From owner-l2tp@diameter.org  Fri Sep  1 12:07:32 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA15321
	for <l2tpext-archive@odin.ietf.org>; Fri, 1 Sep 2000 12:07:32 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e81C6XU03438
	for l2tp-list; Fri, 1 Sep 2000 05:06:33 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from eins.siemens.at (eins.siemens.at [193.81.246.11])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e81C5El03432
	for <l2tp@diameter.org>; Fri, 1 Sep 2000 05:05:14 -0700
Received: from scesie13.sie.siemens.at (forix-hme0 [10.1.140.2])
	by eins.siemens.at  with ESMTP id NAA10118;
	Fri, 1 Sep 2000 13:59:04 +0200 (MET DST)
Received: (from smap@localhost)
	by scesie13.sie.siemens.at (8.9.3/8.9.3) id OAA25748;
	Fri, 1 Sep 2000 14:07:28 +0200 (MET DST)
Received: from vies141a.sie.siemens.at(195.1.196.56) by scesie13 via smap (V2.0beta)
	id xma023981; Fri, 1 Sep 00 14:06:53 +0200
Received: by vies141a.sie.siemens.at with Internet Mail Service (5.5.2650.21)
	id <R0DZYZ7J>; Fri, 1 Sep 2000 14:06:52 +0200
Message-ID: <D9F2B9CD7BD5D21196BC0800060D9ED601EFBF3E@vies186a.sie.siemens.at>
From: Klausberger Walter <walter.klausberger@siemens.at>
To: "'streifer@us.ibm.com'" <streifer@us.ibm.com>, l2tp@diameter.org
Subject: AW: L2TP Tunnel Authentication with RADIUS
Date: Fri, 1 Sep 2000 14:06:52 +0200 
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by charizard.diameter.org id e81C5Fl03433
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 8bit

Well, we will do so. I think it depends on which side of the tunnel you
are...

In LAC case you can use the tunnel password to setup an appropriate tunnel
to an LNS. During user authentication (with PPP PAP, CHAP,...) you do an
Access-Request and you get the tunnel password in return. That's ok, so
far...

I think your problem is really on the LNS side. A tunnel comes in and there
may be a tunnel password stored on the RADIUS for the incomming session tied
to the username. You are correct. This will not work.

1.) The user may come at different times from different LACs. So there is no
use to bind the password to the username.

2.) As you stated before, it is too late when you get the username.

A solution would be:
You get a LAC host name together with the incomming tunnel. You can use an
Access-Request with this hostname to get a tunnel password from the RADIUS.
So you can answer the challenge and issue your own challenge at the same
time. Nobody said that RADIUS can only be used together with PPP. We even
use a DNIS String as username to get centrally stored information and
configuration.

with best regards
Walter Klausberger
Siemens AG Austria

-----Ursprüngliche Nachricht-----
Von: streifer@us.ibm.com [mailto:streifer@us.ibm.com]
Gesendet am: Dienstag, 22. August 2000 13:51
An: l2tp@diameter.org
Betreff: L2TP Tunnel Authentication with RADIUS


Are there any L2TP implementations that use RADIUS for L2TP tunnel
authentication ?

Looking through the various RADIUS RFCs it looks like RADIUS can be used
for PPP authentication after the tunnel and call are established but I
can't see a way to get a RADIUS Server to send back a MD5 Hash value or a
password to use for a Challenge-Response for tunnel authentication. There
is a Tunnel-Password attribute but it is only sent back in an Access-Accept
packet. From an LNS standpoint it looks to me that I would need a tunnel
password stored locally in order to generate a Challenge-Response which
would defeat the purpose of using RADIUS to consolidate authentication
administration to a single system.

Am I missing something ?


Thanks,

Kurt Streifert
IBM AS/400 Endicott Programming Lab



From owner-l2tp@diameter.org  Sun Sep  3 23:10:59 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id XAA16079
	for <l2tpext-archive@odin.ietf.org>; Sun, 3 Sep 2000 23:10:58 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e841CPM12551
	for l2tp-list; Sun, 3 Sep 2000 18:12:25 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from center.mshindo.net (center.mshindo.net [210.231.221.221])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e841BOl12545
	for <l2tp@diameter.org>; Sun, 3 Sep 2000 18:11:24 -0700
Received: from cosinecom.com (017.cosine.nttpc.gr.jp [202.229.42.17] (may be forged))
	by center.mshindo.net (8.9.3/3.7W/00041811) with SMTP id KAA37382;
	Mon, 4 Sep 2000 10:09:46 +0900 (JST)
Date: Mon, 04 Sep 2000 10:13:15 +0900 (JST)
Message-Id: <20000904.101315.74751693.mshindo@mshindo.net>
To: walter.klausberger@siemens.at
Cc: streifer@us.ibm.com, l2tp@diameter.org
Subject: Re: AW: L2TP Tunnel Authentication with RADIUS
From: Motonori Shindo <mshindo@mshindo.net>
In-Reply-To: <D9F2B9CD7BD5D21196BC0800060D9ED601EFBF3E@vies186a.sie.siemens.at>
References: <D9F2B9CD7BD5D21196BC0800060D9ED601EFBF3E@vies186a.sie.siemens.at>
X-Mailer: Mew version 1.95b30 on XEmacs 21.1 (Canyonlands)
X-PGP-fingerprint: 06 B0 B1 A4 06 C1 6A 14  63 C0 D7 18 01 CD D9 83
Mime-Version: 1.0
Content-Type: Text/Plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by charizard.diameter.org id e841BPl12546
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 8bit

All,

I do agree with Walter's comment. 

Tunnel authentication and user authentication (associated with an
incoming call) are totally two different things, and more importantly,
the establishment of a tunnel does not necessarily happen at the same
time as the session (ie. PPP) establishment.

From this point of view, a straight-forward LAC implementation should
work in the following way:

 1) LAC gets to know the user name (via PAP/CHAP or whatever) and
    queries to RADIUS to authenticate this user and obtain information
    associated with it. If this user is an L2TP user, then the
    RADIUS entry for this user would contain a Tunnel-Server-Endpoint
    attribute associated with this user.

 2) If a tunnel for the LNS specified by Tunnel-Sever-Endpoint hasn't 
    already existed and if the LAC is configured to do tunnel
    authentication, then a LAC queries to RADIUS again to obtain a tunnel
    passowrd. In this case, the RADIUS entry has a LNS name (identified
    by a Tunnel-Server-Endpoint in the previous query) as a user
    name, and should include a Tuennel-Password attribute in its reply
    item.

Please note that this sequence involves two RADIUS queries, one for
user authentication and another for tunnel authentication. To make it
work in more efficent way, an optimization is possible by piggybacking
a Tunnel-Password attribute on the first RADIUS query for the user
authentication. Since there's a one-to-one relationship between
Tunnel-Server-Endpoint and Tunnel-Password, this optimization works
without any problem.

Howerver, on the LNS side, as Walter pointed out, this optimization
can not be applied and two RADIUS queries are required. That is, when
the LNS receives an SCCRQ from the LAC (along with an L2TP's CHALLENGE
attribute), the LNS first queries to RADIUS about the tunnel password
associated with a name specified in the L2TP's HOSTNAME attribute (the
use of HOSTNAME attribute for this purpose is not explicitly defined
in the RFC but it is a good choice, I believe). After the tunnel
authentication is finished and the tunnel is successfully established,
the ICRQ follws and then user authentication happens through, again,
RADIUS.

Regards,

From: Klausberger Walter <walter.klausberger@siemens.at>
Subject: AW: L2TP Tunnel Authentication with RADIUS
Date: Fri, 1 Sep 2000 14:06:52 +0200 

> Well, we will do so. I think it depends on which side of the tunnel you
> are...
> 
> In LAC case you can use the tunnel password to setup an appropriate tunnel
> to an LNS. During user authentication (with PPP PAP, CHAP,...) you do an
> Access-Request and you get the tunnel password in return. That's ok, so
> far...
> 
> I think your problem is really on the LNS side. A tunnel comes in and there
> may be a tunnel password stored on the RADIUS for the incomming session tied
> to the username. You are correct. This will not work.
> 
> 1.) The user may come at different times from different LACs. So there is no
> use to bind the password to the username.
> 
> 2.) As you stated before, it is too late when you get the username.
> 
> A solution would be:
> You get a LAC host name together with the incomming tunnel. You can use an
> Access-Request with this hostname to get a tunnel password from the RADIUS.
> So you can answer the challenge and issue your own challenge at the same
> time. Nobody said that RADIUS can only be used together with PPP. We even
> use a DNIS String as username to get centrally stored information and
> configuration.
> 
> with best regards
> Walter Klausberger
> Siemens AG Austria
> 
> -----Ursprüngliche Nachricht-----
> Von: streifer@us.ibm.com [mailto:streifer@us.ibm.com]
> Gesendet am: Dienstag, 22. August 2000 13:51
> An: l2tp@diameter.org
> Betreff: L2TP Tunnel Authentication with RADIUS
> 
> 
> Are there any L2TP implementations that use RADIUS for L2TP tunnel
> authentication ?
> 
> Looking through the various RADIUS RFCs it looks like RADIUS can be used
> for PPP authentication after the tunnel and call are established but I
> can't see a way to get a RADIUS Server to send back a MD5 Hash value or a
> password to use for a Challenge-Response for tunnel authentication. There
> is a Tunnel-Password attribute but it is only sent back in an Access-Accept
> packet. From an LNS standpoint it looks to me that I would need a tunnel
> password stored locally in order to generate a Challenge-Response which
> would defeat the purpose of using RADIUS to consolidate authentication
> administration to a single system.
> 
> Am I missing something ?
> 
> 
> Thanks,
> 
> Kurt Streifert
> IBM AS/400 Endicott Programming Lab
> 
> 

=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=
 +----+----+     
 |.. .|    |     Motonori Shindo
 |_~__|    |     
 | .. |~~_~|     Sr. Systems Engineer
 | .  |    |     CoSine Communications Inc.
 +----+----+     
 C o S i n e     e-mail:  mshindo@cosinecom.com 
Communications
=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=


From owner-l2tp@diameter.org  Mon Sep  4 14:36:45 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA05895
	for <l2tpext-archive@odin.ietf.org>; Mon, 4 Sep 2000 14:36:38 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e84GEJO13355
	for l2tp-list; Mon, 4 Sep 2000 09:14:19 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from relay03.netaddress.usa.net (relay03.netaddress.usa.net [204.68.24.183])
	by charizard.diameter.org (8.10.0/8.10.0) with SMTP id e84GEFl13349
	for <l2tp@diameter.org>; Mon, 4 Sep 2000 09:14:15 -0700
Received: (qmail 18077 invoked from network); 4 Sep 2000 16:16:42 -0000
Received: from beta.postoffice.net (HELO pb151.postoffice.net) (204.68.24.251)
  by relay03.netaddress.usa.net with SMTP; 4 Sep 2000 16:16:42 -0000
Received: (qmail 7436 invoked by uid 60001); 4 Sep 2000 16:20:01 -0000
Message-ID: <20000904162001.7433.qmail@pb151.postoffice.net>
Received: from 204.68.26.151 by pb151 for [63.197.3.168] via web-mailer(MaintM3.3.0.77) on Mon Sep  4 16:20:01 GMT 2000
Date:  4 Sep 00 09:20:01 PDT
From: Jeff Enderwick <jeff@allegrosys.com>
To: l2tp@diameter.org
Subject: L2F ethereal mods - who has 'em?
X-Mailer: USANET web-mailer (MaintM3.3.0.77)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by charizard.diameter.org id e84GEGl13350
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 8bit

In a prior life, I added L2F support to ethereal
and handed the mods off to a guy who was going to
pull across the inner-packet display functionality
to the L2TP support. He was going to check both
code sets in.

If you're this guy, please email me: jeff@allegrosys.com

Thanks!

(Sorry for the non-L2TP content)


From owner-l2tp@diameter.org  Mon Sep  4 22:22:52 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA10190
	for <l2tpext-archive@odin.ietf.org>; Mon, 4 Sep 2000 22:22:51 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e85092G13651
	for l2tp-list; Mon, 4 Sep 2000 17:09:02 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from center.mshindo.net (center.mshindo.net [210.231.221.221])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8508vl13645
	for <l2tp@diameter.org>; Mon, 4 Sep 2000 17:08:58 -0700
Received: from cosinecom.com (E175147.dion.ne.jp [210.238.175.147])
	by center.mshindo.net (8.9.3/3.7W/00041811) with SMTP id JAA39047;
	Tue, 5 Sep 2000 09:09:36 +0900 (JST)
Date: Tue, 05 Sep 2000 09:12:56 +0900 (JST)
Message-Id: <20000905.091256.71081791.mshindo@mshindo.net>
To: jeff@allegrosys.com
Cc: l2tp@diameter.org
Subject: Re: L2F ethereal mods - who has 'em?
From: Motonori Shindo <mshindo@mshindo.net>
In-Reply-To: <20000904162001.7433.qmail@pb151.postoffice.net>
References: <20000904162001.7433.qmail@pb151.postoffice.net>
X-Mailer: Mew version 1.95b30 on XEmacs 21.1 (Canyonlands)
X-PGP-fingerprint: 06 B0 B1 A4 06 C1 6A 14  63 C0 D7 18 01 CD D9 83
Mime-Version: 1.0
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

All,

From: Jeff Enderwick <jeff@allegrosys.com>
Subject: L2F ethereal mods - who has 'em?
Date: 4 Sep 00 09:20:01 PDT

> In a prior life, I added L2F support to ethereal
> and handed the mods off to a guy who was going to
> pull across the inner-packet display functionality
> to the L2TP support. He was going to check both
> code sets in.
> 
> If you're this guy, please email me: jeff@allegrosys.com
> 
> Thanks!
> 
> (Sorry for the non-L2TP content)

Sorry for continuing a non-L2TP topic. I'm not the man you're looking
for, but I did basically the same thing for tcpdump. It not only
decodes the L2TP but it also decodes the inner PPP frame. I enhanced
the PPP decode function that used to be in the old tcpdump code and it
now has a comprehensive PPP support, I believe. You can get the latest
code from http://www.tcpdump.org.  Hope this help for everyone here:-)

Regards,

=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=
 +----+----+     
 |.. .|    |     Motonori Shindo
 |_~__|    |     
 | .. |~~_~|     Sr. Systems Engineer
 | .  |    |     CoSine Communications Inc.
 +----+----+     
 C o S i n e     e-mail:  mshindo@cosinecom.com 
Communications
=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=





