From mailman-admin@ietf.org  Tue Apr  1 09:26:55 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA13505
	for <l2tpext-archive@lists.ietf.org>; Tue, 1 Apr 2003 09:26:55 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h31EofK18351
	for <l2tpext-archive@lists.ietf.org>; Tue, 1 Apr 2003 09:50:41 -0500
Date: Tue, 01 Apr 2003 09:50:40 -0500
Message-ID: <20030401145040.29326.82123.Mailman@www1.ietf.org>
Subject: ietf.org mailing list memberships reminder
From: mailman-owner@www1.ietf.org
To: l2tpext-archive@ietf.org
X-No-Archive: yes
X-Ack: no
Sender: mailman-admin@ietf.org
Errors-To: mailman-admin@ietf.org
X-BeenThere: mailman@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk

This is a reminder, sent out once a month, about your ietf.org mailing
list memberships.  It includes your subscription info and how to use
it to change it or unsubscribe from a list.

You can visit the URLs to change your membership status or
configuration, including unsubscribing, setting digest-style delivery
or disabling delivery altogether (e.g., for a vacation), and so on.

In addition to the URL interfaces, you can also use email to make such
changes.  For more info, send a message to the '-request' address of
the list (for example, l2tpext-request@ietf.org) containing just the
word 'help' in the message body, and an email message will be sent to
you with instructions.

***************************************************************************


                              Note Well

All statements related to the activities of the IETF and addressed to
the IETF are subject to all provisions of Section 10 of RFC 2026,
which grants to the IETF and its participants certain licenses and
rights in such statements. Such statements include verbal statements
in IETF meetings, as well as written and electronic communications
made at any time or place, which are addressed to

        * the IETF plenary session,
        * any IETF working group or portion thereof,
        * the IESG, or any member thereof on behalf of the IESG,
        * the IAB or any member thereof on behalf of the IAB,
        * any IETF mailing list, including the IETF list itself, any
working
            group or design team list, or any other list functioning
under IETF
            auspices,
        * the RFC Editor or the Internet-Drafts function

Statements made outside of an IETF meeting, mailing list or other
function, that are clearly not intended to be input to an IETF
activity, group or function, are not subject to these provisions.

   
***************************************************************************


If you have questions, problems, comments, etc, send them to
mailman-owner@www1.ietf.org.  Thanks!

Passwords for l2tpext-archive@lists.ietf.org:

List                                     Password // URL
----                                     --------  
l2tpext@ietf.org                         ruNF      
https://www1.ietf.org/mailman/options/l2tpext/l2tpext-archive%40lists.ietf.org


From l2tpext-admin@ietf.org  Tue Apr  1 09:34:09 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA14039
	for <l2tpext-archive@lists.ietf.org>; Tue, 1 Apr 2003 09:34:09 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h31EqaK19589;
	Tue, 1 Apr 2003 09:52:36 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h31EpiK18497
	for <l2tpext@optimus.ietf.org>; Tue, 1 Apr 2003 09:51:44 -0500
Received: from LONMSW03.uk.eu.colt (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA13543
	for <l2tpext@ietf.org>; Tue, 1 Apr 2003 09:27:22 -0500 (EST)
From: Andreas.Waibel@colt.de
Message-ID: <4403667CFF70F84283D8521D50CF4CF3036E30F1@franoa03.de.eu.colt>
To: l2tpext@ietf.org
Date: Tue, 1 Apr 2003 15:29:45 +0100 
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2653.19)
Content-Type: text/plain;
	charset="iso-8859-1"
Subject: [L2tpext] fragmentation and MRU setting in L2TPv2 scenarios
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

Colleagues,

I have observed problems regarding the treatment of packets 
with DF bit set in L2TPv2 architectures. Unfortunately the
statement in the RFC 2661 is not too verbose:

"IP fragmentation may occur as the L2TP packet travels over the IP
   substrate. L2TP makes no special efforts to optimize this. A LAC
   implementation MAY cause its LCP to negotiate for a specific MRU,
   which could optimize for LAC environments in which the MTU's of the
   path over which the L2TP packets are likely to travel have a
   consistent value."

I also saw a document "draft-ietf-pwe3-fragmentation-00.txt"
but this also appears not to answer my questions.

I have formulated several questions below, that might help
to clarify the problem and possible solutions to it.

a) consider a dial setup similar to the following:


client ----- LAC ------ LNS
  |---------ppp---------|

Due to required L2TP overhead fragmentation
of packets larger than 1460 bytes should happen
no matter what the MRU/MTU settings on client, LAC
or LNS are. This is true, if the DF bit is not set
in the packet. Normally only testing applications
should actively prevent fragmention using the DF Bit.

However there are some applications (e.g ftp server) 
and clients in the internet that set the DF bit in requests 
or responses. I think they should not do that (except if 
it is for testing), but they do and customers use those 
applications. This generally does not cause problems if the 
packet size transported without fragmentation
is 1500. But we've seen problems when L2TP is used as
some of the data packets are between 1460 and 1500 
bytes and some applications stop communication, if they 
are lost.

Q: Can You confirm that L2TP implementations should always 
care for the DF bit and drop too large packets (+ return an
ICMP)? 

Q: Can You confirm that the setting of the MRU on the
LNS does not matter with respect to fragmentation as 
long as it is => the maximum tunnel width for data of 
1460 bytes?

Q: Are there any best-practice workarounds if a broken
application continues to set the DF bit in its packets
and uses packet sizes between 1460 and 1500?

b) consider a DSL setup similar to the following:

client ---atmPVC--- LAC ------ LNS
  |---------ppp-----------------|

In this scenario we might have considerably larger
maximum packet sizes between client and LNS than
1500, e.g. 1572. Then appearantly the MRU setting 
on the LNS has impact as it might be lower than the 
otherwise maximum width of the pipe. 

Q: Can L2TP transport packet sizes > 1500?

Q: Are there any reasons to set a lower MRU than
1500 on the LNS, if the overall width of the tunnel pipe
is e.g. 1572? 

best regards & thank You for Your help!

Andreas
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext


From mailnull@www1.ietf.org  Tue Apr  1 09:35:13 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA14077
	for <l2tpext-archive@odin.ietf.org>; Tue, 1 Apr 2003 09:35:13 -0500 (EST)
Received: (from mailnull@localhost)
	by www1.ietf.org (8.11.6/8.11.6) id h31Ewx120414
	for l2tpext-archive@odin.ietf.org; Tue, 1 Apr 2003 09:58:59 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h31EwxK20411
	for <l2tpext-web-archive@optimus.ietf.org>; Tue, 1 Apr 2003 09:58:59 -0500
Received: from www1.ietf.org (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA14061
	for <l2tpext-web-archive@ietf.org>; Tue, 1 Apr 2003 09:34:42 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h31EqaK19589;
	Tue, 1 Apr 2003 09:52:36 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h31EpiK18497
	for <l2tpext@optimus.ietf.org>; Tue, 1 Apr 2003 09:51:44 -0500
Received: from LONMSW03.uk.eu.colt (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA13543
	for <l2tpext@ietf.org>; Tue, 1 Apr 2003 09:27:22 -0500 (EST)
From: Andreas.Waibel@colt.de
Message-ID: <4403667CFF70F84283D8521D50CF4CF3036E30F1@franoa03.de.eu.colt>
To: l2tpext@ietf.org
Date: Tue, 1 Apr 2003 15:29:45 +0100 
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2653.19)
Content-Type: text/plain;
	charset="iso-8859-1"
Subject: [L2tpext] fragmentation and MRU setting in L2TPv2 scenarios
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

Colleagues,

I have observed problems regarding the treatment of packets 
with DF bit set in L2TPv2 architectures. Unfortunately the
statement in the RFC 2661 is not too verbose:

"IP fragmentation may occur as the L2TP packet travels over the IP
   substrate. L2TP makes no special efforts to optimize this. A LAC
   implementation MAY cause its LCP to negotiate for a specific MRU,
   which could optimize for LAC environments in which the MTU's of the
   path over which the L2TP packets are likely to travel have a
   consistent value."

I also saw a document "draft-ietf-pwe3-fragmentation-00.txt"
but this also appears not to answer my questions.

I have formulated several questions below, that might help
to clarify the problem and possible solutions to it.

a) consider a dial setup similar to the following:


client ----- LAC ------ LNS
  |---------ppp---------|

Due to required L2TP overhead fragmentation
of packets larger than 1460 bytes should happen
no matter what the MRU/MTU settings on client, LAC
or LNS are. This is true, if the DF bit is not set
in the packet. Normally only testing applications
should actively prevent fragmention using the DF Bit.

However there are some applications (e.g ftp server) 
and clients in the internet that set the DF bit in requests 
or responses. I think they should not do that (except if 
it is for testing), but they do and customers use those 
applications. This generally does not cause problems if the 
packet size transported without fragmentation
is 1500. But we've seen problems when L2TP is used as
some of the data packets are between 1460 and 1500 
bytes and some applications stop communication, if they 
are lost.

Q: Can You confirm that L2TP implementations should always 
care for the DF bit and drop too large packets (+ return an
ICMP)? 

Q: Can You confirm that the setting of the MRU on the
LNS does not matter with respect to fragmentation as 
long as it is => the maximum tunnel width for data of 
1460 bytes?

Q: Are there any best-practice workarounds if a broken
application continues to set the DF bit in its packets
and uses packet sizes between 1460 and 1500?

b) consider a DSL setup similar to the following:

client ---atmPVC--- LAC ------ LNS
  |---------ppp-----------------|

In this scenario we might have considerably larger
maximum packet sizes between client and LNS than
1500, e.g. 1572. Then appearantly the MRU setting 
on the LNS has impact as it might be lower than the 
otherwise maximum width of the pipe. 

Q: Can L2TP transport packet sizes > 1500?

Q: Are there any reasons to set a lower MRU than
1500 on the LNS, if the overall width of the tunnel pipe
is e.g. 1572? 

best regards & thank You for Your help!

Andreas
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext



From l2tpext-admin@ietf.org  Wed Apr  2 04:44:25 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id EAA01944
	for <l2tpext-archive@lists.ietf.org>; Wed, 2 Apr 2003 04:44:25 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32A4bK21384;
	Wed, 2 Apr 2003 05:04:37 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32A0OK21213
	for <l2tpext@optimus.ietf.org>; Wed, 2 Apr 2003 05:00:24 -0500
Received: from LONMSW03.uk.eu.colt (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id EAA01804
	for <l2tpext@ietf.org>; Wed, 2 Apr 2003 04:35:43 -0500 (EST)
From: Andreas.Waibel@colt.de
Message-ID: <4403667CFF70F84283D8521D50CF4CF3036E3100@franoa03.de.eu.colt>
To: l2tpext@ietf.org
Subject: AW: [L2tpext] fragmentation and MRU setting in L2TPv2 scenarios
Date: Wed, 2 Apr 2003 10:38:07 +0100 
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2653.19)
Content-Type: text/plain;
	charset="iso-8859-1"
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

Archie,

thank You for helping to reduce my confusion ;-)
it's getting less but still a few things:

> The DF flag is set as a normal part of TCP's path-MTU
> discovery process:

sure, but I thought it should not be set in any normal
application data packet, once the path is discovered
or if PMTU was not enabled at all.

> If peer A's MRU is < peer B's MTU, then there is a bug
> in somebody's PPP stack, as PPP negotiation should never
> allow this to happen.

The negotiated MRU/MTU value agreed on by both sides could 
still be larger than the actual maximum packet size for 
unfragmented transport over the wire. My interpretation of 
the PPP logs I have seen is that negotiation usally only depends 
on the (manual) settings on both sides. The actual maximum 
unfragmented packet size in the L2TP pipe may differ from 
the negotiated value.
The question is now whether it makes sense to adapt the settings
of the MRU on the LNS and/or LAC (used as proposal for PPP negotiation) 
to the maximum unfragmented packet size. Or are there any caveats
or side effects when doing this?

> > Q: Are there any best-practice workarounds if a broken
> > application continues to set the DF bit in its packets
> > and uses packet sizes between 1460 and 1500?
> The application is not broken. Probably a firewall somewhere
> is blocking the ICMP "need fragmentation" packets from getting
> back to the application. See http://www.faqs.org/rfcs/rfc2923.html

I understand that an application using PMTU sends periodically large 
packets with the DF bit set. However I'm not sure, whether these testing
packets are already normal data packets or just ICMP trials BEFORE
actually sending data. One way or another I understand now, that
the problem only happens if the application has PMTU enabled and 
does not have a timeout mechanism as described in RFC 2923

   How to fix
      TCP should notice that the connection is timing out.  After
      several timeouts, TCP should attempt to send smaller packets,
      perhaps turning off the DF flag for each packet.  If this
      succeeds, it should continue to turn off PMTUD for the connection
      for some reasonable period of time, after which it should probe
      again to try to determine if the path has changed.

best regards

Andreas



_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext


From mailnull@www1.ietf.org  Wed Apr  2 04:45:14 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id EAA01970
	for <l2tpext-archive@odin.ietf.org>; Wed, 2 Apr 2003 04:45:14 -0500 (EST)
Received: (from mailnull@localhost)
	by www1.ietf.org (8.11.6/8.11.6) id h32A9Ng22406
	for l2tpext-archive@odin.ietf.org; Wed, 2 Apr 2003 05:09:23 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32A9NK22403
	for <l2tpext-web-archive@optimus.ietf.org>; Wed, 2 Apr 2003 05:09:23 -0500
Received: from www1.ietf.org (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id EAA01963
	for <l2tpext-web-archive@ietf.org>; Wed, 2 Apr 2003 04:44:43 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32A4bK21384;
	Wed, 2 Apr 2003 05:04:37 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32A0OK21213
	for <l2tpext@optimus.ietf.org>; Wed, 2 Apr 2003 05:00:24 -0500
Received: from LONMSW03.uk.eu.colt (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id EAA01804
	for <l2tpext@ietf.org>; Wed, 2 Apr 2003 04:35:43 -0500 (EST)
From: Andreas.Waibel@colt.de
Message-ID: <4403667CFF70F84283D8521D50CF4CF3036E3100@franoa03.de.eu.colt>
To: l2tpext@ietf.org
Subject: AW: [L2tpext] fragmentation and MRU setting in L2TPv2 scenarios
Date: Wed, 2 Apr 2003 10:38:07 +0100 
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2653.19)
Content-Type: text/plain;
	charset="iso-8859-1"
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

Archie,

thank You for helping to reduce my confusion ;-)
it's getting less but still a few things:

> The DF flag is set as a normal part of TCP's path-MTU
> discovery process:

sure, but I thought it should not be set in any normal
application data packet, once the path is discovered
or if PMTU was not enabled at all.

> If peer A's MRU is < peer B's MTU, then there is a bug
> in somebody's PPP stack, as PPP negotiation should never
> allow this to happen.

The negotiated MRU/MTU value agreed on by both sides could 
still be larger than the actual maximum packet size for 
unfragmented transport over the wire. My interpretation of 
the PPP logs I have seen is that negotiation usally only depends 
on the (manual) settings on both sides. The actual maximum 
unfragmented packet size in the L2TP pipe may differ from 
the negotiated value.
The question is now whether it makes sense to adapt the settings
of the MRU on the LNS and/or LAC (used as proposal for PPP negotiation) 
to the maximum unfragmented packet size. Or are there any caveats
or side effects when doing this?

