From owner-l2tp@diameter.org  Fri Dec  1 09:28:39 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id JAA18650
	for <l2tpext-archive@odin.ietf.org>; Fri, 1 Dec 2000 09:28:38 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eB1CQWX15659
	for l2tp-list; Fri, 1 Dec 2000 04:26:32 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eB1CQRH15653
	for <l2tp@l2tp.net>; Fri, 1 Dec 2000 04:26:28 -0800
Received: from ietf.org (odin.ietf.org [132.151.1.176]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id DAA14852 for <l2tp@ipsec.org>; Fri, 1 Dec 2000 03:33:48 -0800 (PST)
Received: from CNRI.Reston.VA.US (localhost [127.0.0.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA25388;
	Fri, 1 Dec 2000 06:32:20 -0500 (EST)
Message-Id: <200012011132.GAA25388@ietf.org>
Mime-Version: 1.0
Content-Type: Multipart/Mixed; Boundary="NextPart"
To: IETF-Announce:;
Cc: l2tp@ipsec.org
From: Internet-Drafts@ietf.org
Reply-to: Internet-Drafts@ietf.org
Subject: I-D ACTION:draft-ietf-l2tpext-l2tphc-03.txt
Date: Fri, 01 Dec 2000 06:32:19 -0500
Sender: owner-l2tp@diameter.org
Precedence: bulk

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the Layer Two Tunneling Protocol Extensions Working Group of the IETF.

	Title		: L2TP Header Compression ('L2TPHC')
	Author(s)	: A. Valencia
	Filename	: draft-ietf-l2tpext-l2tphc-03.txt
	Pages		: 8
	Date		: 30-Nov-00
	
The Layer 2 Tunneling Protocol ('L2TP') [RFC2661] defines a
mechanism for tunneling PPP sessions over arbitrary media.  There
exists a class of specific media applications for which protocol
overhead may be optimized, and where such reduction results in
improved operation.  This document describes the solution space
addressed, its underlying motivations, and the protocol modifications
required.  The enhancement to the L2TP protocol is called L2TP Header
Compression, or 'L2TPHC'.

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-l2tpext-l2tphc-03.txt

Internet-Drafts are also available by anonymous FTP. Login with the username
"anonymous" and a password of your e-mail address. After logging in,
type "cd internet-drafts" and then
	"get draft-ietf-l2tpext-l2tphc-03.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt


Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-l2tpext-l2tphc-03.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.
		
		
Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body;
	access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID:	<20001130115120.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-l2tpext-l2tphc-03.txt

--OtherAccess
Content-Type: Message/External-body;
	name="draft-ietf-l2tpext-l2tphc-03.txt";
	site="ftp.ietf.org";
	access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID:	<20001130115120.I-D@ietf.org>

--OtherAccess--

--NextPart--




From owner-l2tp@diameter.org  Sat Dec  2 04:35:56 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id EAA16130
	for <l2tpext-archive@odin.ietf.org>; Sat, 2 Dec 2000 04:35:55 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eB26q9716940
	for l2tp-list; Fri, 1 Dec 2000 22:52:09 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eB26ocH16932
	for <l2tp@l2tp.net>; Fri, 1 Dec 2000 22:50:39 -0800
Received: from occamnetworks.com (mail.occamnetworks.com [216.64.159.194]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id VAA00268 for <l2tp@ipsec.org>; Fri, 1 Dec 2000 21:58:01 -0800 (PST)
Received: from occamnetworks.com (honeycombs.occamnetworks.com [192.168.2.202])
	by occamnetworks.com (8.10.1/8.10.1) with ESMTP id eB1Nb3Q27993;
	Fri, 1 Dec 2000 15:37:03 -0800
Message-ID: <3A283624.A592723F@occamnetworks.com>
Date: Fri, 01 Dec 2000 15:37:08 -0800
From: Evan Caves <evan@occamnetworks.com>
Organization: Occam Networks, Inc.
X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.2.16-3 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: Klausberger Walter <walter.klausberger@siemens.at>, dthaler@MICROSOFT.com
CC: l2tp@ipsec.org
Subject: Re: ifStackTable layering in L2TP
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Klausberger Walter wrote:
> 
> Hi everybody,
> 
> I started to read the new L2TP-MIB draft and I already have several
> questions...
> 
> As I understood stacking should be made visible up to the 'forwarding
> layer'(e.g. IP). When looking at common ifStackTables I never saw an IP
> layer until yet. 

Not sure what you mean by 'saw'.

> So there should be no layer above Ethernet when
> transporting IP packets...

Unless you have a tunnel sourced from that ethernet interface.

> 
> As an example in <draft-ietf-l2tpext-l2tp-mib-01.txt> the tunnel interface
> in the LNS is set on top of an Ethernet interface. But L2TP is never set
> directly above Ethernet, but will be transported via UDP/IP. So there is no
> direct context. The connection depends on the routing table.

There is always a direct context. Every L2TP packet has a source address
that corresponds to one of the L2TP devices configured IP addresses.
That IP 
address is associated with a particular ifStack on the device of which
the L2TP tunnel interface then gets layered on.

> 
> Assume there is an LNS with an Ethernet and an unnumbered Frame Relay
> interface. First the tunnel is setup via Ethernet IF. Due to the entries in
> the routing table, the tunnel IF is set on top of the ethernet IF.

Not sure what your saying here but I assume you mean "we will choose the
source IP address based on which egress interface is closest to the
tunnel endpoint".

> Now a
> problem occurs in the backbone and the connection is switched over to the
> unnumbered Frame Relay IF (same IP address). Would this need a change in the
> ifStackTable?

The layering model (at least from the tunnel I/F on down) is straight
from
the "IP Tunnel MIB". I agree that in pratice tunnel packets could be
routed
out of a different egress interface based on shortest path routing. As a
result of that the layering model as depicted breaks (inconsistant
statistics
and possibly state). I am cc'ing Dave Thaler to see if he can provide
some
input into this matter.

> 
> It is possible to transport L2TP directly over ATM or via UDP/IP/ATM. 

Yes.

> Would
> there be no difference in the ifStackTable?

There should not be.

> 
> In case of LAC it is even more unclear to me how the ifStackTable will look
> like. In former L2TP-MIB drafts it was stated that the tunnel IF sits on top
> of the interfaces that forward the PPP packets (e.g. DS0). 

It was changed because conceptually there has to be a switching decision
made to move packets from an 'ingress' interface (the DS0 for example)
to 
an 'egress' interface (the L2TP tunnel interface). This seemed to make
the 
most sense otherwise you end up layering egress interfaces on top of
ingress
interfaces in the ifStackTable (which I personally have never seen 
before).

> Now there is a
> reference to the tunnel MIB (RFC 2667) and everything seems to be different
> now. 

There has always been a reference to the "IP Tunnel MIB".

> Do we have a problem here with unnumbered IP interfaces, 

I don't think it is just a problem with unnumbered interfaces. If
topology
changes you can still route a packet out a numbered egress interface
while
the source address of the packet has an IP address of another interface.
 
> because...
>    Tunnels are handled by creating a logical interface (ifEntry) for each
> tunnel. These are then correlated, using the ifStack table of the Interfaces
> MIB, to those interfaces on which the local IPv4 addresses of the tunnels
> are configured.
> ...or do I have to take the numbered Ethernet IF for the ifStackTable, even
> if the tunnel is connected via Frame Relay IF?

No. That would not make sense.

> 
> Or am I totally wrong?

Nope. I think you have brought up a legitimate concern but to resolve it
we have to consult with the IP Tunnel MIB folks.

> 
> just some questions, I hope someone can answer...

Sorry for the delay (I have just moved to a different employer). Give me
some more time to think on it and hopefully Dave will have some
suggestions.

cheers,

evan
-

> 
> with best regards
> Walter


From owner-l2tp@diameter.org  Wed Dec  6 15:38:58 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id PAA12364
	for <l2tpext-archive@odin.ietf.org>; Wed, 6 Dec 2000 15:38:57 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eB6Jo0830839
	for l2tp-list; Wed, 6 Dec 2000 11:50:00 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from sj-msg-core-2.cisco.com (sj-msg-core-2.cisco.com [171.69.43.88])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eB6JnqH30828
	for <l2tp@l2tp.net>; Wed, 6 Dec 2000 11:49:52 -0800
Received: from iwan-view6.cisco.com (iwan-view6.cisco.com [171.69.24.141])
	by sj-msg-core-2.cisco.com (8.9.3/8.9.1) with ESMTP id KAA29020;
	Wed, 6 Dec 2000 10:57:29 -0800 (PST)
Received: from cisco.com (sjck-dial-gw5-214.cisco.com [10.19.238.215]) by iwan-view6.cisco.com (8.8.8-Cisco List Logging/CISCO.WS.1.2) with ESMTP id KAA16073; Wed, 6 Dec 2000 10:57:25 -0800 (PST)
Message-ID: <3A2F09B8.E0B8899D@cisco.com>
Date: Wed, 06 Dec 2000 22:53:28 -0500
From: "W. Mark Townsley" <townsley@cisco.com>
X-Mailer: Mozilla 4.7 [en] (Win95; U)
X-Accept-Language: en,pdf
MIME-Version: 1.0
To: agenda@ietf.org, l2tp@l2tp.net
Subject: Agenda for L2TPEXT
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit


Barring any surprises, we should have time to spare.

Agenda Bashing, 5 minutes

Danny McPherson, Amber Networks 15 minutes
draft-ietf-l2tpext-svctype-00.txt

Danny McPherson, Amber Networks, 15 minutes
L2TP Diffserv draft (new version did not get posted before cutoff) 

Bill Palter, Redback Networks 15 minutes
draft-ietf-l2tpext-eth-00.txt

Mark Townsley, cisco Systems 15 minutes
draft-ietf-l2tpext-l2tpbis-01.txt


From owner-l2tp@diameter.org  Wed Dec  6 16:15:18 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id QAA22937
	for <l2tpext-archive@odin.ietf.org>; Wed, 6 Dec 2000 16:15:18 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eB6KinO30902
	for l2tp-list; Wed, 6 Dec 2000 12:44:49 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eB6KijH30896
	for <l2tp@l2tp.net>; Wed, 6 Dec 2000 12:44:45 -0800
Received: from occamnetworks.com (mail.occamnetworks.com [216.64.159.194]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id LAA05858 for <l2tp@ipsec.org>; Wed, 6 Dec 2000 11:52:23 -0800 (PST)
Received: from occamnetworks.com (honeycombs.occamnetworks.com [192.168.2.202])
	by occamnetworks.com (8.10.1/8.10.1) with ESMTP id eB6JpTQ25279;
	Wed, 6 Dec 2000 11:51:29 -0800
Message-ID: <3A2E98D8.C03888A@occamnetworks.com>
Date: Wed, 06 Dec 2000 11:51:52 -0800
From: Evan Caves <evan@occamnetworks.com>
Organization: Occam Networks, Inc.
X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.2.16-3 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: "'Dave Thaler'" <dthaler@Exchange.Microsoft.com>
CC: Klausberger Walter <walter.klausberger@siemens.at>, l2tp@ipsec.org,
        kzm@cisco.com
Subject: Re: AW: ifStackTable layering in L2TP
References: <D9F2B9CD7BD5D21196BC0800060D9ED604373012@vies186a.sie.siemens.at>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Dave,

Dave Thaler wrote:

> Klausberger Walter wrote:
>[snip...]
> 
> > -----Original Message-----
> > From: Klausberger Walter [mailto:walter.klausberger@siemens.at]
> > Sent: Wednesday, November 29, 2000 2:06 AM
> > To: l2tp@ipsec.org
> > Subject: ifStackTable layering in L2TP
> >
> > Hi everybody,
> >
> > I started to read the new L2TP-MIB draft and I already have several
> > questions...
> >
> > As I understood stacking should be made visible up to the 'forwarding
> > layer'(e.g. IP). When looking at common ifStackTables I never
> > saw an IP
> > layer until yet. So there should be no layer above Ethernet when
> > transporting IP packets...
> 
> Correct.

??? Then why in the tunnel mib is their a tunnel interface on top of
a physical interface (ie: ethernet).

> 
> > As an example in <draft-ietf-l2tpext-l2tp-mib-01.txt> the
> > tunnel interface
> > in the LNS is set on top of an Ethernet interface. But L2TP
> > is never set
> > directly above Ethernet, but will be transported via UDP/IP.
> > So there is no
> > direct context. The connection depends on the routing table.
> 
> Correct.

The tunnel mib infers that there is a direct context (ie: the 
tunnel sits on top of the ifStack associated with one of the 
devices locally configured IP addresses) and completely ignores
any discussion on ifStack layering issues based on destination 
route lookups.

As an aside it would seem that, at least to me, the tunnel
interface should be at the bottom of the ifStack (after all it is
a virtual wire to a remote network). Attempting to stack the tunnel
I/F on ifStacks based on the current topology to me is problematic
(especially if their are multiple paths to a particular destination
which means an understanding of how the forwarder is choosing nexthops
for tunneled packets) and provides little or no benefit. Is their
something in the ifStack architecture that says the bottom of every
stack must terminate at a physical interface?


> 
> > Assume there is an LNS with an Ethernet and an unnumbered Frame Relay
> > interface. First the tunnel is setup via Ethernet IF. Due to
> > the entries in
> > the routing table, the tunnel IF is set on top of the
> > ethernet IF. Now a
> > problem occurs in the backbone and the connection is switched
> > over to the
> > unnumbered Frame Relay IF (same IP address). Would this need
> > a change in the
> > ifStackTable?
> 
> Yes.  Keep in mind that the ifStackTable is only a logical
> table.  It doesn't necessarily correspond to any state in
> the device.

I do not understand this statement. Of course it represents the
state of the device at the time of the query.

> It just means that if you query the ifStackTable,
> that the agent can at least determind which IF the tunnel
> is currently being routed over.

This could be many.

> 
> > It is possible to transport L2TP directly over ATM or via
> > UDP/IP/ATM. Would
> > there be no difference in the ifStackTable?
> 
> No difference.  There would be a difference in the
> addresses used on either end of the tunnel though
> (for the outer header), as in one case they're IP addresses
> and in the other case they're not.
> 
> > In case of LAC it is even more unclear to me how the
> > ifStackTable will look
> > like. In former L2TP-MIB drafts it was stated that the tunnel
> > IF sits on top
> > of the interfaces that forward the PPP packets (e.g. DS0).
> > Now there is a
> > reference to the tunnel MIB (RFC 2667) and everything seems
> > to be different
> > now. Do we have a problem here with unnumbered IP interfaces,
> > because...
> >    Tunnels are handled by creating a logical interface
> > (ifEntry) for each
> > tunnel. These are then correlated, using the ifStack table of
> > the Interfaces
> > MIB, to those interfaces on which the local IPv4 addresses of
> > the tunnels
> > are configured.
> 
> The correlation should be to the interface used to reach
> the remote address, which is not necessarily the one
> which has the local IPv4 address assigned to it.

This statement is completely missing from the tunnel mib, and if 
you ask me, a very important one because as the mib reads today it
is describing a very different layering model.

evan
-


From owner-l2tp@diameter.org  Thu Dec  7 00:09:30 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id AAA08558
	for <l2tpext-archive@odin.ietf.org>; Thu, 7 Dec 2000 00:09:29 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eB74mWx31263
	for l2tp-list; Wed, 6 Dec 2000 20:48:32 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eB74mRH31257
	for <l2tp@l2tp.net>; Wed, 6 Dec 2000 20:48:27 -0800
Received: from occamnetworks.com (mail.occamnetworks.com [216.64.159.194]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id TAA06353 for <l2tp@ipsec.org>; Wed, 6 Dec 2000 19:56:06 -0800 (PST)
Received: from occamnetworks.com (honeycombs.occamnetworks.com [192.168.2.202])
	by occamnetworks.com (8.10.1/8.10.1) with ESMTP id eB73sMQ10458;
	Wed, 6 Dec 2000 19:54:22 -0800
Message-ID: <3A2F0A02.D37DF364@occamnetworks.com>
Date: Wed, 06 Dec 2000 19:54:42 -0800
From: Evan Caves <evan@occamnetworks.com>
Organization: Occam Networks, Inc.
X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.2.16-3 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: Dave Thaler <dthaler@Exchange.Microsoft.com>
CC: Klausberger Walter <walter.klausberger@siemens.at>, l2tp@ipsec.org,
        kzm@cisco.com
Subject: Re: AW: ifStackTable layering in L2TP
References: <5B90AD65A9E8934987DB0C8C0762657420FCD4@DF-BOWWOW.platinum.corp.microsoft.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Dave Thaler wrote:
> 
> > -----Original Message-----
> > From: Evan Caves [mailto:evan@occamnetworks.com]
> > Sent: Wednesday, December 06, 2000 11:52 AM
> > To: Dave Thaler
> > Cc: Klausberger Walter; l2tp@ipsec.org; kzm@cisco.com
> > Subject: Re: AW: ifStackTable layering in L2TP
> >
> >
> > Dave,
> >
> > Dave Thaler wrote:
> > 
> > > Klausberger Walter wrote:
> > >[snip...]
> > >
> > > > -----Original Message-----
> > > > From: Klausberger Walter [mailto:walter.klausberger@siemens.at]
> > > > Sent: Wednesday, November 29, 2000 2:06 AM
> > > > To: l2tp@ipsec.org
> > > > Subject: ifStackTable layering in L2TP
> > > >
> > > > Hi everybody,
> > > >
> > > > I started to read the new L2TP-MIB draft and I already
> > have several
> > > > questions...
> > > >
> > > > As I understood stacking should be made visible up to the
> > 'forwarding
> > > > layer'(e.g. IP). When looking at common ifStackTables I never
> > > > saw an IP
> > > > layer until yet. So there should be no layer above Ethernet when
> > > > transporting IP packets...
> > >
> > > Correct.
> >
> > ??? Then why in the tunnel mib is their a tunnel interface on top of
> > a physical interface (ie: ethernet).
> 
> Because a tunnel isn't an IP packet (e.g. a TCP or UDP packet).
> It's a Foo-over-IP packet (e.g. IP inside IP).  There's no
> interface corresponding to the IP or TDP or UDP layers
> when you just send IP packets over an Ethernet.
> 
> When you run a tunnel, then you have another interface.

I don't think this answers either my or Walters question.

> 
> > > > As an example in <draft-ietf-l2tpext-l2tp-mib-01.txt> the
> > > > tunnel interface
> > > > in the LNS is set on top of an Ethernet interface. But L2TP
> > > > is never set
> > > > directly above Ethernet, but will be transported via UDP/IP.
> > > > So there is no
> > > > direct context. The connection depends on the routing table.
> > >
> > > Correct.
> >
> > The tunnel mib infers that there is a direct context (ie: the
> > tunnel sits on top of the ifStack associated with one of the
> > devices locally configured IP addresses) and completely ignores
> > any discussion on ifStack layering issues based on destination
> > route lookups.
> 
> There's no object called an "ifStack".  

