Re: [homenet] Let's make in-home ULA presence a MUST !?

JF Tremblay <> Tue, 14 October 2014 15:21 UTC

Return-Path: <>
Received: from localhost ( []) by (Postfix) with ESMTP id D3EA41A8907 for <>; Tue, 14 Oct 2014 08:21:20 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -1.693
X-Spam-Status: No, score=-1.693 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, PLING_QUERY=0.994, RP_MATCHES_RCVD=-0.786, SPF_PASS=-0.001] autolearn=ham
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id myYNAIklXJ2H for <>; Tue, 14 Oct 2014 08:21:15 -0700 (PDT)
Received: from ( [IPv6:2620:0:230:8000::2]) by (Postfix) with ESMTP id 84EA51A88D0 for <>; Tue, 14 Oct 2014 08:21:15 -0700 (PDT)
Received: from ( []) by (Postfix) with ESMTPSA id 64B4F403B5 for <>; Tue, 14 Oct 2014 11:21:14 -0400 (EDT)
Content-Type: text/plain; charset=windows-1252
Mime-Version: 1.0 (Mac OS X Mail 7.3 \(1878.6\))
From: JF Tremblay <>
In-Reply-To: <20141014145930.GY31092@Space.Net>
Date: Tue, 14 Oct 2014 11:21:13 -0400
Content-Transfer-Encoding: quoted-printable
Message-Id: <>
References: <> <> <> <20141014142746.GX31092@Space.Net> <> <20141014145930.GY31092@Space.Net>
To: HOMENET Working Group <>
X-Mailer: Apple Mail (2.1878.6)
Subject: Re: [homenet] Let's make in-home ULA presence a MUST !?
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, 14 Oct 2014 15:21:21 -0000

On Oct 14, 2014, at 10:59 AM, Gert Doering <> wrote:
> On Tue, Oct 14, 2014 at 09:41:55AM -0500, Ted Lemon wrote:
>> On Oct 14, 2014, at 9:27 AM, Gert Doering <> wrote:
>>> "flash renumber is a problem" is pretty much a non-argument, as flash 
>>> renumbering *will* happen, and devices in the home *will* have to handle it.
>> Indeed.   The question is, should we increase the number of instances in which they are forced to handle it, or no?
> Yes.
> Because this is the only way that application developers will learn to
> handle it.

+1, renumbering will happen and should be properly handled. 

A good example here is IPv6 renumbering from docsis domain splits in the cable world. In a typical case, a client would see a loss of upstream for 5-10 minutes followed with restoration with a new /56 GUA. Although techniques are available to limit renumbering during these events, they aren’t always deployed, for practical, technical or political reasons. From various discussions between cable operators, the rule of thumb is that this would typically happen roughly every 12 months for any single residential user. Cable users in areas seeing rapid growth or new technology deployments could expect shorter timeframes between renumbering events (every 4-6 months for example, or even shorter).     

I like the approach Mikael described earlier. Keep the GUA until it expires, then generate a ULA if required, if nothing else exists. Keeping the GUA beyond expiration should be avoided IMHO. I’ve seen broken implementations do this and causing a lot of headaches. The removal of the old GUA prefix before it’s expiration is also a problem in many implementations, but that’s probably a battle for another day.