> > Q: Are there any best-practice workarounds if a broken
> > application continues to set the DF bit in its packets
> > and uses packet sizes between 1460 and 1500?
> The application is not broken. Probably a firewall somewhere
> is blocking the ICMP "need fragmentation" packets from getting
> back to the application. See http://www.faqs.org/rfcs/rfc2923.html

I understand that an application using PMTU sends periodically large 
packets with the DF bit set. However I'm not sure, whether these testing
packets are already normal data packets or just ICMP trials BEFORE
actually sending data. One way or another I understand now, that
the problem only happens if the application has PMTU enabled and 
does not have a timeout mechanism as described in RFC 2923

   How to fix
      TCP should notice that the connection is timing out.  After
      several timeouts, TCP should attempt to send smaller packets,
      perhaps turning off the DF flag for each packet.  If this
      succeeds, it should continue to turn off PMTUD for the connection
      for some reasonable period of time, after which it should probe
      again to try to determine if the path has changed.

best regards

Andreas



_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext



From l2tpext-admin@ietf.org  Wed Apr  2 06:28:02 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA04089
	for <l2tpext-archive@lists.ietf.org>; Wed, 2 Apr 2003 06:28:02 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32Bo4K29689;
	Wed, 2 Apr 2003 06:50:04 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h31GuiK31110
	for <l2tpext@optimus.ietf.org>; Tue, 1 Apr 2003 11:56:44 -0500
Received: from mailman.precisionio.com (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA27267
	for <l2tpext@ietf.org>; Tue, 1 Apr 2003 11:32:23 -0500 (EST)
Received: from precisionio.com (bubba.precisionio.com [172.16.0.223])
	by mailman.precisionio.com (8.12.8/8.12.6) with ESMTP id h31GYlnr095477;
	Tue, 1 Apr 2003 08:34:48 -0800 (PST)
	(envelope-from archie@precisionio.com)
Message-ID: <3E89BFA7.7080308@precisionio.com>
Date: Tue, 01 Apr 2003 08:34:47 -0800
From: Archie Cobbs <archie@precisionio.com>
Organization: Precision I/O
User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.1) Gecko/20021106
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Andreas.Waibel@colt.de
CC: l2tpext@ietf.org
Subject: Re: [L2tpext] fragmentation and MRU setting in L2TPv2 scenarios
References: <4403667CFF70F84283D8521D50CF4CF3036E30F1@franoa03.de.eu.colt>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Content-Transfer-Encoding: 7bit
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>
Content-Transfer-Encoding: 7bit

Andreas.Waibel@colt.de wrote:
> a) consider a dial setup similar to the following:
> 
> 
> client ----- LAC ------ LNS
>   |---------ppp---------|
> 
> Due to required L2TP overhead fragmentation
> of packets larger than 1460 bytes should happen
> no matter what the MRU/MTU settings on client, LAC
> or LNS are. This is true, if the DF bit is not set
> in the packet. Normally only testing applications
> should actively prevent fragmention using the DF Bit.

The DF flag is set as a normal part of TCP's path-MTU
discovery process:

http://www.google.com/search?q=path+MTU+discovery+TCP&sourceid=mozilla-search&start=0&start=0

> Q: Can You confirm that L2TP implementations should always 
> care for the DF bit and drop too large packets (+ return an
> ICMP)? 

Yes; more generally, this is required for all IP routers.

> Q: Can You confirm that the setting of the MRU on the
> LNS does not matter with respect to fragmentation as 
> long as it is => the maximum tunnel width for data of 
> 1460 bytes?

If peer A's MRU is < peer B's MTU, then there is a bug
in somebody's PPP stack, as PPP negotiation should never
allow this to happen.

Otherwise, yes there should be no problem receiving packets because
they'll all be <= MRU.

> Q: Are there any best-practice workarounds if a broken
> application continues to set the DF bit in its packets
> and uses packet sizes between 1460 and 1500?

The application is not broken. Probably a firewall somewhere
is blocking the ICMP "need fragmentation" packets from getting
back to the application. See http://www.faqs.org/rfcs/rfc2923.html

> Q: Can L2TP transport packet sizes > 1500?

Sure, use multi-link PPP and use whatever MRRU setting you want.

-Archie

__________________________________________________________________________
Archie Cobbs     *     Precision I/O      *     http://www.precisionio.com

_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext


From mailnull@www1.ietf.org  Wed Apr  2 06:29:23 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA04225
	for <l2tpext-archive@odin.ietf.org>; Wed, 2 Apr 2003 06:29:23 -0500 (EST)
Received: (from mailnull@localhost)
	by www1.ietf.org (8.11.6/8.11.6) id h32BrZQ29953
	for l2tpext-archive@odin.ietf.org; Wed, 2 Apr 2003 06:53:35 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32BrZK29950
	for <l2tpext-web-archive@optimus.ietf.org>; Wed, 2 Apr 2003 06:53:35 -0500
Received: from www1.ietf.org (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA04214
	for <l2tpext-web-archive@ietf.org>; Wed, 2 Apr 2003 06:28:52 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32Bo4K29689;
	Wed, 2 Apr 2003 06:50:04 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h31GuiK31110
	for <l2tpext@optimus.ietf.org>; Tue, 1 Apr 2003 11:56:44 -0500
Received: from mailman.precisionio.com (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA27267
	for <l2tpext@ietf.org>; Tue, 1 Apr 2003 11:32:23 -0500 (EST)
Received: from precisionio.com (bubba.precisionio.com [172.16.0.223])
	by mailman.precisionio.com (8.12.8/8.12.6) with ESMTP id h31GYlnr095477;
	Tue, 1 Apr 2003 08:34:48 -0800 (PST)
	(envelope-from archie@precisionio.com)
Message-ID: <3E89BFA7.7080308@precisionio.com>
Date: Tue, 01 Apr 2003 08:34:47 -0800
From: Archie Cobbs <archie@precisionio.com>
Organization: Precision I/O
User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.1) Gecko/20021106
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Andreas.Waibel@colt.de
CC: l2tpext@ietf.org
Subject: Re: [L2tpext] fragmentation and MRU setting in L2TPv2 scenarios
References: <4403667CFF70F84283D8521D50CF4CF3036E30F1@franoa03.de.eu.colt>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Content-Transfer-Encoding: 7bit
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>
Content-Transfer-Encoding: 7bit
Content-Transfer-Encoding: 7bit

Andreas.Waibel@colt.de wrote:
> a) consider a dial setup similar to the following:
> 
> 
> client ----- LAC ------ LNS
>   |---------ppp---------|
> 
> Due to required L2TP overhead fragmentation
> of packets larger than 1460 bytes should happen
> no matter what the MRU/MTU settings on client, LAC
> or LNS are. This is true, if the DF bit is not set
> in the packet. Normally only testing applications
> should actively prevent fragmention using the DF Bit.

The DF flag is set as a normal part of TCP's path-MTU
discovery process:

http://www.google.com/search?q=path+MTU+discovery+TCP&sourceid=mozilla-search&start=0&start=0

> Q: Can You confirm that L2TP implementations should always 
> care for the DF bit and drop too large packets (+ return an
> ICMP)? 

Yes; more generally, this is required for all IP routers.

> Q: Can You confirm that the setting of the MRU on the
> LNS does not matter with respect to fragmentation as 
> long as it is => the maximum tunnel width for data of 
> 1460 bytes?

If peer A's MRU is < peer B's MTU, then there is a bug
in somebody's PPP stack, as PPP negotiation should never
allow this to happen.

Otherwise, yes there should be no problem receiving packets because
they'll all be <= MRU.

> Q: Are there any best-practice workarounds if a broken
> application continues to set the DF bit in its packets
> and uses packet sizes between 1460 and 1500?

The application is not broken. Probably a firewall somewhere
is blocking the ICMP "need fragmentation" packets from getting
back to the application. See http://www.faqs.org/rfcs/rfc2923.html

> Q: Can L2TP transport packet sizes > 1500?

Sure, use multi-link PPP and use whatever MRRU setting you want.

-Archie

__________________________________________________________________________
Archie Cobbs     *     Precision I/O      *     http://www.precisionio.com

_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext



From l2tpext-admin@ietf.org  Wed Apr  2 11:46:56 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA17483
	for <l2tpext-archive@lists.ietf.org>; Wed, 2 Apr 2003 11:46:56 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32H8tK25835;
	Wed, 2 Apr 2003 12:08:55 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32H7dK25667
	for <l2tpext@optimus.ietf.org>; Wed, 2 Apr 2003 12:07:39 -0500
Received: from InterJet.dellroad.org (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA17193
	for <l2tpext@ietf.org>; Wed, 2 Apr 2003 11:42:48 -0500 (EST)
Received: from precisionio.com (arch20m.dellroad.org [10.1.1.20])
	by InterJet.dellroad.org (8.9.1a/8.9.1) with ESMTP id IAA71618;
	Wed, 2 Apr 2003 08:34:10 -0800 (PST)
Message-ID: <3E8B1102.1080502@precisionio.com>
Date: Wed, 02 Apr 2003 08:34:10 -0800
From: Archie Cobbs <archie@precisionio.com>
User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.1) Gecko/20021106
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Andreas.Waibel@colt.de
CC: l2tpext@ietf.org
Subject: Re: AW: [L2tpext] fragmentation and MRU setting in L2TPv2 scenarios
References: <4403667CFF70F84283D8521D50CF4CF3036E3100@franoa03.de.eu.colt>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Content-Transfer-Encoding: 7bit
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>
Content-Transfer-Encoding: 7bit

Andreas.Waibel@colt.de wrote:
>>The DF flag is set as a normal part of TCP's path-MTU
>>discovery process:
> 
> sure, but I thought it should not be set in any normal
> application data packet, once the path is discovered
> or if PMTU was not enabled at all.

That's true.. although this falls under "what normally happens"
rather than "what is required". E.g., there's nothing that
says that a host can't initiate another PMTU discovery
every 3 minutes if it wants to.

>>If peer A's MRU is < peer B's MTU, then there is a bug
>>in somebody's PPP stack, as PPP negotiation should never
>>allow this to happen.
> 
> The negotiated MRU/MTU value agreed on by both sides could 
> still be larger than the actual maximum packet size for 
> unfragmented transport over the wire. My interpretation of 
> the PPP logs I have seen is that negotiation usally only depends 
> on the (manual) settings on both sides. The actual maximum 
> unfragmented packet size in the L2TP pipe may differ from 
> the negotiated value.

The problem is that you don't control your PPP MTU, the peer
does, and the peer is allowed to reject any values greater
than 1500 bytes if it wants to. So the local configuration
only sets an upper limit, not a lower limit in effect.

> The question is now whether it makes sense to adapt the settings
> of the MRU on the LNS and/or LAC (used as proposal for PPP negotiation) 
> to the maximum unfragmented packet size. Or are there any caveats
> or side effects when doing this?

There are two levels at which IP fragmentation can happen, on the
inner IP packet and on the outer IP packet. Outer packet fragments
are a potential performance hit but otherwise not a problem. But
if the inner IP packet must be fragmented - because you're not
doing multi-link and your MTU for the L2TP link is too small for
the original packet plus overhead - then your L2TP router is
required to fragment the IP packet or send back an ICMP "must
fragment" response if DF is set.

You could avoid "inner" IP fragmentation if your L2TP link MTU
were large enough, but your MTU setting is at the mercy of the
remote PPP peer, because it is allowed to reject anything larger
than 1500 bytes (and typically the inner IP packet is up to 1500
bytes because it came from an Ethernet port). So you can't avoid
this possibility in general.

If you're talking about a UNIX type device, then you would normally
configure the L2TP interface to have an MTU of 1460 (or whatever)
so that the kernel handles the fragmentation and ICMP stuff for you.

With multi-link PPP, inner IP fragmentation is no longer necessary
as you can fragment at the PPP frame level instead.

So in summary:

- You must be prepared to fragment and/or send ICMP "must fragment"
   packets (don't forget to decrement TTL's too)
- Life gets easier if you enable multi-link PPP (even with just
   one link)
- Life gets easier if you can get the peer to accept a large
   enough MTU for the L2TP link, but don't count on it.

FYI, WinXP is "friendlier" than earlier versions of Windows
on these points...

> I understand that an application using PMTU sends periodically large 
> packets with the DF bit set. However I'm not sure, whether these testing
> packets are already normal data packets or just ICMP trials BEFORE
> actually sending data. One way or another I understand now, that
> the problem only happens if the application has PMTU enabled and 
> does not have a timeout mechanism as described in RFC 2923
> 
>    How to fix
>       TCP should notice that the connection is timing out.  After
>       several timeouts, TCP should attempt to send smaller packets,
>       perhaps turning off the DF flag for each packet.  If this
>       succeeds, it should continue to turn off PMTUD for the connection
>       for some reasonable period of time, after which it should probe
>       again to try to determine if the path has changed.

You want to avoid having the application timeout.. it's much better to
have all routers able to send ICMP "need fragmentation" packets and
all intermediate firewalls not block these packets.

Hope this helps.

-Archie

__________________________________________________________________________
Archie Cobbs     *     Precision I/O      *     http://www.precisionio.com

_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext


From mailnull@www1.ietf.org  Wed Apr  2 11:47:46 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA17549
	for <l2tpext-archive@odin.ietf.org>; Wed, 2 Apr 2003 11:47:46 -0500 (EST)
Received: (from mailnull@localhost)
	by www1.ietf.org (8.11.6/8.11.6) id h32HC4n26082
	for l2tpext-archive@odin.ietf.org; Wed, 2 Apr 2003 12:12:04 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32HC4K26079
	for <l2tpext-web-archive@optimus.ietf.org>; Wed, 2 Apr 2003 12:12:04 -0500
Received: from www1.ietf.org (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA17523
	for <l2tpext-web-archive@ietf.org>; Wed, 2 Apr 2003 11:47:14 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32H8tK25835;
	Wed, 2 Apr 2003 12:08:55 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h32H7dK25667
	for <l2tpext@optimus.ietf.org>; Wed, 2 Apr 2003 12:07:39 -0500
Received: from InterJet.dellroad.org (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA17193
	for <l2tpext@ietf.org>; Wed, 2 Apr 2003 11:42:48 -0500 (EST)
Received: from precisionio.com (arch20m.dellroad.org [10.1.1.20])
	by InterJet.dellroad.org (8.9.1a/8.9.1) with ESMTP id IAA71618;
	Wed, 2 Apr 2003 08:34:10 -0800 (PST)
Message-ID: <3E8B1102.1080502@precisionio.com>
Date: Wed, 02 Apr 2003 08:34:10 -0800
From: Archie Cobbs <archie@precisionio.com>
User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.1) Gecko/20021106
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Andreas.Waibel@colt.de
CC: l2tpext@ietf.org
Subject: Re: AW: [L2tpext] fragmentation and MRU setting in L2TPv2 scenarios
References: <4403667CFF70F84283D8521D50CF4CF3036E3100@franoa03.de.eu.colt>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Content-Transfer-Encoding: 7bit
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>
Content-Transfer-Encoding: 7bit
Content-Transfer-Encoding: 7bit

Andreas.Waibel@colt.de wrote:
>>The DF flag is set as a normal part of TCP's path-MTU
>>discovery process:
> 
> sure, but I thought it should not be set in any normal
> application data packet, once the path is discovered
> or if PMTU was not enabled at all.

