11.4 says:

   C-multicast routes are "de facto" aggregated by BGP. This
   is because the MCAST-VPN NLRIs advertised by multiple PEs, for a C-
   multicast route for a particular C-S and C-G (or a particular C-* and
   C-G) of a given MVPN are identical.

The use of "de facto" means that no new procedures are required, and that
the "aggregation" is the same as BGP automatically does in all cases where
two different routes have the same NLRI.  It is a very creative
interpretation to suggest this requires the RR to originate a new route
rather than to distribute a route received from a PE. 

> two vendors with an inter-operable implementation have had no
> trouble interpreting section 11.4

Note that different interpretations of section 11.4 will still
interoperate.  It's a bit odd to say that no one has had trouble
interpreting it when at the same time you're telling folks that they are
misinterpreting it ;-)

I think what we're really talking about here is an implementation
optimization that is interesting, but that is not required for
interoperability, and that might or might not be worth doing.  Ordinarily,
this sort of optimization would be considered to be a local matter, not
something about which there has to be WG consensus.  It's only the presence
of Appendix A that forces us to get bogged down in unproductive discussions
of such details, as we now have to argue about whether the so-called
analysis there is to be based only on what the specifications require or
also on what implementations are likely to do.






















From mn1921@att.com  Thu Dec 24 06:56:58 2009
Return-Path: <mn1921@att.com>
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 2DE1A3A69F4 for <l3vpn@core3.amsl.com>; Thu, 24 Dec 2009 06:56:58 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -106.412
X-Spam-Level: 
X-Spam-Status: No, score=-106.412 tagged_above=-999 required=5 tests=[AWL=0.187, BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Xxdj1hSbIaOo for <l3vpn@core3.amsl.com>; Thu, 24 Dec 2009 06:56:57 -0800 (PST)
Received: from mail146.messagelabs.com (mail146.messagelabs.com [216.82.241.147]) by core3.amsl.com (Postfix) with ESMTP id 449893A69E5 for <l3vpn@ietf.org>; Thu, 24 Dec 2009 06:56:56 -0800 (PST)
X-VirusChecked: Checked
X-Env-Sender: mn1921@att.com
X-Msg-Ref: server-3.tower-146.messagelabs.com!1261666595!8078878!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [144.160.20.146]
Received: (qmail 32672 invoked from network); 24 Dec 2009 14:56:36 -0000
Received: from sbcsmtp7.sbc.com (HELO mlpd194.enaf.sfdc.sbc.com) (144.160.20.146) by server-3.tower-146.messagelabs.com with DHE-RSA-AES256-SHA encrypted SMTP; 24 Dec 2009 14:56:36 -0000
Received: from enaf.sfdc.sbc.com (localhost.localdomain [127.0.0.1]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBOEuZJX007005; Thu, 24 Dec 2009 09:56:35 -0500
Received: from misout7msgusr7e.ugd.att.com (misout7msgusr7e.ugd.att.com [144.155.43.107]) by mlpd194.enaf.sfdc.sbc.com (8.14.3/8.14.3) with ESMTP id nBOEuUmf006965; Thu, 24 Dec 2009 09:56:30 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /sectionA.1.1.3 
Date: Thu, 24 Dec 2009 09:56:31 -0500
Message-ID: <2F1DE4DFCFF32144B771BD2C246E6A200463371C@misout7msgusr7e.ugd.att.com>
In-Reply-To: <14291.1261596671@erosen-linux>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: draft-ietf-l3vpn-mvpn-considerations-05 / Appendix A /sectionA.1.1.3 
Thread-Index: AcqEBoE679mG8MeUSzyMDp4OxehK1gAom9HQ
References: Your message of Tue, 22 Dec 2009 15:03:31 -0800. <20091222144315.T2504@sapphire.juniper.net> <14291.1261596671@erosen-linux>
From: "NAPIERALA, MARIA H (ATTLABS)" <mn1921@att.com>
To: <erosen@cisco.com>, "Rahul Aggarwal" <rahul@juniper.net>
Cc: Robert Raszuk <raszuk@cisco.com>, Thomas Morin <thomas.morin@orange-ftgroup.com>, L3VPN <l3vpn@ietf.org>
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 24 Dec 2009 14:56:58 -0000

> It's only the presence
> of Appendix A that forces us to get bogged down in unproductive
discussions
> of such details, as we now have to argue about whether the so-called
> analysis there is to be based only on what the specifications require
> or also on what implementations are likely to do.
=20
Ditto!

Maria