From owner-l2tp@diameter.org  Thu Sep  7 15:52:14 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id PAA25110
	for <l2tpext-archive@odin.ietf.org>; Thu, 7 Sep 2000 15:52:13 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e87GxxR16365
	for l2tp-list; Thu, 7 Sep 2000 09:59:59 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from drawbridge.ascend.com (drawbridge.ascend.com [198.4.92.1])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e87Gxvl16359
	for <l2tp@diameter.org>; Thu, 7 Sep 2000 09:59:57 -0700
Received: from fw-ext.ascend.com (fw-ext [198.4.92.5])
	by drawbridge.ascend.com (8.9.1a/8.9.1) with SMTP id JAA23090
	for <l2tp@diameter.org>; Thu, 7 Sep 2000 09:57:01 -0700 (PDT)
Received: from russet.ascend.com by fw-ext.ascend.com
          via smtpd (for drawbridge.ascend.com [198.4.92.1]) with SMTP; 7 Sep 2000 17:04:38 UT
Received: from wopr.eng.ascend.com (wopr.eng.ascend.com [206.65.212.178])
	by russet.ascend.com (8.9.1a/8.9.1) with ESMTP id KAA16270
	for <l2tp@diameter.org>; Thu, 7 Sep 2000 10:04:37 -0700 (PDT)
Received: from scamp.eng.ascend.com (scamp.eng.ascend.com [135.140.53.42])
	by wopr.eng.ascend.com (8.9.1/8.9.1) with ESMTP id KAA27822
	for <l2tp@diameter.org>; Thu, 7 Sep 2000 10:04:38 -0700 (PDT)
Received: from igoyret-pc.eng.ascend.com by scamp.eng.ascend.com (8.8.8+Sun/SMI-SVR4)
	id KAA13978; Thu, 7 Sep 2000 10:04:37 -0700 (PDT)
Message-Id: <3.0.5.32.20000907100431.013614d0@scamp.eng.ascend.com>
X-Sender: igoyret@scamp.eng.ascend.com (Unverified)
X-Mailer: QUALCOMM Windows Eudora Pro Version 3.0.5 (32)
Date: Thu, 07 Sep 2000 10:04:31 -0700
To: l2tp@diameter.org
From: Ignacio Goyret <igoyret@ascend.com>
Subject: draft-l2tpext-l2tp-mib-00.txt
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-l2tp@diameter.org
Precedence: bulk

Finally, I figured out how to send messages to the list. :-)
Cheers,
-Ignacio

>Date: Tue, 05 Sep 2000 16:04:47 -0700
>To: evan.caves@ericsson.com
>From: Ignacio Goyret <igoyret@lucent.com>
>Subject: draft-l2tpext-l2tp-mib-00.txt
>Cc: l2tp@diameter.org
>
>Hi Evan,
>I sent this email a few weeks ago but for some reason, I haven't seen it
>posted on the L2TP list. Sending it again just in case.
>
>-----------------------------------
>Hi Evan,
>I've just noticed a little discrepancy between your draft and rfc2661.
>
>* 'l2tpSessionStatsAuthMethod' has a value for EAP (pppEap, 5) but no
>   value for Microsoft CHAP v1.
>
>   l2tpSessionStatsAuthMethod OBJECT-TYPE
>           SYNTAX          INTEGER {
>                               none(1),
>                               text(2),
>                               pppChap(3),
>                               pppPap(4),
>                               pppEap(5)
>                           }
>     ...
>
>Since we are talking about missing values, rfc2661 failed to mention EAP:
>
>   Proxy Authen Type (ICCN)
>
>      The Proxy Authen Type AVP, Attribute Type 29, determines if proxy
>      authentication should be used.
>   ...
>
>      Defined Authen Type values are:
>         0 - Reserved
>         1 - Textual username/password exchange
>         2 - PPP CHAP
>         3 - PPP PAP
>         4 - No Authentication
>         5 - Microsoft CHAP Version 1 (MSCHAPv1)
>
>Of course, there are other authentication methods that are not mentioned
>such as DNIS or CLID authentication. From conversations with customers and
>other manufacturers, it looks that everyone has done it just a little bit
>different.
>
>Let me suggest a solution:
>
>* On the L2TP MIB draft, change:
>      l2tpSessionStatsAuthMethod OBJECT-TYPE
>           SYNTAX          INTEGER {
>                               none(1),
>                               text(2),
>                               pppChap(3),
>                               pppPap(4),
>                               pppEap(5)
>                           }
>  to:
>      l2tpSessionStatsAuthMethod OBJECT-TYPE
>           SYNTAX          INTEGER {
>                               none(1),
>                               text(2),
>                               pppChap(3),
>                               pppPap(4),
>                               pppMsChap(5),        <<<
>                               pppEap(6)            <<<
>                           }
>
>
>* For RFC 2661-bis, add:
>
>      Defined Authen Type values are:
>         0 - Reserved
>         1 - Textual username/password exchange
>         2 - PPP CHAP
>         3 - PPP PAP
>         4 - No Authentication
>         5 - Microsoft CHAP Version 1 (MSCHAPv1)
>         6 - PPP EAP                                 <<<<
>
>Any thoughts?

Needless to say, I'll be happy with pppMsChap(6) and pppEap(5) as well. :-)


>Cheers,
>-Ignacio



From owner-l2tp@diameter.org  Thu Sep  7 16:10:39 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id QAA26112
	for <l2tpext-archive@odin.ietf.org>; Thu, 7 Sep 2000 16:10:39 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e87GkK116347
	for l2tp-list; Thu, 7 Sep 2000 09:46:20 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from drawbridge.ascend.com (drawbridge.ascend.com [198.4.92.1])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e87GkHl16341
	for <l2tp@diameter.org>; Thu, 7 Sep 2000 09:46:17 -0700
Received: from fw-ext.ascend.com (fw-ext [198.4.92.5])
	by drawbridge.ascend.com (8.9.1a/8.9.1) with SMTP id JAA21272
	for <l2tp@diameter.org>; Thu, 7 Sep 2000 09:43:16 -0700 (PDT)
Received: from russet.ascend.com by fw-ext.ascend.com
          via smtpd (for drawbridge.ascend.com [198.4.92.1]) with SMTP; 7 Sep 2000 16:50:53 UT
Received: from wopr.eng.ascend.com (wopr.eng.ascend.com [206.65.212.178])
	by russet.ascend.com (8.9.1a/8.9.1) with ESMTP id JAA11303;
	Thu, 7 Sep 2000 09:50:52 -0700 (PDT)
Received: from scamp.eng.ascend.com (scamp.eng.ascend.com [135.140.53.42])
	by wopr.eng.ascend.com (8.9.1/8.9.1) with ESMTP id JAA27113;
	Thu, 7 Sep 2000 09:50:52 -0700 (PDT)
Received: from igoyret-pc.eng.ascend.com by scamp.eng.ascend.com (8.8.8+Sun/SMI-SVR4)
	id JAA13527; Thu, 7 Sep 2000 09:50:48 -0700 (PDT)
Message-Id: <3.0.5.32.20000907095042.0142f5a0@scamp.eng.ascend.com>
X-Sender: igoyret@scamp.eng.ascend.com (Unverified)
X-Mailer: QUALCOMM Windows Eudora Pro Version 3.0.5 (32)
Date: Thu, 07 Sep 2000 09:50:42 -0700
To: l2tp@diameter.org
From: Ignacio Goyret <igoyret@ascend.com>
Subject: test - please ignore
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-l2tp@diameter.org
Precedence: bulk

test - please ignore


From owner-l2tp@diameter.org  Fri Sep  8 01:04:45 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id BAA05187
	for <l2tpext-archive@odin.ietf.org>; Fri, 8 Sep 2000 01:04:44 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e881Boi16650
	for l2tp-list; Thu, 7 Sep 2000 18:11:50 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from omega.cisco.com (omega.cisco.com [171.69.63.141])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e881Apl16644
	for <l2tp@diameter.org>; Thu, 7 Sep 2000 18:10:51 -0700
Received: from tmima-nt.cisco.com (dhcp-171-70-57-27.cisco.com [171.70.57.27])
	by omega.cisco.com (8.8.8-Cisco List Logging/8.8.8) with ESMTP id SAA22285;
	Thu, 7 Sep 2000 18:08:41 -0700 (PDT)
Message-Id: <4.3.1.2.20000907161936.029ba8e0@omega.cisco.com>
X-Sender: tmima@omega.cisco.com
X-Mailer: QUALCOMM Windows Eudora Version 4.3.1
Date: Thu, 07 Sep 2000 18:03:29 -0700
To: Andy Valencia <vandys@zendo.com>, "W. Mark Townsley" <townsley@cisco.com>
From: Tmima Koren <tmima@cisco.com>
Subject: 0 byte L2TPHC header and implied PPPMUX
Cc: brucet@cisco.com, dwing@cisco.com,
        Stephen Casner <casner@packetdesign.com>, l2tp@diameter.org,
        rem-conf@es.net
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-l2tp@diameter.org
Precedence: bulk


I suggest a few minor changes to the L2TPHC draft draft-ietf-l2tpext-l2tphc-02.txt
to improve bandwidth utilization in special cases.


When the L2TPHC header is 1 byte long, actually only 1 bit is significant: the P (Priority) bit.

If all the packets in the tunnel will have the same priority, the L2TPHC byte is redundant.

We could add to the L2TPHC AVP one more byte (byte 9)
We really need only 2 flag bits:

flag 1:  value 1 means: L2TPHC byte will be omitted
            value 0 means: L2TPHC byte is there, get P (Priority) from the L2TPHC byte

flag 2:  value 1 means all packets have implied P=1
            value 0 means all packets have implied P=0

consult flag 2 only when flag 1 is 1


If the packet in the L2TPHC tunnel is always PPPMUX, there is another useful flag:

flag 3: value 1 means 'PPPMUX only'
           value 0 means: any PPP packet


When flags 1 and 3 are on (=1), both L2TPHC header and PPPMUX protocol id can be omitted 
PPP packets that have a different priority or that are not PPPMUX can still be sent in the tunnel using the UDP/L2TP format, same as the L2TP control packets

TCRTP (Tunneled Compressed multiplexed RTP) can benefit from these changes: the tunneling overhead will be reduced by 2 bytes.

Tmima



From owner-l2tp@diameter.org  Fri Sep  8 05:13:16 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id FAA19225
	for <l2tpext-archive@odin.ietf.org>; Fri, 8 Sep 2000 05:13:15 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e886TWY16786
	for l2tp-list; Thu, 7 Sep 2000 23:29:32 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from arianne.in.ishoni.com (IDENT:root@[164.164.83.132])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e886TAl16780
	for <l2tp@diameter.org>; Thu, 7 Sep 2000 23:29:11 -0700
Received: from rohit (rashmi [192.168.1.16])
	by arianne.in.ishoni.com (8.9.3/8.9.3) with SMTP id MAA15207
	for <l2tp@diameter.org>; Fri, 8 Sep 2000 12:07:39 +0530
Reply-To: <rkhosla@ishoni.com>
From: "Rohit Khosla" <rkhosla@ishoni.com>
To: <l2tp@diameter.org>
Subject: IPSec interop workshop Schedule
Date: Fri, 8 Sep 2000 12:14:41 +0530
Message-ID: <C182A5918209D41190DE00C04F0CCD254BADE1@LEONOID>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook CWS, Build 9.0.2416 (9.0.2911.0)
In-reply-to: <C182A5918209D41190DE00C04F0CCD254CC2CF@LEONOID>
Importance: Normal
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit


Hi All,

	would appreciate any info regarding IPSec interop workshop schedule.

Thanks,
Rohit



From owner-l2tp@diameter.org  Mon Sep 11 15:30:04 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id PAA08369
	for <l2tpext-archive@odin.ietf.org>; Mon, 11 Sep 2000 15:30:01 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8BH2nI24912
	for l2tp-list; Mon, 11 Sep 2000 10:02:49 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from mailsrv.acc.com (mailsrv.acc.com [129.192.64.128])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8BH2jl24906
	for <l2tp@diameter.org>; Mon, 11 Sep 2000 10:02:46 -0700
Received: from ericsson.com (calcium.dev.acc.am.ericsson.se [129.192.57.200])
	by mailsrv.acc.com (8.9.3/8.9.1) with ESMTP id KAA02495;
	Mon, 11 Sep 2000 10:11:14 -0700 (PDT)
Message-ID: <39BD1287.A9D08E25@ericsson.com>
Date: Mon, 11 Sep 2000 10:12:39 -0700
From: Evan Caves <evan.caves@ericsson.com>
X-Mailer: Mozilla 4.61 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Ignacio Goyret <igoyret@ascend.com>
CC: l2tp@diameter.org
Subject: Re: draft-l2tpext-l2tp-mib-00.txt
References: <3.0.5.32.20000907100431.013614d0@scamp.eng.ascend.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Hi Ignacio,

   No problem. I can add a enum for pppMsChap to the statsAuthMethod
object however should the enum reflect 'v1' (ie: pppMsChapV1)? At
the same time I should also add a 'other' enum as a place holder for
other types of authentication methods that may be employed but not
enmumerated.

evan
-

Ignacio Goyret wrote:
> 
> Finally, I figured out how to send messages to the list. :-)
> Cheers,
> -Ignacio
> 
> >Date: Tue, 05 Sep 2000 16:04:47 -0700
> >To: evan.caves@ericsson.com
> >From: Ignacio Goyret <igoyret@lucent.com>
> >Subject: draft-l2tpext-l2tp-mib-00.txt
> >Cc: l2tp@diameter.org
> >
> >Hi Evan,
> >I sent this email a few weeks ago but for some reason, I haven't seen it
> >posted on the L2TP list. Sending it again just in case.
> >
> >-----------------------------------
> >Hi Evan,
> >I've just noticed a little discrepancy between your draft and rfc2661.
> >
> >* 'l2tpSessionStatsAuthMethod' has a value for EAP (pppEap, 5) but no
> >   value for Microsoft CHAP v1.
> >
> >   l2tpSessionStatsAuthMethod OBJECT-TYPE
> >           SYNTAX          INTEGER {
> >                               none(1),
> >                               text(2),
> >                               pppChap(3),
> >                               pppPap(4),
> >                               pppEap(5)
> >                           }
> >     ...
> >
> >Since we are talking about missing values, rfc2661 failed to mention EAP:
> >
> >   Proxy Authen Type (ICCN)
> >
> >      The Proxy Authen Type AVP, Attribute Type 29, determines if proxy
> >      authentication should be used.
> >   ...
> >
> >      Defined Authen Type values are:
> >         0 - Reserved
> >         1 - Textual username/password exchange
> >         2 - PPP CHAP
> >         3 - PPP PAP
> >         4 - No Authentication
> >         5 - Microsoft CHAP Version 1 (MSCHAPv1)
> >
> >Of course, there are other authentication methods that are not mentioned
> >such as DNIS or CLID authentication. From conversations with customers and
> >other manufacturers, it looks that everyone has done it just a little bit
> >different.
> >
> >Let me suggest a solution:
> >
> >* On the L2TP MIB draft, change:
> >      l2tpSessionStatsAuthMethod OBJECT-TYPE
> >           SYNTAX          INTEGER {
> >                               none(1),
> >                               text(2),
> >                               pppChap(3),
> >                               pppPap(4),
> >                               pppEap(5)
> >                           }
> >  to:
> >      l2tpSessionStatsAuthMethod OBJECT-TYPE
> >           SYNTAX          INTEGER {
> >                               none(1),
> >                               text(2),
> >                               pppChap(3),
> >                               pppPap(4),
> >                               pppMsChap(5),        <<<
> >                               pppEap(6)            <<<
> >                           }
> >
> >
> >* For RFC 2661-bis, add:
> >
> >      Defined Authen Type values are:
> >         0 - Reserved
> >         1 - Textual username/password exchange
> >         2 - PPP CHAP
> >         3 - PPP PAP
> >         4 - No Authentication
> >         5 - Microsoft CHAP Version 1 (MSCHAPv1)
> >         6 - PPP EAP                                 <<<<
> >
> >Any thoughts?
> 
> Needless to say, I'll be happy with pppMsChap(6) and pppEap(5) as well. :-)
> 
> >Cheers,
> >-Ignacio