That's true.. although this falls under "what normally happens"
rather than "what is required". E.g., there's nothing that
says that a host can't initiate another PMTU discovery
every 3 minutes if it wants to.

>>If peer A's MRU is < peer B's MTU, then there is a bug
>>in somebody's PPP stack, as PPP negotiation should never
>>allow this to happen.
> 
> The negotiated MRU/MTU value agreed on by both sides could 
> still be larger than the actual maximum packet size for 
> unfragmented transport over the wire. My interpretation of 
> the PPP logs I have seen is that negotiation usally only depends 
> on the (manual) settings on both sides. The actual maximum 
> unfragmented packet size in the L2TP pipe may differ from 
> the negotiated value.

The problem is that you don't control your PPP MTU, the peer
does, and the peer is allowed to reject any values greater
than 1500 bytes if it wants to. So the local configuration
only sets an upper limit, not a lower limit in effect.

> The question is now whether it makes sense to adapt the settings
> of the MRU on the LNS and/or LAC (used as proposal for PPP negotiation) 
> to the maximum unfragmented packet size. Or are there any caveats
> or side effects when doing this?

There are two levels at which IP fragmentation can happen, on the
inner IP packet and on the outer IP packet. Outer packet fragments
are a potential performance hit but otherwise not a problem. But
if the inner IP packet must be fragmented - because you're not
doing multi-link and your MTU for the L2TP link is too small for
the original packet plus overhead - then your L2TP router is
required to fragment the IP packet or send back an ICMP "must
fragment" response if DF is set.

You could avoid "inner" IP fragmentation if your L2TP link MTU
were large enough, but your MTU setting is at the mercy of the
remote PPP peer, because it is allowed to reject anything larger
than 1500 bytes (and typically the inner IP packet is up to 1500
bytes because it came from an Ethernet port). So you can't avoid
this possibility in general.

If you're talking about a UNIX type device, then you would normally
configure the L2TP interface to have an MTU of 1460 (or whatever)
so that the kernel handles the fragmentation and ICMP stuff for you.

With multi-link PPP, inner IP fragmentation is no longer necessary
as you can fragment at the PPP frame level instead.

So in summary:

- You must be prepared to fragment and/or send ICMP "must fragment"
   packets (don't forget to decrement TTL's too)
- Life gets easier if you enable multi-link PPP (even with just
   one link)
- Life gets easier if you can get the peer to accept a large
   enough MTU for the L2TP link, but don't count on it.

FYI, WinXP is "friendlier" than earlier versions of Windows
on these points...

> I understand that an application using PMTU sends periodically large 
> packets with the DF bit set. However I'm not sure, whether these testing
> packets are already normal data packets or just ICMP trials BEFORE
> actually sending data. One way or another I understand now, that
> the problem only happens if the application has PMTU enabled and 
> does not have a timeout mechanism as described in RFC 2923
> 
>    How to fix
>       TCP should notice that the connection is timing out.  After
>       several timeouts, TCP should attempt to send smaller packets,
>       perhaps turning off the DF flag for each packet.  If this
>       succeeds, it should continue to turn off PMTUD for the connection
>       for some reasonable period of time, after which it should probe
>       again to try to determine if the path has changed.

You want to avoid having the application timeout.. it's much better to
have all routers able to send ICMP "need fragmentation" packets and
all intermediate firewalls not block these packets.

Hope this helps.

-Archie

__________________________________________________________________________
Archie Cobbs     *     Precision I/O      *     http://www.precisionio.com

_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext



From l2tpext-admin@ietf.org  Fri Apr  4 06:52:15 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA16853
	for <l2tpext-archive@lists.ietf.org>; Fri, 4 Apr 2003 06:52:15 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h34BqXK05938;
	Fri, 4 Apr 2003 06:52:33 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h34BomK05670
	for <l2tpext@optimus.ietf.org>; Fri, 4 Apr 2003 06:50:48 -0500
Received: from LONMSW03.uk.eu.colt (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA16046
	for <l2tpext@ietf.org>; Fri, 4 Apr 2003 06:47:30 -0500 (EST)
From: Andreas.Waibel@colt.de
Message-ID: <4403667CFF70F84283D8521D50CF4CF3036E3119@franoa03.de.eu.colt>
To: l2tpext@ietf.org
Subject: AW: AW: [L2tpext] fragmentation and MRU setting in L2TPv2 scenari
	os
Date: Fri, 4 Apr 2003 11:39:42 +0100 
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2653.19)
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by www1.ietf.org id h34BomK05671
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>
Content-Transfer-Encoding: 8bit

Archie,

one thing puzzles me a bit in Your answer. You are
talking about negotiating a MTU value with the PPP
peer LARGER than 1500, while I am considering to propose
a MRU value (LNS) respectively MTU value (peer) for the
PPP connection SMALLER than 1500 (=equal to the maximum
unfragmented packet size in the pipe).
While I'm quite sure, that it would amend performance if
peer agrees on such a smaller value for the PPP link, I 
wonder whether it might make the PMTU problem worse for 
large packets from the server side.

best regards

Andreas

> > The question is now whether it makes sense to adapt the settings
> > of the MRU on the LNS and/or LAC (used as proposal for PPP 
> negotiation) 
> > to the maximum unfragmented packet size. Or are there any caveats
> > or side effects when doing this?
> 
> There are two levels at which IP fragmentation can happen, on the
> inner IP packet and on the outer IP packet. Outer packet fragments
> are a potential performance hit but otherwise not a problem. But
> if the inner IP packet must be fragmented - because you're not
> doing multi-link and your MTU for the L2TP link is too small for
> the original packet plus overhead - then your L2TP router is
> required to fragment the IP packet or send back an ICMP "must
> fragment" response if DF is set.
> 

> -----Ursprüngliche Nachricht-----
> Von: Archie Cobbs [mailto:archie@precisionio.com]
> Gesendet am: Mittwoch, 2. April 2003 18:34
> An: Waibel, Andreas
> Cc: l2tpext@ietf.org
> Betreff: Re: AW: [L2tpext] fragmentation and MRU setting in L2TPv2
> scenarios
> 
> Andreas.Waibel@colt.de wrote:
> >>The DF flag is set as a normal part of TCP's path-MTU
> >>discovery process:
> > 
> > sure, but I thought it should not be set in any normal
> > application data packet, once the path is discovered
> > or if PMTU was not enabled at all.
> 
> That's true.. although this falls under "what normally happens"
> rather than "what is required". E.g., there's nothing that
> says that a host can't initiate another PMTU discovery
> every 3 minutes if it wants to.
> 
> >>If peer A's MRU is < peer B's MTU, then there is a bug
> >>in somebody's PPP stack, as PPP negotiation should never
> >>allow this to happen.
> > 
> > The negotiated MRU/MTU value agreed on by both sides could 
> > still be larger than the actual maximum packet size for 
> > unfragmented transport over the wire. My interpretation of 
> > the PPP logs I have seen is that negotiation usally only depends 
> > on the (manual) settings on both sides. The actual maximum 
> > unfragmented packet size in the L2TP pipe may differ from 
> > the negotiated value.
> 
> The problem is that you don't control your PPP MTU, the peer
> does, and the peer is allowed to reject any values greater
> than 1500 bytes if it wants to. So the local configuration
> only sets an upper limit, not a lower limit in effect.
> 
> > The question is now whether it makes sense to adapt the settings
> > of the MRU on the LNS and/or LAC (used as proposal for PPP 
> negotiation) 
> > to the maximum unfragmented packet size. Or are there any caveats
> > or side effects when doing this?
> 
> There are two levels at which IP fragmentation can happen, on the
> inner IP packet and on the outer IP packet. Outer packet fragments
> are a potential performance hit but otherwise not a problem. But
> if the inner IP packet must be fragmented - because you're not
> doing multi-link and your MTU for the L2TP link is too small for
> the original packet plus overhead - then your L2TP router is
> required to fragment the IP packet or send back an ICMP "must
> fragment" response if DF is set.
> 
> You could avoid "inner" IP fragmentation if your L2TP link MTU
> were large enough, but your MTU setting is at the mercy of the
> remote PPP peer, because it is allowed to reject anything larger
> than 1500 bytes (and typically the inner IP packet is up to 1500
> bytes because it came from an Ethernet port). So you can't avoid
> this possibility in general.
> 
> If you're talking about a UNIX type device, then you would normally
> configure the L2TP interface to have an MTU of 1460 (or whatever)
> so that the kernel handles the fragmentation and ICMP stuff for you.
> 
> With multi-link PPP, inner IP fragmentation is no longer necessary
> as you can fragment at the PPP frame level instead.
> 
> So in summary:
> 
> - You must be prepared to fragment and/or send ICMP "must fragment"
>    packets (don't forget to decrement TTL's too)
> - Life gets easier if you enable multi-link PPP (even with just
>    one link)
> - Life gets easier if you can get the peer to accept a large
>    enough MTU for the L2TP link, but don't count on it.
> 
> FYI, WinXP is "friendlier" than earlier versions of Windows
> on these points...
> 
> > I understand that an application using PMTU sends 
> periodically large 
> > packets with the DF bit set. However I'm not sure, whether 
> these testing
> > packets are already normal data packets or just ICMP trials BEFORE
> > actually sending data. One way or another I understand now, that
> > the problem only happens if the application has PMTU enabled and 
> > does not have a timeout mechanism as described in RFC 2923
> > 
> >    How to fix
> >       TCP should notice that the connection is timing out.  After
> >       several timeouts, TCP should attempt to send smaller packets,
> >       perhaps turning off the DF flag for each packet.  If this
> >       succeeds, it should continue to turn off PMTUD for 
> the connection
> >       for some reasonable period of time, after which it 
> should probe
> >       again to try to determine if the path has changed.
> 
> You want to avoid having the application timeout.. it's much better to
> have all routers able to send ICMP "need fragmentation" packets and
> all intermediate firewalls not block these packets.
> 
> Hope this helps.
> 
> -Archie
> 
> ______________________________________________________________
> ____________
> Archie Cobbs     *     Precision I/O      *     
http://www.precisionio.com
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext


From mailnull@www1.ietf.org  Fri Apr  4 06:59:23 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA17474
	for <l2tpext-archive@odin.ietf.org>; Fri, 4 Apr 2003 06:59:23 -0500 (EST)
Received: (from mailnull@localhost)
	by www1.ietf.org (8.11.6/8.11.6) id h34C28r08545
	for l2tpext-archive@odin.ietf.org; Fri, 4 Apr 2003 07:02:08 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h34C28K08542
	for <l2tpext-web-archive@optimus.ietf.org>; Fri, 4 Apr 2003 07:02:08 -0500
Received: from www1.ietf.org (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA17466
	for <l2tpext-web-archive@ietf.org>; Fri, 4 Apr 2003 06:58:51 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h34BqXK05938;
	Fri, 4 Apr 2003 06:52:33 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h34BomK05670
	for <l2tpext@optimus.ietf.org>; Fri, 4 Apr 2003 06:50:48 -0500
Received: from LONMSW03.uk.eu.colt (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id GAA16046
	for <l2tpext@ietf.org>; Fri, 4 Apr 2003 06:47:30 -0500 (EST)
From: Andreas.Waibel@colt.de
Message-ID: <4403667CFF70F84283D8521D50CF4CF3036E3119@franoa03.de.eu.colt>
To: l2tpext@ietf.org
Subject: AW: AW: [L2tpext] fragmentation and MRU setting in L2TPv2 scenari
	os
Date: Fri, 4 Apr 2003 11:39:42 +0100 
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2653.19)
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by www1.ietf.org id h34BomK05671
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>
Content-Transfer-Encoding: 8bit
Content-Transfer-Encoding: 8bit

Archie,

one thing puzzles me a bit in Your answer. You are
talking about negotiating a MTU value with the PPP
peer LARGER than 1500, while I am considering to propose
a MRU value (LNS) respectively MTU value (peer) for the
PPP connection SMALLER than 1500 (=equal to the maximum
unfragmented packet size in the pipe).
While I'm quite sure, that it would amend performance if
peer agrees on such a smaller value for the PPP link, I 
wonder whether it might make the PMTU problem worse for 
large packets from the server side.

best regards

Andreas

> > The question is now whether it makes sense to adapt the settings
> > of the MRU on the LNS and/or LAC (used as proposal for PPP 
> negotiation) 
> > to the maximum unfragmented packet size. Or are there any caveats
> > or side effects when doing this?
> 
> There are two levels at which IP fragmentation can happen, on the
> inner IP packet and on the outer IP packet. Outer packet fragments
> are a potential performance hit but otherwise not a problem. But
> if the inner IP packet must be fragmented - because you're not
> doing multi-link and your MTU for the L2TP link is too small for
> the original packet plus overhead - then your L2TP router is
> required to fragment the IP packet or send back an ICMP "must
> fragment" response if DF is set.
> 

> -----Ursprüngliche Nachricht-----
> Von: Archie Cobbs [mailto:archie@precisionio.com]
> Gesendet am: Mittwoch, 2. April 2003 18:34
> An: Waibel, Andreas
> Cc: l2tpext@ietf.org
> Betreff: Re: AW: [L2tpext] fragmentation and MRU setting in L2TPv2
> scenarios
> 
> Andreas.Waibel@colt.de wrote:
> >>The DF flag is set as a normal part of TCP's path-MTU
> >>discovery process:
> > 
> > sure, but I thought it should not be set in any normal
> > application data packet, once the path is discovered
> > or if PMTU was not enabled at all.
> 
> That's true.. although this falls under "what normally happens"
> rather than "what is required". E.g., there's nothing that
> says that a host can't initiate another PMTU discovery
> every 3 minutes if it wants to.
> 
> >>If peer A's MRU is < peer B's MTU, then there is a bug
> >>in somebody's PPP stack, as PPP negotiation should never
> >>allow this to happen.
> > 
> > The negotiated MRU/MTU value agreed on by both sides could 
> > still be larger than the actual maximum packet size for 
> > unfragmented transport over the wire. My interpretation of 
> > the PPP logs I have seen is that negotiation usally only depends 
> > on the (manual) settings on both sides. The actual maximum 
> > unfragmented packet size in the L2TP pipe may differ from 
> > the negotiated value.
> 
> The problem is that you don't control your PPP MTU, the peer
> does, and the peer is allowed to reject any values greater
> than 1500 bytes if it wants to. So the local configuration
> only sets an upper limit, not a lower limit in effect.
> 
> > The question is now whether it makes sense to adapt the settings
> > of the MRU on the LNS and/or LAC (used as proposal for PPP 
> negotiation) 
> > to the maximum unfragmented packet size. Or are there any caveats
> > or side effects when doing this?
> 
> There are two levels at which IP fragmentation can happen, on the
> inner IP packet and on the outer IP packet. Outer packet fragments
> are a potential performance hit but otherwise not a problem. But
> if the inner IP packet must be fragmented - because you're not
> doing multi-link and your MTU for the L2TP link is too small for
> the original packet plus overhead - then your L2TP router is
> required to fragment the IP packet or send back an ICMP "must
> fragment" response if DF is set.
> 
> You could avoid "inner" IP fragmentation if your L2TP link MTU
> were large enough, but your MTU setting is at the mercy of the
> remote PPP peer, because it is allowed to reject anything larger
> than 1500 bytes (and typically the inner IP packet is up to 1500
> bytes because it came from an Ethernet port). So you can't avoid
> this possibility in general.
> 
> If you're talking about a UNIX type device, then you would normally
> configure the L2TP interface to have an MTU of 1460 (or whatever)
> so that the kernel handles the fragmentation and ICMP stuff for you.
> 
> With multi-link PPP, inner IP fragmentation is no longer necessary
> as you can fragment at the PPP frame level instead.
> 
> So in summary:
> 
> - You must be prepared to fragment and/or send ICMP "must fragment"
>    packets (don't forget to decrement TTL's too)
> - Life gets easier if you enable multi-link PPP (even with just
>    one link)
> - Life gets easier if you can get the peer to accept a large
>    enough MTU for the L2TP link, but don't count on it.
> 
> FYI, WinXP is "friendlier" than earlier versions of Windows
> on these points...
> 
> > I understand that an application using PMTU sends 
> periodically large 
> > packets with the DF bit set. However I'm not sure, whether 
> these testing
> > packets are already normal data packets or just ICMP trials BEFORE
> > actually sending data. One way or another I understand now, that
> > the problem only happens if the application has PMTU enabled and 
> > does not have a timeout mechanism as described in RFC 2923
> > 
> >    How to fix
> >       TCP should notice that the connection is timing out.  After
> >       several timeouts, TCP should attempt to send smaller packets,
> >       perhaps turning off the DF flag for each packet.  If this
> >       succeeds, it should continue to turn off PMTUD for 
> the connection
> >       for some reasonable period of time, after which it 
> should probe
> >       again to try to determine if the path has changed.
> 
> You want to avoid having the application timeout.. it's much better to
> have all routers able to send ICMP "need fragmentation" packets and
> all intermediate firewalls not block these packets.
> 
> Hope this helps.
> 
> -Archie
> 
> ______________________________________________________________
> ____________
> Archie Cobbs     *     Precision I/O      *     
http://www.precisionio.com
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext



