Re: [rrg] London

Patrick Frejborg <> Mon, 13 January 2014 19:13 UTC

Return-Path: <>
Received: from localhost ( []) by (Postfix) with ESMTP id 3C4CA1ADFD7 for <>; Mon, 13 Jan 2014 11:13:14 -0800 (PST)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -2
X-Spam-Status: No, score=-2 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FREEMAIL_FROM=0.001, SPF_PASS=-0.001] autolearn=ham
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id O19exHHF6I1D for <>; Mon, 13 Jan 2014 11:13:12 -0800 (PST)
Received: from ( [IPv6:2607:f8b0:400d:c00::236]) by (Postfix) with ESMTP id D5D471ADF48 for <>; Mon, 13 Jan 2014 11:13:11 -0800 (PST)
Received: by with SMTP id i13so1987537qae.27 for <>; Mon, 13 Jan 2014 11:13:00 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type:content-transfer-encoding; bh=A6e9fncfKxCGtC0m1Q2NlqKWRbvvM8/oOA74tsxLLp0=; b=0A/0pICf2p0ac32Th7EqLq9QQiKMaIQO3Ystl/Y1lslXVzrhAfjgfHWESG1xFOixu9 G7tKMq07vij59NI4rinJsWOifyddcj3NH8MnxPlfXZI6HqANuoq0suwttv1sKu8D/d4Y J8VPMqRSwwMICzQXYg7pYysOyhwnYvdGgtB6zINNxZP8e/cpdK7Z6hIIEgwXowd77IKv 6UwEkrGEUODSnfdsw3696WxFdDjmTb6/skFqeebNXmXK6VFE7Psg6bGGAIWGgD8puELQ y2JZ3sbDljhPt+dLz9TE9I+CpPy6yffcDSSPmu6qWKtiXAsFtgaVS6YgjGEF6wRQQWTi XiwA==
MIME-Version: 1.0
X-Received: by with SMTP id v3mr42920964qar.55.1389640380443; Mon, 13 Jan 2014 11:13:00 -0800 (PST)
Received: by with HTTP; Mon, 13 Jan 2014 11:13:00 -0800 (PST)
In-Reply-To: <>
References: <> <> <> <>
Date: Mon, 13 Jan 2014 21:13:00 +0200
Message-ID: <>
From: Patrick Frejborg <>
To: Tony Li <>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: quoted-printable
Cc: RRG <>
Subject: Re: [rrg] London
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: IRTF Routing Research Group <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Mon, 13 Jan 2014 19:13:14 -0000


here is a shotshell for your shotgun to consider, and also for the
group to consider as a research topic.

Currently there is no papers or presentations available, it is an
early idea based on a combination of several research works - if it is
doable it could produce some new research papers from this group.

The basic idea is:

1) Take ILNPv6 as a baseline.

2) Add a new routing header extension to ILNPv6, it will contain some
locator fields. It must be compatible with RFC2460, some tweaks are
needed but it has to obey the spirit and format of RFC2460.

3) From RFC6306, take the concept of realms and for the "segments
left" field in the new routing header field, take the conpets of exit
routing, DFZ routing and approach routing - apply those on ILNPv6 with
the new routing header extension. However, "segments left" field has
to be applied in such way that it do obey RFC5095 in DFZ routing path
of a packet's journey (this is the tweak).

4) The DFZ must remain backwards compatible.

5) With steps 1-3 in place we can introduce a new type of identifier,
a service identifier with a "late binding", see

6) Multipath transport protocols, such as MPTCP and SCTP will add a
new option of multi-homing and mobility to ILNPv6 ("new option" is the
wrong semantics here, but anyway) - ILNPv6 with the host identifier is
great for business-to-business sessions (e.g. LAN-to-LAN VPN tunnels)
but I think that a multipath transport protocol with a service
identifier solution is better suited for the cloud and mobility era of
the Internet.

So here is the shotshell in a nutshell ;-)
It is up to the group if we can think this might be doable and willing
to spend some time researching on this specific topic.


On Mon, Jan 13, 2014 at 7:01 AM, Tony Li <> wrote:
> Lixia,
> The whole point of London is a broad survey of routing topics with exactly the goal of trying to clarify the topic(s) that the RRG is interested in.
> In some circles, this is known as the 'shotgun approach'.  ;-)
> Tony
> On Jan 12, 2014, at 10:12 AM, Lixia Zhang <> wrote:
>> I may not necessarily agree with the technical specifics of Heiner's msg, but I do think the msg made a good point: to provide useful input into RRG London agenda, we need to first have a clear goal(s) for that meeting.
>> Just volunteering a talk for whatever latest results multiple individuals have achieved may or may not help RRG to move forward.
>> So what do we as a group want to achieve?
>> Reactivating RRG would be the results of identifying that goal (the group would march towards that clearly articulated goal).
>> Or maybe the goal of London meeting could be to identify/clarify the goal?
>> (in that case, Heiner's msg did make one position statement as an input, for the group to consider)
>> Lixia
>> PS: I must admit that I have NOT had time to go thru the 40 or so RRG msgs since Vancouver IETF, ignore my comments if the goal had been clarified earlier (other than "reactivating RRG").
>> On Jan 11, 2014, at 12:24 PM, Heiner Hummel <> wrote:
>>> Tony,
>>> Given that the IESG doesn't care about IPv4 anymore, what kind of papers are you looking for? IPv6 papers ? IPv7 papers, based on WHAT ?
>>> For anyone it would be easier to outline a routing architecture while 16 octets are available instead of 4. But IPv6 is grounded from the start. Not even the fact that  IPv4 runs out of addresses will help to unleash IPv6.
>>> Combined with NAT,  IPv4 won't effectively have a deletion problem. Not for the next thousand years.
>>> Technologically NAT is a sin  fall, it hampers internet services  enormously. But this is true with IPv6 as well. So what!
>>> Backward compatibility and incremental deployability should be the two MUSTs for any architecture, no matter who proposes it. IPv6 isn't backward compatible due to the dual stack issue - right? or is there any way around?
>>> What is expected from any new architecture? To be backward compatible with IPv4 ( I would think so), to be backward compatible with IPv6 (I wouldn't think so)? So what is expected?
>>> And of course what do you expect to be solved, respectively to be enabled at least ?
>>> ----
>>> Another Proposal:  This group should come up with ideas about what a networking layer should do/support/enable WITHOUT being restricted by any existing technical solution.
>>> Two examples: 1): Prefix building per se is not a service. It is only a subdue instrument to cope with problems of a minor design.
>>> 2): State-less multicast would enable an enormous progress ( Imagine TV-broadcast to billions of spectators). But this is disabled due to the crazy belief that there is only one way  a FIB can be designed (namely as the existing FIB looks like)
>>> and because a Multicast-FIB cannot reasonable be cast like the existing Unicast-FIB there cannot/must not be state-less multicast. Period.
>>> Heiner
>>> Am 09.01.2014 um 10:12 schrieb Tony Li <>li>:
>>>> Hi all,
>>>> We are still looking for excellent papers for our workshop in London.  We really need your help in collecting these.  If you know of interesting research, could you please contact the authors and ask if they would be willing to present in London?  It's coming very soon…
>>>> Thanks,
>>>> Tony
>>>> _______________________________________________
>>>> rrg mailing list
>>> _______________________________________________
>>> rrg mailing list
> _______________________________________________
> rrg mailing list