Re: [Isis-wg] FW: New Version Notification for draft-ginsberg-isis-sbfd-discriminator-00.txt

Jeffrey Haas <> Thu, 08 May 2014 20:27 UTC

Return-Path: <>
Received: from localhost ( []) by (Postfix) with ESMTP id 06C901A010D; Thu, 8 May 2014 13:27:25 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -2.219
X-Spam-Status: No, score=-2.219 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, IP_NOT_FRIENDLY=0.334, RP_MATCHES_RCVD=-0.651, SPF_HELO_PASS=-0.001, SPF_PASS=-0.001] autolearn=ham
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id 3F9idPAmRLEF; Thu, 8 May 2014 13:27:17 -0700 (PDT)
Received: from ( []) by (Postfix) with ESMTP id F0B5A1A0067; Thu, 8 May 2014 13:27:16 -0700 (PDT)
Received: by (Postfix, from userid 1001) id 60523C2A8; Thu, 8 May 2014 16:27:12 -0400 (EDT)
Date: Thu, 8 May 2014 16:27:12 -0400
From: Jeffrey Haas <>
To: "Les Ginsberg (ginsberg)" <>
Message-ID: <20140508202712.GF3935@pfrc>
References: <> <> <20140508192534.GA24787@pfrc> <>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <>
User-Agent: Mutt/1.5.21 (2010-09-15)
X-Mailman-Approved-At: Fri, 09 May 2014 09:24:56 -0700
Cc: Jeffrey Haas <>, "" <>, "" <>
Subject: Re: [Isis-wg] FW: New Version Notification for draft-ginsberg-isis-sbfd-discriminator-00.txt
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: IETF IS-IS working group <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Thu, 08 May 2014 20:27:25 -0000


On Thu, May 08, 2014 at 08:10:35PM +0000, Les Ginsberg (ginsberg) wrote:
> How S-BFD Discriminators get used is a subject that no doubt the BFD WG
> will discuss at length in the near future. However, the additional
> information that may be required to differentiate how - when multiple
> discriminators are advertised by a node - each discriminator should be
> used is out of scope for the IGP advertisement. This gets into the area of
> application information which has to be defined/standardized elsewhere and
> does not belong in the IGP advertisements. So the deliberate intent here
> as far as the IGP advertisement is to limit it to simply a list of
> assigned discriminators - nothing more.

I don't buy the argument, even though the underlying reasoning may not be
your responsibility. :-)

IGPs distribute node and link reachability.  The underlying mechanism that
is being used isn't even binding the information so much to the node as it
is to protocol plumbing.

If the feature is intended to be node-addressable or potentially
link-addressable is very much relevant to how the feature is developed and
used.  If we're not to the point of answering those two very basic
questions, then the IGP drafts are premature.  And almost certainly,
squatting on a code point even tentatively is premature.  I strongly suggest
moving it to TBD in the next version of the draft.

That said, I suspect the use cases will shake out very shortly and hopefully
the point will be moot in -01.

-- Jeff