From l2tpext-admin@ietf.org  Sat Apr  5 18:03:18 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA16400
	for <l2tpext-archive@lists.ietf.org>; Sat, 5 Apr 2003 18:03:18 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h35N3o827098;
	Sat, 5 Apr 2003 18:03:51 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h34NAD829746
	for <l2tpext@optimus.ietf.org>; Fri, 4 Apr 2003 18:10:13 -0500
Received: from CNRI.Reston.VA.US (localhost [127.0.0.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA12046;
	Fri, 4 Apr 2003 18:06:42 -0500 (EST)
Message-Id: <200304042306.SAA12046@ietf.org>
To: IETF-Announce: ;
Cc: RFC Editor <rfc-editor@isi.edu>, Internet Architecture Board <iab@iab.org>,
        l2tpext@ietf.org
From: The IESG <iesg-secretary@ietf.org>
Date: Fri, 04 Apr 2003 18:06:42 -0500
Subject: [L2tpext] Protocol Action: Signalling of Modem-On-Hold status in Layer 2
 Tunneling Protocol (L2TP) to Proposed Standard
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>


The IESG has approved the Internet-Draft 'Signalling of Modem-On-Hold
status in Layer 2 Tunneling Protocol (L2TP)' 
<draft-ietf-l2tpext-v92-moh-05.txt> as a Proposed Standard. This 
document is the product of the Layer Two Tunneling Protocol Extensions
Working Group. The IESG contact persons are Thomas Narten and Erik
Nordmark.
   
   
Technical Summary
   
The Layer 2 Tunneling Protocol (L2TP) defines a mechanism for
tunneling Point-to-Point Protocol (PPP) sessions. It is common for
these PPP sessions to be established using modems connected over the
public switched telephone network.

One of the standards governing modem operation is the ITU-T
Recommendation V.92, which defines procedures that enable a client
modem to put the call on hold and later, re-establish the modem link
with minimal delay and without having to redial. While the modem
call is on hold, the client phone line can be used to place or
receive other calls.

The L2TP base protocol does not provide any means to signal these
events from the L2TP Access Controller (LAC), where the modem is
physically connected, to the L2TP Network Server (LNS), where the PPP
session is handled.

This document describes a method to let the LNS know when a client
modem connected to a LAC has placed the call on hold.
   
Working Group Summary
   
There was support for this extension in the WG and no issues were
raised during the Last Call.
   
Protocol Quality
   
This document has been reviewed for the IESG by Thomas Narten.
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext


From mailnull@www1.ietf.org  Sat Apr  5 18:04:22 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA16551
	for <l2tpext-archive@odin.ietf.org>; Sat, 5 Apr 2003 18:04:22 -0500 (EST)
Received: (from mailnull@localhost)
	by www1.ietf.org (8.11.6/8.11.6) id h35N7n028011
	for l2tpext-archive@odin.ietf.org; Sat, 5 Apr 2003 18:07:49 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h35N7n828008
	for <l2tpext-web-archive@optimus.ietf.org>; Sat, 5 Apr 2003 18:07:49 -0500
Received: from www1.ietf.org (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA16482
	for <l2tpext-web-archive@ietf.org>; Sat, 5 Apr 2003 18:03:51 -0500 (EST)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h35N3o827098;
	Sat, 5 Apr 2003 18:03:51 -0500
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h34NAD829746
	for <l2tpext@optimus.ietf.org>; Fri, 4 Apr 2003 18:10:13 -0500
Received: from CNRI.Reston.VA.US (localhost [127.0.0.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA12046;
	Fri, 4 Apr 2003 18:06:42 -0500 (EST)
Message-Id: <200304042306.SAA12046@ietf.org>
To: IETF-Announce: ;
Cc: RFC Editor <rfc-editor@isi.edu>, Internet Architecture Board <iab@iab.org>,
        l2tpext@ietf.org
From: The IESG <iesg-secretary@ietf.org>
Date: Fri, 04 Apr 2003 18:06:42 -0500
Subject: [L2tpext] Protocol Action: Signalling of Modem-On-Hold status in Layer 2
 Tunneling Protocol (L2TP) to Proposed Standard
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>


The IESG has approved the Internet-Draft 'Signalling of Modem-On-Hold
status in Layer 2 Tunneling Protocol (L2TP)' 
<draft-ietf-l2tpext-v92-moh-05.txt> as a Proposed Standard. This 
document is the product of the Layer Two Tunneling Protocol Extensions
Working Group. The IESG contact persons are Thomas Narten and Erik
Nordmark.
   
   
Technical Summary
   
The Layer 2 Tunneling Protocol (L2TP) defines a mechanism for
tunneling Point-to-Point Protocol (PPP) sessions. It is common for
these PPP sessions to be established using modems connected over the
public switched telephone network.

One of the standards governing modem operation is the ITU-T
Recommendation V.92, which defines procedures that enable a client
modem to put the call on hold and later, re-establish the modem link
with minimal delay and without having to redial. While the modem
call is on hold, the client phone line can be used to place or
receive other calls.

The L2TP base protocol does not provide any means to signal these
events from the L2TP Access Controller (LAC), where the modem is
physically connected, to the L2TP Network Server (LNS), where the PPP
session is handled.

This document describes a method to let the LNS know when a client
modem connected to a LAC has placed the call on hold.
   
Working Group Summary
   
There was support for this extension in the WG and no issues were
raised during the Last Call.
   
Protocol Quality
   
This document has been reviewed for the IESG by Thomas Narten.
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext



From l2tpext-admin@ietf.org  Tue Apr 22 12:05:23 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA14745
	for <l2tpext-archive@lists.ietf.org>; Tue, 22 Apr 2003 12:05:23 -0400 (EDT)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3MGDK817792;
	Tue, 22 Apr 2003 12:13:20 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3MG1d816119
	for <l2tpext@optimus.ietf.org>; Tue, 22 Apr 2003 12:01:39 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA14344
	for <l2tpext@ietf.org>; Tue, 22 Apr 2003 11:49:33 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 19803l-0002et-00
	for l2tpext@ietf.org; Tue, 22 Apr 2003 11:51:53 -0400
Received: from e33.co.us.ibm.com ([32.97.110.131])
	by ietf-mx with esmtp (Exim 4.12)
	id 19803j-0002ei-00
	for l2tpext@ietf.org; Tue, 22 Apr 2003 11:51:52 -0400
Received: from westrelay01.boulder.ibm.com (westrelay01.boulder.ibm.com [9.17.195.10])
	by e33.co.us.ibm.com (8.12.9/8.12.2) with ESMTP id h3MFpUXD088086
	for <l2tpext@ietf.org>; Tue, 22 Apr 2003 11:51:30 -0400
Received: from cichlid.adsl.duke.edu (sig-9-65-244-156.mts.ibm.com [9.65.244.156])
	by westrelay01.boulder.ibm.com (8.12.8/NCO/VER6.5) with ESMTP id h3MFpT98034076
	for <l2tpext@ietf.org>; Tue, 22 Apr 2003 09:51:29 -0600
Received: from cichlid.adsl.duke.edu (narten@localhost)
	by cichlid.adsl.duke.edu (8.11.6/8.9.3) with ESMTP id h3MFnbm02335
	for <l2tpext@ietf.org>; Tue, 22 Apr 2003 11:49:38 -0400
Message-Id: <200304221549.h3MFnbm02335@cichlid.adsl.duke.edu>
To: l2tpext@ietf.org
Date: Tue, 22 Apr 2003 11:49:37 -0400
From: Thomas Narten <narten@us.ibm.com>
Subject: [L2tpext] AD comments on draft-ietf-l2tpext-l2tp-base-07.txt
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

High level issues first, followed by editorial/detailed comments.

Running l2tp directly over IP:
  - no checksum for l2tp fields. This seems a bad idea, especially for
    the control connection.
  - not clear that IPsec has been specified enough for running l2tp
    directly over IP; this needs to get fixed before the IESG will
    advance.

The document should enable UDP checksum by default. It should require
explicit action to disable them, rather than the other way around.

Note: by staring sequence numbers at 0, the protocol is vulnerable to
problems if control connections are retarted frequently, and one has
delayed messages  in the network.

MUST use exponential backoff on retransmission (SHOULD is not
sufficient). Same for congestion control/slow start.

Killing a connection after only 5 retransmission seems aggressive. Is
this the right default value to recommend??

negotiating whether to use v2 or v3 could be better. There is text
that suggests trying v3, and if no response, falling back to v2. But
this is a bad idea because you'll get false fall backs if there are
connectivity problems. In some parts of the document, it is suggested
how to try and negotiate v3 but have it fall back to v2
transparently. Shouldn't that be the preferred method in all cases?

Does this spec have adequate support for IPv6? Seems like there is at
least one AVP that holds an IPv4 address, but not an IPv6 address.

IANA considerations could be made better my making it more clear what
IANA needs to assign for this document (the document includes existing
assignments plus some new ones for new AVPs defined in this document.)

My sense is that a new revision is appropriate before going to IETF
Last Call.

Specific comments follow.

move acknowledgments to end?

Is there enough clarity with regards to what happens to v2 (historic?
stays at proposed?)

UTF-8 issues need a bit more work I think.

The sequence number AVPs for resequencing traffic have fields that are
too small. There will be problems with sequence number wrapping, I
suspect.

> 1.  Introduction

Could be better.

> 
>    The Layer Two Tunneling Protocol (L2TP) provides a dynamic tunneling
>    mechanism for multiple Layer 2 (L2) circuits across a packet-oriented
>    data network.

Better:

    The Layer Two Tunneling Protocol (L2TP) provides a dynamic
    mechanism for tunneling Layer 2 (L2) "circuits" across a
    packet-oriented data network (e.g., over IP).

>  L2TP, as originally defined in RFC 2661, is a standard
>    method for tunneling Point to Point Protocol (PPP) [RFC1661]
>    sessions.  L2TP has since been adopted for tunneling a number of
>    other L2 protocols.  In order to provide greater modularity, this
>    document describes the base L2TP protocol, independent of the L2
>    payload that is being tunneled.
> 
>    The base L2TP protocol defined here consists of (1) the control
>    protocol for dynamic creation, maintenance, and teardown of L2TP
>    sessions, and (2) the L2TP data encapsulation to multiplex and
>    demultiplex L2 data streams between two L2TP peers.

Across IP, I would assume, so please say so.

> Additional
>    pseudowire-specific documents (e.g. for PPP, Ethernet, Frame Relay,
>    etc.) are expected to be published for each pseudowire-type supported
>    by L2TP, containing individual details outside the scope of this
>    specification. Whenever possible, these documents will utilize the
>    base constructs defined in this specification.

Better:

 Additional documents are expected to describe the details of using
 L2TP to tunnel L2 technologies other than PPP. Likewise, L2TP can be
 carried across other network types besides IP (e.g., ATM); other
 documents describe the details.


or something like that

>    Attribute Value Pair (AVP)
> 
>       The variable-length concatenation of a unique Attribute
>       (represented by an integer) and a Value containing the actual
>       value identified by the attribute.  Zero or more AVPs make up the
>       body of control messages, which are used in the establishment,
>       maintenance, and teardown of control connections.  This basic
>       construct is sometimes referred to as a Type-Length-Value (TLV) in
>       some specifications.  (See also: Control Connection, Control
>       Message.)

Why do you not mention the Length field in the definition of the AVP?
Seems odd/incomplete.

>       point to point ethernet, or an L2TP session), or it may have

point-to-point

>      two LCCEs over a PSN.  (See also: Control Connection, Data

LCCE used before defined. ditto for PSN

>      The action of receiving a call (circuit up event) on an LAC.  The

is it "an LAC" or "a LAC"? I would assume the latter, as the former
isn't the way one would actually say it. (this occurs throughout)


>    L2TP Network Server (LNS)
> 
>       An LCCE that logically terminates a tunneled circuit locally and
>       that processes the tunneled traffic as though the circuit were
>       physically connected to the device.  The LNS may tunnel to either
>       an LAC or another LNS.  (See also: LCCE, LAC.)

an LNS can tunnel to another LNS? as in tunnel switching?

>    Outgoing Call
> 
>       The action of placing a call on an LAC, typically in response to

"on an LAC" or "by an LAC"?


>    Outgoing Call Request
> 
>       A request sent to an LAC to place an outgoing call.  The request
>       contains specific information for the LAC in placing the call,
>       information that is typically not known a priori by the LAC.

Add "(e.g., the number to dial)" ??

>  (See
>       also: Call, Incoming Call, Outgoing Call.)

>      L2TP control connection (i.e. the remote LCCE).  An LAC's peer may

nit, it's i.e., (with a comma). e.g. should be handled the same
way. Fix throughout?

>    Session
> 
>       An L2TP session is created by a particular L2TP control connection
>       between two LCCEs when a circuit is successfully established.  The
>       circuit may either pass through (LAC) or terminate locally (LNS)
>       on the LCCEs, which maintain state for the circuit.  There is a
>       one-to-one relationship between established L2TP sessions and
>       their associated circuits.  (See also: Circuit, LAC, LCCE, LNS.)

the referencs to (LAC) and (LNS) is a bit cryptic and maybe not proper
english. Indeed, does it make sense to define "tunnel switching"  to
make more clear the difference between LAC and LNS as you describe
above? Or add more text to teh LAC/LNS definition to explain how you
can chain LACs/LNSs?