Huh? "ifStack" in the above sentence means "interface stack".

> An ifStackEntry corresponds
> to a particular combination of one specific interface being layered
> over another.
> 
> > As an aside it would seem that, at least to me, the tunnel
> > interface should be at the bottom of the ifStack (after all it is
> > a virtual wire to a remote network). Attempting to stack the tunnel
> > I/F on ifStacks based on the current topology to me is problematic
> > (especially if their are multiple paths to a particular destination
> > which means an understanding of how the forwarder is choosing nexthops
> > for tunneled packets) and provides little or no benefit. Is their
> > something in the ifStack architecture that says the bottom of every
> > stack must terminate at a physical interface?
> 
> Yes, read RFC 2233.  For example, "where the bottom of the stack is a
> physical interface (ifConnectorPresent has the value 'true')"

Unfortunalely I was looking at its predecessor rfc1573. This paragragh 
seems to have been added in this later version. Do you know why this has 
to be the case?

> 
> What you say is "problematic" is not problematic and the ability
> to handle that situation is one of the features of the ifStackTable
> as discussed in RFC 2233 (e.g. "An example of upward multiplexing is MLP
> (Multi-Link-Procedure) which provides load-sharing over several
> serial lines by appearing as a single point-to-point link to the
> sub-layer(s) above.")

OK let me try again. In traditional ifStacks there is a well defined 
boundary or service interface between interface sub-layers where each 
sub-layer does its protocol processing and forwards 'foo' to the next 
sub-layer. We could model an ifStack up to a fowarder as

		   Forwarder
                -----------------
		I/F sub-layer n		(PPP)
		-----------------
		I/F sub-layer n-1	(HDLC)
		-----------------
		I/F sub-layer n-2	(RS232)
		-----------------


In the IP tunnel case the picture looks more like
		
		-----------------
		 I/F sub-layer n	(Tunnel)
		-----------------
		   Forwarder
                -----------------
		I/F sub-layer n-1	(PPP)
		-----------------
		I/F sub-layer n-2	(HDLC)
		-----------------
		I/F sub-layer n-3	(RS232)
		-----------------

in which case the fowarder is not part of any interface sub-layer.
So the client to layer n-1 is a fowarder and the service provided 
to layer n is a forwarder. The policies implemented by the forwarder
are not known by either layer so it becomes difficult to determine
who is below layer n without actually forwarding a tunneled packet
as the packet contents will determine how the packet is forwarded
(based on the forwarders local policies like filters, multipath
criteria, QOS and route table lookups).

[snip...]
> > > The correlation should be to the interface used to reach
> > > the remote address, which is not necessarily the one
> > > which has the local IPv4 address assigned to it.
> >
> > This statement is completely missing from the tunnel mib, and if
> > you ask me, a very important one because as the mib reads today it
> > is describing a very different layering model.

Don't you think that this should be described in the tunnel mib?

evan
-


From owner-l2tp@diameter.org  Thu Dec  7 00:19:07 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id AAA11916
	for <l2tpext-archive@odin.ietf.org>; Thu, 7 Dec 2000 00:19:07 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eB74m4831255
	for l2tp-list; Wed, 6 Dec 2000 20:48:04 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from franklin.cisco.com (franklin.cisco.com [171.70.156.17])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eB74lxH31249
	for <l2tp@diameter.org>; Wed, 6 Dec 2000 20:47:59 -0800
Received: from gwzpc (sj-dial-3-115.cisco.com [171.68.180.116]) by franklin.cisco.com (8.8.6 (PHNE_17190)/CISCO.SERVER.1.2) with SMTP id TAA27519; Wed, 6 Dec 2000 19:55:33 -0800 (PST)
Reply-To: <gwz@cisco.com>
From: "Glen Zorn" <gwz@cisco.com>
To: "W. Mark Townsley" <townsley@cisco.com>
Cc: "l2tpext" <l2tp@diameter.org>
Subject: RE: Agenda for L2TPEXT
Date: Wed, 6 Dec 2000 19:56:25 -0800
Message-ID: <NDBBIHMPILAAGDHPCIOPKENGCPAA.gwz@cisco.com>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0)
Importance: Normal
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2615.200
In-Reply-To: <3A2F09B8.E0B8899D@cisco.com>
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit


> Barring any surprises, we should have time to spare.
> 
> Agenda Bashing, 5 minutes
> 
> Danny McPherson, Amber Networks 15 minutes
> draft-ietf-l2tpext-svctype-00.txt
> 
> Danny McPherson, Amber Networks, 15 minutes
> L2TP Diffserv draft (new version did not get posted before cutoff) 

How are we supposed to talk about something we've never seen?

> 
> Bill Palter, Redback Networks 15 minutes
> draft-ietf-l2tpext-eth-00.txt
> 
> Mark Townsley, cisco Systems 15 minutes
> draft-ietf-l2tpext-l2tpbis-01.txt
> 


From owner-l2tp@diameter.org  Thu Dec  7 03:44:23 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id DAA01741
	for <l2tpext-archive@odin.ietf.org>; Thu, 7 Dec 2000 03:44:22 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eB77WaT31426
	for l2tp-list; Wed, 6 Dec 2000 23:32:36 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from cisco.com (shako.cisco.com [161.44.3.76])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eB77WKH31420
	for <l2tp@diameter.org>; Wed, 6 Dec 2000 23:32:20 -0800
Received: from cisco.com (sjck-dial-gw5-3.cisco.com [10.19.238.4])
	by cisco.com (8.8.8/2.6/Cisco List Logging/8.8.8) with ESMTP id BAA29257;
	Thu, 7 Dec 2000 01:39:39 -0500 (EST)
Message-ID: <3A308141.82C81459@cisco.com>
Date: Thu, 07 Dec 2000 22:35:45 -0800
From: "W. Mark Townsley" <townsley@cisco.com>
X-Mailer: Mozilla 4.7 [en] (Win95; U)
X-Accept-Language: en,pdf
MIME-Version: 1.0
To: gwz@cisco.com
CC: l2tpext <l2tp@diameter.org>
Subject: Re: Agenda for L2TPEXT
References: <NDBBIHMPILAAGDHPCIOPKENGCPAA.gwz@cisco.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit



Glen Zorn wrote:
> 
> > Barring any surprises, we should have time to spare.
> >
> > Agenda Bashing, 5 minutes
> >
> > Danny McPherson, Amber Networks 15 minutes
> > draft-ietf-l2tpext-svctype-00.txt
> >
> > Danny McPherson, Amber Networks, 15 minutes
> > L2TP Diffserv draft (new version did not get posted before cutoff)
> 
> How are we supposed to talk about something we've never seen?

He promised to present something, so I imagine he will at least have
some slides. Danny?

> 
> >
> > Bill Palter, Redback Networks 15 minutes
> > draft-ietf-l2tpext-eth-00.txt
> >
> > Mark Townsley, cisco Systems 15 minutes
> > draft-ietf-l2tpext-l2tpbis-01.txt
> >


From owner-l2tp@diameter.org  Thu Dec  7 23:43:43 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id XAA12423
	for <l2tpext-archive@odin.ietf.org>; Thu, 7 Dec 2000 23:43:42 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eB83ej000363
	for l2tp-list; Thu, 7 Dec 2000 19:40:45 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from tcb.net (tcb.net [205.168.100.1])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eB83e6H00357
	for <l2tp@diameter.org>; Thu, 7 Dec 2000 19:40:06 -0800
Received: from sofos.tcb.net (sofos.tcb.net [127.0.0.1])
	by tcb.net (8.9.3/8.9.3) with ESMTP id TAA07558
	for <l2tp@diameter.org>; Thu, 7 Dec 2000 19:51:39 -0700
Message-Id: <200012080251.TAA07558@tcb.net>
X-Mailer: exmh version 2.0.3
To: l2tpext <l2tp@diameter.org>
From: Danny McPherson <danny@tcb.net>
Reply-To: danny@tcb.net
Subject: Re: Agenda for L2TPEXT 
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Thu, 07 Dec 2000 19:51:39 -0700
Sender: owner-l2tp@diameter.org
Precedence: bulk

> 
> 
> Glen Zorn wrote:
> > 
> > > Barring any surprises, we should have time to spare.
> > >
> > > Agenda Bashing, 5 minutes
> > >
> > > Danny McPherson, Amber Networks 15 minutes
> > > draft-ietf-l2tpext-svctype-00.txt
> > >
> > > Danny McPherson, Amber Networks, 15 minutes
> > > L2TP Diffserv draft (new version did not get posted before cutoff)
> > 
> > How are we supposed to talk about something we've never seen?
> 
> He promised to present something, so I imagine he will at least have
> some slides. Danny?

Yeah, it's mostly pretty trivial.  The slides will [hopefully] address the key 
& questionable attributes of the draft.

I'll try to post the latest version of the draft to the list before Monday.  

Apologies for not getting it out already, I certainly understand your 
reasoning.

-danny





From owner-l2tp@diameter.org  Fri Dec  8 20:07:23 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id UAA28474
	for <l2tpext-archive@odin.ietf.org>; Fri, 8 Dec 2000 20:07:22 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eB8Mujk02140
	for l2tp-list; Fri, 8 Dec 2000 14:56:45 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from tcb.net (tcb.net [205.168.100.1])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eB8MufH02134
	for <l2tp@l2tp.net>; Fri, 8 Dec 2000 14:56:41 -0800
Received: from sofos.tcb.net (sofos.tcb.net [127.0.0.1])
	by tcb.net (8.9.3/8.9.3) with ESMTP id PAA12015
	for <l2tp@l2tp.net>; Fri, 8 Dec 2000 15:08:22 -0700
Message-Id: <200012082208.PAA12015@tcb.net>
X-Mailer: exmh version 2.0.3
To: l2tp@l2tp.net
From: Danny McPherson <danny@tcb.net>
Reply-To: danny@tcb.net
Subject: draft-ietf-l2tpext-ds-01.txt
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Fri, 08 Dec 2000 15:08:22 -0700
Sender: owner-l2tp@diameter.org
Precedence: bulk


I'll post it to internet-drafts@ietf.org after San Diego is over.  

Note especially the TODO section.  Comments welcome, I'll be discussing the 
highlights Thursday.

-danny

========================================================================
Network Working Group                                     Pat R. Calhoun
INTERNET DRAFT                                    Sun Microsystems, Inc.
                                                         Danny McPherson
                                                    Amber Networks, Inc.
                                                              Ken Peirce
December 2000                                      Malibu Networks, Inc.



                 L2TP Differentiated Services Extension
                     <draft-ietf-l2tpext-ds-01.txt>


1. Status of this Memo

   This document is an Internet-Draft and is in full conformance with
   all provisions of Section 10 of RFC2026.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-
   Drafts.

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

   The list of current Internet-Drafts can be accessed at
   http://www.ietf.org/ietf/1id-abstracts.txt.

   The list of Internet-Draft Shadow Directories can be accessed at
   http://www.ietf.org/shadow.html.


2. Abstract

   The Layer Two Tunneling Protocol (L2TP) [RFC 2661] provides a
   standard method for tunneling PPP [RFC 1661] packets.  The current
   specification provides no provisions for supporting Differentiated
   Services (diffserv) [RFC 2474, RFC 2475] over the L2TP control
   connection or subsequent data sessions.  As a result, no standard
   mechanism currently exists within L2TP to provide L2TP protocol
   negotiations for service discrimination.

   This document describes mechanisms which enable L2TP to negotiate
   desired DS values for the L2TP control connection, as well as



Calhoun, McPherson, Peirce                             	[Page 1]





INTERNET DRAFT                                            December 2000


   individual sessions within an L2TP tunnel.


3. Specification of Requirements

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED",  "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in [RFC 2119].


4. Introduction

   The L2TP specification currently provides no mechanism for supporting
   diffserv (DS).  This document describes mechanisms that enable L2TP
   to indicate desired DS values to be associated with an L2TP control
   connection, as well as individual sessions within an L2TP tunnel.

   This document will describe how a set of L2TP peers MAY negotiate a
   set of differential services indicators for a tunnel control
   connection, as well as for individual sessions within the tunnel.

   The actual bit interpretation of the DS field is beyond the scope of
   this document, and is purposefully omitted.  This document is
   concerned only with defining a uniform exchange and subsequent
   mapping mechanism for the DS AVPs.


5. Control Connection Operation

   As defined in [RFC 2661], a control connection operates in-band over
   a tunnel to control the establishment, release, and maintenance of
   sessions and of the tunnel itself.  As such, this document provides a
   mechanism to enable discrimination of L2TP control messages from
   other packets.  For this purpose, we introduce the Control Connection
   DS (CCDS) AVP.

   The presence of the CCDS AVP serves as an indication to the peer (LAC
   or LNS) that the tunnel initiator wishes to use the specified DS
   value on all packets within the tunnel's control connection.

   Upon receipt of a Start-Control-Connection-Request (SCCRQ) containing
   the CCDS AVP, if the tunnel terminator provides no support for the
   CCDS AVP it SHOULD ignore the AVP and send a SCCRP to the tunnel
   initiator without the CCDS AVP.  The tunnel initiator interprets the
   absence of the CCDS AVP in the SCCRP as as an indication that the
   tunnel terminator is incapable of supporting CCDS.

   Upon receipt of a SCCRP that contains no CCDS AVP in response to a



Calhoun, McPherson, Peirce                             	[Page 2]





INTERNET DRAFT                                            December 2000


   SCCRQ that contained a CCDS AVP, if the tunnel initiator wants to
   continue tunnel establishment it sends a SCCCN; otherwise, it sends a
   StopCCN to the tunnel terminator to end the connection.  The StopCCN
   control message MUST contain a Result Code AVP that indicates CCDS
   AVP value [TBD] as the reason for sending the StopCCN.

   If the tunnel terminator provides support for CCDS but is unwilling
   or unable to support the requested DS value the tunnel terminator
   MUST send a SCCRP control message containing a CCDS AVP indicating
   the value it's willing to use.  If the CCDS AVP value is the same as
   the one in the SCCRQ, it signals the acceptence of the requested DS
   value.  If the value is different it serves as a counter-offer by the
   tunnel terminator.

   If the tunnel initiator receives an SCCRP that contains a CCDS AVP
   with a value other than that requested in the SCCRQ, and the tunnel
   initiator is unwilling to use the value, the tunnel initiator MUST
   send a StopCCN control message containing a Result Code AVP that
   indicates CCDS AVP value [TBD] as the reason for sending the StopCCN.


5.1. Control Connection DS AVP (SCCRQ, SCCRP)

   The CCDS AVP is encoded as Vendor ID 43, and the Attribute Value is
   the 16-bit quantity 1 (the ID 43 reflects 3Com Corporation, it should
   be changed to 0 and an official Attribute Value chosen should this
   specification advance on as standards track).

   Each CCDS AVP is encoded as follows:

     Vendor ID = 43
     Attribute = 1

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |M|H|0|0|0|0|    Length         |              43               |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 
   |                1              |           DS Value            |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   This AVP MAY be present in the following message types:  SCCRQ and
   SCCRP.  This AVP MAY be hidden (the H-bit set to 0 or 1) and is
   optional (M-bit not set).  The length (before hiding) of this AVP
   MUST be 8 octets.






Calhoun, McPherson, Peirce                             	[Page 3]





INTERNET DRAFT                                            December 2000


6. Session Operation

   As defined in [RFC 2661], a L2TP session is connection-oriented. The
   LNS and LAC maintain state for each Call that is initiated or
   answered by an LAC. An L2TP Session is created between the LAC and
   LNS when an end-to-end PPP connection is established between a Remote
   System and the LNS.  Datagrams related to the PPP connection are sent
   over the Tunnel between the LAC and LNS. There is a one to one
   relationship between established L2TP Sessions and their associated
   Calls.  As such, this document provides a mechanism to enable
   discrimination for packets within an particular session from other
   packets.  For this purpose, we introduce the Session DS (SDS) AVP.

   The presence of the SDS AVP serves as an indication to the peer (LAC
   or LNS) that the session initiator wishes to use the specified DS
   value on all packets within the session.

   Upon receipt of a Incoming-Call-Request (ICRQ) or Outgoing-Call-
   Request (OCRQ) containing the SDS AVP if the session terminator
   provides no support for the requested DS value a Call-Disconnect-
   Notify (CDN) message MUST be sent to the peer.  The CDN message MUST
   contain a Result Code AVP that indicates SDS AVP value [TBD] as the
   reason for sending the CDN.

   If the session terminator provides support for SDS but is unwilling
   or unable to support the requested DS value the session terminator
   MUST do one of the following:

   1) Send a CDN message containing a Result Code AVP that indicates SDS
   AVP value [TBD] as the reason for sending the CDN.

   2) Send an Incoming-Call-Reply (ICRP) or Outgoing-Call-Reply (OCRP)
   message containing a CDN AVP indicating the DS value the terminator
   is willing to use for the session.

   3) If the session terminator supports the DS value in the SDS AVP
   session establishment MUST continue as defined in [RFC 2661].

   If the session initiator receives an ICRP or OCRP that contains an
   SDS AVP with a value other than that requested in the ICRQ or OCRQ,
   and the session initiator is unwilling to use the value, the session
   initiator MUST send a CDN message containing a Result Code AVP that
   indicates SDS AVP value [TBD] as the reason for sending the CDN.

   If the session initiator receives an ICRP or OCRP that contains a SDS
   AVP with a value other than that requested in the ICRP or OCRP, and
   the session initiator is willing to use the value, the session
   initiator MUST proceed as indicated in [RFC 2661].



Calhoun, McPherson, Peirce                             	[Page 4]





INTERNET DRAFT                                            December 2000


6.1. Session DS AVP (ICRQ, ICRP, OCRQ, OCRP)

   The SDS AVP is encoded as Vendor ID 43, and the Attribute Value is a
   16-bit quantity 2 (the ID 43 reflects 3Com Corporation, it should be
   changed to 0 and an official Attribute Value chosen should this
   specification advance on as standards track).

   Each SDS AVP is encoded as follows:

     Vendor ID = 43
     Attribute = 2

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |M|H|0|0|0|0|    Length         |              43               |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 
   |                2              |           DS Value            |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   This AVP MAY be present in the following message types:  ICRQ, ICRP,
   OCRQ and OCRP.  This AVP MAY be hidden (the H-bit set to 0 or 1) and
   is optional (M-bit is not set 0).  The length (before hiding) of this
   AVP MUST be 8 octets.


7. DS Value Encoding

   The DS value is a left-justified 16-bit field (i.e., the leftmost bit
   signifies bit 0 of the DS value, the right-most bit signifies bit
   15).

   Each DS value is encoded as follows:

    0                   1
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 
   +--+--+--+--+--+--+--+--+--+--+-+ 
   |    DSCP   | | |0|0|0|0|0|0|0|0| 
   +--+--+--+--+--+--+--+--+--+--+-+

   The 6-bit DSCP field represents the codepoint used to select the per-
   hop behavior (PHB), as defined in [RFC 2474].  Bits 6 and 7 are used
   by ECN [RFC 2481] and SHOULD not be mapped.  Bits 8 through 15 are
   reserved for future use and MUST be set to zero.