From owner-l2tp@diameter.org  Mon Sep 11 16:47:40 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id QAA08880
	for <l2tpext-archive@odin.ietf.org>; Mon, 11 Sep 2000 16:47:39 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8BI6bp24961
	for l2tp-list; Mon, 11 Sep 2000 11:06:37 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from drawbridge.ascend.com (drawbridge.ascend.com [198.4.92.1])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8BI6Yl24955
	for <l2tp@diameter.org>; Mon, 11 Sep 2000 11:06:34 -0700
Received: from fw-ext.ascend.com (fw-ext [198.4.92.5])
	by drawbridge.ascend.com (8.9.1a/8.9.1) with SMTP id LAA09273;
	Mon, 11 Sep 2000 11:00:28 -0700 (PDT)
Received: from russet.ascend.com by fw-ext.ascend.com
          via smtpd (for drawbridge.ascend.com [198.4.92.1]) with SMTP; 11 Sep 2000 18:08:07 UT
Received: from wopr.eng.ascend.com (wopr.eng.ascend.com [206.65.212.178])
	by russet.ascend.com (8.9.1a/8.9.1) with ESMTP id LAA01263;
	Mon, 11 Sep 2000 11:08:06 -0700 (PDT)
Received: from scamp.eng.ascend.com (scamp.eng.ascend.com [135.140.53.42])
	by wopr.eng.ascend.com (8.9.1/8.9.1) with ESMTP id LAA11289;
	Mon, 11 Sep 2000 11:08:06 -0700 (PDT)
Received: from igoyret-pc.eng.ascend.com by scamp.eng.ascend.com (8.8.8+Sun/SMI-SVR4)
	id LAA28261; Mon, 11 Sep 2000 11:08:05 -0700 (PDT)
Message-Id: <3.0.5.32.20000911110804.03871190@scamp.eng.ascend.com>
X-Sender: igoyret@scamp.eng.ascend.com (Unverified)
X-Mailer: QUALCOMM Windows Eudora Pro Version 3.0.5 (32)
Date: Mon, 11 Sep 2000 11:08:04 -0700
To: Evan Caves <evan.caves@ericsson.com>
From: Ignacio Goyret <igoyret@ascend.com>
Subject: Re: draft-l2tpext-l2tp-mib-00.txt
Cc: Ignacio Goyret <igoyret@ascend.com>, l2tp@diameter.org
In-Reply-To: <39BD1287.A9D08E25@ericsson.com>
References: <3.0.5.32.20000907100431.013614d0@scamp.eng.ascend.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-l2tp@diameter.org
Precedence: bulk

Hi Evan,

At 10:12 AM 9/11/00 -0700, Evan Caves wrote:
>Hi Ignacio,
>
>   No problem. I can add a enum for pppMsChap to the statsAuthMethod
>object however should the enum reflect 'v1' (ie: pppMsChapV1)?

That's probably a good idea. Since you are at it, you might as well
add a value for v2 (ie, pppMsChapV2). The RFC references are 2433 (v1)
and 2759 (v2).


>At
>the same time I should also add a 'other' enum as a place holder for
>other types of authentication methods that may be employed but not
>enmumerated.

Sounds good to me.

Thanks,
-Ignacio


>Ignacio Goyret wrote:
>> 
>> >Hi Evan,
>> >I've just noticed a little discrepancy between your draft and rfc2661.
>> >
>> >* 'l2tpSessionStatsAuthMethod' has a value for EAP (pppEap, 5) but no
>> >   value for Microsoft CHAP v1.
>> >
>> >   l2tpSessionStatsAuthMethod OBJECT-TYPE
>> >           SYNTAX          INTEGER {
>> >                               none(1),
>> >                               text(2),
>> >                               pppChap(3),
>> >                               pppPap(4),
>> >                               pppEap(5)
>> >                           }
>> >     ...
>> >
>> >Since we are talking about missing values, rfc2661 failed to mention EAP:
>> >
>> >   Proxy Authen Type (ICCN)
>> >
>> >      The Proxy Authen Type AVP, Attribute Type 29, determines if proxy
>> >      authentication should be used.
>> >   ...
>> >
>> >      Defined Authen Type values are:
>> >         0 - Reserved
>> >         1 - Textual username/password exchange
>> >         2 - PPP CHAP
>> >         3 - PPP PAP
>> >         4 - No Authentication
>> >         5 - Microsoft CHAP Version 1 (MSCHAPv1)
>> >
>> >Of course, there are other authentication methods that are not mentioned
>> >such as DNIS or CLID authentication. From conversations with customers and
>> >other manufacturers, it looks that everyone has done it just a little bit
>> >different.
>> >
>> >Let me suggest a solution:
>> >
>> >* On the L2TP MIB draft, change:
>> >      l2tpSessionStatsAuthMethod OBJECT-TYPE
>> >           SYNTAX          INTEGER {
>> >                               none(1),
>> >                               text(2),
>> >                               pppChap(3),
>> >                               pppPap(4),
>> >                               pppEap(5)
>> >                           }
>> >  to:
>> >      l2tpSessionStatsAuthMethod OBJECT-TYPE
>> >           SYNTAX          INTEGER {
>> >                               none(1),
>> >                               text(2),
>> >                               pppChap(3),
>> >                               pppPap(4),
>> >                               pppMsChap(5),        <<<
>> >                               pppEap(6)            <<<
>> >                           }
>> >
>> >
>> >* For RFC 2661-bis, add:
>> >
>> >      Defined Authen Type values are:
>> >         0 - Reserved
>> >         1 - Textual username/password exchange
>> >         2 - PPP CHAP
>> >         3 - PPP PAP
>> >         4 - No Authentication
>> >         5 - Microsoft CHAP Version 1 (MSCHAPv1)
>> >         6 - PPP EAP                                 <<<<
>> >
>> >Any thoughts?
>> 
>> Needless to say, I'll be happy with pppMsChap(6) and pppEap(5) as well. :-)



From owner-l2tp@diameter.org  Mon Sep 11 17:17:02 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id RAA09130
	for <l2tpext-archive@odin.ietf.org>; Mon, 11 Sep 2000 17:17:01 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8BJCAn25042
	for l2tp-list; Mon, 11 Sep 2000 12:12:10 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from mailsrv.acc.com (mailsrv.acc.com [129.192.64.128])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8BJC6l25036
	for <l2tp@diameter.org>; Mon, 11 Sep 2000 12:12:06 -0700
Received: from ericsson.com (calcium.dev.acc.am.ericsson.se [129.192.57.200])
	by mailsrv.acc.com (8.9.3/8.9.1) with ESMTP id MAA11667;
	Mon, 11 Sep 2000 12:18:54 -0700 (PDT)
Message-ID: <39BD3073.8DF7014A@ericsson.com>
Date: Mon, 11 Sep 2000 12:20:19 -0700
From: Evan Caves <evan.caves@ericsson.com>
X-Mailer: Mozilla 4.61 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Ignacio Goyret <igoyret@ascend.com>
CC: l2tp@diameter.org
Subject: Re: draft-l2tpext-l2tp-mib-00.txt
References: <39BD1287.A9D08E25@ericsson.com>
	 <3.0.5.32.20000907100431.013614d0@scamp.eng.ascend.com> <3.0.5.32.20000911120524.04b37620@scamp.eng.ascend.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Ignacio Goyret wrote:
> 
> Just to round up the issue, is this what you have in mind for numbers and names?
> 
>    l2tpSessionStatsAuthMethod OBJECT-TYPE
>         SYNTAX          INTEGER {
>                              none(1),
>                              text(2),
>                              pppChap(3),
>                              pppPap(4),
>                              pppEap(5),
>                              pppMsChapV1(6),        <<<
>                              pppMsChapV2(7),        <<<
>                              other(8),              <<<

I think I prefer this one, that is add to the end of the
enum list, instead of reordering enums at this stage of MIB
development.

evan
-
>                          }
> 
> or this one (that agrees a little better with the values in the Authen Type AVP):
> 
>    l2tpSessionStatsAuthMethod OBJECT-TYPE
>         SYNTAX          INTEGER {
>                              none(1),
>                              text(2),
>                              pppChap(3),
>                              pppPap(4),
>                              pppMsChapV1(5),        <<<  (in Authen Type AVP)
>                              pppMsChapV2(6),        <<<  (not in Authen Type AVP)
>                              pppEap(7),             <<<  (not in Authen Type AVP)
>                              other(8),              <<<  (not in Authen Type AVP)
>                          }
> 
> Sorry to bother you, but I need to finish this today :-(
> -Ignacio
> 
> At 11:08 AM 9/11/00 -0700, Ignacio Goyret wrote:
> >Hi Evan,
> >
> >At 10:12 AM 9/11/00 -0700, Evan Caves wrote:
> >>Hi Ignacio,
> >>
> >>   No problem. I can add a enum for pppMsChap to the statsAuthMethod
> >>object however should the enum reflect 'v1' (ie: pppMsChapV1)?
> >
> >That's probably a good idea. Since you are at it, you might as well
> >add a value for v2 (ie, pppMsChapV2). The RFC references are 2433 (v1)
> >and 2759 (v2).
> >
> >
> >>At
> >>the same time I should also add a 'other' enum as a place holder for
> >>other types of authentication methods that may be employed but not
> >>enmumerated.
> >
> >Sounds good to me.
> >
> >Thanks,
> >-Ignacio
> >
> >
> >>Ignacio Goyret wrote:
> >>>
> >>> >Hi Evan,
> >>> >I've just noticed a little discrepancy between your draft and rfc2661.
> >>> >
> >>> >* 'l2tpSessionStatsAuthMethod' has a value for EAP (pppEap, 5) but no
> >>> >   value for Microsoft CHAP v1.
> >>> >
> >>> >   l2tpSessionStatsAuthMethod OBJECT-TYPE
> >>> >           SYNTAX          INTEGER {
> >>> >                               none(1),
> >>> >                               text(2),
> >>> >                               pppChap(3),
> >>> >                               pppPap(4),
> >>> >                               pppEap(5)
> >>> >                           }
> >>> >     ...
> >>> >
> >>> >Since we are talking about missing values, rfc2661 failed to mention EAP:
> >>> >
> >>> >   Proxy Authen Type (ICCN)
> >>> >
> >>> >      The Proxy Authen Type AVP, Attribute Type 29, determines if proxy
> >>> >      authentication should be used.
> >>> >   ...
> >>> >
> >>> >      Defined Authen Type values are:
> >>> >         0 - Reserved
> >>> >         1 - Textual username/password exchange
> >>> >         2 - PPP CHAP
> >>> >         3 - PPP PAP
> >>> >         4 - No Authentication
> >>> >         5 - Microsoft CHAP Version 1 (MSCHAPv1)
> >>> >
> >>> >Of course, there are other authentication methods that are not mentioned
> >>> >such as DNIS or CLID authentication. From conversations with customers and
> >>> >other manufacturers, it looks that everyone has done it just a little bit
> >>> >different.
> >>> >
> >>> >Let me suggest a solution:
> >>> >
> >>> >* On the L2TP MIB draft, change:
> >>> >      l2tpSessionStatsAuthMethod OBJECT-TYPE
> >>> >           SYNTAX          INTEGER {
> >>> >                               none(1),
> >>> >                               text(2),
> >>> >                               pppChap(3),
> >>> >                               pppPap(4),
> >>> >                               pppEap(5)
> >>> >                           }
> >>> >  to:
> >>> >      l2tpSessionStatsAuthMethod OBJECT-TYPE
> >>> >           SYNTAX          INTEGER {
> >>> >                               none(1),
> >>> >                               text(2),
> >>> >                               pppChap(3),
> >>> >                               pppPap(4),
> >>> >                               pppMsChap(5),        <<<
> >>> >                               pppEap(6)            <<<
> >>> >                           }
> >>> >
> >>> >
> >>> >* For RFC 2661-bis, add:
> >>> >
> >>> >      Defined Authen Type values are:
> >>> >         0 - Reserved
> >>> >         1 - Textual username/password exchange
> >>> >         2 - PPP CHAP
> >>> >         3 - PPP PAP
> >>> >         4 - No Authentication
> >>> >         5 - Microsoft CHAP Version 1 (MSCHAPv1)
> >>> >         6 - PPP EAP                                 <<<<
> >>> >
> >>> >Any thoughts?
> >>>
> >>> Needless to say, I'll be happy with pppMsChap(6) and pppEap(5) as well. :-)
> >
> >


From owner-l2tp@diameter.org  Mon Sep 11 17:21:38 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id RAA09174
	for <l2tpext-archive@odin.ietf.org>; Mon, 11 Sep 2000 17:21:37 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8BJ0ZS25019
	for l2tp-list; Mon, 11 Sep 2000 12:00:35 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from drawbridge.ascend.com (drawbridge.ascend.com [198.4.92.1])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8BJ0Wl25013
	for <l2tp@diameter.org>; Mon, 11 Sep 2000 12:00:32 -0700
Received: from fw-ext.ascend.com (fw-ext [198.4.92.5])
	by drawbridge.ascend.com (8.9.1a/8.9.1) with SMTP id LAA14768;
	Mon, 11 Sep 2000 11:57:48 -0700 (PDT)
Received: from russet.ascend.com by fw-ext.ascend.com
          via smtpd (for drawbridge.ascend.com [198.4.92.1]) with SMTP; 11 Sep 2000 19:05:27 UT
Received: from wopr.eng.ascend.com (wopr.eng.ascend.com [206.65.212.178])
	by russet.ascend.com (8.9.1a/8.9.1) with ESMTP id MAA26015;
	Mon, 11 Sep 2000 12:05:26 -0700 (PDT)
Received: from scamp.eng.ascend.com (scamp.eng.ascend.com [135.140.53.42])
	by wopr.eng.ascend.com (8.9.1/8.9.1) with ESMTP id MAA14647;
	Mon, 11 Sep 2000 12:05:26 -0700 (PDT)
Received: from igoyret-pc.eng.ascend.com by scamp.eng.ascend.com (8.8.8+Sun/SMI-SVR4)
	id MAA00338; Mon, 11 Sep 2000 12:05:25 -0700 (PDT)
Message-Id: <3.0.5.32.20000911120524.04b37620@scamp.eng.ascend.com>
X-Sender: igoyret@scamp.eng.ascend.com (Unverified)
X-Mailer: QUALCOMM Windows Eudora Pro Version 3.0.5 (32)
Date: Mon, 11 Sep 2000 12:05:24 -0700
To: Evan Caves <evan.caves@ericsson.com>
From: Ignacio Goyret <igoyret@ascend.com>
Subject: Re: draft-l2tpext-l2tp-mib-00.txt
Cc: Ignacio Goyret <igoyret@ascend.com>, l2tp@diameter.org
In-Reply-To: <3.0.5.32.20000911110804.03871190@scamp.eng.ascend.com>
References: <39BD1287.A9D08E25@ericsson.com>
 <3.0.5.32.20000907100431.013614d0@scamp.eng.ascend.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-l2tp@diameter.org
Precedence: bulk

