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

James Woodyatt <> Fri, 17 October 2014 21:41 UTC

Return-Path: <>
Received: from localhost ( []) by (Postfix) with ESMTP id A6E0C1A86DE for <>; Fri, 17 Oct 2014 14:41:30 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -0.984
X-Spam-Status: No, score=-0.984 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, PLING_QUERY=0.994, RCVD_IN_DNSWL_LOW=-0.7, SPF_PASS=-0.001] autolearn=no
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id BEMayPVkbQIK for <>; Fri, 17 Oct 2014 14:41:29 -0700 (PDT)
Received: from ( []) (using TLSv1 with cipher ECDHE-RSA-RC4-SHA (128/128 bits)) (No client certificate requested) by (Postfix) with ESMTPS id 2EAD81A702A for <>; Fri, 17 Oct 2014 14:41:29 -0700 (PDT)
Received: by with SMTP id id10so1254367vcb.34 for <>; Fri, 17 Oct 2014 14:41:28 -0700 (PDT)
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:content-type; bh=zw41ho1vEo3VCYdU+1cABOyDCgQO9oPHssvGvI4k0gE=; b=g7BTgpLx15o2LzG+afSdkuQDB5UyfF8dKjv5TAjU3dPK5SnNfo4vpTdn7VewMpyrd2 qTLHP+OfZkxzr2j3GOrZU9kDXwOI5/z7Kb4XA9oOpEi7ELiCphRXAVWCdBfp94RohejH HIVmL+m8/FwbM0SSIvRh82Y8t6a1JH9/QzRvqGsg6Bxrc0VWYKRe4Z4d/z2SSvacGDX8 YuELXbnAZs8sTKE0f6N0n8ZP2V57dFc2LXzNlmUGfEddH49H9fNQnmW7Li5vKOvjkAUs LykW6ve0fkvvVM0tB0LteNq9vKGrK3pR4tnXefMS9JtQPxpQkAWz9mfwri8qyq7hykQx DKgw==
X-Gm-Message-State: ALoCoQkNc0BaCzu1L65zOPrHo2whnknSmyfQhCJRY5av6F+ZC21D+9FYWWNnc+D/NFD620ahvRDG
MIME-Version: 1.0
X-Received: by with SMTP id su8mr8044290vdb.8.1413582088218; Fri, 17 Oct 2014 14:41:28 -0700 (PDT)
Received: by with HTTP; Fri, 17 Oct 2014 14:41:28 -0700 (PDT)
In-Reply-To: <>
References: <> <> <20141014142746.GX31092@Space.Net> <> <20141014145930.GY31092@Space.Net> <> <20141014154111.GZ31092@Space.Net> <> <20141015150422.GW31092@Space.Net> <> <20141015154841.GY31092@Space.Net> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <>
Date: Fri, 17 Oct 2014 14:41:28 -0700
Message-ID: <>
From: James Woodyatt <>
To: HOMENET Working Group <>
Content-Type: multipart/alternative; boundary=bcaec52d571d3ac0d30505a53a85
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: Fri, 17 Oct 2014 21:41:30 -0000

On Fri, Oct 17, 2014 at 12:54 PM, Ted Lemon <> wrote:

> On Oct 17, 2014, at 2:49 PM, James Woodyatt <> wrote:
> > As I recall, the proposals in your response were less than concrete and
> didn't solve the problems. In particular, I remain curious about how to
> expire the locally generated ULA prefixes that accumulate over repeated
> network joins and splits.  I remember explaining how those events could be
> rather more frequent than people might be assuming, and that's where the
> discourse seemed to stop.
> If you don't burn a ULA prefix every time you split and rejoin, then I
> don't see why there would be any kind of significant growth in prefixes.
>  That's why I suggested the algorithm that I suggested.

I explained why you must generate a new ULA prefix every time you
commission a new network.

It's true that not every split entails commissioning a new network,
especially when the AAA service for the previously commissioned network is
still available on both sides of the split, but that isn't always true, and
besides, human user behavior must still be accounted. Whenever you have AAA
service, you are attempting to automate what humans think about who is and
who isn't allowed to do what. Sometimes people change their plans.
Sometimes people don't plan at all. Sometimes people make plans and don't
tell anyone, much less the silicon beasties in their house. Sometimes they
make plans without even recognizing it. Routers have to operate mostly in
the dark when a split happens. They may not be able to determine whether
they are still expected to be members of the previously commissioned
network or if they have been repurposed into a new network until well after
they have been operating split for some time, and perhaps joined with other
networks in the meantime. After a split, they may need to operate in a mode
that allows them to commission a new standalone network at any time when
they are split from a previously commissioned network.

In any case, we have defined a way for locally generated ULA prefixes
distributed in persistent storage of HOMENET interior routers to accumulate
without limit when independently commissioned networks join. Offering the
view that sensible people don't commission new networks very frequently is
not a solution to the problem. It's a way of implying the problem isn't
relevant to your interests. The problem may not be relevant to your
interests, but it is to mine.

james woodyatt <>
Nest Labs, Communications Engineering