Multilink subnet issues and proxy/relay DAD

["Templin, Fred L" <Fred.L.Templin at boeing.com>]

The question of proxy/relay DAD for multilink networks that
comprise shared links has come up on the NETLMM and Autoconf
mailing lists. Proxy/relay DAD is a mechanism whereby NS(DAD)
messages are relayed to the link on which a node with a
colliding address resides, with the colliding node's NA(DAD)
being relayed back to the link on which the soliciting node
resides. (The Target Link Layer Address Option in the relayed
NA(DAD) must not be changed by the relays to ensure proper
SEND operation in this process.)

James Kempf has twice indicated that there have been prior
discussions on this subject between Jari Arkko and Dave
Thaler, and that they should be consulted for further
information (see below), but several attempts to contact
them have so far produced no results.

To ensure a fair and open dialogue at IETF67, I am requesting
that Jari, Dave, James and others with firsthand information
address the following questions on the lists prior to the
meetings:

  1. What are the issues wrt proxy/relay DAD that would
     interfere with its adoption as a standard mechanism?
  2. What harmful on-link assumptions could there be for
     IPv6 Prefix Information Options that advertise a
     shared prefix with 'L=0'?
  3. Does the RFC1812 "subnet forwarding model" still apply
     to IPv6, when there are no IPv6 documents that reference
     RFC1812 normatively?
  4. What other non-obvious issues relating to multilink
     subnets for shared links need to be observed for NETLMM,
     Autoconf and other contexts?

Fred Templin
fred.l.templin at boeing.com     

-----Original Message-----
From: James Kempf [mailto:kempf at docomolabs-usa.com] 
Sent: Wednesday, November 01, 2006 1:50 PM
To: Templin, Fred L; Behcet Sarikaya; Alexandru Petrescu
Cc: netlmm at ietf.org
Subject: Re: How does "per-MN prefix model" simplify things? (was:
RE:[netlmm]Re: PMIP follow-up questions)

>You mentioned a while back that there were some discussions
>involving IAB and/or IESG members that indicated there might
>be some barrier to acceptance of a proxy/relay DAD solution.
>Can you say anything more about that?

Primarily DaveT's multi-link subnet draft. I think it would be prudent
to 
discuss the issue with him before the WG makes a decision. Also, I think
we 
should check with Jari to find out what he thinks. It is necessary that
Jari 
agree with anything the WG comes up with if we are to get IESG approval.

            jak 


--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6 at ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------