Just to round up the issue, is this what you have in mind for numbers and names?

   l2tpSessionStatsAuthMethod OBJECT-TYPE
        SYNTAX          INTEGER {
                             none(1),
                             text(2),
                             pppChap(3),
                             pppPap(4),
                             pppEap(5),
                             pppMsChapV1(6),        <<<
                             pppMsChapV2(7),        <<<
                             other(8),              <<<
                         }

or this one (that agrees a little better with the values in the Authen Type AVP):

   l2tpSessionStatsAuthMethod OBJECT-TYPE
        SYNTAX          INTEGER {
                             none(1),
                             text(2),
                             pppChap(3),
                             pppPap(4),
                             pppMsChapV1(5),        <<<  (in Authen Type AVP)
                             pppMsChapV2(6),        <<<  (not in Authen Type AVP)
                             pppEap(7),             <<<  (not in Authen Type AVP)
                             other(8),              <<<  (not in Authen Type AVP)
                         }


Sorry to bother you, but I need to finish this today :-(
-Ignacio


At 11:08 AM 9/11/00 -0700, Ignacio Goyret wrote:
>Hi Evan,
>
>At 10:12 AM 9/11/00 -0700, Evan Caves wrote:
>>Hi Ignacio,
>>
>>   No problem. I can add a enum for pppMsChap to the statsAuthMethod
>>object however should the enum reflect 'v1' (ie: pppMsChapV1)?
>
>That's probably a good idea. Since you are at it, you might as well
>add a value for v2 (ie, pppMsChapV2). The RFC references are 2433 (v1)
>and 2759 (v2).
>
>
>>At
>>the same time I should also add a 'other' enum as a place holder for
>>other types of authentication methods that may be employed but not
>>enmumerated.
>
>Sounds good to me.
>
>Thanks,
>-Ignacio
>
>
>>Ignacio Goyret wrote:
>>> 
>>> >Hi Evan,
>>> >I've just noticed a little discrepancy between your draft and rfc2661.
>>> >
>>> >* 'l2tpSessionStatsAuthMethod' has a value for EAP (pppEap, 5) but no
>>> >   value for Microsoft CHAP v1.
>>> >
>>> >   l2tpSessionStatsAuthMethod OBJECT-TYPE
>>> >           SYNTAX          INTEGER {
>>> >                               none(1),
>>> >                               text(2),
>>> >                               pppChap(3),
>>> >                               pppPap(4),
>>> >                               pppEap(5)
>>> >                           }
>>> >     ...
>>> >
>>> >Since we are talking about missing values, rfc2661 failed to mention EAP:
>>> >
>>> >   Proxy Authen Type (ICCN)
>>> >
>>> >      The Proxy Authen Type AVP, Attribute Type 29, determines if proxy
>>> >      authentication should be used.
>>> >   ...
>>> >
>>> >      Defined Authen Type values are:
>>> >         0 - Reserved
>>> >         1 - Textual username/password exchange
>>> >         2 - PPP CHAP
>>> >         3 - PPP PAP
>>> >         4 - No Authentication
>>> >         5 - Microsoft CHAP Version 1 (MSCHAPv1)
>>> >
>>> >Of course, there are other authentication methods that are not mentioned
>>> >such as DNIS or CLID authentication. From conversations with customers and
>>> >other manufacturers, it looks that everyone has done it just a little bit
>>> >different.
>>> >
>>> >Let me suggest a solution:
>>> >
>>> >* On the L2TP MIB draft, change:
>>> >      l2tpSessionStatsAuthMethod OBJECT-TYPE
>>> >           SYNTAX          INTEGER {
>>> >                               none(1),
>>> >                               text(2),
>>> >                               pppChap(3),
>>> >                               pppPap(4),
>>> >                               pppEap(5)
>>> >                           }
>>> >  to:
>>> >      l2tpSessionStatsAuthMethod OBJECT-TYPE
>>> >           SYNTAX          INTEGER {
>>> >                               none(1),
>>> >                               text(2),
>>> >                               pppChap(3),
>>> >                               pppPap(4),
>>> >                               pppMsChap(5),        <<<
>>> >                               pppEap(6)            <<<
>>> >                           }
>>> >
>>> >
>>> >* For RFC 2661-bis, add:
>>> >
>>> >      Defined Authen Type values are:
>>> >         0 - Reserved
>>> >         1 - Textual username/password exchange
>>> >         2 - PPP CHAP
>>> >         3 - PPP PAP
>>> >         4 - No Authentication
>>> >         5 - Microsoft CHAP Version 1 (MSCHAPv1)
>>> >         6 - PPP EAP                                 <<<<
>>> >
>>> >Any thoughts?
>>> 
>>> Needless to say, I'll be happy with pppMsChap(6) and pppEap(5) as well. :-)
>
>


From owner-l2tp@diameter.org  Mon Sep 11 18:20:53 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id SAA09575
	for <l2tpext-archive@odin.ietf.org>; Mon, 11 Sep 2000 18:20:51 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8BJCpI25050
	for l2tp-list; Mon, 11 Sep 2000 12:12:51 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from drawbridge.ascend.com (drawbridge.ascend.com [198.4.92.1])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8BJCml25044
	for <l2tp@diameter.org>; Mon, 11 Sep 2000 12:12:48 -0700
Received: from fw-ext.ascend.com (fw-ext [198.4.92.5])
	by drawbridge.ascend.com (8.9.1a/8.9.1) with SMTP id MAA15883;
	Mon, 11 Sep 2000 12:10:04 -0700 (PDT)
Received: from russet.ascend.com by fw-ext.ascend.com
          via smtpd (for drawbridge.ascend.com [198.4.92.1]) with SMTP; 11 Sep 2000 19:17:43 UT
Received: from wopr.eng.ascend.com (wopr.eng.ascend.com [206.65.212.178])
	by russet.ascend.com (8.9.1a/8.9.1) with ESMTP id MAA00691;
	Mon, 11 Sep 2000 12:17:42 -0700 (PDT)
Received: from scamp.eng.ascend.com (scamp.eng.ascend.com [135.140.53.42])
	by wopr.eng.ascend.com (8.9.1/8.9.1) with ESMTP id MAA15203;
	Mon, 11 Sep 2000 12:17:42 -0700 (PDT)
Received: from igoyret-pc.eng.ascend.com by scamp.eng.ascend.com (8.8.8+Sun/SMI-SVR4)
	id MAA00710; Mon, 11 Sep 2000 12:17:41 -0700 (PDT)
Message-Id: <3.0.5.32.20000911121739.03877c10@scamp.eng.ascend.com>
X-Sender: igoyret@scamp.eng.ascend.com (Unverified)
X-Mailer: QUALCOMM Windows Eudora Pro Version 3.0.5 (32)
Date: Mon, 11 Sep 2000 12:17:39 -0700
To: Evan Caves <evan.caves@ericsson.com>
From: Ignacio Goyret <igoyret@ascend.com>
Subject: Re: draft-l2tpext-l2tp-mib-00.txt
Cc: Ignacio Goyret <igoyret@ascend.com>, l2tp@diameter.org
In-Reply-To: <39BD3073.8DF7014A@ericsson.com>
References: <39BD1287.A9D08E25@ericsson.com>
 <3.0.5.32.20000907100431.013614d0@scamp.eng.ascend.com>
 <3.0.5.32.20000911120524.04b37620@scamp.eng.ascend.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-l2tp@diameter.org
Precedence: bulk

At 12:20 PM 9/11/00 -0700, Evan Caves wrote:
>Ignacio Goyret wrote:
>> 
>> Just to round up the issue, is this what you have in mind for numbers and names?
>> 
>>    l2tpSessionStatsAuthMethod OBJECT-TYPE
>>         SYNTAX          INTEGER {
>>                              none(1),
>>                              text(2),
>>                              pppChap(3),
>>                              pppPap(4),
>>                              pppEap(5),
>>                              pppMsChapV1(6),        <<<
>>                              pppMsChapV2(7),        <<<
>>                              other(8),              <<<
>
>I think I prefer this one, that is add to the end of the
>enum list, instead of reordering enums at this stage of MIB
>development.
>
>evan

Ok, it sounds good to me.
Thanks for the prompt answer :-)
-Ignacio


From owner-l2tp@diameter.org  Wed Sep 13 08:58:41 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id IAA06843
	for <l2tpext-archive@odin.ietf.org>; Wed, 13 Sep 2000 08:58:39 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8D9jBB26834
	for l2tp-list; Wed, 13 Sep 2000 02:45:11 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from cvis29.marconicomms.com (cvis29.marconicomms.com [195.99.244.61])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8D9j7l26828
	for <l2tp@diameter.org>; Wed, 13 Sep 2000 02:45:07 -0700
Received: from cvis01.gpt.co.uk (unverified) by cvis29.marconicomms.com
 (Content Technologies SMTPRS 4.1.5) with ESMTP id <Tc363f43df84e9fe71416@cvis29.marconicomms.com>;
 Wed, 13 Sep 2000 10:49:30 +0100
Received: from marconicomms.com by cvis01.gpt.co.uk with SMTP
 (8.8.8+Sun/cvms-28) id KAA09543; Wed, 13 Sep 2000 10:49:27 +0100 (BST)
Received: by marconicomms.com(Lotus SMTP MTA v4.6.3  (733.2 10-16-1998))  id 80256959.0035EA52 ; Wed, 13 Sep 2000 10:48:53 +0100
X-Lotus-FromDomain: MCMAIN@MCEXT
From: "Massimiliano Piastra" <Massimiliano.Piastra@marconi.com>
To: evan.caves@ericsson.com
cc: l2tp@diameter.org
Message-ID: <80256959.00358242.00@marconicomms.com>
Date: Wed, 13 Sep 2000 11:44:27 +0200
Subject: L2TP MIB
Mime-Version: 1.0
Content-type: text/plain; charset=us-ascii
Content-Disposition: inline
Sender: owner-l2tp@diameter.org
Precedence: bulk



Hi Evan,

I hope not to bother you.

I have a doubt regarding the RowStatus field that I found on other MIBs.

Why L2TP MIB doesn't contain this field in any table ?

It will be present in the next issue of the draft ?

Thank you in advance.

Massimiliano Piastra
R & D Laboratories
Marconi Communications
Italy.




From owner-l2tp@diameter.org  Wed Sep 13 15:29:04 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id PAA16445
	for <l2tpext-archive@odin.ietf.org>; Wed, 13 Sep 2000 15:29:02 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8DGWUb27239
	for l2tp-list; Wed, 13 Sep 2000 09:32:30 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from mailsrv.acc.com (mailsrv.acc.com [129.192.64.128])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8DGWQl27233
	for <l2tp@diameter.org>; Wed, 13 Sep 2000 09:32:26 -0700
Received: from ericsson.com (calcium.dev.acc.am.ericsson.se [129.192.57.200])
	by mailsrv.acc.com (8.9.3/8.9.1) with ESMTP id JAA25485;
	Wed, 13 Sep 2000 09:38:49 -0700 (PDT)
Message-ID: <39BFADF0.5001E92F@ericsson.com>
Date: Wed, 13 Sep 2000 09:40:16 -0700
From: Evan Caves <evan.caves@ericsson.com>
X-Mailer: Mozilla 4.61 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Massimiliano Piastra <Massimiliano.Piastra@marconi.com>
CC: l2tp@diameter.org
Subject: Re: L2TP MIB
References: <80256959.00358242.00@marconicomms.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Can you be more specific about which tables do not contain RowStatus
fields?

evan
-

Massimiliano Piastra wrote:
> 
> Hi Evan,
> 
> I hope not to bother you.
> 
> I have a doubt regarding the RowStatus field that I found on other MIBs.
> 
> Why L2TP MIB doesn't contain this field in any table ?
> 
> It will be present in the next issue of the draft ?
> 
> Thank you in advance.
> 
> Massimiliano Piastra
> R & D Laboratories
> Marconi Communications
> Italy.