   Upon successful establishment of an L2TP tunnel control connection or
   individual L2TP session employing the appropriate DS AVP defined in
   this document, a direct mapping of bits 0-7 into the DS field of the
   IP header of packets transmitted on the connection SHOULD occur.




Calhoun, McPherson, Peirce                             	[Page 5]





INTERNET DRAFT                                            December 2000


8. DSCP Selection

   The requirements or rules of each service and DSCP mapping MUST be
   set through administrative policy mechanisms which are outside the
   scope of this document.


9. Packet Reordering and Sequence Numbers

   Sequence numbers are required to be present in all control messages
   and are used to provide reliable delivery on the control connection,
   as defined in [RFC 2661].  While packet reordering is inevitably as
   much a function of the network as it is local traffic conditioning,
   the probability of it occuring when employing the CCDS AVP is same as
   when not employing the AVP.

   This is because the control connection is contained within a single
   behavior aggregate (BA), and as such, all packets within the BA
   SHOULD be provided with similar per-hop behaviors (PHBs) throughout
   the DS domain.

   Data messages MAY use sequence numbers to reorder packets and detect
   lost packets.  Data messages within a given session employing the SDS
   AVP SHOULD not be reordered as all packets within the session are
   contained within a single BA.


10. Crossing Differentiated Services Boundaries

   With an arbitrary number of domains, signaling DSCPs via L2TP poses
   some problems.  Other protocols such as RSVP [RFC 2205] can do this
   because RSVP is supposed to be processed at every node, and hence a
   boundary node can rewrite the DSCPs as it goes by, so that the
   signalled DSCPs are always with respect to whatever DS domain the
   packet happens to be in.

   Note in Section "DS Value Encoding" that a direct mapping of the
   reported DS value to the DS field of IP packets transmitted on the
   connection is NOT required.  By not requiring this flexibility is
   provided in the event that a uni-directional section of a tunnel or
   session wants to use varying DSCP values in each direction.  This is
   especially important when considering multi-DS domains with varying
   PHBs associated with a given BA.

   As a result, it is perfectly acceptable that the outermost DS Field
   of packets arriving on a given control connection or session are not
   marked with a DSCP value that was negotiated during call setup.




Calhoun, McPherson, Peirce                             	[Page 6]





INTERNET DRAFT                                            December 2000


   Currently, however, the preferred model for accommodating multi-
   domain DS seems to be that of deploying traffic (re)classifiers at
   the edges of DS domains and allowing the administrators of those
   domains to coordinate DSCP mappings and associated PHBs as
   contractually determined.


11. TODO

   The intent of this version of the draft is to solicit feedback on how
   L2TP DS should behave.  The following areas have been identified as
   requiring additional work in this document:

    o Add User ID AVP to be used in conjunction with SDS AVP so
      that the LNS can determine which user is requesting the
      connection, and it can lookup it's local or remote database
      for proper authorization parameters.

    o Add support for 'default Session DS'.  Will likely employ
      SDS AVP in SCCRQ/SCCRP for this purpose.

    o Need to add support for multiple DSCPs (similar to RSVP
      DCLASS/RFC

    o Need to provide capability for differing DSCP values to be used
      in each direction.

    o Need to provide consistency between Control Connection and
      Session behavior.

    o Need to generalize wording so that Session payload types other
      than PPP are included as well.

   Presumably, additional issues that require attention will be
   uncovered during the San Diego WG meeting.
















Calhoun, McPherson, Peirce                             	[Page 7]





INTERNET DRAFT                                            December 2000


12. IANA Considerations

   Should this document advance on as standards track official Attribute
   Values need to be assigned for the CCDS and SDS AVPs.


13. Security Considerations

   This encoding in itself raises no security issues. However, users of
   this encoding should consider that modifying a DSCP MAY constitute
   theft or denial of service, so protocols using this encoding MUST be
   adequately protected.  No new security issues beyond those discussed
   in [RFC 2474] and [RFC 2475] are introduced here.


14. Acknowledgements

   Many thanks to David Black, W. Mark Townsley, Wei Luo, Nishit
   Vasavada, Andy Koscinski and John Shriver for their review and
   insightful feedback.


15. References

     [RFC 1661] Simpson, W., "The Point-to-Point Protocol (PPP)", STD
                51, RFC 1661, July 1994.

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

     [RFC 2205] Braden, R., Zhang, L., Berson, S., Herzog, S. and S.
                Jamin, "Resource ReSerVation Protocol (RSVP) -- Version
                1 Functional Specification", RFC 2205, September 1997.

     [RFC 2474] Nichols et al., "Definition of the Differentiated
                Services Field (DS Field) in the IPv4 and IPv6 Headers",
                RFC 2474, December 1998.

     [RFC 2475] Blake et al., "An Architecture for Differentiated
                Services", RFC 2475, December 1998.

     [RFC 2481] Ramakrishnan, K., and Floyd, S., "A Proposal to add
                Explicit Congestion Notification (ECN) to IP", RFC 2481,
                January 1999.

     [RFC 2661] W. Townsley, A. Valencia, A. Rubens, G. Pall, G. Zorn,
                B. Palter, "Layer 2 Tunnel Protocol (L2TP)", RFC 2661,
                August 1999.



Calhoun, McPherson, Peirce                             	[Page 8]





INTERNET DRAFT                                            December 2000


16. Authors' Address

   Pat R. Calhoun
   Network and Security Research Center, Sun Labs
   Sun Microsystems, Inc.
   15 Network Circle
   Menlo Park, California, 94025
   Phone: +1 650.786.7733
   Email: pcalhoun@eng.sun.com

   Danny McPherson
   Amber Networks, Inc.
   2465 Augustine Drive
   Santa Clara, CA  95054
   Phone: +1 408.486.6336
   Email: danny@ambernetworks.com

   Ken Peirce
   Malibu Networks, Inc.
   1035 Suncast Lane, Suite 130
   El Dorado Hills, CA, 95762
   Phone: +1 916.941.8814
   Email: Ken@malibunetworks.com




























Calhoun, McPherson, Peirce                             	[Page 9]



From owner-l2tp@diameter.org  Fri Dec  8 22:56:12 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id WAA29768
	for <l2tpext-archive@odin.ietf.org>; Fri, 8 Dec 2000 22:56:11 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eB92BeK02471
	for l2tp-list; Fri, 8 Dec 2000 18:11:40 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from tcb.net (tcb.net [205.168.100.1])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eB92AxH02461
	for <l2tp@diameter.org>; Fri, 8 Dec 2000 18:11:00 -0800
Received: from sofos.tcb.net (sofos.tcb.net [127.0.0.1])
	by tcb.net (8.9.3/8.9.3) with ESMTP id SAA12689
	for <l2tp@diameter.org>; Fri, 8 Dec 2000 18:22:40 -0700
Message-Id: <200012090122.SAA12689@tcb.net>
X-Mailer: exmh version 2.0.3
To: l2tp@l2tp.net
From: Danny McPherson <danny@tcb.net>
Reply-To: danny@tcb.net
Subject: draft-ietf-l2tpext-ds-01.txt
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Fri, 08 Dec 2000 15:08:22 -0700
Sender: owner-l2tp@diameter.org
Precedence: bulk


I'll post it to internet-drafts@ietf.org after San Diego is over.  

Note especially the TODO section.  Comments welcome, I'll be discussing the 
highlights Thursday.

-danny

========================================================================
Network Working Group                                     Pat R. Calhoun
INTERNET DRAFT                                    Sun Microsystems, Inc.
                                                         Danny McPherson
                                                    Amber Networks, Inc.
                                                              Ken Peirce
December 2000                                      Malibu Networks, Inc.



                 L2TP Differentiated Services Extension
                     <draft-ietf-l2tpext-ds-01.txt>


1. Status of this Memo

   This document is an Internet-Draft and is in full conformance with
   all provisions of Section 10 of RFC2026.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-
   Drafts.

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

   The list of current Internet-Drafts can be accessed at
   http://www.ietf.org/ietf/1id-abstracts.txt.

   The list of Internet-Draft Shadow Directories can be accessed at
   http://www.ietf.org/shadow.html.


2. Abstract

   The Layer Two Tunneling Protocol (L2TP) [RFC 2661] provides a
   standard method for tunneling PPP [RFC 1661] packets.  The current
   specification provides no provisions for supporting Differentiated
   Services (diffserv) [RFC 2474, RFC 2475] over the L2TP control
   connection or subsequent data sessions.  As a result, no standard
   mechanism currently exists within L2TP to provide L2TP protocol
   negotiations for service discrimination.

   This document describes mechanisms which enable L2TP to negotiate
   desired DS values for the L2TP control connection, as well as



Calhoun, McPherson, Peirce                             	[Page 1]





INTERNET DRAFT                                            December 2000


   individual sessions within an L2TP tunnel.


3. Specification of Requirements

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED",  "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in [RFC 2119].


4. Introduction

   The L2TP specification currently provides no mechanism for supporting
   diffserv (DS).  This document describes mechanisms that enable L2TP
   to indicate desired DS values to be associated with an L2TP control
   connection, as well as individual sessions within an L2TP tunnel.

   This document will describe how a set of L2TP peers MAY negotiate a
   set of differential services indicators for a tunnel control
   connection, as well as for individual sessions within the tunnel.

   The actual bit interpretation of the DS field is beyond the scope of
   this document, and is purposefully omitted.  This document is
   concerned only with defining a uniform exchange and subsequent
   mapping mechanism for the DS AVPs.


5. Control Connection Operation

   As defined in [RFC 2661], a control connection operates in-band over
   a tunnel to control the establishment, release, and maintenance of
   sessions and of the tunnel itself.  As such, this document provides a
   mechanism to enable discrimination of L2TP control messages from
   other packets.  For this purpose, we introduce the Control Connection
   DS (CCDS) AVP.

   The presence of the CCDS AVP serves as an indication to the peer (LAC
   or LNS) that the tunnel initiator wishes to use the specified DS
   value on all packets within the tunnel's control connection.

   Upon receipt of a Start-Control-Connection-Request (SCCRQ) containing
   the CCDS AVP, if the tunnel terminator provides no support for the
   CCDS AVP it SHOULD ignore the AVP and send a SCCRP to the tunnel
   initiator without the CCDS AVP.  The tunnel initiator interprets the
   absence of the CCDS AVP in the SCCRP as as an indication that the
   tunnel terminator is incapable of supporting CCDS.

   Upon receipt of a SCCRP that contains no CCDS AVP in response to a



Calhoun, McPherson, Peirce                             	[Page 2]





INTERNET DRAFT                                            December 2000


   SCCRQ that contained a CCDS AVP, if the tunnel initiator wants to
   continue tunnel establishment it sends a SCCCN; otherwise, it sends a
   StopCCN to the tunnel terminator to end the connection.  The StopCCN
   control message MUST contain a Result Code AVP that indicates CCDS
   AVP value [TBD] as the reason for sending the StopCCN.

   If the tunnel terminator provides support for CCDS but is unwilling
   or unable to support the requested DS value the tunnel terminator
   MUST send a SCCRP control message containing a CCDS AVP indicating
   the value it's willing to use.  If the CCDS AVP value is the same as
   the one in the SCCRQ, it signals the acceptence of the requested DS
   value.  If the value is different it serves as a counter-offer by the
   tunnel terminator.

   If the tunnel initiator receives an SCCRP that contains a CCDS AVP
   with a value other than that requested in the SCCRQ, and the tunnel
   initiator is unwilling to use the value, the tunnel initiator MUST
   send a StopCCN control message containing a Result Code AVP that
   indicates CCDS AVP value [TBD] as the reason for sending the StopCCN.


5.1. Control Connection DS AVP (SCCRQ, SCCRP)

   The CCDS AVP is encoded as Vendor ID 43, and the Attribute Value is
   the 16-bit quantity 1 (the ID 43 reflects 3Com Corporation, it should
   be changed to 0 and an official Attribute Value chosen should this
   specification advance on as standards track).

   Each CCDS AVP is encoded as follows:

     Vendor ID = 43
     Attribute = 1

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |M|H|0|0|0|0|    Length         |              43               |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 
   |                1              |           DS Value            |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   This AVP MAY be present in the following message types:  SCCRQ and
   SCCRP.  This AVP MAY be hidden (the H-bit set to 0 or 1) and is
   optional (M-bit not set).  The length (before hiding) of this AVP
   MUST be 8 octets.






Calhoun, McPherson, Peirce                             	[Page 3]





INTERNET DRAFT                                            December 2000


6. Session Operation

   As defined in [RFC 2661], a L2TP session is connection-oriented. The
   LNS and LAC maintain state for each Call that is initiated or
   answered by an LAC. An L2TP Session is created between the LAC and
   LNS when an end-to-end PPP connection is established between a Remote
   System and the LNS.  Datagrams related to the PPP connection are sent
   over the Tunnel between the LAC and LNS. There is a one to one
   relationship between established L2TP Sessions and their associated
   Calls.  As such, this document provides a mechanism to enable
   discrimination for packets within an particular session from other
   packets.  For this purpose, we introduce the Session DS (SDS) AVP.

   The presence of the SDS AVP serves as an indication to the peer (LAC
   or LNS) that the session initiator wishes to use the specified DS
   value on all packets within the session.

   Upon receipt of a Incoming-Call-Request (ICRQ) or Outgoing-Call-
   Request (OCRQ) containing the SDS AVP if the session terminator
   provides no support for the requested DS value a Call-Disconnect-
   Notify (CDN) message MUST be sent to the peer.  The CDN message MUST
   contain a Result Code AVP that indicates SDS AVP value [TBD] as the
   reason for sending the CDN.

   If the session terminator provides support for SDS but is unwilling
   or unable to support the requested DS value the session terminator
   MUST do one of the following:

   1) Send a CDN message containing a Result Code AVP that indicates SDS
   AVP value [TBD] as the reason for sending the CDN.

   2) Send an Incoming-Call-Reply (ICRP) or Outgoing-Call-Reply (OCRP)
   message containing a CDN AVP indicating the DS value the terminator
   is willing to use for the session.

   3) If the session terminator supports the DS value in the SDS AVP
   session establishment MUST continue as defined in [RFC 2661].

   If the session initiator receives an ICRP or OCRP that contains an
   SDS AVP with a value other than that requested in the ICRQ or OCRQ,
   and the session initiator is unwilling to use the value, the session
   initiator MUST send a CDN message containing a Result Code AVP that
   indicates SDS AVP value [TBD] as the reason for sending the CDN.

   If the session initiator receives an ICRP or OCRP that contains a SDS
   AVP with a value other than that requested in the ICRP or OCRP, and
   the session initiator is willing to use the value, the session
   initiator MUST proceed as indicated in [RFC 2661].



Calhoun, McPherson, Peirce                             	[Page 4]





INTERNET DRAFT                                            December 2000


6.1. Session DS AVP (ICRQ, ICRP, OCRQ, OCRP)

   The SDS AVP is encoded as Vendor ID 43, and the Attribute Value is a
   16-bit quantity 2 (the ID 43 reflects 3Com Corporation, it should be
   changed to 0 and an official Attribute Value chosen should this
   specification advance on as standards track).

   Each SDS AVP is encoded as follows:

     Vendor ID = 43
     Attribute = 2

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |M|H|0|0|0|0|    Length         |              43               |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 
   |                2              |           DS Value            |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   This AVP MAY be present in the following message types:  ICRQ, ICRP,
   OCRQ and OCRP.  This AVP MAY be hidden (the H-bit set to 0 or 1) and
   is optional (M-bit is not set 0).  The length (before hiding) of this
   AVP MUST be 8 octets.


7. DS Value Encoding

   The DS value is a left-justified 16-bit field (i.e., the leftmost bit
   signifies bit 0 of the DS value, the right-most bit signifies bit
   15).

   Each DS value is encoded as follows:

    0                   1
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 
   +--+--+--+--+--+--+--+--+--+--+-+ 
   |    DSCP   | | |0|0|0|0|0|0|0|0| 
   +--+--+--+--+--+--+--+--+--+--+-+

   The 6-bit DSCP field represents the codepoint used to select the per-
   hop behavior (PHB), as defined in [RFC 2474].  Bits 6 and 7 are used
   by ECN [RFC 2481] and SHOULD not be mapped.  Bits 8 through 15 are
   reserved for future use and MUST be set to zero.

   Upon successful establishment of an L2TP tunnel control connection or
   individual L2TP session employing the appropriate DS AVP defined in
   this document, a direct mapping of bits 0-7 into the DS field of the
   IP header of packets transmitted on the connection SHOULD occur.




Calhoun, McPherson, Peirce                             	[Page 5]





INTERNET DRAFT                                            December 2000


8. DSCP Selection

   The requirements or rules of each service and DSCP mapping MUST be
   set through administrative policy mechanisms which are outside the
   scope of this document.


9. Packet Reordering and Sequence Numbers

   Sequence numbers are required to be present in all control messages
   and are used to provide reliable delivery on the control connection,
   as defined in [RFC 2661].  While packet reordering is inevitably as
   much a function of the network as it is local traffic conditioning,
   the probability of it occuring when employing the CCDS AVP is same as
   when not employing the AVP.

   This is because the control connection is contained within a single
   behavior aggregate (BA), and as such, all packets within the BA
   SHOULD be provided with similar per-hop behaviors (PHBs) throughout
   the DS domain.

   Data messages MAY use sequence numbers to reorder packets and detect
   lost packets.  Data messages within a given session employing the SDS
   AVP SHOULD not be reordered as all packets within the session are
   contained within a single BA.


10. Crossing Differentiated Services Boundaries

   With an arbitrary number of domains, signaling DSCPs via L2TP poses
   some problems.  Other protocols such as RSVP [RFC 2205] can do this
   because RSVP is supposed to be processed at every node, and hence a
   boundary node can rewrite the DSCPs as it goes by, so that the
   signalled DSCPs are always with respect to whatever DS domain the
   packet happens to be in.

   Note in Section "DS Value Encoding" that a direct mapping of the
   reported DS value to the DS field of IP packets transmitted on the
   connection is NOT required.  By not requiring this flexibility is
   provided in the event that a uni-directional section of a tunnel or
   session wants to use varying DSCP values in each direction.  This is
   especially important when considering multi-DS domains with varying
   PHBs associated with a given BA.

   As a result, it is perfectly acceptable that the outermost DS Field
   of packets arriving on a given control connection or session are not
   marked with a DSCP value that was negotiated during call setup.




Calhoun, McPherson, Peirce                             	[Page 6]





INTERNET DRAFT                                            December 2000


   Currently, however, the preferred model for accommodating multi-
   domain DS seems to be that of deploying traffic (re)classifiers at
   the edges of DS domains and allowing the administrators of those
   domains to coordinate DSCP mappings and associated PHBs as
   contractually determined.


