Re: Address Family Support in OSPFv3

Acee Lindem <acee@REDBACK.COM> Wed, 09 July 2003 20:07 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 QAA19603 for <ospf-archive@LISTS.IETF.ORG>; Wed, 9 Jul 2003 16:07:19 -0400 (EDT)
Received: from PEAR.EASE.LSOFT.COM (209.119.0.19) by cherry.ease.lsoft.com (LSMTP for Digital Unix v1.1b) with SMTP id <9.00A5BD07@cherry.ease.lsoft.com>; Wed, 9 Jul 2003 16:07:16 -0400
Received: from PEACH.EASE.LSOFT.COM by PEACH.EASE.LSOFT.COM (LISTSERV-TCP/IP release 1.8e) with spool id 47949424 for OSPF@PEACH.EASE.LSOFT.COM; Wed, 9 Jul 2003 16:06:56 -0400
Received: from 155.53.12.9 by WALNUT.EASE.LSOFT.COM (SMTPL release 1.0i) with TCP; Wed, 9 Jul 2003 16:06:56 -0400
Received: from redback.com (login003.redback.com [155.53.12.55]) by prattle.redback.com (Postfix) with ESMTP id 8AF0D7E0193 for <OSPF@PEACH.EASE.LSOFT.COM>; Wed, 9 Jul 2003 13:05:00 -0700 (PDT)
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.0.2) Gecko/20030208 Netscape/7.02
X-Accept-Language: en-us, en
MIME-Version: 1.0
References: <000401c34650$4bfce9e0$386545ab@amer.cisco.com>
Content-Type: text/plain; charset="us-ascii"; format="flowed"
Content-Transfer-Encoding: 7bit
Message-ID: <3F0C756D.6000308@redback.com>
Date: Wed, 09 Jul 2003 16:05:01 -0400
Reply-To: Mailing List <OSPF@PEACH.EASE.LSOFT.COM>
Sender: Mailing List <OSPF@PEACH.EASE.LSOFT.COM>
From: Acee Lindem <acee@REDBACK.COM>
Subject: Re: Address Family Support in OSPFv3
To: OSPF@PEACH.EASE.LSOFT.COM
Precedence: list
Content-Transfer-Encoding: 7bit

Sina Mirtorabi wrote:
> Acee,
>
> ->> The fact of not using AF cannot be counted an advantage for
> ->multiple
> ->> instance-id as the solution to AF in OSPFv3
>
> ->Sina,
> ->
> ->I disagree with you here. Adding a new layer for AF to OSPFv3
> ->requires more base protocol changes as well as the attendant
> ->implementation costs and overhead. Any unbiased evaluation
> ->has to take this into consideration.
>
> If you do not need to have other AF, you can only have the minimum
> required in our proposal that is setting the AF bit in the Hello and
> v6-AF bit in the AdressFamilies field  (Disabled mode)
>
> Once you want to have additional AF then you have to bother about the
> additional two new LSA and potentially other new LSAs in other AF (if
> needed) which even in case of multiple instance you would need to define
> ways to be backward compatibles, define new LSA for new AF etc

Sina,
Let me try and be more clear - What I'm saying is that with multiple
instances you already have the separation between the topologies without
any protocol changes or an additional AF layer in your implementation.

For example, to support the OSPFv3 IPv6 multicast topology with the
multi-AF approach one must implement the two drafts you've posted plus
deal with all the issues related to the AF layer in configuration and
management (i.e., show commands, etc). To do it with multiple instances,
all you have to do is configure the address family and default OSPFv3
interface ID at the instance level. I haven't thought about this at
great depth but on the surface I can't see why it wouldn't work.
There will be no protocol changes and I'll submit to you that the
implementation cost is less.

At this point, I'm not saying that one approach is obviously better or
trying to precisely quantify the relative efficiencies when multiple AFs
are used. All I'm saying is that you can't completely discount the
associated costs.

Thanks,
Acee

> Sina
>
>
> ->
> ->Thanks,
> ->Acee
> ->
> ->>
> ->> thanks
> ->> Sina
> ->>
> ->
> ->
> ->--
> ->Acee
> ->
>


--
Acee