> 2.  Topology
> 
>    L2TP operates between two L2TP Control Connection Endpoints (LCCEs),
>    tunneling circuit traffic across a packet network.  An L2TP Network
>    Server (LNS) is an LCCE that decapsulates tunneled traffic, removes
>    layer 2 which may be present and processes the encapsulated layer 3
>    payload without passing on to another piece of equipment (i.e.
>    routing an IP packet within an HDLC frame tunneled by L2TP). An L2TP

Weird wording about "equipment". Can't you just say and processes it
using normal L3 stuff?

   tunneling circuit traffic across a packet network.  An L2TP Network
   Server (LNS) is an LCCE that decapsulates tunneled traffic, removes
   layer 2 which may be present and processes the encapsulated layer 3
   payload without passing on to another piece of equipment (i.e.
   routing an IP packet within an HDLC frame tunneled by L2TP). An L2TP
   Access Concentrator (LAC) is an LCCE that processes tunneled traffic
   based primarily on layer 2 information, either by directly switching
   session traffic to an attached circuit or another L2TP session, or by
   handing off to a virtual bridging function. A single piece of network
   equipment may serve as both an LAC and LNS.

Can't you move the above sentences to the defintion of LAC/LNS?

   (c) LNS-LNS Reference Model: This model has two LNSs as the LCCEs.  A
   user-level, traffic-generated, or signaled event typically drives
   session establishment from one side of the tunnel.

This model could be better explained/motivated. When does one use it?
how is it different than LAC/LAC? Presumably you mention it in the
document because there are features in the l2tp protocol to support this?

   be negotiated by the L2TP control connection.  Similarly, the L2TPv3
   data message format defined in this document may also be used without
   the L2TP control channel, utilizing manual configuration to
   statically "setup" each L2TP sessions. It is recommended that this be
   limited to relatively small-scale deployments, or deployments in
   which some other form of automatic control information distribution
   is employed.

Does this even need to be mentioned? By saying so, one is blessing it
as compliant with the spec. Is that necessary?

>    steps: (1) Establishing the control connection, if required, and (2)

when is establishing the control connection NOT required?

>    In general, an L2TP data message consists of a (1) Session Header,
>    (2) an optional L2-Specific Sublayer, and (3) the Tunneled L2 Frame,
>    as depicted below.
> 
>                   Figure 3.2.2: L2TP Data Message Header
> 
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>    |                      L2TP Session Header                      |
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>    |                      L2-Specific Sublayer                     |
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Picture only lists 2 of the three mentioned parts.


>    The L2 Specific Sublayer is an intermediary layer between the L2TP
>    session header and the start of the tunneled frame.  It contains
>    control fields that are used to facilitate the tunneling of each
>    frame (e.g. sequence numbers or flags).  The default L2-Specific
>    Sublayer for L2TPv3 is defined in Section 4.6.

Could be better. Is the intent here to say something like:

   The L2 Specific Sublayer is an intermediary layer between the L2TP
   session header and the start of the tunneled frame.  It contains
   control fields that are used to facilitate the reconstruction of
   the pseudo-wire stream that is begin carried in cases other than
   when PPP is carried.  The default L2-Specific Sublayer for L2TPv3
   is defined in Section 4.6.

> 4.  Protocol Operation
> 
>    This section addresses defines the control and data protocol
>    operation for L2TP.

double word

   If necessary, L2TP may operate over a variety of Packet Switched
   Networks.  The L2TP Session Header encapsulation MAY vary for a given
   PSN, though the default methods for operation over IPv4 are defined
   in this document.
   
s/though the default methods /though only the methods/

>    While there are efficiencies gained by running L2TP directly over IP,
>    there are possible side effects as well.  For instance, L2TP over IP
>    is not as NAT-friendly as L2TP over UDP.  Also, control messages
>    transmitted over IP are not protected by a network-layer checksum as
>    they are with UDP.

hmmm....

>    When negotiating a control connection over UDP, control messages must
>    be sent as UDP datagrams using the registered UDP port 1701
>    [RFC1700].  The initiator of an L2TP control connection picks an
>    available source UDP port (which may or may not be 1701), and sends
>    to the desired destination address at port 1701.  The recipient picks
>    a free port on its own system (which may or may not be 1701) and
>    sends its reply to the initiator's UDP port and address, setting its
>    own source port to the free port it found.

Very un-NAT/firewall  friendly. I know I raised this with the original
L2TP spec... Why allow the responder to change the source port? what
value does this have? (I guess I have to ask...)

>    choosing an arbitrary source port.  Any NAT device that can pass TFTP
>    traffic should be able to pass L2TP UDP traffic since both protocols
>    employ similar policies with regard to UDP port selection.

I'd drop this, as the claim is somewhat dubious...

> 4.1.2.3  UDP Checksum
> 
>    UDP checksums MUST be enabled for control messages and MAY be enabled
>    for data messages.  It should be noted that enabling checksums on
>    data packets may significantly increase the data packet processing
>    burden.

UDP checksums should be enabled by default, with clueful people
perhaps disabling them.

>    The message sequence number, Ns, begins at 0.  Each subsequent
>    message is sent with the next increment of the sequence number.  The
>    sequence number is thus a free-running counter represented modulo
>    65536.  The sequence number in the header of a received message is

Note: this algorithm is flawwed, in that the protocol becomes subject
vulnerabilities from delayed or out-of-order packets. It would be
better to let each end of the connection choose the starting sequence
number OR require the cookie be used to diambiguate old from new
packets.

   the Ns of the ZLB.  As a precaution, Nr should be sanity-checked
   before flushing the retransmit queue.  For instance, if the Nr
   received in a control message is greater than the last Ns sent plus 1
   modulo 65536, the control message is clearly invalid.

the "as a precaution" wording is weird. Isn't this required from
protocol correctness? (as opposed to being just a "precaution").


>    The reliable delivery mechanism at a receiving peer is responsible
>    for making sure that control messages are delivered in order and
>    without duplication to the upper level.  Messages arriving out of
>    order may be queued for in-order delivery when the missing messages
>    are received.  Alternatively, they may be discarded, thus requiring a
>    retransmission by the peer.  When dropping out of order control
>    packets, Nr MAY be updated before the packet is discarded.

Don't follow last sentence. You can't update Nr if you are discarding
packets... What would you update Nr to?


>    Each subsequent retransmission of a message MUST employ an
>    exponential backoff interval.  Thus, if the first retransmission
>    occurred after 1 second, the next retransmission should occur after 2
>    seconds has elapsed, then 4 seconds, etc.  An implementation MAY
>    place a cap upon the maximum interval between retransmissions.  This
>    cap MUST be no less than 8 seconds per retransmission.  If no peer
>    response is detected after several retransmissions (a recommended
>    default is 5, but SHOULD be configurable), the control connection and
>    all associated sessions MUST be cleared.

Is the default of 5 reasonable in practice? Seems pretty aggressive to
me.

>    When a control connection is being shut down for reasons other than
>    loss of connectivity, the state and reliable delivery mechanisms MUST
>    be maintained and operated for the full retransmission interval after
>    the final message exchange has occurred.

And what is that time exactly?

>    When retransmitting control messages, a slow start and congestion
>    avoidance window adjustment procedure SHOULD be utilized.  A
>    recommended procedure is described in Appendix A.

MUST?


>    In addition, a peer MUST NOT withhold acknowledgment of messages in
>    order to maintain state in the L2TP state machine.  Conversely, the
>    L2TP state machine MUST be capable of maintaining state if a ZLB ACK
>    is received in response to a control message.  However, determining
>    when a state should no longer be maintained (e.g. how long to wait in
>    wait-reply state for an ICRP from the peer) before destroying a
>    session or control connection is an issue that is left to each
>    implementation.

Leaving this implementation dependent is inconsistent with previous
text about MUST keep state around for the full retransmission
interval.

>    The sending of Hello messages and the policy for sending them are
>    left up to the implementation.  A peer MUST NOT expect Hello messages

Makes it sound like Hellos are optional. Don't you want to (just) say
that the policy for when to send them is implemenation dependent? But
that they must be implemented?

>    An L2TP implementation may first attempt to operate in L2TPv3 over IP
>    mode, then fall back to L2TPv2 (over UDP) if L2TPv3 over IP is
>    unavailable.  It does so by first sending an L2TPv3-formatted SCCRQ
>    over IP to try to initiate an L2TPv3 control connection.  If the
>    SCCRQ elicits no response, the implementation may fall back to L2TPv2
>    operation, as defined in [RFC2661].  Fallback to L2TPv2 should be
>    seamless and occur automatically.  (See Section 4.7.3 for further
>    details.)

suboptimal to have no response be used to determine lack of
support. Temporary connectivity problems can cause the wrong type of
connection to become established.

>    4.3).  Hidden values MUST NOT be unhidden until after LCCE
>    authentication has completed successfully (perhaps requiring the
>    hidden value to be stored until after receipt of additional setup
>    messages).  To do otherwise runs the risk of AVP data being utilized

What does this mean?

> 5.3  Hiding of AVP Attribute Values

verify with security ADs that this is OK.


>    The Attribute Value field for this AVP has the following format:
> 
>     0                   1
>     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>    |         Message Type          |
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

This AVP is just a message type, with no Length or value? (Picture
seems incomplete)

> Random Vector (ICRQ, ICRP, ICCN, OCRQ, OCRP, OCCN, CDN, WEN, SLI,
> StopCCN)

Is this a subsection heading? Why doesn't it then have a subsection
number? (this occurs on several...)


>    9 - Try another directed.  If an LAC or LNS is aware of other possible
>        destinations, it should inform the initiator of the control
>        connection or session.  The Error Message MUST contain a
>        comma-separated list of addresses from which the initiator may
>        choose.  If the L2TP data channel runs over IPv4, then this would
>        be a comma-separated list of IP addresses in the canonical
>        dotted-decimal format (e.g. "10.0.0.1, 10.0.0.2, 10.0.0.3") in the
>        UTF-8 charset using the Default Language [RFC2277].  If there are
>        no servers for the LAC or LNS to suggest, then Error Code 7 should
>        be used.  The delimiter between addresses MUST be precisely a
>        single comma and a single space.

Hmm. Does this spec support IPv6 at all? Shouldn't it?

> Control Connection Tie Breaker (SCCRQ)

I don't quite see the need for this. If running over IP, why not just
use the IP addresses to break ties?

> Host Name (SCCRQ, SCCRP)
> 
>    The Host Name AVP, Attribute Type 7, indicates the name of the
>    issuing LAC or LNS.
> 
>    The Attribute Value field for this AVP has the following format:
> 
>     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
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>    | Host Name ... (arbitrary number of octets)
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
> 
>    The Host Name is of arbitrary length, but MUST be at least 1 octet.
> 
>    This name should be as broadly unique as possible; for hosts
>    participating in DNS [RFC1034], a host name with fully qualified
>    domain would be appropriate.  The Host Name AVP and/or Router ID AVP
>    MUST be used to identify an LCCE as described in Section 3.3.

need to be more clear about the encoding of the DNS name. Just using
strings can lead to interoperability issues.

>    The Vendor Name is the indicated number of octets representing the
>    vendor string.  Human-readable text for this AVP MUST be provided in
>    the UTF-8 charset using the Default Language [RFC2277].

not sure this is the  right way to do this.

>   The Router ID AVP, Attribute Type TBA, is an identifier used to

clear enough to IANA?

> Router ID (SCCRQ, SCCRP)

Is 4 bytes. OK for IPv4, but for IPv6???

>    The Assigned Control Connection ID AVP, Attribute Type TBA, encodes
>    the ID being assigned to this control connection by the sender.

s/encodes/contains/?

> Pseudowire Capabilities List (SCCRQ, SCCRP)


iana name space?

> Remote End ID (ICRQ, OCRQ)

need to understand better.

> Application ID (ICRQ, OCRQ)

What is this for?

> Data Sequencing (ICRQ, ICRP, ICCN, OCRQ, OCRP, OCCN)

The sequence number is only 2 bytes. THis seems too small for lots of
uses (i.e., sequence wrap will occur very quickly..)

>    The following values are valid data sequencing levels:
> 
>       0 - No incoming data packets require sequencing.
>       1 - Only non-IP data packets require sequencing.
>       2 - All incoming data packets require sequencing.

is 2 well specified? What is PPP? Non-IP? What is IP exactly in this
context?

>    Data sequencing may only be requested when there is a L2-Specific
>    Sublayer that can provide sequence numbers present. If sequencing is

doesn't parse well.

>    The N (New) bit indicates whether the circuit status indication is
>    for a new circuit (1) or an existing circuit (0).

what is a new vs. old circuit? what is this distinction needed for?

>    accept the call.  The peer may choose not to accept the call if, for
>    instance, there are insufficient resources to handle an additional
>    session.

Section 4 could say more about how a peer refuses to complete a
call. What messages get sent?


> 7.6  Termination of a Control Connection
> 
>    The termination of a control connection consists of either peer
>    issuing a StopCCN.  The sender of this message SHOULD wait a finite
>    period of time for the acknowledgment of this message before
>    releasing the control information associated with the control
>    connection.  The recipient of this message should send an
>    acknowledgment of the message to the peer, then release the
>    associated control information.

need to say what this time is...

>    [RFC3193] defines the recommended method for securing L2TP as defined
>    in [RFC2661].  L2TP as defined in this document should possess the
>    same interface to IPsec as [RFC2661] when running on UDP/IP.  When
>    operating over IP, some default security association parameters
>    defined in [RFC3193] will have to change for IP protocol 115, vs. UDP
>    and the associated ports for L2TPv2.

details need to be specified...

> 9.1  AVP Attributes
> 
>    As defined in Section 5.1, AVPs contain Vendor ID, Attribute, and
>    Value fields.  For a Vendor ID value of 0, IANA will maintain a
>    registry of assigned Attributes and, in some cases, Values.
>    Attributes 0-39 are assigned as defined in Section 5.4.  The
>    remaining values are available for assignment upon Expert Review
>    [RFC2434].

mention that the name space was originally created via RFC
2661. (otherwise IANA may think this is a different registry).

> 9.6 Pseudowire Types
> 
>    The Pseudowire Type (PW Type) is used in the Pseudowire Type AVP and
>    Pseudowire Capabilities List AVP defined in Section 5.4.3.  0 to
>    32767 are assignable by Expert Review [RFC2434].  There are no
>    specific pseudowire types assigned within this document, however each
>    pseudowire-specific document MUST allocate its own PW types as
>    necessary according to these guidelines.

none are defined? But isn't the PW Type Required, e.g.,  in ICRQ?

Note: the iana considerations section needs to be more clear about
which values IANA needs to fill in. I'd suggest including a table for
each name space from which  values are needed, and fill them in with
"TBDx" "Section XXX" type references. That makes it easy for IANA to
fill things in. Also TBDx would have a different x for each distinct
value needed, to be sure that the document gets updated correctly when
IANA assigns the values.

>    The 24-bit field of the Default L2-Specific Sublayer contains a
>    free-running counter, including zero. Each sequenced data packet that
>    is sent must contain the sequence number, incremented by one, of the
>    previous sequenced packet sent on a given L2TP session. Upon receipt,
>    any packet with a sequence number equal to or greater than the
>    current expected packet (the last received in-order packet plus one)
>    should be considered "new" and accepted. All other packets are
>    considered "old" and discarded.

need to say something about sequence wrap here.

>    To mitigate this additional packet loss, one MUST inspect the
>    sequence numbers of packets dropped due to being classified as "old"
>    and reset the expected sequence number accordingly. This may be
>    accomplished by counting the number of "old" packets dropped that
>    were in sequence among themselves and upon reaching a threshold,
>    resetting the next expected sequence number to that seen in the
>    arriving data packets. Packet timestamps may also be used as an
>    indicator to reset the expected sequence number by detecting a period
>    of time over which "old" packets have been received in-sequence. The
>    ideal thresholds will vary depending on link speed, sequence number

This is suspect...