From owner-l2tp@diameter.org  Thu Sep 14 07:19:14 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id HAA08475
	for <l2tpext-archive@odin.ietf.org>; Thu, 14 Sep 2000 07:19:13 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8E8grQ27757
	for l2tp-list; Thu, 14 Sep 2000 01:42:53 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from omega.cisco.com (omega.cisco.com [171.69.63.141])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8E8gol27751
	for <l2tp@diameter.org>; Thu, 14 Sep 2000 01:42:50 -0700
Received: from tmima-homepc.cisco.com (tmima-dsl5.cisco.com [144.254.211.46])
	by omega.cisco.com (8.8.8-Cisco List Logging/8.8.8) with ESMTP id BAA10176;
	Thu, 14 Sep 2000 01:39:55 -0700 (PDT)
Message-Id: <4.3.1.2.20000914010737.01f5b750@omega.cisco.com>
X-Sender: tmima@omega.cisco.com
X-Mailer: QUALCOMM Windows Eudora Version 4.3.1
Date: Thu, 14 Sep 2000 01:40:08 -0700
To: Andy Valencia <vandys@zendo.com>, "W. Mark Townsley" <townsley@cisco.com>
From: Tmima Koren <tmima@cisco.com>
Subject: Re: 0 byte L2TPHC header and implied PPPMUX
Cc: brucet@cisco.com, dwing@cisco.com,
        Stephen Casner <casner@packetdesign.com>, gurdeep@MICROSOFT.com,
        palter@zev.net, l2tp@diameter.org, rem-conf@es.net
In-Reply-To: <4.3.1.2.20000907161936.029ba8e0@omega.cisco.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-l2tp@diameter.org
Precedence: bulk


Hi
Since there was no feedback from the lists about this issue, can we update the draft with the suggested changes?
(3 flags in byte 9 of L2TPHC AVP)
Thanks,
Tmima


At 06:03 PM 9/7/00 -0700, Tmima Koren wrote:

>I suggest a few minor changes to the L2TPHC draft draft-ietf-l2tpext-l2tphc-02.txt
>to improve bandwidth utilization in special cases.
>
>
>When the L2TPHC header is 1 byte long, actually only 1 bit is significant: the P (Priority) bit.
>
>If all the packets in the tunnel will have the same priority, the L2TPHC byte is redundant.
>
>We could add to the L2TPHC AVP one more byte (byte 9)
>We really need only 2 flag bits:
>
>flag 1:  value 1 means: L2TPHC byte will be omitted
>             value 0 means: L2TPHC byte is there, get P (Priority) from the L2TPHC byte
>
>flag 2:  value 1 means all packets have implied P=1
>             value 0 means all packets have implied P=0
>
>consult flag 2 only when flag 1 is 1
>
>
>If the packet in the L2TPHC tunnel is always PPPMUX, there is another useful flag:
>
>flag 3: value 1 means 'PPPMUX only'
>            value 0 means: any PPP packet
>
>
>When flags 1 and 3 are on (=1), both L2TPHC header and PPPMUX protocol id can be omitted 
>PPP packets that have a different priority or that are not PPPMUX can still be sent in the tunnel using the UDP/L2TP format, same as the L2TP control packets
>
>TCRTP (Tunneled Compressed multiplexed RTP) can benefit from these changes: the tunneling overhead will be reduced by 2 bytes.
>
>Tmima
>



From owner-l2tp@diameter.org  Thu Sep 14 16:23:14 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id QAA23481
	for <l2tpext-archive@odin.ietf.org>; Thu, 14 Sep 2000 16:23:13 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8EHEB928630
	for l2tp-list; Thu, 14 Sep 2000 10:14:11 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from omega.cisco.com (omega.cisco.com [171.69.63.141])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8EHE8l28624
	for <l2tp@diameter.org>; Thu, 14 Sep 2000 10:14:08 -0700
Received: from tmima-homepc.cisco.com (tmima-dsl5.cisco.com [144.254.211.46])
	by omega.cisco.com (8.8.8-Cisco List Logging/8.8.8) with ESMTP id KAA15944;
	Thu, 14 Sep 2000 10:14:56 -0700 (PDT)
Message-Id: <4.3.1.2.20000914093751.027ac100@omega.cisco.com>
X-Sender: tmima@omega.cisco.com
X-Mailer: QUALCOMM Windows Eudora Version 4.3.1
Date: Thu, 14 Sep 2000 10:15:15 -0700
To: Andy Valencia <vandys@zendo.com>
From: Tmima Koren <tmima@cisco.com>
Subject: Re: 0 byte L2TPHC header and implied PPPMUX 
Cc: "W. Mark Townsley" <townsley@cisco.com>, brucet@cisco.com, dwing@cisco.com,
        Stephen Casner <casner@packetdesign.com>, gurdeep@MICROSOFT.com,
        palter@zev.net, tmima@cisco.com, l2tp@diameter.org, rem-conf@es.net
In-Reply-To: <200009141544.IAA00468@vandys-pc.zendo.com>
References: <Your message of "Thu, 14 Sep 2000 01:40:08 PDT." <4.3.1.2.20000914010737.01f5b750@omega.cisco.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-l2tp@diameter.org
Precedence: bulk

At 08:44 AM 9/14/00 -0700, Andy Valencia wrote:
>[Tmima Koren <tmima@cisco.com> writes:]
>
> >>flag 2:  value 1 means all packets have implied P=1
> >>             value 0 means all packets have implied P=0
>
>I'm concerned with this one.  Since L2TPHC is basically useless except for
>payload, what you've achieved with this flag is to ensure that your control
>traffic will be interrupted if you *ever* permit all payload traffic to be
>given preference by choosing P=1.  My experience is that this will make your
>link unworkable.  And no, using the UDP side doesn't help, since priority is
>(and should be) global to the PPP connection.
>
>I would suggest that when omitting the L2TPHC header entirely, that P=0
>always be the mode of operation.  This shouldn't affect your TCRTP mode of
>operation, because the only thing aside from those payload packets would be
>control packets which should be given priority anyway (and they can go via
>UDP).

In the typical case of a TCRTP tunnel both ends of the tunnel are tied to virtual PPP interfaces.
RTP packets arrive to one side, compressed and tunneled to the other side where they are decompressed and routed to their destination.
Those RTP packets carry voice and may have certain priority. Those packets should regain their priority when they leave the tunnel.
Can this be done with P=0?


> >>When flags 1 and 3 are on (=1), both L2TPHC header and PPPMUX protocol id can
> > be omitted 
>
>How about A/C fields?  Isn't it a little weird that a field in the middle of
>the PPP packet is coming and going based on this, but the bytes in front
>depend on PPP negotiation?
>
>Also, why hard-wire it to PPPMUX?  Why not include the implicit protocol
>value, so that these savings could apply to IP-only connections which
>weren't using PPPMUX and/or TCRTP?

I thought about this option. The difference is that the PPPMux protocol ID at the head of the packet really acts as a flag to indicate that there is more then one packet in this packet..
but each multiplex carries its own PPP protocol ID.
If you think it's ok to generalize for any protocol, I support that.

Tmima


> >>PPP packets that have a different priority or that are not PPPMUX can still b
> >e sent in the tunnel using the UDP/L2TP format, same as the L2TP control packe
> >ts
> >>TCRTP (Tunneled Compressed multiplexed RTP) can benefit from these changes: t
> >he tunneling overhead will be reduced by 2 bytes.
>
>I guess I'm OK with a 0-byte L2TPHC header option (subject to a nod from the
>chair).  I'd sure like to hear from some people (not just infer something
>from silence) on the PPP header aspect.
>
>Andy Valencia



From owner-l2tp@diameter.org  Fri Sep 15 16:02:12 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id QAA24464
	for <l2tpext-archive@odin.ietf.org>; Fri, 15 Sep 2000 16:02:11 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8FGk2E29507
	for l2tp-list; Fri, 15 Sep 2000 09:46:02 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from mailsrv.acc.com (mailsrv.acc.com [129.192.64.128])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8FGiil29501
	for <l2tp@diameter.org>; Fri, 15 Sep 2000 09:44:44 -0700
Received: from ericsson.com (calcium.dev.acc.am.ericsson.se [129.192.57.200])
	by mailsrv.acc.com (8.9.3/8.9.1) with ESMTP id JAA06272
	for <l2tp@diameter.org>; Fri, 15 Sep 2000 09:54:03 -0700 (PDT)
Message-ID: <39C25484.59A3B464@ericsson.com>
Date: Fri, 15 Sep 2000 09:55:32 -0700
From: Evan Caves <evan.caves@ericsson.com>
X-Mailer: Mozilla 4.61 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: l2tp@diameter.org
Subject: [Fwd: L2TP Diffserv Draft]
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-l2tp@diameter.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

I sent this out several weeks ago but got no response. This time can
anyone respond (not just the authors :-).

evan
-

Evan Caves wrote:
> 
> I have a number of question regarding this draft which I hope one
> of the authors can comment on.
> 
> 1. Is the DSCP AVP a six or eight bit value?
> 
> 2. What DSCP value is used for tunnel control messages? If the
>    session DSCP has a 'higher priority' than the tunnels then
>    payload packets may arrive before session setup completion.
> 
> 3. Why is this AVP restricted to just sessions? One could conceive
>    bringing up a tunnel for each PHB and forwarding all tunnel
>    traffic (control and payload) with one of the PHB class DSCPs.
> 
> 4. This draft makes a big assumption that the tunnel traverses a
>    single DS domain. If the tunnel traverses multiple DS domains
>    the DSCP to PHB mappings could be different (this may mean
>    that each tunnel endpoint uses different DSCP marks that get
>    translated at some DS boundary). How is this resolved?
> 
> 5. Related to 4 it would seem that it may be more appropriate that
>    the tunnel initiator ask for a PHB rather than a DSCP and then
>    let each endpoint decide how that PHB maps to a DSCP.
> 
> evan
> -


From owner-l2tp@diameter.org  Fri Sep 15 22:32:48 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id WAA28888
	for <l2tpext-archive@odin.ietf.org>; Fri, 15 Sep 2000 22:32:42 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8FMiTR29713
	for l2tp-list; Fri, 15 Sep 2000 15:44:29 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from tcb.net (tcb.net [205.168.100.1])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8FMiPl29707
	for <l2tp@diameter.org>; Fri, 15 Sep 2000 15:44:25 -0700
Received: from sofos.tcb.net (sofos.tcb.net [127.0.0.1])
	by tcb.net (8.9.3/8.9.3) with ESMTP id QAA31968;
	Fri, 15 Sep 2000 16:48:44 -0600
Message-Id: <200009152248.QAA31968@tcb.net>
X-Mailer: exmh version 2.0.3
To: Evan Caves <evan.caves@ericsson.com>
cc: l2tp@diameter.org
From: Danny McPherson <danny@tcb.net>
Reply-To: danny@tcb.net
Subject: Re: [Fwd: L2TP Diffserv Draft] 
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Fri, 15 Sep 2000 16:48:44 -0600
Sender: owner-l2tp@diameter.org
Precedence: bulk


> I sent this out several weeks ago but got no response. This time can
> anyone respond (not just the authors :-).

I just picked it up from Ken & Pat so I'll give it a shot.

> > 1. Is the DSCP AVP a six or eight bit value?

The format as currently specified is as follows:

       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
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
      |1|1|0|0|        Length         |              43               |
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
      |                1              |        DSCP  Value            |
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

This means that the total length is 8 octets, with 2 octets representing the 
DSCP value.  The value will remain at least one octet, as two octets seem to 
enable little more than tidy packets encoding pictures.


> > 2. What DSCP value is used for tunnel control messages? If the
> >    session DSCP has a 'higher priority' than the tunnels then
> >    payload packets may arrive before session setup completion.

The next version of the draft, which is hopefully no more than a couple weeks 
out, will specify the behavior in this scenario.

> > 3. Why is this AVP restricted to just sessions? One could conceive
> >    bringing up a tunnel for each PHB and forwarding all tunnel
> >    traffic (control and payload) with one of the PHB class DSCPs.

Agreed.  And this is something that we were just discussing.  I'll ensure that 
the upcoming draft provides provisions for this as well (comments welcome!).

> > 4. This draft makes a big assumption that the tunnel traverses a
> >    single DS domain. If the tunnel traverses multiple DS domains
> >    the DSCP to PHB mappings could be different (this may mean
> >    that each tunnel endpoint uses different DSCP marks that get
> >    translated at some DS boundary). How is this resolved?

Multiple DS domains is of course is a very well-know and interesting problem, 
and again, the next version of the draft will take a stab at addressing this 
as well.  We plan to do our best to be in alignment with similar work being 
done in the diffserv WG.  (Again, comments welcome!).

> > 5. Related to 4 it would seem that it may be more appropriate that
> >    the tunnel initiator ask for a PHB rather than a DSCP and then
> >    let each endpoint decide how that PHB maps to a DSCP.

Though you still have lots of problems with this model as well when traverses 
multiple DS domains.  Again, we'll take a stab at it and work with the DS 
folks to ensure that we're aligned.  Comments here are welcome!

Thanks for the input, hope to have something useful within a couple of weeks 
at most.

-danny






From owner-l2tp@diameter.org  Thu Sep 21 12:39:29 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id MAA09618
	for <l2tpext-archive@odin.ietf.org>; Thu, 21 Sep 2000 12:39:28 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8LEi3608437
	for l2tp-list; Thu, 21 Sep 2000 07:44:03 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8LEhwl08431
	for <l2tp@l2tp.net>; Thu, 21 Sep 2000 07:43:59 -0700
Received: from cvis29.marconicomms.com (cvis29.marconicomms.com [195.99.244.61]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id HAA06885 for <l2tp@ipsec.org>; Thu, 21 Sep 2000 07:49:14 -0700 (PDT)
Received: from cvis01.gpt.co.uk (unverified) by cvis29.marconicomms.com
 (Content Technologies SMTPRS 4.1.5) with ESMTP id <Tc363f43dde4eca2bd8e0@cvis29.marconicomms.com> for <l2tp@ipsec.org>;
 Thu, 21 Sep 2000 15:48:40 +0100
Received: from marconicomms.com by cvis01.gpt.co.uk with SMTP
 (8.8.8+Sun/cvms-28) id PAA17839; Thu, 21 Sep 2000 15:48:39 +0100 (BST)
Received: by marconicomms.com(Lotus SMTP MTA v4.6.3  (733.2 10-16-1998))  id 80256961.00514A6C ; Thu, 21 Sep 2000 15:47:54 +0100
X-Lotus-FromDomain: MCMAIN@MCEXT
From: "Fabio Poggi" <Fabio.Poggi@marconi.com>
To: l2tp@ipsec.org
Message-ID: <80256961.0051452F.00@marconicomms.com>
Date: Thu, 21 Sep 2000 16:48:13 +0200
Subject: L2TP question
Mime-Version: 1.0
Content-type: text/plain; charset=us-ascii
Content-Disposition: inline
Sender: owner-l2tp@diameter.org
Precedence: bulk



Hi all

I have a question about the relationships between PPP and L2TP in the LNS
system.
It seems to me that this argument isn't treat of in none documents (is it true,
first of all ?).

In particular, what is the behaviour of the LNS system when an incoming call is
received from the LAC.

   A PPP peer (a remote dial-up user) starts to send LCP and authentication
   packets to a LAC.
   After receiving an authentication request message, the LAC sets up a new L2TP
    tunnel/session by the FQDN "user@domain" (in order to understand the
   destination LNS).
   After the tunnel/session set up, the PPP remote user and the LNS will
   exchange PPP packets directly (the LAC forwards them transparently).
   The PPP state machine of the LNS is not started yet (it's still in its "PPP
   dead phase"), while the PPP user is in its authentication phase.

What happens at this time ?
(I've not found the answer in any documents !)

I supposed (but, is it correct ??) that the LNS, after tunnel/session
establishment, "must" be forced to send an LCP configure request to its PPP peer
 (the user)
in order to set up a PPP connection (the previous PPP connection of the user
will be restarted).
Is this true?
Are there other methods?

If the LNS does't make any actions, the remote user will retry to send an
authentication request message again,
this message will arrive directly to the LNS (after the tunnel/session
establishment, the LAC is transparently) and, here, it will silent discard
(the LNS is still in its "PPP dead phase").

Could anyone help me?

Thanks a lot in advance.