11. TODO

   The intent of this version of the draft is to solicit feedback on how
   L2TP DS should behave.  The following areas have been identified as
   requiring additional work in this document:

    o Add User ID AVP to be used in conjunction with SDS AVP so
      that the LNS can determine which user is requesting the
      connection, and it can lookup it's local or remote database
      for proper authorization parameters.

    o Add support for 'default Session DS'.  Will likely employ
      SDS AVP in SCCRQ/SCCRP for this purpose.

    o Need to add support for multiple DSCPs (similar to RSVP
      DCLASS/RFC

    o Need to provide capability for differing DSCP values to be used
      in each direction.

    o Need to provide consistency between Control Connection and
      Session behavior.

    o Need to generalize wording so that Session payload types other
      than PPP are included as well.

   Presumably, additional issues that require attention will be
   uncovered during the San Diego WG meeting.
















Calhoun, McPherson, Peirce                             	[Page 7]





INTERNET DRAFT                                            December 2000


12. IANA Considerations

   Should this document advance on as standards track official Attribute
   Values need to be assigned for the CCDS and SDS AVPs.


13. Security Considerations

   This encoding in itself raises no security issues. However, users of
   this encoding should consider that modifying a DSCP MAY constitute
   theft or denial of service, so protocols using this encoding MUST be
   adequately protected.  No new security issues beyond those discussed
   in [RFC 2474] and [RFC 2475] are introduced here.


14. Acknowledgements

   Many thanks to David Black, W. Mark Townsley, Wei Luo, Nishit
   Vasavada, Andy Koscinski and John Shriver for their review and
   insightful feedback.


15. References

     [RFC 1661] Simpson, W., "The Point-to-Point Protocol (PPP)", STD
                51, RFC 1661, July 1994.

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

     [RFC 2205] Braden, R., Zhang, L., Berson, S., Herzog, S. and S.
                Jamin, "Resource ReSerVation Protocol (RSVP) -- Version
                1 Functional Specification", RFC 2205, September 1997.

     [RFC 2474] Nichols et al., "Definition of the Differentiated
                Services Field (DS Field) in the IPv4 and IPv6 Headers",
                RFC 2474, December 1998.

     [RFC 2475] Blake et al., "An Architecture for Differentiated
                Services", RFC 2475, December 1998.

     [RFC 2481] Ramakrishnan, K., and Floyd, S., "A Proposal to add
                Explicit Congestion Notification (ECN) to IP", RFC 2481,
                January 1999.

     [RFC 2661] W. Townsley, A. Valencia, A. Rubens, G. Pall, G. Zorn,
                B. Palter, "Layer 2 Tunnel Protocol (L2TP)", RFC 2661,
                August 1999.



Calhoun, McPherson, Peirce                             	[Page 8]





INTERNET DRAFT                                            December 2000


16. Authors' Address

   Pat R. Calhoun
   Network and Security Research Center, Sun Labs
   Sun Microsystems, Inc.
   15 Network Circle
   Menlo Park, California, 94025
   Phone: +1 650.786.7733
   Email: pcalhoun@eng.sun.com

   Danny McPherson
   Amber Networks, Inc.
   2465 Augustine Drive
   Santa Clara, CA  95054
   Phone: +1 408.486.6336
   Email: danny@ambernetworks.com

   Ken Peirce
   Malibu Networks, Inc.
   1035 Suncast Lane, Suite 130
   El Dorado Hills, CA, 95762
   Phone: +1 916.941.8814
   Email: Ken@malibunetworks.com




























Calhoun, McPherson, Peirce                             	[Page 9]



From owner-l2tp@diameter.org  Sun Dec 10 20:54:35 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id UAA29150
	for <l2tpext-archive@odin.ietf.org>; Sun, 10 Dec 2000 20:54:34 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBB0Dgo10581
	for l2tp-list; Sun, 10 Dec 2000 16:13:42 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBB0DVH10575
	for <l2tp@l2tp.net>; Sun, 10 Dec 2000 16:13:31 -0800
Received: from eins.siemens.at (eins.siemens.at [193.81.246.11]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id PAA16792 for <l2tp@ipsec.org>; Sun, 10 Dec 2000 15:21:22 -0800 (PST)
Received: from scesie13.sie.siemens.at (forix [10.1.140.2])
	by eins.siemens.at  with ESMTP id eBANKnk25415;
	Mon, 11 Dec 2000 00:20:49 +0100
Received: (from smap@localhost)
	by scesie13.sie.siemens.at (8.9.3/8.9.3) id AAA07753;
	Mon, 11 Dec 2000 00:20:48 +0100 (MET)
Received: from vies141a.sie.siemens.at(195.1.196.56) by scesie13 via smap (V2.0beta)
	id xma007708; Mon, 11 Dec 00 00:20:45 +0100
Received: by vies141a.sie.siemens.at with Internet Mail Service (5.5.2650.21)
	id <YNL1C90V>; Mon, 11 Dec 2000 00:20:42 +0100
Message-ID: <D9F2B9CD7BD5D21196BC0800060D9ED604373018@vies186a.sie.siemens.at>
From: Klausberger Walter <walter.klausberger@siemens.at>
To: "'Dave Thaler '" <dthaler@Exchange.Microsoft.com>,
        "'Evan Caves '"
	 <evan@occamnetworks.com>
Cc: "'l2tp@ipsec.org '" <l2tp@ipsec.org>, "'kzm@cisco.com '"
	 <kzm@cisco.com>
Subject: AW: AW: ifStackTable layering in L2TP
Date: Mon, 11 Dec 2000 00:20:42 +0100
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C062FF.D00F47BE"
Sender: owner-l2tp@diameter.org
Precedence: bulk

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_001_01C062FF.D00F47BE
Content-Type: text/plain;
	charset="iso-8859-1"

Hi all,

sorry for my absence, but I was on a short vacation and now I'm in San
Diego. I must admit, I'm a little confused and I'm not shure if everyone
understands it the way Dave does.

Are there any responses from already existing implementations?

I do not fully agree to the argument that a forwarder (like IP router) is
the same as a multiplexer (like MLP). A multiplexer normally works with a
fixed set of connections (well I know they can be teared on and so on). It
multiplexes packets comming from a higher layer to several lower layers or
the other way round. A forwarder just takes packets from a lower layer and
puts them to another lower layer depending on some rules. In my opinion
Daves approach shows new aspects. I hope all others see ist the same way!? 

On the other hand I agree with Dave that a tunnel can be seen as a virtual
interface.

So in L2TP case the tunnel IF would sit on top of an invisible forwarder and
PPP would be the forwarding layer (in LAC case), e.g.

        ------PPP-------
        |              |
    DS0(phys. IF)    tunnel
                       |
             (IP router, invisible)
                       |
                    ethernet

or should it be


       PPP               tunnel
        |                   |
    DS0(phys. IF) (IP router, invisible)
                            |
                         ethernet

or anything else?

We are loosing the context between the physical interface, where the PPP
session comes in (DS0) and the tunnel this session goes to. This information
is now provided with the L2TP MIB only.

As PPP now is the forwarding layer should it be visible on top of the DS0
layer, or only if the PPP connection is not tunneled.

>In the multipath example, the tunnel would be layered over ALL
>the interfaces which have potential routes over them.  In this
>sense, it's very analogous to multilink PPP, among other
>things.

Dave, do I understand it correct that regardless of the routing entry a
tunnel IF should sit on all possible interfaces it may be routed over? That
would also be a new aspect to me.

I think these are quite basic questions and I'm unsure how to proceed. Maybe
there should be a clarification in the L2TP MIB draft(I don't want to change
the tunnel MIB RFC, though it would be a good idea to add a clarification in
case of a new version)

with best regards
Walter

------_=_NextPart_001_01C062FF.D00F47BE
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Diso-8859-1">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
5.5.2650.12">
<TITLE>AW: AW: ifStackTable layering in L2TP</TITLE>
</HEAD>
<BODY>

<P><FONT SIZE=3D2>Hi all,</FONT>
</P>

<P><FONT SIZE=3D2>sorry for my absence, but I was on a short vacation =
and now I'm in San Diego. I must admit, I'm a little confused and I'm =
not shure if everyone understands it the way Dave does.</FONT></P>

<P><FONT SIZE=3D2>Are there any responses from already existing =
implementations?</FONT>
</P>

<P><FONT SIZE=3D2>I do not fully agree to the argument that a forwarder =
(like IP router) is the same as a multiplexer (like MLP). A multiplexer =
normally works with a fixed set of connections (well I know they can be =
teared on and so on). It multiplexes packets comming from a higher =
layer to several lower layers or the other way round. A forwarder just =
takes packets from a lower layer and puts them to another lower layer =
depending on some rules. In my opinion Daves approach shows new =
aspects. I hope all others see ist the same way!? </FONT></P>

<P><FONT SIZE=3D2>On the other hand I agree with Dave that a tunnel can =
be seen as a virtual interface.</FONT>
</P>

<P><FONT SIZE=3D2>So in L2TP case the tunnel IF would sit on top of an =
invisible forwarder and PPP would be the forwarding layer (in LAC =
case), e.g.</FONT></P>

<P><FONT SIZE=3D2>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
------PPP-------</FONT>
<BR><FONT SIZE=3D2>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
|&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp; |</FONT>
<BR><FONT SIZE=3D2>&nbsp;&nbsp;&nbsp; DS0(phys. IF)&nbsp;&nbsp;&nbsp; =
tunnel</FONT>
<BR><FONT =
SIZE=3D2>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
|</FONT>
<BR><FONT =
SIZE=3D2>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp; (IP router, invisible)</FONT>
<BR><FONT =
SIZE=3D2>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
|</FONT>
<BR><FONT =
SIZE=3D2>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ethernet</FONT>
</P>

<P><FONT SIZE=3D2>or should it be</FONT>
</P>
<BR>

<P><FONT SIZE=3D2>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
PPP&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp; tunnel</FONT>
<BR><FONT SIZE=3D2>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
|&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |</FONT>
<BR><FONT SIZE=3D2>&nbsp;&nbsp;&nbsp; DS0(phys. IF) (IP router, =
invisible)</FONT>
<BR><FONT =
SIZE=3D2>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp; |</FONT>
<BR><FONT =
SIZE=3D2>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp; ethernet</FONT>
</P>

<P><FONT SIZE=3D2>or anything else?</FONT>
</P>

<P><FONT SIZE=3D2>We are loosing the context between the physical =
interface, where the PPP session comes in (DS0) and the tunnel this =
session goes to. This information is now provided with the L2TP MIB =
only.</FONT></P>

<P><FONT SIZE=3D2>As PPP now is the forwarding layer should it be =
visible on top of the DS0 layer, or only if the PPP connection is not =
tunneled.</FONT></P>

<P><FONT SIZE=3D2>&gt;In the multipath example, the tunnel would be =
layered over ALL</FONT>
<BR><FONT SIZE=3D2>&gt;the interfaces which have potential routes over =
them.&nbsp; In this</FONT>
<BR><FONT SIZE=3D2>&gt;sense, it's very analogous to multilink PPP, =
among other</FONT>
<BR><FONT SIZE=3D2>&gt;things.</FONT>
</P>

<P><FONT SIZE=3D2>Dave, do I understand it correct that regardless of =
the routing entry a tunnel IF should sit on all possible interfaces it =
may be routed over? That would also be a new aspect to me.</FONT></P>

<P><FONT SIZE=3D2>I think these are quite basic questions and I'm =
unsure how to proceed. Maybe there should be a clarification in the =
L2TP MIB draft(I don't want to change the tunnel MIB RFC, though it =
would be a good idea to add a clarification in case of a new =
version)</FONT></P>

<P><FONT SIZE=3D2>with best regards</FONT>
<BR><FONT SIZE=3D2>Walter</FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C062FF.D00F47BE--


From owner-l2tp@diameter.org  Mon Dec 11 14:07:13 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id OAA23176
	for <l2tpext-archive@odin.ietf.org>; Mon, 11 Dec 2000 14:07:12 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBBHof811837
	for l2tp-list; Mon, 11 Dec 2000 09:50:41 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from eins.siemens.at (eins.siemens.at [193.81.246.11])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBBHoaH11831
	for <l2tp@l2tp.net>; Mon, 11 Dec 2000 09:50:37 -0800
Received: from scesie13.sie.siemens.at (forix [10.1.140.2])
	by eins.siemens.at  with ESMTP id eBBGwVk10167
	for <l2tp@l2tp.net>; Mon, 11 Dec 2000 17:58:31 +0100
Received: (from smap@localhost)
	by scesie13.sie.siemens.at (8.9.3/8.9.3) id RAA23329
	for <l2tp@l2tp.net>; Mon, 11 Dec 2000 17:58:30 +0100 (MET)
Received: from vies141a.sie.siemens.at(195.1.196.56) by scesie13 via smap (V2.0beta)
	id xma023152; Mon, 11 Dec 00 17:58:01 +0100
Received: by vies141a.sie.siemens.at with Internet Mail Service (5.5.2650.21)
	id <YNL11B6F>; Mon, 11 Dec 2000 17:57:59 +0100
Message-ID: <D9F2B9CD7BD5D21196BC0800060D9ED60437301B@vies186a.sie.siemens.at>
From: Klausberger Walter <walter.klausberger@siemens.at>
To: "'l2tp@l2tp.net '" <l2tp@l2tp.net>
Subject: RE: draft-ietf-l2tpext-ds-01.txt
Date: Mon, 11 Dec 2000 17:57:58 +0100
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C06393.832F7D26"
Sender: owner-l2tp@diameter.org
Precedence: bulk

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_001_01C06393.832F7D26
Content-Type: text/plain

Hi folks,

I like the idea of negotiating QoS parameters during tunnel/session
establishment, but I asked myself all the time, where the DS parameters will
be configured for a special session.

There are parameters planned in the L2TP MIB, but in LAC case the parameters
often come frome an AAA server like RADIUS. Shouldn't there be an extension
in the tunneling parameters?

Another thing is the Tequila group. They want to provide SLA negotiation for
RAS user via PPP IPCP extension. This should also be mapped onto tunnel
sessions (if granted). To grant such features again new AAA attributes could
be used.

Is anyone already thinking about that, or is someone willing to discuss
these matters in more detail?

see you at the IETF
with best regards
Walter

------_=_NextPart_001_01C06393.832F7D26
Content-Type: text/html
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3DUS-ASCII">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
5.5.2650.12">
<TITLE>RE: draft-ietf-l2tpext-ds-01.txt</TITLE>
</HEAD>
<BODY>

<P><FONT SIZE=3D2>Hi folks,</FONT>
</P>

<P><FONT SIZE=3D2>I like the idea of negotiating QoS parameters during =
tunnel/session establishment, but I asked myself all the time, where =
the DS parameters will be configured for a special session.</FONT></P>

<P><FONT SIZE=3D2>There are parameters planned in the L2TP MIB, but in =
LAC case the parameters often come frome an AAA server like RADIUS. =
Shouldn't there be an extension in the tunneling parameters?</FONT></P>

<P><FONT SIZE=3D2>Another thing is the Tequila group. They want to =
provide SLA negotiation for RAS user via PPP IPCP extension. This =
should also be mapped onto tunnel sessions (if granted). To grant such =
features again new AAA attributes could be used.</FONT></P>

<P><FONT SIZE=3D2>Is anyone already thinking about that, or is someone =
willing to discuss these matters in more detail?</FONT>
</P>

<P><FONT SIZE=3D2>see you at the IETF</FONT>
<BR><FONT SIZE=3D2>with best regards</FONT>
<BR><FONT SIZE=3D2>Walter</FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C06393.832F7D26--


From owner-l2tp@diameter.org  Mon Dec 11 19:25:51 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id TAA14016
	for <l2tpext-archive@odin.ietf.org>; Mon, 11 Dec 2000 19:25:51 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBBMirv12238
	for l2tp-list; Mon, 11 Dec 2000 14:44:53 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com ([155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBBMimH12232
	for <l2tp@l2tp.net>; Mon, 11 Dec 2000 14:44:48 -0800
Received: from occamnetworks.com (mail.occamnetworks.com [216.64.159.194]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id NAA19596 for <l2tp@ipsec.org>; Mon, 11 Dec 2000 13:51:28 -0800 (PST)
Received: from occamnetworks.com (honeycombs.occamnetworks.com [192.168.2.202])
	by occamnetworks.com (8.10.1/8.10.1) with ESMTP id eBBLnVQ10890;
	Mon, 11 Dec 2000 13:49:31 -0800
Message-ID: <3A354BD1.82D0B232@occamnetworks.com>
Date: Mon, 11 Dec 2000 13:49:05 -0800
From: Evan Caves <evan@occamnetworks.com>
Organization: Occam Networks, Inc.
X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.2.16-3 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: Klausberger Walter <walter.klausberger@siemens.at>
CC: "'Dave Thaler '" <dthaler@Exchange.Microsoft.com>,
        "'l2tp@ipsec.org '" <l2tp@ipsec.org>,
        "'kzm@cisco.com '" <kzm@cisco.com>
Subject: Re: AW: AW: ifStackTable layering in L2TP
References: <D9F2B9CD7BD5D21196BC0800060D9ED604373018@vies186a.sie.siemens.at>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> Klausberger Walter wrote:
> 
> Hi all,
> 
> sorry for my absence, but I was on a short vacation and now I'm in San
> Diego. I must admit, I'm a little confused and I'm not shure if
> everyone understands it the way Dave does.
> 
> Are there any responses from already existing implementations?
> 
> I do not fully agree to the argument that a forwarder (like IP router)
> is the same as a multiplexer (like MLP). A multiplexer normally works
> with a fixed set of connections (well I know they can be teared on and
> so on). It multiplexes packets comming from a higher layer to several
> lower layers or the other way round. A forwarder just takes packets
> from a lower layer and puts them to another lower layer depending on
> some rules. In my opinion Daves approach shows new aspects. I hope all
> others see ist the same way!?
> 
> On the other hand I agree with Dave that a tunnel can be seen as a
> virtual interface.
> 
> So in L2TP case the tunnel IF would sit on top of an invisible
> forwarder and PPP would be the forwarding layer (in LAC case), e.g.
> 
>         ------PPP-------
>         |              |
>     DS0(phys. IF)    tunnel
>                        |
>              (IP router, invisible)
>                        |
>                     ethernet

This is what is currently presented in the MIB except in its model
we do not call the forwarding layer a 'PPP forwarding layer', it
is modeled as a switching element. The reason being that on the LAC
side one could argue that PPP isn't really there anymore once the 
tunnel is up (it sort of steps aside).

> 
> or should it be
> 
>        PPP               tunnel
>         |                   |
>     DS0(phys. IF) (IP router, invisible)
>                             |
>                          ethernet
> 
> or anything else?

I would argue against this. And besides in this picture how do PPP
packets end up in the tunnel?

> 
> We are loosing the context between the physical interface, where the
> PPP session comes in (DS0) and the tunnel this session goes to. This
> information is now provided with the L2TP MIB only.

Yup. Just like you need the CIDR MIB (or whatever it is called now)
to dump your routing tables.