Thomas
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext


From mailnull@www1.ietf.org  Tue Apr 22 12:06:51 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA14800
	for <l2tpext-archive@odin.ietf.org>; Tue, 22 Apr 2003 12:06:51 -0400 (EDT)
Received: (from mailnull@localhost)
	by www1.ietf.org (8.11.6/8.11.6) id h3MGIRY18132
	for l2tpext-archive@odin.ietf.org; Tue, 22 Apr 2003 12:18:27 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3MGIR818129
	for <l2tpext-web-archive@optimus.ietf.org>; Tue, 22 Apr 2003 12:18:27 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA14791
	for <l2tpext-web-archive@ietf.org>; Tue, 22 Apr 2003 12:06:20 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 1980K0-0002mF-00
	for l2tpext-web-archive@ietf.org; Tue, 22 Apr 2003 12:08:40 -0400
Received: from ietf.org ([132.151.1.19] helo=www1.ietf.org)
	by ietf-mx with esmtp (Exim 4.12)
	id 1980K0-0002mA-00
	for l2tpext-web-archive@ietf.org; Tue, 22 Apr 2003 12:08:40 -0400
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3MGDK817792;
	Tue, 22 Apr 2003 12:13:20 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3MG1d816119
	for <l2tpext@optimus.ietf.org>; Tue, 22 Apr 2003 12:01:39 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA14344
	for <l2tpext@ietf.org>; Tue, 22 Apr 2003 11:49:33 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 19803l-0002et-00
	for l2tpext@ietf.org; Tue, 22 Apr 2003 11:51:53 -0400
Received: from e33.co.us.ibm.com ([32.97.110.131])
	by ietf-mx with esmtp (Exim 4.12)
	id 19803j-0002ei-00
	for l2tpext@ietf.org; Tue, 22 Apr 2003 11:51:52 -0400
Received: from westrelay01.boulder.ibm.com (westrelay01.boulder.ibm.com [9.17.195.10])
	by e33.co.us.ibm.com (8.12.9/8.12.2) with ESMTP id h3MFpUXD088086
	for <l2tpext@ietf.org>; Tue, 22 Apr 2003 11:51:30 -0400
Received: from cichlid.adsl.duke.edu (sig-9-65-244-156.mts.ibm.com [9.65.244.156])
	by westrelay01.boulder.ibm.com (8.12.8/NCO/VER6.5) with ESMTP id h3MFpT98034076
	for <l2tpext@ietf.org>; Tue, 22 Apr 2003 09:51:29 -0600
Received: from cichlid.adsl.duke.edu (narten@localhost)
	by cichlid.adsl.duke.edu (8.11.6/8.9.3) with ESMTP id h3MFnbm02335
	for <l2tpext@ietf.org>; Tue, 22 Apr 2003 11:49:38 -0400
Message-Id: <200304221549.h3MFnbm02335@cichlid.adsl.duke.edu>
To: l2tpext@ietf.org
Date: Tue, 22 Apr 2003 11:49:37 -0400
From: Thomas Narten <narten@us.ibm.com>
Subject: [L2tpext] AD comments on draft-ietf-l2tpext-l2tp-base-07.txt
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

High level issues first, followed by editorial/detailed comments.

Running l2tp directly over IP:
  - no checksum for l2tp fields. This seems a bad idea, especially for
    the control connection.
  - not clear that IPsec has been specified enough for running l2tp
    directly over IP; this needs to get fixed before the IESG will
    advance.

The document should enable UDP checksum by default. It should require
explicit action to disable them, rather than the other way around.

Note: by staring sequence numbers at 0, the protocol is vulnerable to
problems if control connections are retarted frequently, and one has
delayed messages  in the network.

MUST use exponential backoff on retransmission (SHOULD is not
sufficient). Same for congestion control/slow start.

Killing a connection after only 5 retransmission seems aggressive. Is
this the right default value to recommend??

negotiating whether to use v2 or v3 could be better. There is text
that suggests trying v3, and if no response, falling back to v2. But
this is a bad idea because you'll get false fall backs if there are
connectivity problems. In some parts of the document, it is suggested
how to try and negotiate v3 but have it fall back to v2
transparently. Shouldn't that be the preferred method in all cases?

Does this spec have adequate support for IPv6? Seems like there is at
least one AVP that holds an IPv4 address, but not an IPv6 address.

IANA considerations could be made better my making it more clear what
IANA needs to assign for this document (the document includes existing
assignments plus some new ones for new AVPs defined in this document.)

My sense is that a new revision is appropriate before going to IETF
Last Call.

Specific comments follow.

move acknowledgments to end?

Is there enough clarity with regards to what happens to v2 (historic?
stays at proposed?)

UTF-8 issues need a bit more work I think.

The sequence number AVPs for resequencing traffic have fields that are
too small. There will be problems with sequence number wrapping, I
suspect.

> 1.  Introduction

Could be better.

> 
>    The Layer Two Tunneling Protocol (L2TP) provides a dynamic tunneling
>    mechanism for multiple Layer 2 (L2) circuits across a packet-oriented
>    data network.

Better:

    The Layer Two Tunneling Protocol (L2TP) provides a dynamic
    mechanism for tunneling Layer 2 (L2) "circuits" across a
    packet-oriented data network (e.g., over IP).

>  L2TP, as originally defined in RFC 2661, is a standard
>    method for tunneling Point to Point Protocol (PPP) [RFC1661]
>    sessions.  L2TP has since been adopted for tunneling a number of
>    other L2 protocols.  In order to provide greater modularity, this
>    document describes the base L2TP protocol, independent of the L2
>    payload that is being tunneled.
> 
>    The base L2TP protocol defined here consists of (1) the control
>    protocol for dynamic creation, maintenance, and teardown of L2TP
>    sessions, and (2) the L2TP data encapsulation to multiplex and
>    demultiplex L2 data streams between two L2TP peers.

Across IP, I would assume, so please say so.

> Additional
>    pseudowire-specific documents (e.g. for PPP, Ethernet, Frame Relay,
>    etc.) are expected to be published for each pseudowire-type supported
>    by L2TP, containing individual details outside the scope of this
>    specification. Whenever possible, these documents will utilize the
>    base constructs defined in this specification.

Better:

 Additional documents are expected to describe the details of using
 L2TP to tunnel L2 technologies other than PPP. Likewise, L2TP can be
 carried across other network types besides IP (e.g., ATM); other
 documents describe the details.


or something like that

>    Attribute Value Pair (AVP)
> 
>       The variable-length concatenation of a unique Attribute
>       (represented by an integer) and a Value containing the actual
>       value identified by the attribute.  Zero or more AVPs make up the
>       body of control messages, which are used in the establishment,
>       maintenance, and teardown of control connections.  This basic
>       construct is sometimes referred to as a Type-Length-Value (TLV) in
>       some specifications.  (See also: Control Connection, Control
>       Message.)

Why do you not mention the Length field in the definition of the AVP?
Seems odd/incomplete.

>       point to point ethernet, or an L2TP session), or it may have

point-to-point

>      two LCCEs over a PSN.  (See also: Control Connection, Data

LCCE used before defined. ditto for PSN

>      The action of receiving a call (circuit up event) on an LAC.  The

is it "an LAC" or "a LAC"? I would assume the latter, as the former
isn't the way one would actually say it. (this occurs throughout)


>    L2TP Network Server (LNS)
> 
>       An LCCE that logically terminates a tunneled circuit locally and
>       that processes the tunneled traffic as though the circuit were
>       physically connected to the device.  The LNS may tunnel to either
>       an LAC or another LNS.  (See also: LCCE, LAC.)

an LNS can tunnel to another LNS? as in tunnel switching?

>    Outgoing Call
> 
>       The action of placing a call on an LAC, typically in response to

"on an LAC" or "by an LAC"?


>    Outgoing Call Request
> 
>       A request sent to an LAC to place an outgoing call.  The request
>       contains specific information for the LAC in placing the call,
>       information that is typically not known a priori by the LAC.

Add "(e.g., the number to dial)" ??

>  (See
>       also: Call, Incoming Call, Outgoing Call.)

>      L2TP control connection (i.e. the remote LCCE).  An LAC's peer may

nit, it's i.e., (with a comma). e.g. should be handled the same
way. Fix throughout?

>    Session
> 
>       An L2TP session is created by a particular L2TP control connection
>       between two LCCEs when a circuit is successfully established.  The
>       circuit may either pass through (LAC) or terminate locally (LNS)
>       on the LCCEs, which maintain state for the circuit.  There is a
>       one-to-one relationship between established L2TP sessions and
>       their associated circuits.  (See also: Circuit, LAC, LCCE, LNS.)

the referencs to (LAC) and (LNS) is a bit cryptic and maybe not proper
english. Indeed, does it make sense to define "tunnel switching"  to
make more clear the difference between LAC and LNS as you describe
above? Or add more text to teh LAC/LNS definition to explain how you
can chain LACs/LNSs?


> 2.  Topology
> 
>    L2TP operates between two L2TP Control Connection Endpoints (LCCEs),
>    tunneling circuit traffic across a packet network.  An L2TP Network
>    Server (LNS) is an LCCE that decapsulates tunneled traffic, removes
>    layer 2 which may be present and processes the encapsulated layer 3
>    payload without passing on to another piece of equipment (i.e.
>    routing an IP packet within an HDLC frame tunneled by L2TP). An L2TP

Weird wording about "equipment". Can't you just say and processes it
using normal L3 stuff?

   tunneling circuit traffic across a packet network.  An L2TP Network
   Server (LNS) is an LCCE that decapsulates tunneled traffic, removes
   layer 2 which may be present and processes the encapsulated layer 3
   payload without passing on to another piece of equipment (i.e.
   routing an IP packet within an HDLC frame tunneled by L2TP). An L2TP
   Access Concentrator (LAC) is an LCCE that processes tunneled traffic
   based primarily on layer 2 information, either by directly switching
   session traffic to an attached circuit or another L2TP session, or by
   handing off to a virtual bridging function. A single piece of network
   equipment may serve as both an LAC and LNS.

Can't you move the above sentences to the defintion of LAC/LNS?

   (c) LNS-LNS Reference Model: This model has two LNSs as the LCCEs.  A
   user-level, traffic-generated, or signaled event typically drives
   session establishment from one side of the tunnel.

This model could be better explained/motivated. When does one use it?
how is it different than LAC/LAC? Presumably you mention it in the
document because there are features in the l2tp protocol to support this?

   be negotiated by the L2TP control connection.  Similarly, the L2TPv3
   data message format defined in this document may also be used without
   the L2TP control channel, utilizing manual configuration to
   statically "setup" each L2TP sessions. It is recommended that this be
   limited to relatively small-scale deployments, or deployments in
   which some other form of automatic control information distribution
   is employed.

Does this even need to be mentioned? By saying so, one is blessing it
as compliant with the spec. Is that necessary?

>    steps: (1) Establishing the control connection, if required, and (2)

when is establishing the control connection NOT required?

>    In general, an L2TP data message consists of a (1) Session Header,
>    (2) an optional L2-Specific Sublayer, and (3) the Tunneled L2 Frame,
>    as depicted below.
> 
>                   Figure 3.2.2: L2TP Data Message Header
> 
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>    |                      L2TP Session Header                      |
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>    |                      L2-Specific Sublayer                     |
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Picture only lists 2 of the three mentioned parts.


>    The L2 Specific Sublayer is an intermediary layer between the L2TP
>    session header and the start of the tunneled frame.  It contains
>    control fields that are used to facilitate the tunneling of each
>    frame (e.g. sequence numbers or flags).  The default L2-Specific
>    Sublayer for L2TPv3 is defined in Section 4.6.

Could be better. Is the intent here to say something like:

   The L2 Specific Sublayer is an intermediary layer between the L2TP
   session header and the start of the tunneled frame.  It contains
   control fields that are used to facilitate the reconstruction of
   the pseudo-wire stream that is begin carried in cases other than
   when PPP is carried.  The default L2-Specific Sublayer for L2TPv3
   is defined in Section 4.6.

> 4.  Protocol Operation
> 
>    This section addresses defines the control and data protocol
>    operation for L2TP.

double word

   If necessary, L2TP may operate over a variety of Packet Switched
   Networks.  The L2TP Session Header encapsulation MAY vary for a given
   PSN, though the default methods for operation over IPv4 are defined
   in this document.
   
s/though the default methods /though only the methods/

>    While there are efficiencies gained by running L2TP directly over IP,
>    there are possible side effects as well.  For instance, L2TP over IP
>    is not as NAT-friendly as L2TP over UDP.  Also, control messages
>    transmitted over IP are not protected by a network-layer checksum as
>    they are with UDP.

hmmm....

>    When negotiating a control connection over UDP, control messages must
>    be sent as UDP datagrams using the registered UDP port 1701
>    [RFC1700].  The initiator of an L2TP control connection picks an
>    available source UDP port (which may or may not be 1701), and sends
>    to the desired destination address at port 1701.  The recipient picks
>    a free port on its own system (which may or may not be 1701) and
>    sends its reply to the initiator's UDP port and address, setting its
>    own source port to the free port it found.

Very un-NAT/firewall  friendly. I know I raised this with the original
L2TP spec... Why allow the responder to change the source port? what
value does this have? (I guess I have to ask...)

>    choosing an arbitrary source port.  Any NAT device that can pass TFTP
>    traffic should be able to pass L2TP UDP traffic since both protocols
>    employ similar policies with regard to UDP port selection.

I'd drop this, as the claim is somewhat dubious...

> 4.1.2.3  UDP Checksum
> 
>    UDP checksums MUST be enabled for control messages and MAY be enabled
>    for data messages.  It should be noted that enabling checksums on
>    data packets may significantly increase the data packet processing
>    burden.

UDP checksums should be enabled by default, with clueful people
perhaps disabling them.

>    The message sequence number, Ns, begins at 0.  Each subsequent
>    message is sent with the next increment of the sequence number.  The
>    sequence number is thus a free-running counter represented modulo
>    65536.  The sequence number in the header of a received message is

Note: this algorithm is flawwed, in that the protocol becomes subject
vulnerabilities from delayed or out-of-order packets. It would be
better to let each end of the connection choose the starting sequence
number OR require the cookie be used to diambiguate old from new
packets.

   the Ns of the ZLB.  As a precaution, Nr should be sanity-checked
   before flushing the retransmit queue.  For instance, if the Nr
   received in a control message is greater than the last Ns sent plus 1
   modulo 65536, the control message is clearly invalid.

the "as a precaution" wording is weird. Isn't this required from
protocol correctness? (as opposed to being just a "precaution").


>    The reliable delivery mechanism at a receiving peer is responsible
>    for making sure that control messages are delivered in order and
>    without duplication to the upper level.  Messages arriving out of
>    order may be queued for in-order delivery when the missing messages
>    are received.  Alternatively, they may be discarded, thus requiring a
>    retransmission by the peer.  When dropping out of order control
>    packets, Nr MAY be updated before the packet is discarded.

Don't follow last sentence. You can't update Nr if you are discarding
packets... What would you update Nr to?