Fabio Poggi




From owner-l2tp@diameter.org  Thu Sep 21 14:06:49 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id OAA11379
	for <l2tpext-archive@odin.ietf.org>; Thu, 21 Sep 2000 14:06:48 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8LFBBW08468
	for l2tp-list; Thu, 21 Sep 2000 08:11:11 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8LFB8l08462
	for <l2tp@l2tp.net>; Thu, 21 Sep 2000 08:11:08 -0700
Received: from cvis29.marconicomms.com (cvis29.marconicomms.com [195.99.244.61]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id IAA06902 for <l2tp@ipsec.org>; Thu, 21 Sep 2000 08:16:33 -0700 (PDT)
Received: from cvis01.gpt.co.uk (unverified) by cvis29.marconicomms.com
 (Content Technologies SMTPRS 4.1.5) with ESMTP id <Tc363f43dde4eca449814@cvis29.marconicomms.com> for <l2tp@ipsec.org>;
 Thu, 21 Sep 2000 16:15:42 +0100
Received: from marconicomms.com by cvis01.gpt.co.uk with SMTP
 (8.8.8+Sun/cvms-28) id QAA28329; Thu, 21 Sep 2000 16:15:40 +0100 (BST)
Received: by marconicomms.com(Lotus SMTP MTA v4.6.3  (733.2 10-16-1998))  id 80256961.0053B392 ; Thu, 21 Sep 2000 16:14:14 +0100
X-Lotus-FromDomain: MCMAIN@MCEXT
From: "Fabio Poggi" <Fabio.Poggi@marconi.com>
To: l2tp@ipsec.org
Message-ID: <80256961.0053ADC0.00@marconicomms.com>
Date: Thu, 21 Sep 2000 17:14:09 +0200
Subject: L2TP transport
Mime-Version: 1.0
Content-type: text/plain; charset=us-ascii
Content-Disposition: inline
Sender: owner-l2tp@diameter.org
Precedence: bulk



Hi all,

I've just finished to read your document <draft-ietf-l2tpext-l2tp-mib-00.txt>.

In the l2tpTunnelConfigTable there is an entry reporting the value for
underlying transport media for L2TP.
This value forecast only 'udpip' and 'frameRelay', other than 'other' and 'none'
value.

Since the L2TP can have ATM as its underlying transport layer
(<draft-ietf-l2tpext-atmext-03.txt> is ready for last call), is it possible to
modify the draft as follow to reflect this option in transport ?

OLD VERSION:
   l2tpTunnelConfigTransport OBJECT-TYPE
           SYNTAX          INTEGER {
                              other(1),
                              none(2),
                              udpIp(3),
                              frameRelay(4)
                           }

MODIFIED VERSION:
   l2tpTunnelConfigTransport OBJECT-TYPE
           SYNTAX          INTEGER {
                              other(1),
                              none(2),
                              udpIp(3),
                              frameRelay(4)
                              atm(5)
                           }

What do you think about ?

Thanks in advance for any kind of response.

Fabio





From owner-l2tp@diameter.org  Thu Sep 21 15:37:24 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id PAA13000
	for <l2tpext-archive@odin.ietf.org>; Thu, 21 Sep 2000 15:37:22 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8LGPtQ08537
	for l2tp-list; Thu, 21 Sep 2000 09:25:55 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8LGPol08531
	for <l2tp@l2tp.net>; Thu, 21 Sep 2000 09:25:51 -0700
Received: from lukla.Sun.COM (lukla.Sun.COM [192.18.98.31]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id JAA06993 for <l2tp@ipsec.org>; Thu, 21 Sep 2000 09:31:15 -0700 (PDT)
Received: from engmail2.Eng.Sun.COM ([129.146.1.25])
	by lukla.Sun.COM (8.9.3+Sun/8.9.3) with ESMTP id KAA02031;
	Thu, 21 Sep 2000 10:30:36 -0600 (MDT)
Received: from nasnfs.eng.sun.com (nasnfs.Eng.Sun.COM [10.6.84.20])
	by engmail2.Eng.Sun.COM (8.9.3+Sun/8.9.3/ENSMAIL,v1.7) with ESMTP id JAA22286;
	Thu, 21 Sep 2000 09:30:35 -0700 (PDT)
Received: from mordor (mordor [129.146.120.122])
	by nasnfs.eng.sun.com (8.9.3+Sun/8.9.1) with SMTP id JAA26980;
	Thu, 21 Sep 2000 09:30:33 -0700 (PDT)
Date: Thu, 21 Sep 2000 09:28:38 -0700 (PDT)
From: "pcalhoun@eng.sun.com" <Pat.Calhoun@eng.sun.com>
Reply-To: "pcalhoun@eng.sun.com" <Pat.Calhoun@eng.sun.com>
Subject: Re: L2TP transport
To: Fabio Poggi <Fabio.Poggi@marconi.com>
cc: l2tp@ipsec.org
In-Reply-To: "Your message with ID" <80256961.0053ADC0.00@marconicomms.com>
Message-ID: <Roam.SIMC.2.0.6.969553718.23805.pcalhoun@nasnfs.eng>
Content-Type: text
X-Sun-Text-Type: ascii
Sender: owner-l2tp@diameter.org
Precedence: bulk

Sounds reasonable (of course, I would like to get this thing out soon, though
:)

PatC
> 
> 
> Hi all,
> 
> I've just finished to read your document
> <draft-ietf-l2tpext-l2tp-mib-00.txt>.
> 
> In the l2tpTunnelConfigTable there is an entry reporting the value for
> underlying transport media for L2TP.
> This value forecast only 'udpip' and 'frameRelay', other than 'other' and
> 'none' value.
> 
> Since the L2TP can have ATM as its underlying transport layer
> (<draft-ietf-l2tpext-atmext-03.txt> is ready for last call), is it possible
> to modify the draft as follow to reflect this option in transport ?
> 
> OLD VERSION:
>    l2tpTunnelConfigTransport OBJECT-TYPE
>            SYNTAX          INTEGER {
>                               other(1),
>                               none(2),
>                               udpIp(3),
>                               frameRelay(4)
>                            }
> 
> MODIFIED VERSION:
>    l2tpTunnelConfigTransport OBJECT-TYPE
>            SYNTAX          INTEGER {
>                               other(1),
>                               none(2),
>                               udpIp(3),
>                               frameRelay(4)
>                               atm(5)
>                            }
> 
> What do you think about ?
> 
> Thanks in advance for any kind of response.
> 
> Fabio
> 
> 
> 



From owner-l2tp@diameter.org  Fri Sep 22 05:40:54 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id FAA04557
	for <l2tpext-archive@odin.ietf.org>; Fri, 22 Sep 2000 05:40:54 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8M6r3Q09124
	for l2tp-list; Thu, 21 Sep 2000 23:53:03 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8M6qwl09118
	for <l2tp@l2tp.net>; Thu, 21 Sep 2000 23:52:59 -0700
Received: from mail3.belgacom.be (mail3.belgacom.be [195.13.11.162]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id XAA07801 for <l2tp@ipsec.org>; Thu, 21 Sep 2000 23:58:25 -0700 (PDT)
From: paolo.crivellari@belgacom.be
Received: from exhub01.bc ([45.34.4.231])
	by mail3.belgacom.be  with SMTP id IAA01860;
	Fri, 22 Sep 2000 08:53:03 +0200
Received: from 127.0.0.1 by exhub01.bc (InterScan E-Mail VirusWall NT); Fri, 22 Sep 2000 08:57:41 +0200 (W. Europe Daylight Time)
Received: by exhub01.bc with Internet Mail Service (5.5.2650.21)
	id <SG12HWVN>; Fri, 22 Sep 2000 08:57:41 +0200
Message-ID: <95C94B2F0094D21180710008C75DD6A4046A2A1C@apl541.bc>
To: Fabio.Poggi@marconi.com, l2tp@ipsec.org
Subject: RE: L2TP question
Date: Fri, 22 Sep 2000 08:57:39 +0200
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: text/plain;
	charset="ISO-8859-1"
Sender: owner-l2tp@diameter.org
Precedence: bulk

You have several options:
- initialize ppp stack in the lns with proxy infos from iccn 
- restart lcp 
- restart authentication (if requested)
If possible make different options configurable
		Paolo
-----Original Message-----
From: Fabio Poggi [mailto:Fabio.Poggi@marconi.com]
Sent: Thursday 21 September 2000 16:48
To: l2tp@ipsec.org
Subject: L2TP question




Hi all

I have a question about the relationships between PPP and L2TP in the LNS
system.
It seems to me that this argument isn't treat of in none documents (is it
true,
first of all ?).

In particular, what is the behaviour of the LNS system when an incoming call
is
received from the LAC.

   A PPP peer (a remote dial-up user) starts to send LCP and authentication
   packets to a LAC.
   After receiving an authentication request message, the LAC sets up a new
L2TP
    tunnel/session by the FQDN "user@domain" (in order to understand the
   destination LNS).
   After the tunnel/session set up, the PPP remote user and the LNS will
   exchange PPP packets directly (the LAC forwards them transparently).
   The PPP state machine of the LNS is not started yet (it's still in its
"PPP
   dead phase"), while the PPP user is in its authentication phase.

What happens at this time ?
(I've not found the answer in any documents !)

I supposed (but, is it correct ??) that the LNS, after tunnel/session
establishment, "must" be forced to send an LCP configure request to its PPP
peer
 (the user)
in order to set up a PPP connection (the previous PPP connection of the user
will be restarted).
Is this true?
Are there other methods?

If the LNS does't make any actions, the remote user will retry to send an
authentication request message again,
this message will arrive directly to the LNS (after the tunnel/session
establishment, the LAC is transparently) and, here, it will silent discard
(the LNS is still in its "PPP dead phase").

Could anyone help me?

Thanks a lot in advance.

Fabio Poggi


**** DISCLAIMER **** 
"This e-mail and any attachments thereto may contain information 
which is confidential and/or protected by intellectual property 
rights and are intended for the sole use of the recipient(s) named above. 
Any use of the information contained herein (including, but not limited to, 
total or partial reproduction, communication or distribution in any form) 
by persons other than the designated recipient(s) is prohibited. 
If you have received this e-mail in error, please notify the sender either 
by telephone or by e-mail and delete the material from any computer. 
Thank you for your cooperation."



From owner-l2tp@diameter.org  Fri Sep 22 10:43:07 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id KAA15768
	for <l2tpext-archive@odin.ietf.org>; Fri, 22 Sep 2000 10:43:06 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8MBk5Q09469
	for l2tp-list; Fri, 22 Sep 2000 04:46:05 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8MBk1l09463
	for <l2tp@l2tp.net>; Fri, 22 Sep 2000 04:46:01 -0700
Received: from cvis29.marconicomms.com (cvis29.marconicomms.com [195.99.244.61]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id EAA09235 for <l2tp@ipsec.org>; Fri, 22 Sep 2000 04:51:30 -0700 (PDT)
Received: from cvis01.gpt.co.uk (unverified) by cvis29.marconicomms.com
 (Content Technologies SMTPRS 4.1.5) with ESMTP id <Tc363f43ddd4eceaf81cd@cvis29.marconicomms.com>;
 Fri, 22 Sep 2000 12:50:57 +0100
Received: from marconicomms.com by cvis01.gpt.co.uk with SMTP
 (8.8.8+Sun/cvms-28) id MAA19770; Fri, 22 Sep 2000 12:50:55 +0100 (BST)
Received: by marconicomms.com(Lotus SMTP MTA v4.6.3  (733.2 10-16-1998))  id 80256962.0040ED4B ; Fri, 22 Sep 2000 12:49:10 +0100
X-Lotus-FromDomain: MCMAIN@MCEXT
From: "Fabio Poggi" <Fabio.Poggi@marconi.com>
To: l2tp@ipsec.org
cc: Paolo.crivellari@belgacom.be
Message-ID: <80256962.0040E81F.00@marconicomms.com>
Date: Fri, 22 Sep 2000 13:49:34 +0200
Subject: RE: L2TP question
Mime-Version: 1.0
Content-type: text/plain; charset=us-ascii
Content-Disposition: inline
Sender: owner-l2tp@diameter.org
Precedence: bulk



Hi Paolo,
the first option you suggest involves some optional L2TP AVPs (Proxy LCP AVPs
are not defined as mandatory either in RFC or in
draft-ietf-l2tpext-l2tpbis-00.txt), so could be reasonable that some LAC or LNS
doesn't implement these AVPs (and anyway LAC and/or LNS are not forced to do so
by none RFC).

The second and third options are not addressed in neither in RFC nor in
draft-ietf-l2tpext-l2tpbis-00.txt; these could be implementations caracterizing
some LAC, but not all LAC !!
It is reasonable to think that a LAC doesn't implement any of these mechanism
(since are not forced to do so by the L2TP RFC), wiht the conseguence of not set
up any kind of effective communication.

In the scenario I've explained in my previous mail, there is none indication
from L2TP RFC standard in order to resolve that problem.

The L2TP RFC seems to lack of a good explanation about the problem of L2TP-PPP
interaction and this could generate some interoperability problem.
It seems to me that a standardization, in an L2TP RFC document (for exemple in
draft-ietf-l2tpext-l2tpbis-00.txt ??), of the interaction between L2TP state
machine and the PPP automaton, either in the LNS or in the LAC or in both, is
necessary.

What do you think about ?

Thanks a lot.
Fabio Poggi









paolo.crivellari@belgacom.be on 22/09/2000 08.57.39
                                                                                
                                                                                
                                                                                


                                                              
                                                              
                                                              
 To:      Fabio Poggi/MAIN/MC1@MCMAIN, l2tp@ipsec.org         
                                                              
 cc:                                                          
                                                              
                                                              
                                                              
 Subject: RE: L2TP question                                   
                                                              








You have several options:
- initialize ppp stack in the lns with proxy infos from iccn
- restart lcp
- restart authentication (if requested)
If possible make different options configurable
          Paolo
-----Original Message-----
From: Fabio Poggi [mailto:Fabio.Poggi@marconi.com]
Sent: Thursday 21 September 2000 16:48
To: l2tp@ipsec.org
Subject: L2TP question




Hi all

I have a question about the relationships between PPP and L2TP in the LNS
system.
It seems to me that this argument isn't treat of in none documents (is it
true,
first of all ?).

In particular, what is the behaviour of the LNS system when an incoming call
is
received from the LAC.

   A PPP peer (a remote dial-up user) starts to send LCP and authentication
   packets to a LAC.
   After receiving an authentication request message, the LAC sets up a new
L2TP
    tunnel/session by the FQDN "user@domain" (in order to understand the
   destination LNS).
   After the tunnel/session set up, the PPP remote user and the LNS will
   exchange PPP packets directly (the LAC forwards them transparently).
   The PPP state machine of the LNS is not started yet (it's still in its
"PPP
   dead phase"), while the PPP user is in its authentication phase.

What happens at this time ?
(I've not found the answer in any documents !)

