[Isis-wg] How should we encode TRILL IS-IS to distinguish it from layer 3 IS-IS?
Radia Perlman <Radia.Perlman@sun.com> Sat, 19 May 2007 04:42 UTC
Return-path: <isis-wg-bounces@ietf.org>
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com) by megatron.ietf.org with esmtp (Exim 4.43) id 1HpGmG-0002Kr-UC; Sat, 19 May 2007 00:42:48 -0400
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org) by megatron.ietf.org with esmtp (Exim 4.43) id 1HpGmF-0002Jl-MX for isis-wg@ietf.org; Sat, 19 May 2007 00:42:47 -0400
Received: from edge.sunlabs.com ([204.153.12.50] helo=mail-mta.sunlabs.com) by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1HpGmD-0001ai-CA for isis-wg@ietf.org; Sat, 19 May 2007 00:42:47 -0400
Received: from mail.sunlabs.com ([152.70.2.186]) by dps.sfvic.sunlabs.com (Sun Java System Messaging Server 6.1 HotFix 0.02 (built Aug 25 2004)) with ESMTP id <0JI9005RNUF0VS00@dps.sfvic.sunlabs.com> for isis-wg@ietf.org; Fri, 18 May 2007 21:42:36 -0700 (PDT)
Received: from [127.0.0.1] ([129.150.16.160]) by mail.sunlabs.com (Sun Java System Messaging Server 6.1 HotFix 0.02 (built Aug 25 2004)) with ESMTPSA id <0JI900KK8UEYEM00@mail.sunlabs.com> for isis-wg@ietf.org; Fri, 18 May 2007 21:42:35 -0700 (PDT)
Date: Fri, 18 May 2007 21:43:07 -0700
From: Radia Perlman <Radia.Perlman@sun.com>
To: isis-wg@ietf.org
Message-id: <464E805B.3050604@sun.com>
MIME-version: 1.0
Content-type: text/plain; charset="ISO-8859-1"; format="flowed"
Content-transfer-encoding: 7bit
User-Agent: Thunderbird 1.5.0.10 (Windows/20070221)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: cd26b070c2577ac175cd3a6d878c6248
Subject: [Isis-wg] How should we encode TRILL IS-IS to distinguish it from layer 3 IS-IS?
X-BeenThere: isis-wg@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: IETF IS-IS working group <isis-wg.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/isis-wg>, <mailto:isis-wg-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www1.ietf.org/pipermail/isis-wg>
List-Post: <mailto:isis-wg@ietf.org>
List-Help: <mailto:isis-wg-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/isis-wg>, <mailto:isis-wg-request@ietf.org?subject=subscribe>
Errors-To: isis-wg-bounces@ietf.org
The TRILL WG would like to use IS-IS for RBridges to have a routing protocol among themselves. RBridges encapsulate packets with a TRILL header as they are forwarded across the RBridged-cloud (called a "campus"). Which means that layer 3 IS-IS, if routers are connected via a "campus", will look like: inner frame: exactly as the first IS-IS router transmits it, i.e., an Ethernet header with Ethertype="IS-IS", Ethernet source address=IS-IS router that transmitted it, Ethernet destination address usually="all-Level1-routers" or "all-Level2-routers", but sometimes unicast (like in the case of a PSNP). And this frame will be TRILL-encapsulated while travelling across the campus between the RBridges, but will be stripped off at the egress RBridge so layer 3 routers won't know it happened. The layer 3 IS-IS packet, when TRILL-encapsulated look like: outside the inner frame is a "TRILL header", which identifies the ingress and egress RBridges, and a hop count, and (luckily, because I think we'll need it) some reserved bits. outside that is an ordinary looking Ethernet header (in case there are bridges between two RBridges), with Ethertype=TRILL. *********** So far OK. But now -- how do we encode TRILL IS-IS packets so they are not confused with layer 3 IS-IS, and also not confused with TRILL-encapsulated data packets? TRILL will use "area address=0", which I'm assuming is not a legal area address for layer 3. Is this assumption correct? Layer 3 IS-IS routers will not ever see a TRILL IS-IS packet because it will be TRILL-encapsulated. The Ethertype will be TRILL rather than IS-IS, and RBridges will not use the IS-IS layer 2 multicast addresses. But the question is---how does an RBridge know the difference between a layer 3 IS-IS packet that has been encapsulated for transit across the campus, and a TRILL IS-IS packet? Both will have TRILL headers. There's lots of solutions -- I just want to know what will be most convenient for IS-IS implementers. a) We could use an "illegal" inner Ethertype, say Ethertype=0, that would never appear in a data packet. Then RBridges would interpret Ethertype=0 as a TRILL IS-IS packet. b) We could use a flag in the TRILL header (steal a reserved bit) to mean "This is a TRILL IS-IS packet" Any other ideas or advice? Thanks, Radia _______________________________________________ Isis-wg mailing list Isis-wg@ietf.org https://www1.ietf.org/mailman/listinfo/isis-wg
- [Isis-wg] How should we encode TRILL IS-IS to dis… Radia Perlman
- Re: [Isis-wg] How should we encode TRILL IS-IS to… Jeff Learman
- RE: [Isis-wg] How should we encode TRILL IS-IS to… Les Ginsberg (ginsberg)
- Re: [Isis-wg] How should we encode TRILL IS-IS to… mike shand
- Re: [Isis-wg] How should we encode TRILL IS-IS to… Radia Perlman
- Re: [Isis-wg] How should we encode TRILL IS-IS to… mike shand
- RE: [Isis-wg] How should we encode TRILL IS-IS to… Les Ginsberg (ginsberg)
- RE: [Isis-wg] How should we encode TRILL IS-IS to… Eastlake III Donald-LDE008
- Re: [Isis-wg] How should we encode TRILL IS-IS to… Tony Li
- Re: [Isis-wg] How should we encode TRILL IS-IS to… Tony Li
- RE: [Isis-wg] How should we encode TRILL IS-IS to… Eastlake III Donald-LDE008
- RE: [Isis-wg] How should we encode TRILL IS-IS to… Eastlake III Donald-LDE008
- Re: [Isis-wg] How should we encode TRILL IS-IS to… Hannes Gredler
- RE: [Isis-wg] How should we encode TRILL IS-IS to… Les Ginsberg (ginsberg)
- Re: [Isis-wg] How should we encode TRILL IS-IS to… James Carlson
- Re: [Isis-wg] How should we encode TRILL IS-IS to… James Carlson
- RE: [Isis-wg] How should we encode TRILL IS-IS to… Eastlake III Donald-LDE008
- Re: [Isis-wg] How should we encode TRILL IS-IS to… Tony Li
- RE: [Isis-wg] How should we encode TRILL IS-IS to… Les Ginsberg (ginsberg)
- Re: [Isis-wg] How should we encode TRILL IS-IS to… Hannes Gredler
- RE: [Isis-wg] How should we encode TRILL IS-IS to… Eastlake III Donald-LDE008