>    Each subsequent retransmission of a message MUST employ an
>    exponential backoff interval.  Thus, if the first retransmission
>    occurred after 1 second, the next retransmission should occur after 2
>    seconds has elapsed, then 4 seconds, etc.  An implementation MAY
>    place a cap upon the maximum interval between retransmissions.  This
>    cap MUST be no less than 8 seconds per retransmission.  If no peer
>    response is detected after several retransmissions (a recommended
>    default is 5, but SHOULD be configurable), the control connection and
>    all associated sessions MUST be cleared.

Is the default of 5 reasonable in practice? Seems pretty aggressive to
me.

>    When a control connection is being shut down for reasons other than
>    loss of connectivity, the state and reliable delivery mechanisms MUST
>    be maintained and operated for the full retransmission interval after
>    the final message exchange has occurred.

And what is that time exactly?

>    When retransmitting control messages, a slow start and congestion
>    avoidance window adjustment procedure SHOULD be utilized.  A
>    recommended procedure is described in Appendix A.

MUST?


>    In addition, a peer MUST NOT withhold acknowledgment of messages in
>    order to maintain state in the L2TP state machine.  Conversely, the
>    L2TP state machine MUST be capable of maintaining state if a ZLB ACK
>    is received in response to a control message.  However, determining
>    when a state should no longer be maintained (e.g. how long to wait in
>    wait-reply state for an ICRP from the peer) before destroying a
>    session or control connection is an issue that is left to each
>    implementation.

Leaving this implementation dependent is inconsistent with previous
text about MUST keep state around for the full retransmission
interval.

>    The sending of Hello messages and the policy for sending them are
>    left up to the implementation.  A peer MUST NOT expect Hello messages

Makes it sound like Hellos are optional. Don't you want to (just) say
that the policy for when to send them is implemenation dependent? But
that they must be implemented?

>    An L2TP implementation may first attempt to operate in L2TPv3 over IP
>    mode, then fall back to L2TPv2 (over UDP) if L2TPv3 over IP is
>    unavailable.  It does so by first sending an L2TPv3-formatted SCCRQ
>    over IP to try to initiate an L2TPv3 control connection.  If the
>    SCCRQ elicits no response, the implementation may fall back to L2TPv2
>    operation, as defined in [RFC2661].  Fallback to L2TPv2 should be
>    seamless and occur automatically.  (See Section 4.7.3 for further
>    details.)

suboptimal to have no response be used to determine lack of
support. Temporary connectivity problems can cause the wrong type of
connection to become established.

>    4.3).  Hidden values MUST NOT be unhidden until after LCCE
>    authentication has completed successfully (perhaps requiring the
>    hidden value to be stored until after receipt of additional setup
>    messages).  To do otherwise runs the risk of AVP data being utilized

What does this mean?

> 5.3  Hiding of AVP Attribute Values

verify with security ADs that this is OK.


>    The Attribute Value field for this AVP has the following format:
> 
>     0                   1
>     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>    |         Message Type          |
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

This AVP is just a message type, with no Length or value? (Picture
seems incomplete)

> Random Vector (ICRQ, ICRP, ICCN, OCRQ, OCRP, OCCN, CDN, WEN, SLI,
> StopCCN)

Is this a subsection heading? Why doesn't it then have a subsection
number? (this occurs on several...)


>    9 - Try another directed.  If an LAC or LNS is aware of other possible
>        destinations, it should inform the initiator of the control
>        connection or session.  The Error Message MUST contain a
>        comma-separated list of addresses from which the initiator may
>        choose.  If the L2TP data channel runs over IPv4, then this would
>        be a comma-separated list of IP addresses in the canonical
>        dotted-decimal format (e.g. "10.0.0.1, 10.0.0.2, 10.0.0.3") in the
>        UTF-8 charset using the Default Language [RFC2277].  If there are
>        no servers for the LAC or LNS to suggest, then Error Code 7 should
>        be used.  The delimiter between addresses MUST be precisely a
>        single comma and a single space.

Hmm. Does this spec support IPv6 at all? Shouldn't it?

> Control Connection Tie Breaker (SCCRQ)

I don't quite see the need for this. If running over IP, why not just
use the IP addresses to break ties?

> Host Name (SCCRQ, SCCRP)
> 
>    The Host Name AVP, Attribute Type 7, indicates the name of the
>    issuing LAC or LNS.
> 
>    The Attribute Value field for this AVP has the following format:
> 
>     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
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
>    | Host Name ... (arbitrary number of octets)
>    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
> 
>    The Host Name is of arbitrary length, but MUST be at least 1 octet.
> 
>    This name should be as broadly unique as possible; for hosts
>    participating in DNS [RFC1034], a host name with fully qualified
>    domain would be appropriate.  The Host Name AVP and/or Router ID AVP
>    MUST be used to identify an LCCE as described in Section 3.3.

need to be more clear about the encoding of the DNS name. Just using
strings can lead to interoperability issues.

>    The Vendor Name is the indicated number of octets representing the
>    vendor string.  Human-readable text for this AVP MUST be provided in
>    the UTF-8 charset using the Default Language [RFC2277].

not sure this is the  right way to do this.

>   The Router ID AVP, Attribute Type TBA, is an identifier used to

clear enough to IANA?

> Router ID (SCCRQ, SCCRP)

Is 4 bytes. OK for IPv4, but for IPv6???

>    The Assigned Control Connection ID AVP, Attribute Type TBA, encodes
>    the ID being assigned to this control connection by the sender.

s/encodes/contains/?

> Pseudowire Capabilities List (SCCRQ, SCCRP)


iana name space?

> Remote End ID (ICRQ, OCRQ)

need to understand better.

> Application ID (ICRQ, OCRQ)

What is this for?

> Data Sequencing (ICRQ, ICRP, ICCN, OCRQ, OCRP, OCCN)

The sequence number is only 2 bytes. THis seems too small for lots of
uses (i.e., sequence wrap will occur very quickly..)

>    The following values are valid data sequencing levels:
> 
>       0 - No incoming data packets require sequencing.
>       1 - Only non-IP data packets require sequencing.
>       2 - All incoming data packets require sequencing.

is 2 well specified? What is PPP? Non-IP? What is IP exactly in this
context?

>    Data sequencing may only be requested when there is a L2-Specific
>    Sublayer that can provide sequence numbers present. If sequencing is

doesn't parse well.

>    The N (New) bit indicates whether the circuit status indication is
>    for a new circuit (1) or an existing circuit (0).

what is a new vs. old circuit? what is this distinction needed for?

>    accept the call.  The peer may choose not to accept the call if, for
>    instance, there are insufficient resources to handle an additional
>    session.

Section 4 could say more about how a peer refuses to complete a
call. What messages get sent?


> 7.6  Termination of a Control Connection
> 
>    The termination of a control connection consists of either peer
>    issuing a StopCCN.  The sender of this message SHOULD wait a finite
>    period of time for the acknowledgment of this message before
>    releasing the control information associated with the control
>    connection.  The recipient of this message should send an
>    acknowledgment of the message to the peer, then release the
>    associated control information.

need to say what this time is...

>    [RFC3193] defines the recommended method for securing L2TP as defined
>    in [RFC2661].  L2TP as defined in this document should possess the
>    same interface to IPsec as [RFC2661] when running on UDP/IP.  When
>    operating over IP, some default security association parameters
>    defined in [RFC3193] will have to change for IP protocol 115, vs. UDP
>    and the associated ports for L2TPv2.

details need to be specified...

> 9.1  AVP Attributes
> 
>    As defined in Section 5.1, AVPs contain Vendor ID, Attribute, and
>    Value fields.  For a Vendor ID value of 0, IANA will maintain a
>    registry of assigned Attributes and, in some cases, Values.
>    Attributes 0-39 are assigned as defined in Section 5.4.  The
>    remaining values are available for assignment upon Expert Review
>    [RFC2434].

mention that the name space was originally created via RFC
2661. (otherwise IANA may think this is a different registry).

> 9.6 Pseudowire Types
> 
>    The Pseudowire Type (PW Type) is used in the Pseudowire Type AVP and
>    Pseudowire Capabilities List AVP defined in Section 5.4.3.  0 to
>    32767 are assignable by Expert Review [RFC2434].  There are no
>    specific pseudowire types assigned within this document, however each
>    pseudowire-specific document MUST allocate its own PW types as
>    necessary according to these guidelines.

none are defined? But isn't the PW Type Required, e.g.,  in ICRQ?

Note: the iana considerations section needs to be more clear about
which values IANA needs to fill in. I'd suggest including a table for
each name space from which  values are needed, and fill them in with
"TBDx" "Section XXX" type references. That makes it easy for IANA to
fill things in. Also TBDx would have a different x for each distinct
value needed, to be sure that the document gets updated correctly when
IANA assigns the values.

>    The 24-bit field of the Default L2-Specific Sublayer contains a
>    free-running counter, including zero. Each sequenced data packet that
>    is sent must contain the sequence number, incremented by one, of the
>    previous sequenced packet sent on a given L2TP session. Upon receipt,
>    any packet with a sequence number equal to or greater than the
>    current expected packet (the last received in-order packet plus one)
>    should be considered "new" and accepted. All other packets are
>    considered "old" and discarded.

need to say something about sequence wrap here.

>    To mitigate this additional packet loss, one MUST inspect the
>    sequence numbers of packets dropped due to being classified as "old"
>    and reset the expected sequence number accordingly. This may be
>    accomplished by counting the number of "old" packets dropped that
>    were in sequence among themselves and upon reaching a threshold,
>    resetting the next expected sequence number to that seen in the
>    arriving data packets. Packet timestamps may also be used as an
>    indicator to reset the expected sequence number by detecting a period
>    of time over which "old" packets have been received in-sequence. The
>    ideal thresholds will vary depending on link speed, sequence number

This is suspect...

Thomas
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext



From l2tpext-admin@ietf.org  Tue Apr 22 12:49:30 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA15949
	for <l2tpext-archive@lists.ietf.org>; Tue, 22 Apr 2003 12:49:30 -0400 (EDT)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3MGxw820899;
	Tue, 22 Apr 2003 12:59:58 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3MGtM820750
	for <l2tpext@optimus.ietf.org>; Tue, 22 Apr 2003 12:55:22 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA15830
	for <L2tpext@ietf.org>; Tue, 22 Apr 2003 12:43:14 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 1980tj-00031e-00
	for L2tpext@ietf.org; Tue, 22 Apr 2003 12:45:35 -0400
Received: from web41301.mail.yahoo.com ([66.218.93.186])
	by ietf-mx with smtp (Exim 4.12)
	id 1980ti-00031R-00
	for L2tpext@ietf.org; Tue, 22 Apr 2003 12:45:34 -0400
Message-ID: <20030422164525.95606.qmail@web41301.mail.yahoo.com>
Received: from [64.162.95.130] by web41301.mail.yahoo.com via HTTP; Tue, 22 Apr 2003 09:45:25 PDT
Date: Tue, 22 Apr 2003 09:45:25 -0700 (PDT)
From: Vipin Jain <vipinietf@yahoo.com>
To: kchen <kchen@seu.edu.cn>, rpenno@nortelnetworks.com, lll@tahoenetworks.com
Cc: l2tp ietflist <L2tpext@ietf.org>
In-Reply-To: <001601c3087b$b54eb620$c90b77ca@kchen>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Subject: [L2tpext] Re: please help me!!!!!
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

kchen,

I'd recommend you send emails to the mailing list as well.
The new version of ts draft is going to add/clarify following:
- what to regenerate a new value with
- expand more on 'congestion control'
- make the terminology L2TPv2 based
- some editorial changes
- support stack AVPs

As regarding your question: The regeneration need not mean
that the value can't ultimately assume the same value as what
came in. So in case of SCCRQ, yes it would. But REGENERATION,
in this case, emphasizes the fact that TSA explicitly acknowledges
what goes in this AVP.

RELAY means 'no modification'; just pass it on.. There is no IP
address in every AVP, so what IP address to change when relaying
the AVP. The control messages destined to the peer, would of course,
have the correct destination IP, but that doesn't constitute the
definition of RELAY an AVP. So, I am not sure if I answered your
question.

thanks,
-- vipin

--- kchen <kchen@seu.edu.cn> wrote:
> hi all:
>     I read the
>
"http://www.ietf.org/internet-drafts/draft-ietf-l2tpext-tunnel-switching-03.txt"written
> by you.I have some questions to ask you.First,you say there are 3 behaviors
> for AVPs in TSA,but I dont know how to regenerated the AVPs.For example,if
> the avp is Message Type such as SCCRQ,why must it be regenerated?Regerate to
> what?Second I think RELAY is to modify the dest IP,is it true?If you can help
> me,I am very appreciate of you.Thank you very much!!!
>                     a chinese student 
> 


__________________________________________________
Do you Yahoo!?
The New Yahoo! Search - Faster. Easier. Bingo
http://search.yahoo.com
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext


From mailnull@www1.ietf.org  Tue Apr 22 12:50:27 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA15975
	for <l2tpext-archive@odin.ietf.org>; Tue, 22 Apr 2003 12:50:27 -0400 (EDT)
Received: (from mailnull@localhost)
	by www1.ietf.org (8.11.6/8.11.6) id h3MH24F21060
	for l2tpext-archive@odin.ietf.org; Tue, 22 Apr 2003 13:02:04 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3MH24821057
	for <l2tpext-web-archive@optimus.ietf.org>; Tue, 22 Apr 2003 13:02:04 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA15964
	for <l2tpext-web-archive@ietf.org>; Tue, 22 Apr 2003 12:49:56 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 19810D-000335-00
	for l2tpext-web-archive@ietf.org; Tue, 22 Apr 2003 12:52:17 -0400
Received: from ietf.org ([132.151.1.19] helo=www1.ietf.org)
	by ietf-mx with esmtp (Exim 4.12)
	id 19810C-000332-00
	for l2tpext-web-archive@ietf.org; Tue, 22 Apr 2003 12:52:16 -0400
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3MGxw820899;
	Tue, 22 Apr 2003 12:59:58 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3MGtM820750
	for <l2tpext@optimus.ietf.org>; Tue, 22 Apr 2003 12:55:22 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA15830
	for <L2tpext@ietf.org>; Tue, 22 Apr 2003 12:43:14 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 1980tj-00031e-00
	for L2tpext@ietf.org; Tue, 22 Apr 2003 12:45:35 -0400
Received: from web41301.mail.yahoo.com ([66.218.93.186])
	by ietf-mx with smtp (Exim 4.12)
	id 1980ti-00031R-00
	for L2tpext@ietf.org; Tue, 22 Apr 2003 12:45:34 -0400
Message-ID: <20030422164525.95606.qmail@web41301.mail.yahoo.com>
Received: from [64.162.95.130] by web41301.mail.yahoo.com via HTTP; Tue, 22 Apr 2003 09:45:25 PDT
Date: Tue, 22 Apr 2003 09:45:25 -0700 (PDT)
From: Vipin Jain <vipinietf@yahoo.com>
To: kchen <kchen@seu.edu.cn>, rpenno@nortelnetworks.com, lll@tahoenetworks.com
Cc: l2tp ietflist <L2tpext@ietf.org>
In-Reply-To: <001601c3087b$b54eb620$c90b77ca@kchen>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Subject: [L2tpext] Re: please help me!!!!!
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

kchen,

I'd recommend you send emails to the mailing list as well.
The new version of ts draft is going to add/clarify following:
- what to regenerate a new value with
- expand more on 'congestion control'
- make the terminology L2TPv2 based
- some editorial changes
- support stack AVPs

As regarding your question: The regeneration need not mean
that the value can't ultimately assume the same value as what
came in. So in case of SCCRQ, yes it would. But REGENERATION,
in this case, emphasizes the fact that TSA explicitly acknowledges
what goes in this AVP.