I supposed (but, is it correct ??) that the LNS, after tunnel/session
establishment, "must" be forced to send an LCP configure request to its PPP
peer
 (the user)
in order to set up a PPP connection (the previous PPP connection of the user
will be restarted).
Is this true?
Are there other methods?

If the LNS does't make any actions, the remote user will retry to send an
authentication request message again,
this message will arrive directly to the LNS (after the tunnel/session
establishment, the LAC is transparently) and, here, it will silent discard
(the LNS is still in its "PPP dead phase").

Could anyone help me?

Thanks a lot in advance.

Fabio Poggi


**** DISCLAIMER ****
"This e-mail and any attachments thereto may contain information
which is confidential and/or protected by intellectual property
rights and are intended for the sole use of the recipient(s) named above.
Any use of the information contained herein (including, but not limited to,
total or partial reproduction, communication or distribution in any form)
by persons other than the designated recipient(s) is prohibited.
If you have received this e-mail in error, please notify the sender either
by telephone or by e-mail and delete the material from any computer.
Thank you for your cooperation."






From owner-l2tp@diameter.org  Fri Sep 22 10:46:08 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id KAA15861
	for <l2tpext-archive@odin.ietf.org>; Fri, 22 Sep 2000 10:46:07 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8MCGuK09497
	for l2tp-list; Fri, 22 Sep 2000 05:16:56 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from ns1.sailpix.com (ns1.sailpix.com [155.53.1.250])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8MCGrl09491
	for <l2tp@l2tp.net>; Fri, 22 Sep 2000 05:16:53 -0700
Received: from mail3.belgacom.be (mail3.belgacom.be [195.13.11.162]) by ns1.sailpix.com (8.8.8/8.6.9) with ESMTP id FAA09270 for <l2tp@ipsec.org>; Fri, 22 Sep 2000 05:22:21 -0700 (PDT)
From: paolo.crivellari@belgacom.be
Received: from exhub01.bc ([45.34.4.231])
	by mail3.belgacom.be  with SMTP id OAA28604;
	Fri, 22 Sep 2000 14:17:27 +0200
Received: from 127.0.0.1 by exhub01.bc (InterScan E-Mail VirusWall NT); Fri, 22 Sep 2000 14:22:06 +0200 (W. Europe Daylight Time)
Received: by exhub01.bc with Internet Mail Service (5.5.2650.21)
	id <SG122FCR>; Fri, 22 Sep 2000 14:22:05 +0200
Message-ID: <95C94B2F0094D21180710008C75DD6A4046A2A1F@apl541.bc>
To: Fabio.Poggi@marconi.com, l2tp@ipsec.org
Cc: paolo.crivellari@belgacom.be
Subject: RE: L2TP question
Date: Fri, 22 Sep 2000 14:22:06 +0200
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: text/plain;
	charset="ISO-8859-1"
Sender: owner-l2tp@diameter.org
Precedence: bulk

Find my answers below

-----Original Message-----
From: Fabio Poggi [mailto:Fabio.Poggi@marconi.com]
Sent: Friday 22 September 2000 13:50
To: l2tp@ipsec.org
Cc: Paolo.crivellari@belgacom.be
Subject: RE: L2TP question




Hi Paolo,
the first option you suggest involves some optional L2TP AVPs (Proxy LCP
AVPs
are not defined as mandatory either in RFC or in
draft-ietf-l2tpext-l2tpbis-00.txt), so could be reasonable that some LAC or
LNS
doesn't implement these AVPs (and anyway LAC and/or LNS are not forced to do
so
by none RFC).

The second and third options are not addressed in neither in RFC nor in
draft-ietf-l2tpext-l2tpbis-00.txt; these could be implementations
caracterizing
some LAC, but not all LAC !!

I suppose here you mean LNS

It is reasonable to think that a LAC doesn't implement any of these
mechanism

Again LNS

(since are not forced to do so by the L2TP RFC), wiht the conseguence of not
set
up any kind of effective communication.

In the scenario I've explained in my previous mail, there is none indication
from L2TP RFC standard in order to resolve that problem.

Read section 9.5 of rfc2661, it looks pretty clear to me

The L2TP RFC seems to lack of a good explanation about the problem of
L2TP-PPP
interaction and this could generate some interoperability problem.
It seems to me that a standardization, in an L2TP RFC document (for exemple
in
draft-ietf-l2tpext-l2tpbis-00.txt ??), of the interaction between L2TP state
machine and the PPP automaton, either in the LNS or in the LAC or in both,
is
necessary.

What do you think about ?

Thanks a lot.
Fabio Poggi









paolo.crivellari@belgacom.be on 22/09/2000 08.57.39
 

 

 



                                                              
                                                              
                                                              
 To:      Fabio Poggi/MAIN/MC1@MCMAIN, l2tp@ipsec.org         
                                                              
 cc:                                                          
                                                              
                                                              
                                                              
 Subject: RE: L2TP question                                   
                                                              








You have several options:
- initialize ppp stack in the lns with proxy infos from iccn
- restart lcp
- restart authentication (if requested)
If possible make different options configurable
          Paolo
-----Original Message-----
From: Fabio Poggi [mailto:Fabio.Poggi@marconi.com]
Sent: Thursday 21 September 2000 16:48
To: l2tp@ipsec.org
Subject: L2TP question




Hi all

I have a question about the relationships between PPP and L2TP in the LNS
system.
It seems to me that this argument isn't treat of in none documents (is it
true,
first of all ?).

In particular, what is the behaviour of the LNS system when an incoming call
is
received from the LAC.

   A PPP peer (a remote dial-up user) starts to send LCP and authentication
   packets to a LAC.
   After receiving an authentication request message, the LAC sets up a new
L2TP
    tunnel/session by the FQDN "user@domain" (in order to understand the
   destination LNS).
   After the tunnel/session set up, the PPP remote user and the LNS will
   exchange PPP packets directly (the LAC forwards them transparently).
   The PPP state machine of the LNS is not started yet (it's still in its
"PPP
   dead phase"), while the PPP user is in its authentication phase.

What happens at this time ?
(I've not found the answer in any documents !)

I supposed (but, is it correct ??) that the LNS, after tunnel/session
establishment, "must" be forced to send an LCP configure request to its PPP
peer
 (the user)
in order to set up a PPP connection (the previous PPP connection of the user
will be restarted).
Is this true?
Are there other methods?

If the LNS does't make any actions, the remote user will retry to send an
authentication request message again,
this message will arrive directly to the LNS (after the tunnel/session
establishment, the LAC is transparently) and, here, it will silent discard
(the LNS is still in its "PPP dead phase").

Could anyone help me?

Thanks a lot in advance.

Fabio Poggi


**** DISCLAIMER ****
"This e-mail and any attachments thereto may contain information
which is confidential and/or protected by intellectual property
rights and are intended for the sole use of the recipient(s) named above.
Any use of the information contained herein (including, but not limited to,
total or partial reproduction, communication or distribution in any form)
by persons other than the designated recipient(s) is prohibited.
If you have received this e-mail in error, please notify the sender either
by telephone or by e-mail and delete the material from any computer.
Thank you for your cooperation."




**** DISCLAIMER **** 
"This e-mail and any attachments thereto may contain information 
which is confidential and/or protected by intellectual property 
rights and are intended for the sole use of the recipient(s) named above. 
Any use of the information contained herein (including, but not limited to, 
total or partial reproduction, communication or distribution in any form) 
by persons other than the designated recipient(s) is prohibited. 
If you have received this e-mail in error, please notify the sender either 
by telephone or by e-mail and delete the material from any computer. 
Thank you for your cooperation."



From owner-l2tp@diameter.org  Wed Sep 27 07:33:11 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id HAA08784
	for <l2tpext-archive@odin.ietf.org>; Wed, 27 Sep 2000 07:33:06 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8R7xWo19559
	for l2tp-list; Wed, 27 Sep 2000 00:59:32 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from brahma.roc.com ([202.54.67.218])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8R7xOl19553
	for <l2tp@diameter.org>; Wed, 27 Sep 2000 00:59:25 -0700
Received: from sridhar.roc.com ([172.16.5.52])
	by brahma.roc.com (8.8.5/8.8.5) with SMTP id NAA29507
	for <l2tp@diameter.org>; Wed, 27 Sep 2000 13:38:04 +0530 (IST)
Message-Id: <200009270808.NAA29507@brahma.roc.com>
X-Sender: sridhar@172.16.5.5
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.0
Date: Wed, 27 Sep 2000 13:42:00 +0530
To: l2tp@diameter.org
From: Sridhar Malarothu <sridhar@trinc.com>
Subject: Re: AW: L2TP Tunnel Authentication with RADIUS
In-Reply-To: <20000904.101315.74751693.mshindo@mshindo.net>
References: <D9F2B9CD7BD5D21196BC0800060D9ED601EFBF3E@vies186a.sie.siemens.at>
 <D9F2B9CD7BD5D21196BC0800060D9ED601EFBF3E@vies186a.sie.siemens.at>
Mime-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
X-MIME-Autoconverted: from quoted-printable to 8bit by charizard.diameter.org id e8R7xTl19554
Sender: owner-l2tp@diameter.org
Precedence: bulk
X-MIME-Autoconverted: from 8bit to quoted-printable by charizard.diameter.org id e8R7xWo19559
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by ietf.org id HAA08784

Hi All,

   Since the same secret which is used for Tunnel Authentication is also
used in hiding the AVP's, I think its not simple for LNS to get the Tunnel
password(secret) from RADIUS(based on peer's host name in the SCCRQ msg),
if both the tunnel authentication and AVP's hiding are used at the same time.
   Even though the Hostname AVP in the SCCRQ msg is not hidden, the other
AVP's like Assigned Tunnel ID etc, can be hidden. 
   Since Hostname is required to get the Tunnel Password from Radius, we
may just extract the Hostname AVP from SCCRQ and differ the decoding of rest
of the message till the RADIUS response is received. But we cannot ZLB ack the
SCCRQ message, since the Assigned Tunnel ID in SCCRQ may be hidden.
       

Best Regards
Sridhar
   

At 10:13 AM 9/4/00 +0900, you wrote:
>All,
>
>I do agree with Walter's comment. 
>
>Tunnel authentication and user authentication (associated with an
>incoming call) are totally two different things, and more importantly,
>the establishment of a tunnel does not necessarily happen at the same
>time as the session (ie. PPP) establishment.
>
>>From this point of view, a straight-forward LAC implementation should
>work in the following way:
>
> 1) LAC gets to know the user name (via PAP/CHAP or whatever) and
>    queries to RADIUS to authenticate this user and obtain information
>    associated with it. If this user is an L2TP user, then the
>    RADIUS entry for this user would contain a Tunnel-Server-Endpoint
>    attribute associated with this user.
>
> 2) If a tunnel for the LNS specified by Tunnel-Sever-Endpoint hasn't 
>    already existed and if the LAC is configured to do tunnel
>    authentication, then a LAC queries to RADIUS again to obtain a tunnel
>    passowrd. In this case, the RADIUS entry has a LNS name (identified
>    by a Tunnel-Server-Endpoint in the previous query) as a user
>    name, and should include a Tuennel-Password attribute in its reply
>    item.
>
>Please note that this sequence involves two RADIUS queries, one for
>user authentication and another for tunnel authentication. To make it
>work in more efficent way, an optimization is possible by piggybacking
>a Tunnel-Password attribute on the first RADIUS query for the user
>authentication. Since there's a one-to-one relationship between
>Tunnel-Server-Endpoint and Tunnel-Password, this optimization works
>without any problem.
>
>Howerver, on the LNS side, as Walter pointed out, this optimization
>can not be applied and two RADIUS queries are required. That is, when
>the LNS receives an SCCRQ from the LAC (along with an L2TP's CHALLENGE
>attribute), the LNS first queries to RADIUS about the tunnel password
>associated with a name specified in the L2TP's HOSTNAME attribute (the
>use of HOSTNAME attribute for this purpose is not explicitly defined
>in the RFC but it is a good choice, I believe). After the tunnel
>authentication is finished and the tunnel is successfully established,
>the ICRQ follws and then user authentication happens through, again,
>RADIUS.
>
>Regards,
>
>From: Klausberger Walter <walter.klausberger@siemens.at>
>Subject: AW: L2TP Tunnel Authentication with RADIUS
>Date: Fri, 1 Sep 2000 14:06:52 +0200 
>
>> Well, we will do so. I think it depends on which side of the tunnel you
>> are...
>> 
>> In LAC case you can use the tunnel password to setup an appropriate tunnel
>> to an LNS. During user authentication (with PPP PAP, CHAP,...) you do an
>> Access-Request and you get the tunnel password in return. That's ok, so
>> far...
>> 
>> I think your problem is really on the LNS side. A tunnel comes in and there
>> may be a tunnel password stored on the RADIUS for the incomming session
tied
>> to the username. You are correct. This will not work.
>> 
>> 1.) The user may come at different times from different LACs. So there
is no
>> use to bind the password to the username.