> 
> As PPP now is the forwarding layer should it be visible on top of the
> DS0 layer, or only if the PPP connection is not tunneled.

PPP isn't even in the picture on the LAC in my model. If it was
you would expect to be able to manage it like any other PPP interface
(which is definitely not the case).

> 
> >In the multipath example, the tunnel would be layered over ALL
> >the interfaces which have potential routes over them.  In this
> >sense, it's very analogous to multilink PPP, among other
> >things.
> 
> Dave, do I understand it correct that regardless of the routing entry
> a tunnel IF should sit on all possible interfaces it may be routed
> over? That would also be a new aspect to me.
> 
> I think these are quite basic questions and I'm unsure how to proceed.
> Maybe there should be a clarification in the L2TP MIB draft(I don't
> want to change the tunnel MIB RFC, though it would be a good idea to
> add a clarification in case of a new version)
> 

And I don't think that the L2TP MIB should be describing how the IP 
Tunnel MIB works (but I will probably loose that argument).

evan
-


From owner-l2tp@diameter.org  Mon Dec 11 19:52:41 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id TAA17386
	for <l2tpext-archive@odin.ietf.org>; Mon, 11 Dec 2000 19:52:40 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBBMlxA12249
	for l2tp-list; Mon, 11 Dec 2000 14:47:59 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from occamnetworks.com (mail.occamnetworks.com [216.64.159.194])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBBMluH12243
	for <l2tp@l2tp.net>; Mon, 11 Dec 2000 14:47:56 -0800
Received: from occamnetworks.com (honeycombs.occamnetworks.com [192.168.2.202])
	by occamnetworks.com (8.10.1/8.10.1) with ESMTP id eBBLtVQ11114;
	Mon, 11 Dec 2000 13:55:31 -0800
Message-ID: <3A354D39.15A4B973@occamnetworks.com>
Date: Mon, 11 Dec 2000 13:55:05 -0800
From: Evan Caves <evan@occamnetworks.com>
Organization: Occam Networks, Inc.
X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.2.16-3 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: Klausberger Walter <walter.klausberger@siemens.at>
CC: "'l2tp@l2tp.net '" <l2tp@l2tp.net>
Subject: Re: draft-ietf-l2tpext-ds-01.txt
References: <D9F2B9CD7BD5D21196BC0800060D9ED60437301B@vies186a.sie.siemens.at>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> Klausberger Walter wrote:
> 
> Hi folks,
> 
> I like the idea of negotiating QoS parameters during tunnel/session
> establishment, but I asked myself all the time, where the DS
> parameters will be configured for a special session.

They will most likely be returned by a AAA server.

> 
> There are parameters planned in the L2TP MIB,

Remind me again about these parameters :-).

> but in LAC case the
> parameters often come frome an AAA server like RADIUS. Shouldn't there
> be an extension in the tunneling parameters?

Yup.

> 
> Another thing is the Tequila group. 

The what group?

> They want to provide SLA
> negotiation for RAS user via PPP IPCP extension. 

Uggghhh.

> This should also be
> mapped onto tunnel sessions (if granted). To grant such features again
> new AAA attributes could be used.
> 
> Is anyone already thinking about that, or is someone willing to
> discuss these matters in more detail?
> 
> see you at the IETF
> with best regards
> Walter


From owner-l2tp@diameter.org  Tue Dec 12 14:08:04 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id OAA22619
	for <l2tpext-archive@odin.ietf.org>; Tue, 12 Dec 2000 14:08:02 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBCHLQk13366
	for l2tp-list; Tue, 12 Dec 2000 09:21:26 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBCHLMH13360
	for <l2tp@l2tp.net>; Tue, 12 Dec 2000 09:21:22 -0800
Received: from eins.siemens.at (eins.siemens.at [193.81.246.11]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id IAA21777 for <l2tp@ipsec.org>; Tue, 12 Dec 2000 08:29:16 -0800 (PST)
Received: from scesie13.sie.siemens.at (forix [10.1.140.2])
	by eins.siemens.at  with ESMTP id eBCGSck22619;
	Tue, 12 Dec 2000 17:28:38 +0100
Received: (from smap@localhost)
	by scesie13.sie.siemens.at (8.9.3/8.9.3) id RAA15609;
	Tue, 12 Dec 2000 17:28:36 +0100 (MET)
Received: from vies194a.sie.siemens.at(195.1.196.94) by scesie13 via smap (V2.0beta)
	id xmab08071; Tue, 12 Dec 00 15:39:15 +0100
Received: by vies194a.sie.siemens.at with Internet Mail Service (5.5.2650.21)
	id <YYGJVJZ2>; Tue, 12 Dec 2000 15:38:59 +0100
Message-ID: <D9F2B9CD7BD5D21196BC0800060D9ED60437301D@vies186a.sie.siemens.at>
From: Klausberger Walter <walter.klausberger@siemens.at>
To: "'Evan Caves '" <evan@occamnetworks.com>
Cc: "''Dave Thaler ' '" <dthaler@Exchange.Microsoft.com>,
        "''l2tp@ipsec.org ' '" <l2tp@ipsec.org>,
        "''kzm@cisco.com ' '"
	 <kzm@cisco.com>
Subject: AW: AW: AW: ifStackTable layering in L2TP
Date: Tue, 12 Dec 2000 00:43:29 +0100
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C06449.41D798D4"
Sender: owner-l2tp@diameter.org
Precedence: bulk

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_001_01C06449.41D798D4
Content-Type: text/plain

 

-----Originalnachricht-----
Von: Evan Caves
An: Klausberger Walter
Cc: 'Dave Thaler '; 'l2tp@ipsec.org '; 'kzm@cisco.com '
Gesendet: 11.12.00 22:49
Betreff: Re: AW: AW: ifStackTable layering in L2TP

>> Klausberger Walter wrote:
>> 
>> Hi all,
>> 
>> sorry for my absence, but I was on a short vacation and now I'm in San
>> Diego. I must admit, I'm a little confused and I'm not shure if
>> everyone understands it the way Dave does.
>> 
>> Are there any responses from already existing implementations?
>> 
>> I do not fully agree to the argument that a forwarder (like IP router)
>> is the same as a multiplexer (like MLP). A multiplexer normally works
>> with a fixed set of connections (well I know they can be teared on and
>> so on). It multiplexes packets comming from a higher layer to several
>> lower layers or the other way round. A forwarder just takes packets
>> from a lower layer and puts them to another lower layer depending on
>> some rules. In my opinion Daves approach shows new aspects. I hope all
>> others see ist the same way!?
>> 
>> On the other hand I agree with Dave that a tunnel can be seen as a
>> virtual interface.
>> 
>> So in L2TP case the tunnel IF would sit on top of an invisible
>> forwarder and PPP would be the forwarding layer (in LAC case), e.g.
>> 
>>         ------PPP-------
>>         |              |
>>     DS0(phys. IF)    tunnel
>>                        |
>>              (IP router, invisible)
>>                        |
>>                     ethernet

>This is what is currently presented in the MIB except in its model
>we do not call the forwarding layer a 'PPP forwarding layer', it
>is modeled as a switching element. The reason being that on the LAC
>side one could argue that PPP isn't really there anymore once the 
>tunnel is up (it sort of steps aside).

ok, I agree with you on that

>> 
>> or should it be
>> 
>>        PPP               tunnel
>>         |                   |
>>     DS0(phys. IF) (IP router, invisible)
>>                             |
>>                          ethernet
>> 
>> or anything else?

>I would argue against this. And besides in this picture how do PPP
>packets end up in the tunnel?

agreed

>> 
>> We are loosing the context between the physical interface, where the
>> PPP session comes in (DS0) and the tunnel this session goes to. This
>> information is now provided with the L2TP MIB only.

>Yup. Just like you need the CIDR MIB (or whatever it is called now)
>to dump your routing tables.

agreed

>> 
>> As PPP now is the forwarding layer should it be visible on top of the
>> DS0 layer, or only if the PPP connection is not tunneled.

>PPP isn't even in the picture on the LAC in my model. If it was
>you would expect to be able to manage it like any other PPP interface
>(which is definitely not the case).

A LAC does not necessarily tunnel a PPP session (you might say it is not a
LAC then, but that's how it's written in the L2TP RFC). If a PPP session is
terminated in a LAC shouldn't the PPP layer be visible?

Thinking about all this, I have concerns about the stack layering MIB. When
I have an AAL5 layer (assume PPP over AAL5), I know by informations from the
L2TP MIB that a PPP session is coming from the AAL5 interface and going into
a certain tunnel.

How do I know if IP packets come from that tunnel? To me there is something
missing, or does the CIDR MIB provide enough information (I'm not so
familiar with this MIB)?

>> 
>> >In the multipath example, the tunnel would be layered over ALL
>> >the interfaces which have potential routes over them.  In this
>> >sense, it's very analogous to multilink PPP, among other
>> >things.
>> 
>> Dave, do I understand it correct that regardless of the routing entry
>> a tunnel IF should sit on all possible interfaces it may be routed
>> over? That would also be a new aspect to me.
>> 
>> I think these are quite basic questions and I'm unsure how to proceed.
>> Maybe there should be a clarification in the L2TP MIB draft(I don't
>> want to change the tunnel MIB RFC, though it would be a good idea to
>> add a clarification in case of a new version)
>> 

>And I don't think that the L2TP MIB should be describing how the IP 
>Tunnel MIB works (but I will probably loose that argument).

I think it could be an issue, otherwise there might be a lot of other people
asking the same questions later, but I agree it is a matter of the Tunnel
MIB description...

Walter
-

------_=_NextPart_001_01C06449.41D798D4
Content-Type: text/html
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3DUS-ASCII">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
5.5.2650.12">
<TITLE>AW: AW: AW: ifStackTable layering in L2TP</TITLE>
</HEAD>
<BODY>

<P><FONT SIZE=3D2>&nbsp;</FONT>
</P>

<P><FONT SIZE=3D2>-----Originalnachricht-----</FONT>
<BR><FONT SIZE=3D2>Von: Evan Caves</FONT>
<BR><FONT SIZE=3D2>An: Klausberger Walter</FONT>
<BR><FONT SIZE=3D2>Cc: 'Dave Thaler '; 'l2tp@ipsec.org '; =
'kzm@cisco.com '</FONT>
<BR><FONT SIZE=3D2>Gesendet: 11.12.00 22:49</FONT>
<BR><FONT SIZE=3D2>Betreff: Re: AW: AW: ifStackTable layering in =
L2TP</FONT>
</P>

<P><FONT SIZE=3D2>&gt;&gt; Klausberger Walter wrote:</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; Hi all,</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; sorry for my absence, but I was on a short =
vacation and now I'm in San</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; Diego. I must admit, I'm a little confused =
and I'm not shure if</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; everyone understands it the way Dave =
does.</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; Are there any responses from already =
existing implementations?</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; I do not fully agree to the argument that a =
forwarder (like IP router)</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; is the same as a multiplexer (like MLP). A =
multiplexer normally works</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; with a fixed set of connections (well I =
know they can be teared on and</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; so on). It multiplexes packets comming from =
a higher layer to several</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; lower layers or the other way round. A =
forwarder just takes packets</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; from a lower layer and puts them to another =
lower layer depending on</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; some rules. In my opinion Daves approach =
shows new aspects. I hope all</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; others see ist the same way!?</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; On the other hand I agree with Dave that a =
tunnel can be seen as a</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; virtual interface.</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; So in L2TP case the tunnel IF would sit on =
top of an invisible</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; forwarder and PPP would be the forwarding =
layer (in LAC case), e.g.</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT =
SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
------PPP-------</FONT>
<BR><FONT =
SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
|&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp; |</FONT>
<BR><FONT SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp; DS0(phys. =
IF)&nbsp;&nbsp;&nbsp; tunnel</FONT>
<BR><FONT =
SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp; |</FONT>
<BR><FONT =
SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp; (IP router, invisible)</FONT>
<BR><FONT =
SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp; |</FONT>
<BR><FONT =
SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
ethernet</FONT>
</P>

<P><FONT SIZE=3D2>&gt;This is what is currently presented in the MIB =
except in its model</FONT>
<BR><FONT SIZE=3D2>&gt;we do not call the =18forwarding layer a 'PPP =
forwarding layer', it</FONT>
<BR><FONT SIZE=3D2>&gt;is modeled as a switching element. The reason =
being that on the LAC</FONT>
<BR><FONT SIZE=3D2>&gt;side one could argue that PPP isn't really there =
anymore once the </FONT>
<BR><FONT SIZE=3D2>&gt;tunnel is up (it sort of steps aside).</FONT>
</P>

<P><FONT SIZE=3D2>ok, I agree with you on that</FONT>
</P>

<P><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; or should it be</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
PPP&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp; tunnel</FONT>
<BR><FONT =
SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
|&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |</FONT>
<BR><FONT SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp; DS0(phys. IF) (IP =
router, invisible)</FONT>
<BR><FONT =
SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |</FONT>
<BR><FONT =
SIZE=3D2>&gt;&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp; ethernet</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; or anything else?</FONT>
</P>

<P><FONT SIZE=3D2>&gt;I would argue against this. And besides in this =
picture how do PPP</FONT>
<BR><FONT SIZE=3D2>&gt;packets end up in the tunnel?</FONT>
</P>

<P><FONT SIZE=3D2>agreed</FONT>
</P>

<P><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; We are loosing the context between the =
physical interface, where the</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; PPP session comes in (DS0) and the tunnel =
this session goes to. This</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; information is now provided with the L2TP =
MIB only.</FONT>
</P>

<P><FONT SIZE=3D2>&gt;Yup. Just like you need the CIDR MIB (or whatever =
it is called now)</FONT>
<BR><FONT SIZE=3D2>&gt;to dump your routing tables.</FONT>
</P>

<P><FONT SIZE=3D2>agreed</FONT>
</P>

<P><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; As PPP now is the forwarding layer should =
it be visible on top of the</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; DS0 layer, or only if the PPP connection is =
not tunneled.</FONT>
</P>

<P><FONT SIZE=3D2>&gt;PPP isn't even in the picture on the LAC in my =
model. If it was</FONT>
<BR><FONT SIZE=3D2>&gt;you would expect to be able to manage it like =
any other PPP interface</FONT>
<BR><FONT SIZE=3D2>&gt;(which is definitely not the case).</FONT>
</P>

<P><FONT SIZE=3D2>A LAC does not necessarily tunnel a PPP session (you =
might say it is not a LAC then, but that's how it's written in the L2TP =
RFC). If a PPP session is terminated in a LAC shouldn't the PPP layer =
be visible?</FONT></P>

<P><FONT SIZE=3D2>Thinking about all this, I have concerns about the =
stack layering MIB. When I have an AAL5 layer (assume PPP over AAL5), I =
know by informations from the L2TP MIB that a PPP session is coming =
from the AAL5 interface and going into a certain tunnel.</FONT></P>

<P><FONT SIZE=3D2>How do I know if IP packets come from that tunnel? To =
me there is something missing, or does the CIDR MIB provide enough =
information (I'm not so familiar with this MIB)?</FONT></P>

<P><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; &gt;In the multipath example, the tunnel =
would be layered over ALL</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; &gt;the interfaces which have potential =
routes over them.&nbsp; In this</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; &gt;sense, it's very analogous to multilink =
PPP, among other</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; &gt;things.</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; Dave, do I understand it correct that =
regardless of the routing entry</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; a tunnel IF should sit on all possible =
interfaces it may be routed</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; over? That would also be a new aspect to =
me.</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; I think these are quite basic questions and =
I'm unsure how to proceed.</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; Maybe there should be a clarification in =
the L2TP MIB draft(I don't</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; want to change the tunnel MIB RFC, though =
it would be a good idea to</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; add a clarification in case of a new =
version)</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
</P>

<P><FONT SIZE=3D2>&gt;And I don't think that the L2TP MIB should be =
describing how the IP </FONT>
<BR><FONT SIZE=3D2>&gt;Tunnel MIB works (but I will probably loose that =
argument).</FONT>
</P>

<P><FONT SIZE=3D2>I think it could be an issue, otherwise there might =
be a lot of other people asking the same questions later, but I agree =
it is a matter of the Tunnel MIB description...</FONT></P>

<P><FONT SIZE=3D2>Walter</FONT>
<BR><FONT SIZE=3D2>-</FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C06449.41D798D4--


From owner-l2tp@diameter.org  Tue Dec 12 14:08:49 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id OAA22770
	for <l2tpext-archive@odin.ietf.org>; Tue, 12 Dec 2000 14:08:48 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBCHKnB13358
	for l2tp-list; Tue, 12 Dec 2000 09:20:49 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from eins.siemens.at (eins.siemens.at [193.81.246.11])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBCHKTH13348
	for <l2tp@l2tp.net>; Tue, 12 Dec 2000 09:20:30 -0800
Received: from scesie13.sie.siemens.at (forix [10.1.140.2])
	by eins.siemens.at  with ESMTP id eBCGSQk22564;
	Tue, 12 Dec 2000 17:28:26 +0100
Received: (from smap@localhost)
	by scesie13.sie.siemens.at (8.9.3/8.9.3) id RAA15342;
	Tue, 12 Dec 2000 17:28:25 +0100 (MET)
Received: from vies194a.sie.siemens.at(195.1.196.94) by scesie13 via smap (V2.0beta)
	id xmae08061; Tue, 12 Dec 00 15:39:05 +0100
Received: by vies194a.sie.siemens.at with Internet Mail Service (5.5.2650.21)
	id <YYGJVJZF>; Tue, 12 Dec 2000 15:38:58 +0100
Message-ID: <D9F2B9CD7BD5D21196BC0800060D9ED60437301C@vies186a.sie.siemens.at>
From: Klausberger Walter <walter.klausberger@siemens.at>
To: "'Evan Caves '" <evan@occamnetworks.com>,
        Klausberger Walter
	 <walter.klausberger@siemens.at>
Cc: "''l2tp@l2tp.net ' '" <l2tp@l2tp.net>
Subject: AW: draft-ietf-l2tpext-ds-01.txt
Date: Tue, 12 Dec 2000 00:26:04 +0100
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C06449.41BAFC9C"
Sender: owner-l2tp@diameter.org
Precedence: bulk

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_001_01C06449.41BAFC9C
Content-Type: text/plain

More information about Tequila...

http://www.ist-tequila.org/sls.html
read <draft-tequila-diffserv-sls-00.txt> chapter 5 (at the bottom)

I'm not sure if I like it and it's no L2TP issue. But there should be a way
a certain service level agreement is assigned to a user. I would prefer
static configuration via AAA.

with best regards
Walter

-----Originalnachricht-----
Von: Evan Caves
An: Klausberger Walter
Cc: 'l2tp@l2tp.net '
Gesendet: 11.12.00 22:55
Betreff: Re: draft-ietf-l2tpext-ds-01.txt

> Klausberger Walter wrote:
> 
> Hi folks,
> 
> I like the idea of negotiating QoS parameters during tunnel/session
> establishment, but I asked myself all the time, where the DS
> parameters will be configured for a special session.