RELAY means 'no modification'; just pass it on.. There is no IP
address in every AVP, so what IP address to change when relaying
the AVP. The control messages destined to the peer, would of course,
have the correct destination IP, but that doesn't constitute the
definition of RELAY an AVP. So, I am not sure if I answered your
question.

thanks,
-- vipin

--- kchen <kchen@seu.edu.cn> wrote:
> hi all:
>     I read the
>
"http://www.ietf.org/internet-drafts/draft-ietf-l2tpext-tunnel-switching-03.txt"written
> by you.I have some questions to ask you.First,you say there are 3 behaviors
> for AVPs in TSA,but I dont know how to regenerated the AVPs.For example,if
> the avp is Message Type such as SCCRQ,why must it be regenerated?Regerate to
> what?Second I think RELAY is to modify the dest IP,is it true?If you can help
> me,I am very appreciate of you.Thank you very much!!!
>                     a chinese student 
> 


__________________________________________________
Do you Yahoo!?
The New Yahoo! Search - Faster. Easier. Bingo
http://search.yahoo.com
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext



From l2tpext-admin@ietf.org  Tue Apr 22 22:16:51 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA04398
	for <l2tpext-archive@lists.ietf.org>; Tue, 22 Apr 2003 22:16:51 -0400 (EDT)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3N2Rf831556;
	Tue, 22 Apr 2003 22:27:41 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3N2O3831434
	for <l2tpext@optimus.ietf.org>; Tue, 22 Apr 2003 22:24:03 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA04353
	for <L2tpext@ietf.org>; Tue, 22 Apr 2003 22:11:43 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 1989lt-00072T-00
	for L2tpext@ietf.org; Tue, 22 Apr 2003 22:14:05 -0400
Received: from seic33.seu.edu.cn ([202.119.24.43] helo=seu.edu.cn)
	by ietf-mx with smtp (Exim 4.12)
	id 1989ls-00072Q-00
	for L2tpext@ietf.org; Tue, 22 Apr 2003 22:14:04 -0400
Received: (qmail 23486 invoked from network); 23 Apr 2003 02:13:55 -0000
Received: from unknown (HELO kchen) (202.119.11.201)
  by seic33.seu.edu.cn with SMTP; 23 Apr 2003 02:13:55 -0000
Message-ID: <002701c3093e$57ebf650$c90b77ca@kchen>
From: "kchen" <kchen@seu.edu.cn>
To: <L2tpext@ietf.org>
Date: Wed, 23 Apr 2003 10:16:26 +0800
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_NextPart_000_0024_01C30981.660A7B60"
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.50.4807.1700
X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4910.0300
Subject: [L2tpext] how can tunnel switch implement in IPSec?
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

This is a multi-part message in MIME format.

------=_NextPart_000_0024_01C30981.660A7B60
Content-Type: text/plain;
	charset="gb2312"
Content-Transfer-Encoding: base64

Q2FuIHNvbWVvbmUgdGVsbCBtZT8NCg==

------=_NextPart_000_0024_01C30981.660A7B60
Content-Type: text/html;
	charset="gb2312"
Content-Transfer-Encoding: base64

PCFET0NUWVBFIEhUTUwgUFVCTElDICItLy9XM0MvL0RURCBIVE1MIDQuMCBUcmFuc2l0aW9uYWwv
L0VOIj4NCjxIVE1MPjxIRUFEPg0KPE1FVEEgY29udGVudD0idGV4dC9odG1sOyBjaGFyc2V0PWdi
MjMxMiIgaHR0cC1lcXVpdj1Db250ZW50LVR5cGU+DQo8TUVUQSBjb250ZW50PSJNU0hUTUwgNS4w
MC4zNTAyLjUzOTAiIG5hbWU9R0VORVJBVE9SPg0KPFNUWUxFPjwvU1RZTEU+DQo8L0hFQUQ+DQo8
Qk9EWSBiZ0NvbG9yPSNmZmZmZmY+DQo8RElWPjxGT05UIHNpemU9Mj4NCjxESVY+PEZPTlQgc2l6
ZT0yPkNhbiBzb21lb25lIHRlbGwgbWU/PC9GT05UPjwvRElWPjwvRk9OVD48L0RJVj48L0JPRFk+
PC9IVE1MPg0K

------=_NextPart_000_0024_01C30981.660A7B60--


_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext


From mailnull@www1.ietf.org  Tue Apr 22 22:17:32 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA04423
	for <l2tpext-archive@odin.ietf.org>; Tue, 22 Apr 2003 22:17:32 -0400 (EDT)
Received: (from mailnull@localhost)
	by www1.ietf.org (8.11.6/8.11.6) id h3N2TMb31631
	for l2tpext-archive@odin.ietf.org; Tue, 22 Apr 2003 22:29:22 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3N2TM831628
	for <l2tpext-web-archive@optimus.ietf.org>; Tue, 22 Apr 2003 22:29:22 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA04415
	for <l2tpext-web-archive@ietf.org>; Tue, 22 Apr 2003 22:17:02 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 1989r1-00073c-00
	for l2tpext-web-archive@ietf.org; Tue, 22 Apr 2003 22:19:23 -0400
Received: from ietf.org ([132.151.1.19] helo=www1.ietf.org)
	by ietf-mx with esmtp (Exim 4.12)
	id 1989r0-00073Y-00
	for l2tpext-web-archive@ietf.org; Tue, 22 Apr 2003 22:19:22 -0400
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3N2Rf831556;
	Tue, 22 Apr 2003 22:27:41 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3N2O3831434
	for <l2tpext@optimus.ietf.org>; Tue, 22 Apr 2003 22:24:03 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA04353
	for <L2tpext@ietf.org>; Tue, 22 Apr 2003 22:11:43 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 1989lt-00072T-00
	for L2tpext@ietf.org; Tue, 22 Apr 2003 22:14:05 -0400
Received: from seic33.seu.edu.cn ([202.119.24.43] helo=seu.edu.cn)
	by ietf-mx with smtp (Exim 4.12)
	id 1989ls-00072Q-00
	for L2tpext@ietf.org; Tue, 22 Apr 2003 22:14:04 -0400
Received: (qmail 23486 invoked from network); 23 Apr 2003 02:13:55 -0000
Received: from unknown (HELO kchen) (202.119.11.201)
  by seic33.seu.edu.cn with SMTP; 23 Apr 2003 02:13:55 -0000
Message-ID: <002701c3093e$57ebf650$c90b77ca@kchen>
From: "kchen" <kchen@seu.edu.cn>
To: <L2tpext@ietf.org>
Date: Wed, 23 Apr 2003 10:16:26 +0800
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_NextPart_000_0024_01C30981.660A7B60"
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.50.4807.1700
X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4910.0300
Subject: [L2tpext] how can tunnel switch implement in IPSec?
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

This is a multi-part message in MIME format.

------=_NextPart_000_0024_01C30981.660A7B60
Content-Type: text/plain;
	charset="gb2312"
Content-Transfer-Encoding: base64

Q2FuIHNvbWVvbmUgdGVsbCBtZT8NCg==

------=_NextPart_000_0024_01C30981.660A7B60
Content-Type: text/html;
	charset="gb2312"
Content-Transfer-Encoding: base64

PCFET0NUWVBFIEhUTUwgUFVCTElDICItLy9XM0MvL0RURCBIVE1MIDQuMCBUcmFuc2l0aW9uYWwv
L0VOIj4NCjxIVE1MPjxIRUFEPg0KPE1FVEEgY29udGVudD0idGV4dC9odG1sOyBjaGFyc2V0PWdi
MjMxMiIgaHR0cC1lcXVpdj1Db250ZW50LVR5cGU+DQo8TUVUQSBjb250ZW50PSJNU0hUTUwgNS4w
MC4zNTAyLjUzOTAiIG5hbWU9R0VORVJBVE9SPg0KPFNUWUxFPjwvU1RZTEU+DQo8L0hFQUQ+DQo8
Qk9EWSBiZ0NvbG9yPSNmZmZmZmY+DQo8RElWPjxGT05UIHNpemU9Mj4NCjxESVY+PEZPTlQgc2l6
ZT0yPkNhbiBzb21lb25lIHRlbGwgbWU/PC9GT05UPjwvRElWPjwvRk9OVD48L0RJVj48L0JPRFk+
PC9IVE1MPg0K

------=_NextPart_000_0024_01C30981.660A7B60--


_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext



From l2tpext-admin@ietf.org  Wed Apr 23 02:23:59 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id CAA29061
	for <l2tpext-archive@lists.ietf.org>; Wed, 23 Apr 2003 02:23:59 -0400 (EDT)
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3N6Z8826323;
	Wed, 23 Apr 2003 02:35:08 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3N6W6826196
	for <l2tpext@optimus.ietf.org>; Wed, 23 Apr 2003 02:32:06 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id CAA24381
	for <L2tpext@ietf.org>; Wed, 23 Apr 2003 02:19:40 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 198Ddo-0000KJ-00
	for L2tpext@ietf.org; Wed, 23 Apr 2003 02:22:00 -0400
Received: from web41312.mail.yahoo.com ([66.218.93.61])
	by ietf-mx with smtp (Exim 4.12)
	id 198Ddo-0000KG-00
	for L2tpext@ietf.org; Wed, 23 Apr 2003 02:22:00 -0400
Message-ID: <20030423062154.70404.qmail@web41312.mail.yahoo.com>
Received: from [64.162.95.130] by web41312.mail.yahoo.com via HTTP; Tue, 22 Apr 2003 23:21:54 PDT
Date: Tue, 22 Apr 2003 23:21:54 -0700 (PDT)
From: Vipin Jain <vipinietf@yahoo.com>
Subject: Re: [L2tpext] how can tunnel switch implement in IPSec?
To: kchen <kchen@seu.edu.cn>, L2tpext@ietf.org
In-Reply-To: <002701c3093e$57ebf650$c90b77ca@kchen>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

kchen,

Why would it be different from regular tunnel switching (I mean without IPSec)?
The fact that L2TP is encapsulated in IPSec, TSA should have no trouble
interpreting (for example relaying, regenerating, or dropping certain AVPs)
incoming control/data packets.

[L2] <--- PPP ---> [LAC] <-- PPP/L2TP1/IPSec1 --> [ TSA ] <--- PPP/L2TP2/IPSec2
--> [LNS]

i.e. PPP session coming in on IPSec SA (IPSec1) that carries incoming tunnel's
(call it L2TP1) traffic switched over to another L2TP tunnel (L2TP2) over IPSec
SA (IPSec2)?

thanks,
-- vipin

--- kchen <kchen@seu.edu.cn> wrote:
> Can someone tell me?
> 


__________________________________________________
Do you Yahoo!?
The New Yahoo! Search - Faster. Easier. Bingo
http://search.yahoo.com
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext


From mailnull@www1.ietf.org  Wed Apr 23 02:25:02 2003
Received: from www1.ietf.org (ietf.org [132.151.1.19] (may be forged))
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id CAA00293
	for <l2tpext-archive@odin.ietf.org>; Wed, 23 Apr 2003 02:25:02 -0400 (EDT)
Received: (from mailnull@localhost)
	by www1.ietf.org (8.11.6/8.11.6) id h3N6au726503
	for l2tpext-archive@odin.ietf.org; Wed, 23 Apr 2003 02:36:56 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3N6au826500
	for <l2tpext-web-archive@optimus.ietf.org>; Wed, 23 Apr 2003 02:36:56 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id CAA29731
	for <l2tpext-web-archive@ietf.org>; Wed, 23 Apr 2003 02:24:31 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 198DiW-0000N8-00
	for l2tpext-web-archive@ietf.org; Wed, 23 Apr 2003 02:26:52 -0400
Received: from ietf.org ([132.151.1.19] helo=www1.ietf.org)
	by ietf-mx with esmtp (Exim 4.12)
	id 198DiV-0000N5-00
	for l2tpext-web-archive@ietf.org; Wed, 23 Apr 2003 02:26:51 -0400
Received: from www1.ietf.org (localhost.localdomain [127.0.0.1])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3N6Z8826323;
	Wed, 23 Apr 2003 02:35:08 -0400
Received: from ietf.org (odin.ietf.org [132.151.1.176])
	by www1.ietf.org (8.11.6/8.11.6) with ESMTP id h3N6W6826196
	for <l2tpext@optimus.ietf.org>; Wed, 23 Apr 2003 02:32:06 -0400
Received: from ietf-mx (ietf-mx.ietf.org [132.151.6.1])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id CAA24381
	for <L2tpext@ietf.org>; Wed, 23 Apr 2003 02:19:40 -0400 (EDT)
Received: from ietf-mx ([132.151.6.1])
	by ietf-mx with esmtp (Exim 4.12)
	id 198Ddo-0000KJ-00
	for L2tpext@ietf.org; Wed, 23 Apr 2003 02:22:00 -0400
Received: from web41312.mail.yahoo.com ([66.218.93.61])
	by ietf-mx with smtp (Exim 4.12)
	id 198Ddo-0000KG-00
	for L2tpext@ietf.org; Wed, 23 Apr 2003 02:22:00 -0400
Message-ID: <20030423062154.70404.qmail@web41312.mail.yahoo.com>
Received: from [64.162.95.130] by web41312.mail.yahoo.com via HTTP; Tue, 22 Apr 2003 23:21:54 PDT
Date: Tue, 22 Apr 2003 23:21:54 -0700 (PDT)
From: Vipin Jain <vipinietf@yahoo.com>
Subject: Re: [L2tpext] how can tunnel switch implement in IPSec?
To: kchen <kchen@seu.edu.cn>, L2tpext@ietf.org
In-Reply-To: <002701c3093e$57ebf650$c90b77ca@kchen>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: l2tpext-admin@ietf.org
Errors-To: l2tpext-admin@ietf.org
X-BeenThere: l2tpext@ietf.org
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=unsubscribe>
List-Id: Layer Two Tunneling Protocol Extensions <l2tpext.ietf.org>
List-Post: <mailto:l2tpext@ietf.org>
List-Help: <mailto:l2tpext-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/l2tpext>,
	<mailto:l2tpext-request@ietf.org?subject=subscribe>

kchen,

Why would it be different from regular tunnel switching (I mean without IPSec)?
The fact that L2TP is encapsulated in IPSec, TSA should have no trouble
interpreting (for example relaying, regenerating, or dropping certain AVPs)
incoming control/data packets.

[L2] <--- PPP ---> [LAC] <-- PPP/L2TP1/IPSec1 --> [ TSA ] <--- PPP/L2TP2/IPSec2
--> [LNS]

i.e. PPP session coming in on IPSec SA (IPSec1) that carries incoming tunnel's
(call it L2TP1) traffic switched over to another L2TP tunnel (L2TP2) over IPSec
SA (IPSec2)?

thanks,
-- vipin

--- kchen <kchen@seu.edu.cn> wrote:
> Can someone tell me?
> 


__________________________________________________
Do you Yahoo!?
The New Yahoo! Search - Faster. Easier. Bingo
http://search.yahoo.com
_______________________________________________
L2tpext mailing list
L2tpext@ietf.org
https://www1.ietf.org/mailman/listinfo/l2tpext



