Re: about Area Splitting

Tony Przygienda <prz@XEBEO.COM> Tue, 09 December 2003 12:30 UTC

Received: from cherry.ease.lsoft.com (cherry.ease.lsoft.com [209.119.0.109]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id HAA23895 for <ospf-archive@LISTS.IETF.ORG>; Tue, 9 Dec 2003 07:30:05 -0500 (EST)
Received: from PEAR.EASE.LSOFT.COM (209.119.0.19) by cherry.ease.lsoft.com (LSMTP for Digital Unix v1.1b) with SMTP id <10.00C6D37F@cherry.ease.lsoft.com>; Tue, 9 Dec 2003 7:30:18 -0500
Received: from PEACH.EASE.LSOFT.COM by PEACH.EASE.LSOFT.COM (LISTSERV-TCP/IP release 1.8e) with spool id 63654312 for OSPF@PEACH.EASE.LSOFT.COM; Tue, 9 Dec 2003 07:30:17 -0500
Received: from 65.200.123.8 by WALNUT.EASE.LSOFT.COM (SMTPL release 1.0i) with TCP; Tue, 9 Dec 2003 07:30:17 -0500
Received: (qmail 26477 invoked by uid 404); 9 Dec 2003 12:30:16 -0000
Received: from prz@xebeo.com by lxmail by uid 401 with qmail-scanner-1.20rc1 (clamscan: 0.60. spamassassin: 2.55. Clear:RC:1:. Processed in 0.532394 secs); 09 Dec 2003 12:30:16 -0000
Received: from unknown (HELO xebeo.com) (172.16.1.102) by lxmail.nj.us.utstar.com with SMTP; 9 Dec 2003 12:30:15 -0000
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.0.2) Gecko/20030716
X-Accept-Language: en-us, en
MIME-Version: 1.0
References: <31C0F08B0D18D511ACC800508BAE7B4707F47DE7@CI0026EXCH001U>
X-Enigmail-Version: 0.75.1.0
X-Enigmail-Supports: pgp-inline, pgp-mime
Content-Type: text/plain; charset="GB2312"
Content-Transfer-Encoding: 7bit
Message-ID: <3FD5C057.4090404@xebeo.com>
Date: Tue, 09 Dec 2003 13:30:15 +0100
Reply-To: Mailing List <OSPF@PEACH.EASE.LSOFT.COM>
Sender: Mailing List <OSPF@PEACH.EASE.LSOFT.COM>
From: Tony Przygienda <prz@XEBEO.COM>
Subject: Re: about Area Splitting
To: OSPF@PEACH.EASE.LSOFT.COM
Precedence: list
Content-Transfer-Encoding: 7bit

Li, Ke Qin (Peter) wrote:

>Hi,
>
>        I've read the Internet Draft "OSPF Areas Considered Harmful" written
>by Mikkel Thorup
>http://www.ietf.org/internet-drafts/draft-thorup-ospf-harmful-00.txt. In
>Section 8.0 "Summarization and infinite loops", a scenario is described
>where area splitting results in infinite routing loop (although IMHO it is
>black hole). I think it is a basic problem related to OSPF hierarchy and
>area range. Has this problem been discussed in this list before? If yes,
>what is the solution?
>
>        Thank you very much!
>
>Keqin Li
>Bell Labs Research China
>Tel: 86-10-68748088-8438
>
>
>
I discussed and started to write once with John what we considered the
optimal solution (and I know some people
implemented it). Basically, when you realized that you see through the
backbone advertisement of an ABR in your
area that you
don't have intra-area connectivity to, you inject with the summary all
the more specifics into the backbone. Simple
to implement, the only real disadvantage really being that you may put
tons of routes into the backbone. On the other
hand, you're only real other solutions are either for one to stop
advertising (avg. 50% of area goes down) or
maybe have some complicated 'tunnel-through-backbone' solution, which we
didn't think was worth the effort.

If anybody feels like writing that down, I can help a bit, I don't think
I will find my writedown anymore on all my
harddisks lying around ; -)

-- tony