They will most likely be returned by a AAA server.

> 
> There are parameters planned in the L2TP MIB,

Remind me again about these parameters :-).

> but in LAC case the
> parameters often come frome an AAA server like RADIUS. Shouldn't there
> be an extension in the tunneling parameters?

Yup.

> 
> Another thing is the Tequila group. 

The what group?

> They want to provide SLA
> negotiation for RAS user via PPP IPCP extension. 

Uggghhh.

> This should also be
> mapped onto tunnel sessions (if granted). To grant such features again
> new AAA attributes could be used.
> 
> Is anyone already thinking about that, or is someone willing to
> discuss these matters in more detail?
> 
> see you at the IETF
> with best regards
> Walter

------_=_NextPart_001_01C06449.41BAFC9C
Content-Type: text/html
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3DUS-ASCII">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
5.5.2650.12">
<TITLE>AW: draft-ietf-l2tpext-ds-01.txt</TITLE>
</HEAD>
<BODY>

<P><FONT SIZE=3D2>More information about Tequila...</FONT>
</P>

<P><FONT SIZE=3D2><A HREF=3D"http://www.ist-tequila.org/sls.html" =
TARGET=3D"_blank">http://www.ist-tequila.org/sls.html</A></FONT>
<BR><FONT SIZE=3D2>read &lt;draft-tequila-diffserv-sls-00.txt&gt; =
chapter 5 (at the bottom)</FONT>
</P>

<P><FONT SIZE=3D2>I'm not sure if I like it and it's no L2TP issue. But =
there should be a way a certain service level agreement is assigned to =
a user. I would prefer static configuration via AAA.</FONT></P>

<P><FONT SIZE=3D2>with best regards</FONT>
<BR><FONT SIZE=3D2>Walter</FONT>
</P>

<P><FONT SIZE=3D2>-----Originalnachricht-----</FONT>
<BR><FONT SIZE=3D2>Von: Evan Caves</FONT>
<BR><FONT SIZE=3D2>An: Klausberger Walter</FONT>
<BR><FONT SIZE=3D2>Cc: 'l2tp@l2tp.net '</FONT>
<BR><FONT SIZE=3D2>Gesendet: 11.12.00 22:55</FONT>
<BR><FONT SIZE=3D2>Betreff: Re: draft-ietf-l2tpext-ds-01.txt</FONT>
</P>

<P><FONT SIZE=3D2>&gt; Klausberger Walter wrote:</FONT>
<BR><FONT SIZE=3D2>&gt; </FONT>
<BR><FONT SIZE=3D2>&gt; Hi folks,</FONT>
<BR><FONT SIZE=3D2>&gt; </FONT>
<BR><FONT SIZE=3D2>&gt; I like the idea of negotiating QoS parameters =
during tunnel/session</FONT>
<BR><FONT SIZE=3D2>&gt; establishment, but I asked myself all the time, =
where the DS</FONT>
<BR><FONT SIZE=3D2>&gt; parameters will be configured for a special =
session.</FONT>
</P>

<P><FONT SIZE=3D2>They will most likely be returned by a AAA =
server.</FONT>
</P>

<P><FONT SIZE=3D2>&gt; </FONT>
<BR><FONT SIZE=3D2>&gt; There are parameters planned in the L2TP =
MIB,</FONT>
</P>

<P><FONT SIZE=3D2>Remind me again about these parameters :-).</FONT>
</P>

<P><FONT SIZE=3D2>&gt; but in LAC case the</FONT>
<BR><FONT SIZE=3D2>&gt; parameters often come frome an AAA server like =
RADIUS. Shouldn't there</FONT>
<BR><FONT SIZE=3D2>&gt; be an extension in the tunneling =
parameters?</FONT>
</P>

<P><FONT SIZE=3D2>Yup.</FONT>
</P>

<P><FONT SIZE=3D2>&gt; </FONT>
<BR><FONT SIZE=3D2>&gt; Another thing is the Tequila group. </FONT>
</P>

<P><FONT SIZE=3D2>The what group?</FONT>
</P>

<P><FONT SIZE=3D2>&gt; They want to provide SLA</FONT>
<BR><FONT SIZE=3D2>&gt; negotiation for RAS user via PPP IPCP =
extension. </FONT>
</P>

<P><FONT SIZE=3D2>Uggghhh.</FONT>
</P>

<P><FONT SIZE=3D2>&gt; This should also be</FONT>
<BR><FONT SIZE=3D2>&gt; mapped onto tunnel sessions (if granted). To =
grant such features again</FONT>
<BR><FONT SIZE=3D2>&gt; new AAA attributes could be used.</FONT>
<BR><FONT SIZE=3D2>&gt; </FONT>
<BR><FONT SIZE=3D2>&gt; Is anyone already thinking about that, or is =
someone willing to</FONT>
<BR><FONT SIZE=3D2>&gt; discuss these matters in more detail?</FONT>
<BR><FONT SIZE=3D2>&gt; </FONT>
<BR><FONT SIZE=3D2>&gt; see you at the IETF</FONT>
<BR><FONT SIZE=3D2>&gt; with best regards</FONT>
<BR><FONT SIZE=3D2>&gt; Walter</FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C06449.41BAFC9C--


From owner-l2tp@diameter.org  Tue Dec 12 21:32:02 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id VAA12552
	for <l2tpext-archive@odin.ietf.org>; Tue, 12 Dec 2000 21:32:01 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBD0cTu14154
	for l2tp-list; Tue, 12 Dec 2000 16:38:29 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from franklin.cisco.com (franklin.cisco.com [171.70.156.17])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBD0cNH14148
	for <l2tp@l2tp.net>; Tue, 12 Dec 2000 16:38:23 -0800
Received: from gwz (sjc-vpn-161.cisco.com [10.21.64.161]) by franklin.cisco.com (8.8.6 (PHNE_17190)/CISCO.SERVER.1.2) with SMTP id PAA24632; Tue, 12 Dec 2000 15:45:59 -0800 (PST)
Reply-To: <gwz@cisco.com>
From: "Glen Zorn" <gwz@cisco.com>
To: "Evan Caves" <evan@occamnetworks.com>,
        "Klausberger Walter" <walter.klausberger@siemens.at>
Cc: <l2tp@l2tp.net>
Subject: RE: draft-ietf-l2tpext-ds-01.txt
Date: Tue, 12 Dec 2000 15:46:48 -0800
Message-ID: <NDBBIHMPILAAGDHPCIOPCEBNDAAA.gwz@cisco.com>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0)
In-Reply-To: <3A354D39.15A4B973@occamnetworks.com>
Importance: Normal
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2615.200
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Evan Caves [mailto://evan@occamnetworks.com] writes:

> > Klausberger Walter wrote:
> >
> > Hi folks,
> >
> > I like the idea of negotiating QoS parameters during tunnel/session
> > establishment, but I asked myself all the time, where the DS
> > parameters will be configured for a special session.
>
> They will most likely be returned by a AAA server.
>
> >
> > There are parameters planned in the L2TP MIB,
>
> Remind me again about these parameters :-).
>
> > but in LAC case the
> > parameters often come frome an AAA server like RADIUS. Shouldn't there
> > be an extension in the tunneling parameters?
>
> Yup.
>
> >
> > Another thing is the Tequila group.
>
> The what group?
>
> > They want to provide SLA
> > negotiation for RAS user via PPP IPCP extension.
>
> Uggghhh.

I'll second that.

>
> > This should also be
> > mapped onto tunnel sessions (if granted). To grant such features again
> > new AAA attributes could be used.

As part of the process of revising RFC 2868 we have considered adding a
diffserv-related attribute.

> >
> > Is anyone already thinking about that, or is someone willing to
> > discuss these matters in more detail?
> >
> > see you at the IETF
> > with best regards
> > Walter
>



From owner-l2tp@diameter.org  Tue Dec 12 23:50:22 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id XAA02938
	for <l2tpext-archive@odin.ietf.org>; Tue, 12 Dec 2000 23:50:21 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBD3RMq14380
	for l2tp-list; Tue, 12 Dec 2000 19:27:22 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from eins.siemens.at (eins.siemens.at [193.81.246.11])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBD3RIH14374
	for <l2tp@l2tp.net>; Tue, 12 Dec 2000 19:27:18 -0800
Received: from scesie13.sie.siemens.at (forix [10.1.140.2])
	by eins.siemens.at  with ESMTP id eBD2ZIk11414;
	Wed, 13 Dec 2000 03:35:18 +0100
Received: (from smap@localhost)
	by scesie13.sie.siemens.at (8.9.3/8.9.3) id DAA15162;
	Wed, 13 Dec 2000 03:35:18 +0100 (MET)
Received: from vies141a.sie.siemens.at(195.1.196.56) by scesie13 via smap (V2.0beta)
	id xma015126; Wed, 13 Dec 00 03:34:53 +0100
Received: by vies141a.sie.siemens.at with Internet Mail Service (5.5.2650.21)
	id <YNL1FM8B>; Wed, 13 Dec 2000 03:34:52 +0100
Message-ID: <D9F2B9CD7BD5D21196BC0800060D9ED604373023@vies186a.sie.siemens.at>
From: Klausberger Walter <walter.klausberger@siemens.at>
To: "'Glen Zorn '" <gwz@cisco.com>
Cc: "'l2tp@l2tp.net '" <l2tp@l2tp.net>
Subject: AW: draft-ietf-l2tpext-ds-01.txt
Date: Wed, 13 Dec 2000 03:34:51 +0100
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C064AD.4561FE86"
Sender: owner-l2tp@diameter.org
Precedence: bulk

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_001_01C064AD.4561FE86
Content-Type: text/plain;
	charset="iso-8859-1"

Hi Glen,

As I know the RADIUS WG has closed. When and by whom will the RFC 2868
revised? By the AAA WG?

Walter


-----Originalnachricht-----
Von: Glen Zorn
An: Evan Caves; Klausberger Walter
Cc: l2tp@l2tp.net
Gesendet: 13.12.00 00:46
Betreff: RE: draft-ietf-l2tpext-ds-01.txt

Evan Caves [mailto://evan@occamnetworks.com] writes:

> > Klausberger Walter wrote:
> >
> > Hi folks,
> >
> > I like the idea of negotiating QoS parameters during tunnel/session
> > establishment, but I asked myself all the time, where the DS
> > parameters will be configured for a special session.
>
> They will most likely be returned by a AAA server.
>
> >
> > There are parameters planned in the L2TP MIB,
>
> Remind me again about these parameters :-).
>
> > but in LAC case the
> > parameters often come frome an AAA server like RADIUS. Shouldn't
there
> > be an extension in the tunneling parameters?
>
> Yup.
>
> >
> > Another thing is the Tequila group.
>
> The what group?
>
> > They want to provide SLA
> > negotiation for RAS user via PPP IPCP extension.
>
> Uggghhh.

I'll second that.

>
> > This should also be
> > mapped onto tunnel sessions (if granted). To grant such features
again
> > new AAA attributes could be used.

As part of the process of revising RFC 2868 we have considered adding a
diffserv-related attribute.

> >
> > Is anyone already thinking about that, or is someone willing to
> > discuss these matters in more detail?
> >
> > see you at the IETF
> > with best regards
> > Walter
>

------_=_NextPart_001_01C064AD.4561FE86
Content-Type: text/html;
	charset="iso-8859-1"

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 5.5.2650.12">
<TITLE>AW: draft-ietf-l2tpext-ds-01.txt</TITLE>
</HEAD>
<BODY>

<P><FONT SIZE=2>Hi Glen,</FONT>
</P>

<P><FONT SIZE=2>As I know the RADIUS WG has closed. When and by whom will the RFC 2868 revised? By the AAA WG?</FONT>
</P>

<P><FONT SIZE=2>Walter</FONT>
</P>
<BR>

<P><FONT SIZE=2>-----Originalnachricht-----</FONT>
<BR><FONT SIZE=2>Von: Glen Zorn</FONT>
<BR><FONT SIZE=2>An: Evan Caves; Klausberger Walter</FONT>
<BR><FONT SIZE=2>Cc: l2tp@l2tp.net</FONT>
<BR><FONT SIZE=2>Gesendet: 13.12.00 00:46</FONT>
<BR><FONT SIZE=2>Betreff: RE: draft-ietf-l2tpext-ds-01.txt</FONT>
</P>

<P><FONT SIZE=2>Evan Caves [<A HREF="mailto://evan@occamnetworks.com">mailto://evan@occamnetworks.com</A>] writes:</FONT>
</P>

<P><FONT SIZE=2>&gt; &gt; Klausberger Walter wrote:</FONT>
<BR><FONT SIZE=2>&gt; &gt;</FONT>
<BR><FONT SIZE=2>&gt; &gt; Hi folks,</FONT>
<BR><FONT SIZE=2>&gt; &gt;</FONT>
<BR><FONT SIZE=2>&gt; &gt; I like the idea of negotiating QoS parameters during tunnel/session</FONT>
<BR><FONT SIZE=2>&gt; &gt; establishment, but I asked myself all the time, where the DS</FONT>
<BR><FONT SIZE=2>&gt; &gt; parameters will be configured for a special session.</FONT>
<BR><FONT SIZE=2>&gt;</FONT>
<BR><FONT SIZE=2>&gt; They will most likely be returned by a AAA server.</FONT>
<BR><FONT SIZE=2>&gt;</FONT>
<BR><FONT SIZE=2>&gt; &gt;</FONT>
<BR><FONT SIZE=2>&gt; &gt; There are parameters planned in the L2TP MIB,</FONT>
<BR><FONT SIZE=2>&gt;</FONT>
<BR><FONT SIZE=2>&gt; Remind me again about these parameters :-).</FONT>
<BR><FONT SIZE=2>&gt;</FONT>
<BR><FONT SIZE=2>&gt; &gt; but in LAC case the</FONT>
<BR><FONT SIZE=2>&gt; &gt; parameters often come frome an AAA server like RADIUS. Shouldn't</FONT>
<BR><FONT SIZE=2>there</FONT>
<BR><FONT SIZE=2>&gt; &gt; be an extension in the tunneling parameters?</FONT>
<BR><FONT SIZE=2>&gt;</FONT>
<BR><FONT SIZE=2>&gt; Yup.</FONT>
<BR><FONT SIZE=2>&gt;</FONT>
<BR><FONT SIZE=2>&gt; &gt;</FONT>
<BR><FONT SIZE=2>&gt; &gt; Another thing is the Tequila group.</FONT>
<BR><FONT SIZE=2>&gt;</FONT>
<BR><FONT SIZE=2>&gt; The what group?</FONT>
<BR><FONT SIZE=2>&gt;</FONT>
<BR><FONT SIZE=2>&gt; &gt; They want to provide SLA</FONT>
<BR><FONT SIZE=2>&gt; &gt; negotiation for RAS user via PPP IPCP extension.</FONT>
<BR><FONT SIZE=2>&gt;</FONT>
<BR><FONT SIZE=2>&gt; Uggghhh.</FONT>
</P>

<P><FONT SIZE=2>I'll second that.</FONT>
</P>

<P><FONT SIZE=2>&gt;</FONT>
<BR><FONT SIZE=2>&gt; &gt; This should also be</FONT>
<BR><FONT SIZE=2>&gt; &gt; mapped onto tunnel sessions (if granted). To grant such features</FONT>
<BR><FONT SIZE=2>again</FONT>
<BR><FONT SIZE=2>&gt; &gt; new AAA attributes could be used.</FONT>
</P>

<P><FONT SIZE=2>As part of the process of revising RFC 2868 we have considered adding a</FONT>
<BR><FONT SIZE=2>diffserv-related attribute.</FONT>
</P>

<P><FONT SIZE=2>&gt; &gt;</FONT>
<BR><FONT SIZE=2>&gt; &gt; Is anyone already thinking about that, or is someone willing to</FONT>
<BR><FONT SIZE=2>&gt; &gt; discuss these matters in more detail?</FONT>
<BR><FONT SIZE=2>&gt; &gt;</FONT>
<BR><FONT SIZE=2>&gt; &gt; see you at the IETF</FONT>
<BR><FONT SIZE=2>&gt; &gt; with best regards</FONT>
<BR><FONT SIZE=2>&gt; &gt; Walter</FONT>
<BR><FONT SIZE=2>&gt;</FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C064AD.4561FE86--


From owner-l2tp@diameter.org  Wed Dec 13 13:34:27 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id NAA27017
	for <l2tpext-archive@odin.ietf.org>; Wed, 13 Dec 2000 13:34:26 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBDHK4A15375
	for l2tp-list; Wed, 13 Dec 2000 09:20:04 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBDHJwH15367
	for <l2tp@l2tp.net>; Wed, 13 Dec 2000 09:19:59 -0800
Received: from occamnetworks.com (mail.occamnetworks.com [216.64.159.194]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id IAA24520 for <l2tp@ipsec.org>; Wed, 13 Dec 2000 08:28:00 -0800 (PST)
Received: from occamnetworks.com (honeycombs.occamnetworks.com [192.168.2.202])
	by occamnetworks.com (8.10.1/8.10.1) with ESMTP id eBDGRFQ26046;
	Wed, 13 Dec 2000 08:27:15 -0800
Message-ID: <3A37A338.78851E6B@occamnetworks.com>
Date: Wed, 13 Dec 2000 08:26:32 -0800
From: Evan Caves <evan@occamnetworks.com>
Organization: Occam Networks, Inc.
X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.2.16-3 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: Klausberger Walter <walter.klausberger@siemens.at>
CC: "''Dave Thaler ' '" <dthaler@Exchange.Microsoft.com>,
        "''l2tp@ipsec.org ' '" <l2tp@ipsec.org>,
        "''kzm@cisco.com ' '" <kzm@cisco.com>
Subject: Re: AW: AW: AW: ifStackTable layering in L2TP
References: <D9F2B9CD7BD5D21196BC0800060D9ED60437301D@vies186a.sie.siemens.at>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> Klausberger Walter wrote:
> 
> 
> 
> -----Originalnachricht-----
> Von: Evan Caves
> An: Klausberger Walter
> Cc: 'Dave Thaler '; 'l2tp@ipsec.org '; 'kzm@cisco.com '
> Gesendet: 11.12.00 22:49
> Betreff: Re: AW: AW: ifStackTable layering in L2TP
> 
> >> Klausberger Walter wrote:
> >>
> >> Hi all,
> >>
> >> sorry for my absence, but I was on a short vacation and now I'm in
> San
> >> Diego. I must admit, I'm a little confused and I'm not shure if
> >> everyone understands it the way Dave does.
> >>
> >> Are there any responses from already existing implementations?
> >>
> >> I do not fully agree to the argument that a forwarder (like IP
> router)
> >> is the same as a multiplexer (like MLP). A multiplexer normally
> works
> >> with a fixed set of connections (well I know they can be teared on
> and
> >> so on). It multiplexes packets comming from a higher layer to
> several
> >> lower layers or the other way round. A forwarder just takes packets
> 
> >> from a lower layer and puts them to another lower layer depending
> on
> >> some rules. In my opinion Daves approach shows new aspects. I hope
> all
> >> others see ist the same way!?
> >>
> >> On the other hand I agree with Dave that a tunnel can be seen as a
> >> virtual interface.
> >>
> >> So in L2TP case the tunnel IF would sit on top of an invisible
> >> forwarder and PPP would be the forwarding layer (in LAC case), e.g.
> 
> >>
> >>         ------PPP-------
> >>         |              |
> >>     DS0(phys. IF)    tunnel
> >>                        |
> >>              (IP router, invisible)
> >>                        |
> >>                     ethernet
> 
> >This is what is currently presented in the MIB except in its model
> >we do not call the forwarding layer a 'PPP forwarding layer', it
> >is modeled as a switching element. The reason being that on the LAC
> >side one could argue that PPP isn't really there anymore once the
> >tunnel is up (it sort of steps aside).
> 
> ok, I agree with you on that
> 
> >>
> >> or should it be
> >>
> >>        PPP               tunnel
> >>         |                   |
> >>     DS0(phys. IF) (IP router, invisible)
> >>                             |
> >>                          ethernet
> >>
> >> or anything else?
> 
> >I would argue against this. And besides in this picture how do PPP
> >packets end up in the tunnel?
> 
> agreed
> 
> >>
> >> We are loosing the context between the physical interface, where
> the
> >> PPP session comes in (DS0) and the tunnel this session goes to.
> This
> >> information is now provided with the L2TP MIB only.
> 
> >Yup. Just like you need the CIDR MIB (or whatever it is called now)
> >to dump your routing tables.
> 
> agreed
> 
> >>
> >> As PPP now is the forwarding layer should it be visible on top of
> the
> >> DS0 layer, or only if the PPP connection is not tunneled.
> 
> >PPP isn't even in the picture on the LAC in my model. If it was
> >you would expect to be able to manage it like any other PPP interface
> 
> >(which is definitely not the case).
> 
> A LAC does not necessarily tunnel a PPP session (you might say it is
> not a LAC then, but that's how it's written in the L2TP RFC). If a PPP
> session is terminated in a LAC shouldn't the PPP layer be visible?