>> 
>> 2.) As you stated before, it is too late when you get the username.
>> 
>> A solution would be:
>> You get a LAC host name together with the incomming tunnel. You can use an
>> Access-Request with this hostname to get a tunnel password from the RADIUS.
>> So you can answer the challenge and issue your own challenge at the same
>> time. Nobody said that RADIUS can only be used together with PPP. We even
>> use a DNIS String as username to get centrally stored information and
>> configuration.
>> 
>> with best regards
>> Walter Klausberger
>> Siemens AG Austria
>> 
>> -----Ursprüngliche Nachricht-----
>> Von: streifer@us.ibm.com [mailto:streifer@us.ibm.com]
>> Gesendet am: Dienstag, 22. August 2000 13:51
>> An: l2tp@diameter.org
>> Betreff: L2TP Tunnel Authentication with RADIUS
>> 
>> 
>> Are there any L2TP implementations that use RADIUS for L2TP tunnel
>> authentication ?
>> 
>> Looking through the various RADIUS RFCs it looks like RADIUS can be used
>> for PPP authentication after the tunnel and call are established but I
>> can't see a way to get a RADIUS Server to send back a MD5 Hash value or a
>> password to use for a Challenge-Response for tunnel authentication. There
>> is a Tunnel-Password attribute but it is only sent back in an Access-Accept
>> packet. From an LNS standpoint it looks to me that I would need a tunnel
>> password stored locally in order to generate a Challenge-Response which
>> would defeat the purpose of using RADIUS to consolidate authentication
>> administration to a single system.
>> 
>> Am I missing something ?
>> 
>> 
>> Thanks,
>> 
>> Kurt Streifert
>> IBM AS/400 Endicott Programming Lab
>> 
>> 
>
>=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=
> +----+----+     
> |.. .|    |     Motonori Shindo
> |_~__|    |     
> | .. |~~_~|     Sr. Systems Engineer
> | .  |    |     CoSine Communications Inc.
> +----+----+     
> C o S i n e     e-mail:  mshindo@cosinecom.com 
>Communications
>=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=
> 



From owner-l2tp@diameter.org  Wed Sep 27 13:20:46 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id NAA17167
	for <l2tpext-archive@odin.ietf.org>; Wed, 27 Sep 2000 13:20:46 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8RDISi20512
	for l2tp-list; Wed, 27 Sep 2000 06:18:28 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from eins.siemens.at (eins.siemens.at [193.81.246.11])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8RDIPl20506
	for <l2tp@diameter.org>; Wed, 27 Sep 2000 06:18:25 -0700
Received: from scesie13.sie.siemens.at (forix-hme0 [10.1.140.2])
	by eins.siemens.at  with ESMTP id PAA23039;
	Wed, 27 Sep 2000 15:14:24 +0200 (MET DST)
Received: (from smap@localhost)
	by scesie13.sie.siemens.at (8.9.3/8.9.3) id PAA28551;
	Wed, 27 Sep 2000 15:23:07 +0200 (MET DST)
Received: from vies141a.sie.siemens.at(195.1.196.56) by scesie13 via smap (V2.0beta)
	id xma027831; Wed, 27 Sep 00 15:22:47 +0200
Received: by vies141a.sie.siemens.at with Internet Mail Service (5.5.2650.21)
	id <TGPM7DNX>; Wed, 27 Sep 2000 15:22:44 +0200
Message-ID: <D9F2B9CD7BD5D21196BC0800060D9ED601EFBFBE@vies186a.sie.siemens.at>
From: Klausberger Walter <walter.klausberger@siemens.at>
To: "'danny@tcb.net'" <danny@tcb.net>, Evan Caves
	 <evan.caves@ericsson.com>
Cc: l2tp@diameter.org
Subject: AW: [Fwd: L2TP Diffserv Draft] 
Date: Wed, 27 Sep 2000 15:22:42 +0200
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: text/plain;
	charset="iso-8859-1"
X-MIME-Autoconverted: from quoted-printable to 8bit by charizard.diameter.org id e8RDIPl20507
Sender: owner-l2tp@diameter.org
Precedence: bulk
X-MIME-Autoconverted: from 8bit to quoted-printable by charizard.diameter.org id e8RDISi20512
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by ietf.org id NAA17167

Hi,

well I like this draft and the comments, but I think there is still
something missing.

When a better service is granted to someone (from a LAC point of view), I
know the maximum bandwidth this user is capable to send, but when he
downloads a file the prioritized traffic can exceed this maximum (may be a
problem with UDP traffic).

Maybe the LNS should take the link speed in the L2TP session setup message
as a hint, or the granted bandwidth should be added to the DSCP AVP.
Shouldn't there be a hint in this draft? Or am I missing something?

Awaiting comments
with best regards
Walter

-----Ursprüngliche Nachricht-----
Von: Danny McPherson [mailto:danny@tcb.net]
Gesendet am: Samstag, 16. September 2000 00:49
An: Evan Caves
Cc: l2tp@diameter.org
Betreff: Re: [Fwd: L2TP Diffserv Draft] 


> I sent this out several weeks ago but got no response. This time can
> anyone respond (not just the authors :-).

I just picked it up from Ken & Pat so I'll give it a shot.

> > 1. Is the DSCP AVP a six or eight bit value?

The format as currently specified is as follows:

       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
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
      |1|1|0|0|        Length         |              43               |
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
      |                1              |        DSCP  Value            |
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

This means that the total length is 8 octets, with 2 octets representing the

DSCP value.  The value will remain at least one octet, as two octets seem to

enable little more than tidy packets encoding pictures.


> > 2. What DSCP value is used for tunnel control messages? If the
> >    session DSCP has a 'higher priority' than the tunnels then
> >    payload packets may arrive before session setup completion.

The next version of the draft, which is hopefully no more than a couple
weeks 
out, will specify the behavior in this scenario.

> > 3. Why is this AVP restricted to just sessions? One could conceive
> >    bringing up a tunnel for each PHB and forwarding all tunnel
> >    traffic (control and payload) with one of the PHB class DSCPs.

Agreed.  And this is something that we were just discussing.  I'll ensure
that 
the upcoming draft provides provisions for this as well (comments welcome!).

> > 4. This draft makes a big assumption that the tunnel traverses a
> >    single DS domain. If the tunnel traverses multiple DS domains
> >    the DSCP to PHB mappings could be different (this may mean
> >    that each tunnel endpoint uses different DSCP marks that get
> >    translated at some DS boundary). How is this resolved?

Multiple DS domains is of course is a very well-know and interesting
problem, 
and again, the next version of the draft will take a stab at addressing this

as well.  We plan to do our best to be in alignment with similar work being 
done in the diffserv WG.  (Again, comments welcome!).

> > 5. Related to 4 it would seem that it may be more appropriate that
> >    the tunnel initiator ask for a PHB rather than a DSCP and then
> >    let each endpoint decide how that PHB maps to a DSCP.

Though you still have lots of problems with this model as well when
traverses 
multiple DS domains.  Again, we'll take a stab at it and work with the DS 
folks to ensure that we're aligned.  Comments here are welcome!

Thanks for the input, hope to have something useful within a couple of weeks

at most.

-danny





From owner-l2tp@diameter.org  Thu Sep 28 20:20:40 2000
Received: from charizard.diameter.org (nobody@c900656-a.plstn1.sfba.home.com [24.20.167.220])
	by ietf.org (8.9.1a/8.9.1a) with SMTP id UAA05870
	for <l2tpext-archive@odin.ietf.org>; Thu, 28 Sep 2000 20:20:39 -0400 (EDT)
Received: (from majordomo@localhost)
	by charizard.diameter.org (8.10.0/8.10.0) id e8SLQeo22223
	for l2tp-list; Thu, 28 Sep 2000 14:26:40 -0700
X-Authentication-Warning: charizard.diameter.org: majordomo set sender to owner-l2tp@diameter.org using -f
Received: from center.mshindo.net (center.mshindo.net [210.231.221.221])
	by charizard.diameter.org (8.10.0/8.10.0) with ESMTP id e8SLPcl22217
	for <l2tp@diameter.org>; Thu, 28 Sep 2000 14:25:38 -0700
Received: from cosinecom.com (proxy191.cosinecom.com [63.88.104.191])
	by center.mshindo.net (8.9.3/3.7W/00041811) with SMTP id GAA84220;
	Fri, 29 Sep 2000 06:31:50 +0900 (JST)
Date: Fri, 29 Sep 2000 06:28:23 +0900 (JST)
Message-Id: <20000929.062823.85419129.mshindo@mshindo.net>
To: sridhar@trinc.com
Cc: l2tp@diameter.org
Subject: Re: AW: L2TP Tunnel Authentication with RADIUS
From: Motonori Shindo <mshindo@mshindo.net>
In-Reply-To: <200009270808.NAA29507@brahma.roc.com>
References: <D9F2B9CD7BD5D21196BC0800060D9ED601EFBF3E@vies186a.sie.siemens.at>
	<20000904.101315.74751693.mshindo@mshindo.net>
	<200009270808.NAA29507@brahma.roc.com>
X-Mailer: Mew version 1.95b30 on XEmacs 21.1 (Canyonlands)
X-PGP-fingerprint: 06 B0 B1 A4 06 C1 6A 14  63 C0 D7 18 01 CD D9 83
Mime-Version: 1.0
Content-Type: Text/Plain; charset=iso-8859-1
X-MIME-Autoconverted: from quoted-printable to 8bit by charizard.diameter.org id e8SLPdl22218
Sender: owner-l2tp@diameter.org
Precedence: bulk
X-MIME-Autoconverted: from 8bit to quoted-printable by charizard.diameter.org id e8SLQeo22223
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by ietf.org id UAA05870

Sridhar,

Good point. When attribute hiding is used along with a tunnel
authentication, it introduces a little bit of complication
(i.e. necessity of deferring the decoding of hidden attributes until
the tunnel password is retrieved from RADIUS). Yet, tunnel auth at LNS
via RADIUS is doable and almost mandatory from the operational point
of view. Tunnel auth at LNS without RADIUS support is a nightmare! 
Everytime LAC is added to the network, all LNSes have to be
reconfigured to reflect a new tunnel password for the LAC.

Regards,

From: Sridhar Malarothu <sridhar@trinc.com>
Subject: Re: AW: L2TP Tunnel Authentication with RADIUS
Date: Wed, 27 Sep 2000 13:42:00 +0530

> Hi All,
> 
>    Since the same secret which is used for Tunnel Authentication is also
> used in hiding the AVP's, I think its not simple for LNS to get the Tunnel
> password(secret) from RADIUS(based on peer's host name in the SCCRQ msg),
> if both the tunnel authentication and AVP's hiding are used at the same time.
>    Even though the Hostname AVP in the SCCRQ msg is not hidden, the other
> AVP's like Assigned Tunnel ID etc, can be hidden. 
>    Since Hostname is required to get the Tunnel Password from Radius, we
> may just extract the Hostname AVP from SCCRQ and differ the decoding of rest
> of the message till the RADIUS response is received. But we cannot ZLB ack the
> SCCRQ message, since the Assigned Tunnel ID in SCCRQ may be hidden.
>        
> 
> Best Regards
> Sridhar
>    
> 
> At 10:13 AM 9/4/00 +0900, you wrote:
> >All,
> >
> >I do agree with Walter's comment. 
> >
> >Tunnel authentication and user authentication (associated with an
> >incoming call) are totally two different things, and more importantly,
> >the establishment of a tunnel does not necessarily happen at the same
> >time as the session (ie. PPP) establishment.
> >
> >>From this point of view, a straight-forward LAC implementation should
> >work in the following way:
> >
> > 1) LAC gets to know the user name (via PAP/CHAP or whatever) and
> >    queries to RADIUS to authenticate this user and obtain information
> >    associated with it. If this user is an L2TP user, then the
> >    RADIUS entry for this user would contain a Tunnel-Server-Endpoint
> >    attribute associated with this user.
> >
> > 2) If a tunnel for the LNS specified by Tunnel-Sever-Endpoint hasn't 
> >    already existed and if the LAC is configured to do tunnel
> >    authentication, then a LAC queries to RADIUS again to obtain a tunnel
> >    passowrd. In this case, the RADIUS entry has a LNS name (identified
> >    by a Tunnel-Server-Endpoint in the previous query) as a user
> >    name, and should include a Tuennel-Password attribute in its reply
> >    item.
> >
> >Please note that this sequence involves two RADIUS queries, one for
> >user authentication and another for tunnel authentication. To make it
> >work in more efficent way, an optimization is possible by piggybacking
> >a Tunnel-Password attribute on the first RADIUS query for the user
> >authentication. Since there's a one-to-one relationship between
> >Tunnel-Server-Endpoint and Tunnel-Password, this optimization works
> >without any problem.
> >
> >Howerver, on the LNS side, as Walter pointed out, this optimization
> >can not be applied and two RADIUS queries are required. That is, when
> >the LNS receives an SCCRQ from the LAC (along with an L2TP's CHALLENGE
> >attribute), the LNS first queries to RADIUS about the tunnel password
> >associated with a name specified in the L2TP's HOSTNAME attribute (the
> >use of HOSTNAME attribute for this purpose is not explicitly defined
> >in the RFC but it is a good choice, I believe). After the tunnel
> >authentication is finished and the tunnel is successfully established,
> >the ICRQ follws and then user authentication happens through, again,
> >RADIUS.
> >
> >Regards,
> >
> >From: Klausberger Walter <walter.klausberger@siemens.at>
> >Subject: AW: L2TP Tunnel Authentication with RADIUS
> >Date: Fri, 1 Sep 2000 14:06:52 +0200 
> >
> >> Well, we will do so. I think it depends on which side of the tunnel you
> >> are...
> >> 
> >> In LAC case you can use the tunnel password to setup an appropriate tunnel
> >> to an LNS. During user authentication (with PPP PAP, CHAP,...) you do an
> >> Access-Request and you get the tunnel password in return. That's ok, so
> >> far...
> >> 
> >> I think your problem is really on the LNS side. A tunnel comes in and there
> >> may be a tunnel password stored on the RADIUS for the incomming session
> tied
> >> to the username. You are correct. This will not work.
> >> 
> >> 1.) The user may come at different times from different LACs. So there
> is no
> >> use to bind the password to the username.
> 
> >> 
> >> 2.) As you stated before, it is too late when you get the username.
> >> 
> >> A solution would be:
> >> You get a LAC host name together with the incomming tunnel. You can use an
> >> Access-Request with this hostname to get a tunnel password from the RADIUS.
> >> So you can answer the challenge and issue your own challenge at the same
> >> time. Nobody said that RADIUS can only be used together with PPP. We even
> >> use a DNIS String as username to get centrally stored information and
> >> configuration.
> >> 
> >> with best regards
> >> Walter Klausberger
> >> Siemens AG Austria
> >> 
> >> -----Ursprüngliche Nachricht-----
> >> Von: streifer@us.ibm.com [mailto:streifer@us.ibm.com]
> >> Gesendet am: Dienstag, 22. August 2000 13:51
> >> An: l2tp@diameter.org
> >> Betreff: L2TP Tunnel Authentication with RADIUS
> >> 
> >> 
> >> Are there any L2TP implementations that use RADIUS for L2TP tunnel
> >> authentication ?
> >> 
> >> Looking through the various RADIUS RFCs it looks like RADIUS can be used
> >> for PPP authentication after the tunnel and call are established but I
> >> can't see a way to get a RADIUS Server to send back a MD5 Hash value or a
> >> password to use for a Challenge-Response for tunnel authentication. There
> >> is a Tunnel-Password attribute but it is only sent back in an Access-Accept
> >> packet. From an LNS standpoint it looks to me that I would need a tunnel
> >> password stored locally in order to generate a Challenge-Response which
> >> would defeat the purpose of using RADIUS to consolidate authentication
> >> administration to a single system.
> >> 
> >> Am I missing something ?
> >> 
> >> 
> >> Thanks,
> >> 
> >> Kurt Streifert
> >> IBM AS/400 Endicott Programming Lab
> >> 
> >> 
> >
> >=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=
> > +----+----+     
> > |.. .|    |     Motonori Shindo
> > |_~__|    |     
> > | .. |~~_~|     Sr. Systems Engineer
> > | .  |    |     CoSine Communications Inc.
> > +----+----+     
> > C o S i n e     e-mail:  mshindo@cosinecom.com 
> >Communications
> >=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=
> > 
> 
> 

=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=
 +----+----+     
 |.. .|    |     Motonori Shindo
 |_~__|    |     
 | .. |~~_~|     Sr. Systems Engineer
 | .  |    |     CoSine Communications Inc.
 +----+----+     
 C o S i n e     e-mail:  mshindo@cosinecom.com 
Communications
=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=


