Re: [rrg] Terminology - CES & CEE again

Robin Whittle <rw@firstpr.com.au> Thu, 04 February 2010 06:00 UTC

Return-Path: <rw@firstpr.com.au>
X-Original-To: rrg@core3.amsl.com
Delivered-To: rrg@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id B4BDB3A6C9A for <rrg@core3.amsl.com>; Wed, 3 Feb 2010 22:00:47 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.654
X-Spam-Level:
X-Spam-Status: No, score=-1.654 tagged_above=-999 required=5 tests=[AWL=0.241, BAYES_00=-2.599, HELO_EQ_AU=0.377, HOST_EQ_AU=0.327]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id DaBnR-K018pA for <rrg@core3.amsl.com>; Wed, 3 Feb 2010 22:00:46 -0800 (PST)
Received: from gair.firstpr.com.au (gair.firstpr.com.au [150.101.162.123]) by core3.amsl.com (Postfix) with ESMTP id 4215C3A6C76 for <rrg@irtf.org>; Wed, 3 Feb 2010 22:00:45 -0800 (PST)
Received: from [10.0.0.6] (wira.firstpr.com.au [10.0.0.6]) by gair.firstpr.com.au (Postfix) with ESMTP id DB4A2175E39; Thu, 4 Feb 2010 17:01:27 +1100 (EST)
Message-ID: <4B6A62B5.7080904@firstpr.com.au>
Date: Thu, 04 Feb 2010 17:01:25 +1100
From: Robin Whittle <rw@firstpr.com.au>
Organization: First Principles
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
To: Xu Xiaohu <xuxh@huawei.com>
References: <002b01caa54d$e54f3a40$090d6f0a@china.huawei.com>
In-Reply-To: <002b01caa54d$e54f3a40$090d6f0a@china.huawei.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Cc: rrg@irtf.org, Randall Atkinson <rja@extremenetworks.com>, 'Noel Chiappa' <jnc@mercury.lcs.mit.edu>
Subject: Re: [rrg] Terminology - CES & CEE again
X-BeenThere: rrg@irtf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: IRTF Routing Research Group <rrg.irtf.org>
List-Unsubscribe: <http://www.irtf.org/mailman/listinfo/rrg>, <mailto:rrg-request@irtf.org?subject=unsubscribe>
List-Archive: <http://www.irtf.org/mail-archive/web/rrg>
List-Post: <mailto:rrg@irtf.org>
List-Help: <mailto:rrg-request@irtf.org?subject=help>
List-Subscribe: <http://www.irtf.org/mailman/listinfo/rrg>, <mailto:rrg-request@irtf.org?subject=subscribe>
X-List-Received-Date: Thu, 04 Feb 2010 06:00:47 -0000

I am perplexed that Ran Atkinson (msg05940) has joined Joel Halpern
(msg05867) and Lixia Zhang (msg05897) in questioning the
architectural validity and importance of the distinction between CEE
(Core-Edge Elimination) and CES (Core-Edge Separation) proposals.

Ran, you wrote:

> I don't find the "CES" or "CEE" terms to be very meaningful,
> in that they don't really inform one about the important
> properties of any proposal.

As I wrote to Joel and Lixia, I would really appreciate you writing
why you disagree with:

   CES & CEE are completely different (graphs)
   http://www.ietf.org/mail-archive/web/rrg/current/msg05865.html

Neither you, nor Joel, nor Lixia have given any explanation why you
think this.  Perhaps the easiest way of doing so would be to point
out what is wrong with (msg05865), in which the architectural
distinctions are made quite explicit, along with their importance to
the RRG.

> Just as Patrick disagrees with how hIP4 has been characterised,
> I don't really agree with how ILNP has been characterised.

OK - its not much good just stating your opinion.  Enquiring minds
want to know why you think this.

What is wrong with me or other people asserting that ILNP is a CEE
architecture?  Like the others, it enables the elimination of "edge"
addresses - PI addresses - and it also does not require the creation
of a new class of "edge" addresses as is done with a Core-Edge
Separation architecture such as LISP or Ivip.

How do you think ILNP should be characterised?

I wrote that hIPv4 was a CEE after looking briefly at the summary.
Patrick tells me I am mistaken, and he has given some descriptions
which make me think that it has at least some things in common with a
CES architecture, without any resemblance to a CEE architecture
(msg05923).  So at present I don't regard it as a CEE proposal.  I
hope to read the whole hIPv4 proposal soon.


> I don't see the value in trying to continue use of those terms
> within the Routing RG context.

OK - but why do you think they don't reflect important distinctions
between proposals?


> Probably best if we all just move on, and keep trying to find
> meaningful ways to characterise the various ideas floating within
> the RG.

I and others have tried to find meaningful ways of characterising the
proposals.  We think the CEE and CES distinction is important.  I
argued why in (msg05865).  You, Lixia and Joel just wrote that you
disagree, without any indication why.

ILNP and the other proposals which I and others regard as being CEE
architectures involve a completely different naming model from that
of today's IPv4 and IPv6, in that the new model has separate objects,
in separate namespaces, for the roles of Identifier and Locator.
Another way of saying this is that CEE architectures, including ILNP,
implement "Locator / Identifier Separation". (Not to be confused with
LISP - which is a CES architecture and does not implement "Loc/ID
Separation".)

Many people on this list, including Tony Li, consider this naming
model to be "doing it right".   I think Tony - and you - believe the
Internet would be so much better with this model that it is worth
upgrading all the host stacks and rewriting all the applications to
support this model.

I disagree - I think it would be a major step in the wrong direction:

  Today's "IP addr. = ID = Loc" naming model should be retained
  http://www.ietf.org/mail-archive/web/rrg/current/msg05864.html

I don't think everything about CEE is bad.  For instance, it does not
involve tunneling - so there is no problem with Path MTU Discovery,
which I now realise is a terrible mess, thanks to Fred's recent research:

  Fred's IPv4 PMTUD research: RFC1191 support frequently broken
  http://www.ietf.org/mail-archive/web/rrg/current/msg05910.html

Except for Ivip with upgrades to all DFZ routers, any CES
architecture needs to tunnel traffic packets en-masse, and it looks
like it is going to be difficult to handle the PMTUD problems this
raises.


Xu Xiaohu wrote:

> Somebody argues that LISP is an edge-centric (i.e., CES) solution, then what
> about LISP-MN? Meanwhile, somebody argues that HIP and RANGI are
> host-centric (i.e., CEE) solution, then what about HIP proxy and RANGI proxy
> mechanisms? 

The CES vs. CEE distinction does not arise from whether hosts are
altered or not.  It arises from the fundamentally different
mechanisms which are used by these two different types of
architecture to achieve scalable routing.

I don't know enough about HIP proxy or RANGI proxy to comment, but if
you understand them, then my (msg05865) gives you a good basis to
decide how closely they resemble either a CES or CEE architecture.

> IMHO, these two options are not mutually exclusive. On the contrary, they
> should be complementary for each other. Each one is suitable for its
> corresponding stage of the Internet architecture evolution path.

I am not arguing that host changes should not be contemplated.  I
observe that CES architectures do not require any host changes and
that CEE architectures require changes to all host stacks and all
host applications in the world - in order to support any one adopter
having all their communications gain the portability, multihoming and
inbound TE benefits.   But these are observations of two kinds of
architecture - see my (msg05865) for the architectural distinctions
themselves.

 - Robin