I would think the PPP interface is visible in the ifStackTable and
ifTable
on the LAC until it gets tunneled. At the time a tunneling decision is 
made it disappears (and should show up on the LNS). At this point you
can use the l2tpSessionMapTable to see which interfaces (the interface
that was below PPP before tunneling occurred) are having their packets
switched to and from a particular tunnel/session.

> 
> Thinking about all this, I have concerns about the stack layering MIB.
> When I have an AAL5 layer (assume PPP over AAL5), I know by
> informations from the L2TP MIB that a PPP session is coming from the
> AAL5 interface and going into a certain tunnel.
> 
> How do I know if IP packets come from that tunnel? To me there is
> something missing, or does the CIDR MIB provide enough information
> (I'm not so familiar with this MIB)?

I am not sure what you mean. If you are asking whether PPP/IP packets
are being tunneled the only way you can tell is by querying the
LNS using the IPCP MIB. The CIDR MIB has nothing to do with this. It
is used to manage the route table on a device.

evan
-

> 
> >>
> >> >In the multipath example, the tunnel would be layered over ALL
> >> >the interfaces which have potential routes over them.  In this
> >> >sense, it's very analogous to multilink PPP, among other
> >> >things.
> >>
> >> Dave, do I understand it correct that regardless of the routing
> entry
> >> a tunnel IF should sit on all possible interfaces it may be routed
> >> over? That would also be a new aspect to me.
> >>
> >> I think these are quite basic questions and I'm unsure how to
> proceed.
> >> Maybe there should be a clarification in the L2TP MIB draft(I don't
> 
> >> want to change the tunnel MIB RFC, though it would be a good idea
> to
> >> add a clarification in case of a new version)
> >>
> 
> >And I don't think that the L2TP MIB should be describing how the IP
> >Tunnel MIB works (but I will probably loose that argument).
> 
> I think it could be an issue, otherwise there might be a lot of other
> people asking the same questions later, but I agree it is a matter of
> the Tunnel MIB description...
> 
> Walter
> -


From owner-l2tp@diameter.org  Wed Dec 13 14:18:18 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id OAA06799
	for <l2tpext-archive@odin.ietf.org>; Wed, 13 Dec 2000 14:18:18 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBDIWc315472
	for l2tp-list; Wed, 13 Dec 2000 10:32:38 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBDIWVH15462
	for <l2tp@l2tp.net>; Wed, 13 Dec 2000 10:32:31 -0800
Received: from eins.siemens.at (eins.siemens.at [193.81.246.11]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id JAA24689 for <l2tp@ipsec.org>; Wed, 13 Dec 2000 09:40:32 -0800 (PST)
Received: from scesie13.sie.siemens.at (forix [10.1.140.2])
	by eins.siemens.at  with ESMTP id eBDHdvk15094;
	Wed, 13 Dec 2000 18:39:57 +0100
Received: (from smap@localhost)
	by scesie13.sie.siemens.at (8.9.3/8.9.3) id SAA00642;
	Wed, 13 Dec 2000 18:39:56 +0100 (MET)
Received: from vies194a.sie.siemens.at(195.1.196.94) by scesie13 via smap (V2.0beta)
	id xma000427; Wed, 13 Dec 00 18:39:33 +0100
Received: by vies194a.sie.siemens.at with Internet Mail Service (5.5.2650.21)
	id <YYGJWR3T>; Wed, 13 Dec 2000 18:39:31 +0100
Message-ID: <D9F2B9CD7BD5D21196BC0800060D9ED604373027@vies186a.sie.siemens.at>
From: Klausberger Walter <walter.klausberger@siemens.at>
To: "'Evan Caves '" <evan@occamnetworks.com>
Cc: "'''Dave Thaler ' ' '" <dthaler@Exchange.Microsoft.com>,
        "'''l2tp@ipsec.org ' ' '" <l2tp@ipsec.org>,
        "'''kzm@cisco.com ' ' '"
	 <kzm@cisco.com>
Subject: AW: AW: AW: AW: ifStackTable layering in L2TP
Date: Wed, 13 Dec 2000 18:39:31 +0100
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C0652B.A602E7C2"
Sender: owner-l2tp@diameter.org
Precedence: bulk

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_001_01C0652B.A602E7C2
Content-Type: text/plain



>-----Originalnachricht-----
>Von: Evan Caves
>An: Klausberger Walter
>Cc: ''Dave Thaler ' '; ''l2tp@ipsec.org ' '; ''kzm@cisco.com ' '
>Gesendet: 13.12.00 17:26
>Betreff: Re: AW: AW: AW: ifStackTable layering in L2TP

>> A LAC does not necessarily tunnel a PPP session (you might say it is
>> not a LAC then, but that's how it's written in the L2TP RFC). If a PPP
>> session is terminated in a LAC shouldn't the PPP layer be visible?

>I would think the PPP interface is visible in the ifStackTable and
>ifTable
>on the LAC until it gets tunneled. At the time a tunneling decision is 
>made it disappears (and should show up on the LNS). At this point you
>can use the l2tpSessionMapTable to see which interfaces (the interface
>that was below PPP before tunneling occurred) are having their packets
>switched to and from a particular tunnel/session.

That sounds reasonable to me. I see it the same way, but I wanted
confirmation on that.
 
>> Thinking about all this, I have concerns about the stack layering MIB.
>> When I have an AAL5 layer (assume PPP over AAL5), I know by
>> informations from the L2TP MIB that a PPP session is coming from the
>> AAL5 interface and going into a certain tunnel.
>> 
>> How do I know if IP packets come from that tunnel? To me there is
>> something missing, or does the CIDR MIB provide enough information
>> (I'm not so familiar with this MIB)?

>I am not sure what you mean. If you are asking whether PPP/IP packets
>are being tunneled the only way you can tell is by querying the
>LNS using the IPCP MIB. The CIDR MIB has nothing to do with this. It
>is used to manage the route table on a device.

Sorry, this was a mistype. I meant 'How do I know if IP packets come from
that AAL5 interface?'. I only see an interface.

- In case of PPP packets coming from this interface, I find an entry in the
L2TP MIB (in case of tunneling)

- In case of IP packets coming from this interface, do I get the information
via routing table (there should be a route to this IP interface)?

walter 

------_=_NextPart_001_01C0652B.A602E7C2
Content-Type: text/html
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3DUS-ASCII">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
5.5.2650.12">
<TITLE>AW: AW: AW: AW: ifStackTable layering in L2TP</TITLE>
</HEAD>
<BODY>
<BR>
<BR>

<P><FONT SIZE=3D2>&gt;-----Originalnachricht-----</FONT>
<BR><FONT SIZE=3D2>&gt;Von: Evan Caves</FONT>
<BR><FONT SIZE=3D2>&gt;An: Klausberger Walter</FONT>
<BR><FONT SIZE=3D2>&gt;Cc: ''Dave Thaler ' '; ''l2tp@ipsec.org ' '; =
''kzm@cisco.com ' '</FONT>
<BR><FONT SIZE=3D2>&gt;Gesendet: 13.12.00 17:26</FONT>
<BR><FONT SIZE=3D2>&gt;Betreff: Re: AW: AW: AW: ifStackTable layering =
in L2TP</FONT>
</P>

<P><FONT SIZE=3D2>&gt;&gt; A LAC does not necessarily tunnel a PPP =
session (you might say it is</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; not a LAC then, but that's how it's written =
in the L2TP RFC). If a PPP</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; session is terminated in a LAC shouldn't =
the PPP layer be visible?</FONT>
</P>

<P><FONT SIZE=3D2>&gt;I would think the PPP interface is visible in the =
ifStackTable and</FONT>
<BR><FONT SIZE=3D2>&gt;ifTable</FONT>
<BR><FONT SIZE=3D2>&gt;on the LAC until it gets tunneled. At the time a =
tunneling decision is </FONT>
<BR><FONT SIZE=3D2>&gt;made it disappears (and should show up on the =
LNS). At this point you</FONT>
<BR><FONT SIZE=3D2>&gt;can use the l2tpSessionMapTable to see which =
interfaces (the interface</FONT>
<BR><FONT SIZE=3D2>&gt;that was below PPP before tunneling occurred) =
are having their packets</FONT>
<BR><FONT SIZE=3D2>&gt;switched to and from a particular =
tunnel/session.</FONT>
</P>

<P><FONT SIZE=3D2>That sounds reasonable to me. I see it the same way, =
but I wanted confirmation on that.</FONT>
<BR><FONT SIZE=3D2>&nbsp;</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; Thinking about all this, I have concerns =
about the stack layering MIB.</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; When I have an AAL5 layer (assume PPP over =
AAL5), I know by</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; informations from the L2TP MIB that a PPP =
session is coming from the</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; AAL5 interface and going into a certain =
tunnel.</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; </FONT>
<BR><FONT SIZE=3D2>&gt;&gt; How do I know if IP packets come from that =
tunnel? To me there is</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; something missing, or does the CIDR MIB =
provide enough information</FONT>
<BR><FONT SIZE=3D2>&gt;&gt; (I'm not so familiar with this MIB)?</FONT>
</P>

<P><FONT SIZE=3D2>&gt;I am not sure what you mean. If you are asking =
whether PPP/IP packets</FONT>
<BR><FONT SIZE=3D2>&gt;are being tunneled the only way you can tell is =
by querying the</FONT>
<BR><FONT SIZE=3D2>&gt;LNS using the IPCP MIB. The CIDR MIB has nothing =
to do with this. It</FONT>
<BR><FONT SIZE=3D2>&gt;is used to manage the route table on a =
device.</FONT>
</P>

<P><FONT SIZE=3D2>Sorry, this was a mistype. I meant 'How do I know if =
IP packets come from that AAL5 interface?'. I only see an =
interface.</FONT></P>

<P><FONT SIZE=3D2>- In case of PPP packets coming from this interface, =
I find an entry in the L2TP MIB (in case of tunneling)</FONT>
</P>

<P><FONT SIZE=3D2>- In case of IP packets coming from this interface, =
do I get the information via routing table (there should be a route to =
this IP interface)?</FONT></P>

<P><FONT SIZE=3D2>walter </FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C0652B.A602E7C2--


From owner-l2tp@diameter.org  Wed Dec 13 18:28:44 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id SAA14311
	for <l2tpext-archive@odin.ietf.org>; Wed, 13 Dec 2000 18:28:43 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBDMGQT15822
	for l2tp-list; Wed, 13 Dec 2000 14:16:26 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from franklin.cisco.com (franklin.cisco.com [171.70.156.17])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBDMGNH15816
	for <l2tp@l2tp.net>; Wed, 13 Dec 2000 14:16:23 -0800
Received: from gwz (sjc-vpn-66.cisco.com [10.21.64.66]) by franklin.cisco.com (8.8.6 (PHNE_17190)/CISCO.SERVER.1.2) with SMTP id NAA21636; Wed, 13 Dec 2000 13:24:06 -0800 (PST)
Reply-To: <gwz@cisco.com>
From: "Glen Zorn" <gwz@cisco.com>
To: "Klausberger Walter" <walter.klausberger@siemens.at>
Cc: <l2tp@l2tp.net>
Subject: RE: draft-ietf-l2tpext-ds-01.txt
Date: Wed, 13 Dec 2000 13:24:55 -0800
Message-ID: <NDBBIHMPILAAGDHPCIOPOECNDAAA.gwz@cisco.com>
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_NextPart_000_0030_01C06508.1551CCC0"
X-Priority: 3 (Normal)
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0)
In-Reply-To: <D9F2B9CD7BD5D21196BC0800060D9ED604373023@vies186a.sie.siemens.at>
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2615.200
Importance: Normal
Sender: owner-l2tp@diameter.org
Precedence: bulk

This is a multi-part message in MIME format.

------=_NextPart_000_0030_01C06508.1551CCC0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit

AW: draft-ietf-l2tpext-ds-01.txtThe AAA WG has, as far as I know, 0 interest
in RADIUS.  2868 bis being revised by its authors.

  -----Original Message-----
  From: Klausberger Walter [mailto:walter.klausberger@siemens.at]
  Sent: Tuesday, December 12, 2000 6:35 PM
  To: 'Glen Zorn '
  Cc: 'l2tp@l2tp.net '
  Subject: AW: draft-ietf-l2tpext-ds-01.txt


  Hi Glen,

  As I know the RADIUS WG has closed. When and by whom will the RFC 2868
revised? By the AAA WG?

  Walter



  -----Originalnachricht-----
  Von: Glen Zorn
  An: Evan Caves; Klausberger Walter
  Cc: l2tp@l2tp.net
  Gesendet: 13.12.00 00:46
  Betreff: RE: draft-ietf-l2tpext-ds-01.txt

  Evan Caves [mailto://evan@occamnetworks.com] writes:

  > > Klausberger Walter wrote:
  > >
  > > Hi folks,
  > >
  > > I like the idea of negotiating QoS parameters during tunnel/session
  > > establishment, but I asked myself all the time, where the DS
  > > parameters will be configured for a special session.
  >
  > They will most likely be returned by a AAA server.
  >
  > >
  > > There are parameters planned in the L2TP MIB,
  >
  > Remind me again about these parameters :-).
  >
  > > but in LAC case the
  > > parameters often come frome an AAA server like RADIUS. Shouldn't
  there
  > > be an extension in the tunneling parameters?
  >
  > Yup.
  >
  > >
  > > Another thing is the Tequila group.
  >
  > The what group?
  >
  > > They want to provide SLA
  > > negotiation for RAS user via PPP IPCP extension.
  >
  > Uggghhh.

  I'll second that.

  >
  > > This should also be
  > > mapped onto tunnel sessions (if granted). To grant such features
  again
  > > new AAA attributes could be used.

  As part of the process of revising RFC 2868 we have considered adding a
  diffserv-related attribute.

  > >
  > > Is anyone already thinking about that, or is someone willing to
  > > discuss these matters in more detail?
  > >
  > > see you at the IETF
  > > with best regards
  > > Walter
  >


------=_NextPart_000_0030_01C06508.1551CCC0
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD><TITLE>AW: draft-ietf-l2tpext-ds-01.txt</TITLE>
<META content=3D"text/html; charset=3Diso-8859-1" =
http-equiv=3DContent-Type>
<META content=3D"MSHTML 5.00.2919.6307" name=3DGENERATOR></HEAD>
<BODY>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D410302321-13122000>The=20
AAA WG has, as far as I know, 0 interest in RADIUS.&nbsp; 2868 bis being =
revised=20
by its authors.<BR></SPAN></FONT></DIV>
<BLOCKQUOTE=20
style=3D"BORDER-LEFT: #0000ff 2px solid; MARGIN-LEFT: 5px; MARGIN-RIGHT: =
0px; PADDING-LEFT: 5px">
  <DIV align=3Dleft class=3DOutlookMessageHeader dir=3Dltr><FONT =
face=3DTahoma=20
  size=3D2>-----Original Message-----<BR><B>From:</B> Klausberger Walter =

