Re: [homenet] dst/src routing drafts (for IETF-91 rtgwg)

Gert Doering <> Tue, 28 October 2014 13:19 UTC

Return-Path: <gert@Space.Net>
Received: from localhost ( []) by (Postfix) with ESMTP id 1001B1A6EEC for <>; Tue, 28 Oct 2014 06:19:00 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -0.011
X-Spam-Status: No, score=-0.011 tagged_above=-999 required=5 tests=[BAYES_20=-0.001, T_RP_MATCHES_RCVD=-0.01] autolearn=ham
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id CJerxXHk2M48 for <>; Tue, 28 Oct 2014 06:18:53 -0700 (PDT)
Received: from ( [IPv6:2001:608:2:81::67]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by (Postfix) with ESMTPS id 70EA11A877F for <>; Tue, 28 Oct 2014 06:18:05 -0700 (PDT)
Received: from (localhost [IPv6:::1]) by (Postfix) with ESMTP id 77C51602FF for <>; Tue, 28 Oct 2014 14:18:03 +0100 (CET)
X-SpaceNet-Relay: true
Received: from (moebius3.Space.Net [IPv6:2001:608:2:2::250]) by (Postfix) with ESMTPS id 3912560141 for <>; Tue, 28 Oct 2014 14:18:03 +0100 (CET)
Received: (qmail 23745 invoked by uid 1007); 28 Oct 2014 14:18:03 +0100
Date: Tue, 28 Oct 2014 14:18:03 +0100
From: Gert Doering <>
To: Ole Troan <>
Message-ID: <20141028131803.GN31092@Space.Net>
References: <> <> <> <>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <>
User-Agent: Mutt/1.5.23 (2014-03-12)
Cc: David Lamparter <>, "" <>, "" <>, "Fred Baker \(fred\)" <>, Mikael Abrahamsson <>
Subject: Re: [homenet] dst/src routing drafts (for IETF-91 rtgwg)
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Tue, 28 Oct 2014 13:19:00 -0000


On Tue, Oct 28, 2014 at 02:06:26PM +0100, Ole Troan wrote:
> isn't multi-prefix multi-homing one of the most obvious use cases
> for source address dependent routing? that's not restricted with
> homenets, but also any small network. I'm assuming large networks
> will continue with PI addresses and BGP based multihoming.

I see the distinction somewhere in the "what defines policy in a network",
as in:

 - in a "big" company network, there usually is some sort of network access
   policy, which is defined and enforced by the network people - so you'll
   see BGP policies and PI, or in a "dual /48" style shop, I expect to see
   dual-NPT66-with-ULA devices where you can configure your policy ("surfing
   goes out via ISP A, mail goes out via ISP B") in the network device

 - in a homenet, small company ("barber shop") network, there is no "admin",
   and to try to enforce network policy ("surfing via cable, bittorrent via
   DSL") in the network device is futile, because "no admin".  So here you
   need SADR to empower the end device - and by that means, the user - to
   define policy.  Have the bittorrent client use the source address from
   the DSL ISP's /48, the web browser use the source address from the 
   Cable ISP, and things work the way the user wants it.

   ... and this is why I really really like dual-/48 multihoming for the
   "not really managed" SOHO case, as it puts policy decisions where they
   can be made (= user's device) - and of course, why dual-/48 is not going
   to fly for enterprise networks (= can't have the user decide that).

(In case this wasn't obvious, I'm agreeing with Ole here, I just try to
shed light from a slightly different angle on that.)

Gert Doering
        -- NetMaster
have you enabled IPv6 on something today...?

SpaceNet AG                        Vorstand: Sebastian v. Bomhard
Joseph-Dollinger-Bogen 14          Aufsichtsratsvors.: A. Grundner-Culemann
D-80807 Muenchen                   HRB: 136055 (AG Muenchen)
Tel: +49 (0)89/32356-444           USt-IdNr.: DE813185279