  [mailto:walter.klausberger@siemens.at]<BR><B>Sent:</B> Tuesday, =
December 12,=20
  2000 6:35 PM<BR><B>To:</B> 'Glen Zorn '<BR><B>Cc:</B> 'l2tp@l2tp.net=20
  '<BR><B>Subject:</B> AW: =
draft-ietf-l2tpext-ds-01.txt<BR><BR></DIV></FONT>
  <P><FONT size=3D2>Hi Glen,</FONT> </P>
  <P><FONT size=3D2>As I know the RADIUS WG has closed. When and by whom =
will the=20
  RFC 2868 revised? By the AAA WG?</FONT> </P>
  <P><FONT size=3D2>Walter</FONT> </P><BR>
  <P><FONT size=3D2>-----Originalnachricht-----</FONT> <BR><FONT =
size=3D2>Von: Glen=20
  Zorn</FONT> <BR><FONT size=3D2>An: Evan Caves; Klausberger =
Walter</FONT>=20
  <BR><FONT size=3D2>Cc: l2tp@l2tp.net</FONT> <BR><FONT =
size=3D2>Gesendet: 13.12.00=20
  00:46</FONT> <BR><FONT size=3D2>Betreff: RE: =
draft-ietf-l2tpext-ds-01.txt</FONT>=20
  </P>
  <P><FONT size=3D2>Evan Caves [<A=20
  =
href=3D"mailto://evan@occamnetworks.com">mailto://evan@occamnetworks.com<=
/A>]=20
  writes:</FONT> </P>
  <P><FONT size=3D2>&gt; &gt; Klausberger Walter wrote:</FONT> <BR><FONT =

  size=3D2>&gt; &gt;</FONT> <BR><FONT size=3D2>&gt; &gt; Hi =
folks,</FONT> <BR><FONT=20
  size=3D2>&gt; &gt;</FONT> <BR><FONT size=3D2>&gt; &gt; I like the idea =
of=20
  negotiating QoS parameters during tunnel/session</FONT> <BR><FONT =
size=3D2>&gt;=20
  &gt; establishment, but I asked myself all the time, where the =
DS</FONT>=20
  <BR><FONT size=3D2>&gt; &gt; parameters will be configured for a =
special=20
  session.</FONT> <BR><FONT size=3D2>&gt;</FONT> <BR><FONT size=3D2>&gt; =
They will=20
  most likely be returned by a AAA server.</FONT> <BR><FONT =
size=3D2>&gt;</FONT>=20
  <BR><FONT size=3D2>&gt; &gt;</FONT> <BR><FONT size=3D2>&gt; &gt; There =
are=20
  parameters planned in the L2TP MIB,</FONT> <BR><FONT =
size=3D2>&gt;</FONT>=20
  <BR><FONT size=3D2>&gt; Remind me again about these parameters =
:-).</FONT>=20
  <BR><FONT size=3D2>&gt;</FONT> <BR><FONT size=3D2>&gt; &gt; but in LAC =
case=20
  the</FONT> <BR><FONT size=3D2>&gt; &gt; parameters often come frome an =
AAA=20
  server like RADIUS. Shouldn't</FONT> <BR><FONT size=3D2>there</FONT> =
<BR><FONT=20
  size=3D2>&gt; &gt; be an extension in the tunneling parameters?</FONT> =
<BR><FONT=20
  size=3D2>&gt;</FONT> <BR><FONT size=3D2>&gt; Yup.</FONT> <BR><FONT=20
  size=3D2>&gt;</FONT> <BR><FONT size=3D2>&gt; &gt;</FONT> <BR><FONT =
size=3D2>&gt;=20
  &gt; Another thing is the Tequila group.</FONT> <BR><FONT =
size=3D2>&gt;</FONT>=20
  <BR><FONT size=3D2>&gt; The what group?</FONT> <BR><FONT =
size=3D2>&gt;</FONT>=20
  <BR><FONT size=3D2>&gt; &gt; They want to provide SLA</FONT> <BR><FONT =

  size=3D2>&gt; &gt; negotiation for RAS user via PPP IPCP =
extension.</FONT>=20
  <BR><FONT size=3D2>&gt;</FONT> <BR><FONT size=3D2>&gt; Uggghhh.</FONT> =
</P>
  <P><FONT size=3D2>I'll second that.</FONT> </P>
  <P><FONT size=3D2>&gt;</FONT> <BR><FONT size=3D2>&gt; &gt; This should =
also=20
  be</FONT> <BR><FONT size=3D2>&gt; &gt; mapped onto tunnel sessions (if =
granted).=20
  To grant such features</FONT> <BR><FONT size=3D2>again</FONT> =
<BR><FONT=20
  size=3D2>&gt; &gt; new AAA attributes could be used.</FONT> </P>
  <P><FONT size=3D2>As part of the process of revising RFC 2868 we have =
considered=20
  adding a</FONT> <BR><FONT size=3D2>diffserv-related attribute.</FONT> =
</P>
  <P><FONT size=3D2>&gt; &gt;</FONT> <BR><FONT size=3D2>&gt; &gt; Is =
anyone already=20
  thinking about that, or is someone willing to</FONT> <BR><FONT =
size=3D2>&gt;=20
  &gt; discuss these matters in more detail?</FONT> <BR><FONT =
size=3D2>&gt;=20
  &gt;</FONT> <BR><FONT size=3D2>&gt; &gt; see you at the IETF</FONT> =
<BR><FONT=20
  size=3D2>&gt; &gt; with best regards</FONT> <BR><FONT size=3D2>&gt; =
&gt;=20
  Walter</FONT> <BR><FONT size=3D2>&gt;</FONT> =
</P></BLOCKQUOTE></BODY></HTML>

------=_NextPart_000_0030_01C06508.1551CCC0--



From owner-l2tp@diameter.org  Wed Dec 13 21:17:58 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id VAA23368
	for <l2tpext-archive@odin.ietf.org>; Wed, 13 Dec 2000 21:17:57 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBE0gfE16068
	for l2tp-list; Wed, 13 Dec 2000 16:42:41 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBE0gaH16062
	for <l2tp@l2tp.net>; Wed, 13 Dec 2000 16:42:37 -0800
Received: from occamnetworks.com (mail.occamnetworks.com [216.64.159.194]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id PAA25592 for <l2tp@ipsec.org>; Wed, 13 Dec 2000 15:50:39 -0800 (PST)
Received: from occamnetworks.com (honeycombs.occamnetworks.com [192.168.2.202])
	by occamnetworks.com (8.10.1/8.10.1) with ESMTP id eBDNnjQ23923;
	Wed, 13 Dec 2000 15:49:45 -0800
Message-ID: <3A380AEA.4BC10101@occamnetworks.com>
Date: Wed, 13 Dec 2000 15:48:59 -0800
From: Evan Caves <evan@occamnetworks.com>
Organization: Occam Networks, Inc.
X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.2.16-3 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: Klausberger Walter <walter.klausberger@siemens.at>
CC: "'''Dave Thaler ' ' '" <dthaler@Exchange.Microsoft.com>,
        "'''l2tp@ipsec.org ' ' '" <l2tp@ipsec.org>,
        "'''kzm@cisco.com ' ' '" <kzm@cisco.com>
Subject: Re: AW: AW: AW: AW: ifStackTable layering in L2TP
References: <D9F2B9CD7BD5D21196BC0800060D9ED604373027@vies186a.sie.siemens.at>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

walter wrote:

> Sorry, this was a mistype. I meant 'How do I know if IP packets come
> from that AAL5 interface?'. I only see an interface.
> 
> - In case of PPP packets coming from this interface, I find an entry
> in the L2TP MIB (in case of tunneling)
> 
> - In case of IP packets coming from this interface, do I get the
> information via routing table (there should be a route to this IP
> interface)?

The ipAddrTable (RFC 1213) will tell you what ifEntries have an IP
address configured on them.

evan
-


From owner-l2tp@diameter.org  Thu Dec 14 14:50:47 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id OAA28605
	for <l2tpext-archive@odin.ietf.org>; Thu, 14 Dec 2000 14:50:46 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBEHeL017148
	for l2tp-list; Thu, 14 Dec 2000 09:40:21 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBEHeIH17142
	for <l2tp@l2tp.net>; Thu, 14 Dec 2000 09:40:18 -0800
Received: from eins.siemens.at (eins.siemens.at [193.81.246.11]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id IAA28849 for <l2tp@ipsec.org>; Thu, 14 Dec 2000 08:48:17 -0800 (PST)
Received: from scesie13.sie.siemens.at (forix [10.1.140.2])
	by eins.siemens.at  with ESMTP id eBEGlhk26523;
	Thu, 14 Dec 2000 17:47:43 +0100
Received: (from smap@localhost)
	by scesie13.sie.siemens.at (8.9.3/8.9.3) id RAA15681;
	Thu, 14 Dec 2000 17:47:42 +0100 (MET)
Received: from vies194a.sie.siemens.at(195.1.196.94) by scesie13 via smap (V2.0beta)
	id xma015341; Thu, 14 Dec 00 17:47:22 +0100
Received: by vies194a.sie.siemens.at with Internet Mail Service (5.5.2650.21)
	id <YYGJXM30>; Thu, 14 Dec 2000 17:47:20 +0100
Message-ID: <D9F2B9CD7BD5D21196BC0800060D9ED60437302F@vies186a.sie.siemens.at>
From: Klausberger Walter <walter.klausberger@siemens.at>
To: "'Evan Caves '" <evan@occamnetworks.com>,
        Klausberger Walter
	 <walter.klausberger@siemens.at>
Cc: "''''Dave Thaler ' ' ' '" <dthaler@Exchange.Microsoft.com>,
        "''''l2tp@ipsec.org ' ' ' '" <l2tp@ipsec.org>,
        "''''kzm@cisco.com ' ' ' '" <kzm@cisco.com>
Subject: RE: ifStackTable layering in L2TP
Date: Thu, 14 Dec 2000 17:47:20 +0100
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C065ED.868106D2"
Sender: owner-l2tp@diameter.org
Precedence: bulk

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_001_01C065ED.868106D2
Content-Type: text/plain

evan wrote:

>walter wrote:

>> Sorry, this was a mistype. I meant 'How do I know if IP packets come
>> from that AAL5 interface?'. I only see an interface.
>> 
>> - In case of PPP packets coming from this interface, I find an entry
>> in the L2TP MIB (in case of tunneling)
>> 
>> - In case of IP packets coming from this interface, do I get the
>> information via routing table (there should be a route to this IP
>> interface)?

>The ipAddrTable (RFC 1213) will tell you what ifEntries have an IP
>address configured on them.

Not in case of unnumbered IP interfaces (like ATM/AAL5,...).

walter
-

------_=_NextPart_001_01C065ED.868106D2
Content-Type: text/html

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=US-ASCII">
<META NAME="Generator" CONTENT="MS Exchange Server version 5.5.2650.12">
<TITLE>RE: ifStackTable layering in L2TP</TITLE>
</HEAD>
<BODY>

<P><FONT SIZE=2>evan wrote:</FONT>
</P>

<P><FONT SIZE=2>&gt;walter wrote:</FONT>
</P>

<P><FONT SIZE=2>&gt;&gt; Sorry, this was a mistype. I meant 'How do I know if IP packets come</FONT>
<BR><FONT SIZE=2>&gt;&gt; from that AAL5 interface?'. I only see an interface.</FONT>
<BR><FONT SIZE=2>&gt;&gt; </FONT>
<BR><FONT SIZE=2>&gt;&gt; - In case of PPP packets coming from this interface, I find an entry</FONT>
<BR><FONT SIZE=2>&gt;&gt; in the L2TP MIB (in case of tunneling)</FONT>
<BR><FONT SIZE=2>&gt;&gt; </FONT>
<BR><FONT SIZE=2>&gt;&gt; - In case of IP packets coming from this interface, do I get the</FONT>
<BR><FONT SIZE=2>&gt;&gt; information via routing table (there should be a route to this IP</FONT>
<BR><FONT SIZE=2>&gt;&gt; interface)?</FONT>
</P>

<P><FONT SIZE=2>&gt;The ipAddrTable (RFC 1213) will tell you what ifEntries have an IP</FONT>
<BR><FONT SIZE=2>&gt;address configured on them.</FONT>
</P>

<P><FONT SIZE=2>Not in case of unnumbered IP interfaces (like ATM/AAL5,...).</FONT>
</P>

<P><FONT SIZE=2>walter</FONT>
<BR><FONT SIZE=2>-</FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C065ED.868106D2--


From owner-l2tp@diameter.org  Thu Dec 21 12:56:12 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id MAA25886
	for <l2tpext-archive@odin.ietf.org>; Thu, 21 Dec 2000 12:56:11 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBLH6Ev03530
	for l2tp-list; Thu, 21 Dec 2000 09:06:14 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBLH62W03513
	for <l2tp@l2tp.net>; Thu, 21 Dec 2000 09:06:02 -0800
Received: from ietf.org (odin.ietf.org [132.151.1.176]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id IAA01061 for <l2tp@ipsec.org>; Thu, 21 Dec 2000 08:12:48 -0800 (PST)
Received: from CNRI.Reston.VA.US (localhost [127.0.0.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA15131;
	Thu, 21 Dec 2000 06:46:22 -0500 (EST)
Message-Id: <200012211146.GAA15131@ietf.org>
Mime-Version: 1.0
Content-Type: Multipart/Mixed; Boundary="NextPart"
To: IETF-Announce:;
Cc: l2tp@ipsec.org
From: Internet-Drafts@ietf.org
Reply-to: Internet-Drafts@ietf.org
Subject: I-D ACTION:draft-ietf-l2tpext-ds-02.txt
Date: Thu, 21 Dec 2000 06:46:15 -0500
Sender: owner-l2tp@diameter.org
Precedence: bulk

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the Layer Two Tunneling Protocol Extensions Working Group of the IETF.

	Title		: L2TP Differentiated Services Extension
	Author(s)	: P. Calhoun, D. McPherson, K. Peirce
	Filename	: draft-ietf-l2tpext-ds-02.txt
	Pages		: 9
	Date		: 20-Dec-00
	
The Layer Two Tunneling Protocol (L2TP) [RFC 2661] provides a
standard method for tunneling PPP [RFC 1661] packets.  The current
specification provides no provisions for supporting Differentiated
Services (diffserv) [RFC 2474, RFC 2475] over the L2TP control
connection or subsequent data sessions.  As a result, no standard
mechanism currently exists within L2TP to provide L2TP protocol
negotiations for service discrimination.

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-l2tpext-ds-02.txt

Internet-Drafts are also available by anonymous FTP. Login with the username
"anonymous" and a password of your e-mail address. After logging in,
type "cd internet-drafts" and then
	"get draft-ietf-l2tpext-ds-02.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt


Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-l2tpext-ds-02.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.
		
		
Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body;
	access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID:	<20001220160808.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-l2tpext-ds-02.txt

--OtherAccess
Content-Type: Message/External-body;
	name="draft-ietf-l2tpext-ds-02.txt";
	site="ftp.ietf.org";
	access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID:	<20001220160808.I-D@ietf.org>

--OtherAccess--

--NextPart--




From owner-l2tp@diameter.org  Sat Dec 23 07:48:39 2000
Received: from charizard.diameter.org (nobody@[24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id HAA04813
	for <l2tpext-archive@odin.ietf.org>; Sat, 23 Dec 2000 07:48:38 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBNBOLi06444
	for l2tp-list; Sat, 23 Dec 2000 03:24:21 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBNBOEW06434
	for <l2tp@l2tp.net>; Sat, 23 Dec 2000 03:24:14 -0800
Received: from shared1-mail.whowhere.com (shared1-qin.whowhere.com [209.185.123.111]) by ns1.sailpix.com (8.8.8/8.6.9) with SMTP id CAA04852 for <l2tp@ipsec.org>; Sat, 23 Dec 2000 02:31:08 -0800 (PST)
Received: from Unknown/Local ([?.?.?.?]) by shared1-mail.whowhere.com; Sat Dec 23 02:30:23 2000
To: l2tp@ipsec.org
Date: Sat, 23 Dec 2000 04:30:23 -0600
From: "Revati Mariappan" <revati_m@eudoramail.com>
Message-ID: <MOFEJBCFIMLOEAAA@shared1-mail.whowhere.com>
Mime-Version: 1.0
X-Sent-Mail: off
Reply-To: revati_m@eudoramail.com
X-Mailer: MailCity Service
Subject: ACCM
X-Sender-Ip: 203.197.138.200
Organization: QUALCOMM Eudora Web-Mail  (http://www.eudoramail.com:80) 
Content-Type: text/plain; charset=us-ascii
Content-Language: en
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

hi,

In which scenario is ACCM used ?
Is it for something like PPP convertor ?

revati


Join 18 million Eudora users by signing up for a free Eudora Web-Mail account at http://www.eudoramail.com


From owner-l2tp@diameter.org  Sat Dec 23 15:30:41 2000
Received: from charizard.diameter.org (nobody@[24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id PAA06619
	for <l2tpext-archive@odin.ietf.org>; Sat, 23 Dec 2000 15:30:41 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBNJW1006907
	for l2tp-list; Sat, 23 Dec 2000 11:32:01 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBNJVrW06901
	for <l2tp@l2tp.net>; Sat, 23 Dec 2000 11:31:53 -0800
Received: from h006008986325.ne.mediaone.net (h006008986325.ne.mediaone.net [24.218.16.153]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id KAA06415 for <l2tp@ipsec.org>; Sat, 23 Dec 2000 10:38:52 -0800 (PST)
Received: (from carlson@localhost)
	by h006008986325.ne.mediaone.net (8.11.0/8.11.0) id eBNIcfY06628;
	Sat, 23 Dec 2000 13:38:41 -0500
From: James Carlson <carlson@workingcode.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <14916.61745.502611.584600@h006008986325.ne.mediaone.net>
Date: Sat, 23 Dec 2000 13:38:41 -0500 (EST)
To: revati_m@eudoramail.com
Cc: l2tp@ipsec.org
Subject: Re: ACCM
In-Reply-To: Revati Mariappan's message of 23 December 2000 04:30:23
References: <MOFEJBCFIMLOEAAA@shared1-mail.whowhere.com>
X-Mailer: VM 6.75 under Emacs 20.6.1
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Revati Mariappan writes:
> In which scenario is ACCM used ?

It's used for negotiating away the default set of escaped characters
on AHDLC links.

> Is it for something like PPP convertor ?

L2TP itself is something like a PPP converter.

-- 
James Carlson                                  <carlson@workingcode.com>
"PPP Design and Debugging" --- http://people.ne.mediaone.net/carlson/ppp


From owner-l2tp@diameter.org  Thu Dec 28 02:21:37 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id CAA14244
	for <l2tpext-archive@odin.ietf.org>; Thu, 28 Dec 2000 02:21:36 -0500 (EST)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id eBS6SfT16093
	for l2tp-list; Wed, 27 Dec 2000 22:28:41 -0800
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id eBS6SNW16083
	for <l2tp@l2tp.net>; Wed, 27 Dec 2000 22:28:23 -0800
Received: from eve.speakeasy.net (eve.speakeasy.net [216.254.0.17]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id VAA15569 for <l2tp@ipsec.org>; Wed, 27 Dec 2000 21:35:37 -0800 (PST)
Received: from bbaumann-0.dsl.speakeasy.net (bbaumann-0.dsl.speakeasy.net [216.231.48.185])
	by eve.speakeasy.net (8.10.0/8.10.0) with ESMTP id eBS5VEp02128
	for <l2tp@ipsec.org>; Wed, 27 Dec 2000 21:31:14 -0800
Date: Wed, 27 Dec 2000 21:35:31 -0800 (PST)
From: Bill Baumann <bbaumann@speakeasy.org>
X-Sender: bbaumann@winner
To: l2tp@ipsec.org
Subject: L2TP & PPP Endpoint descriminators
In-Reply-To: <14916.61745.502611.584600@h006008986325.ne.mediaone.net>
Message-ID: <Pine.LNX.4.10.10012271433230.21238-100000@winner>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-l2tp@diameter.org
Precedence: bulk


I'm working on an open source implementation of a L2TP Compulsory LAC for
Linux.  Currently, it works minimally with incoming calls only.  We have
taken a stab at the endpoint descriminator issue, but are hoping someone
here could give some expert advice.

Current state:  We create two independent LCP layers between client and
LAC, and LAC and LNS.  The client & LAC negotiate LCP first.  The NULL
endpoint descriminator is sent to the client.  'What the client said' is
passed by LAC to the LNS during their LCP negotiations.  'What the LNS
said' is ignored.

The LNS receives the correct information thus should be ok.  The LAC
doesn't care.  So the question is -- will this work on the client?

We're trying to avoid the complexity and problems of having the LNS
renegotiate LCP with the client and still arrive at a general solution.


Bill Baumann

For those interested in the project, see www.ratdevil.com/l2tp.  The site
is a couple of months out of date, please contact me for current
information.








