On Aug 24, 2006, at 11:11 PM, ext Jari Arkko wrote:
Tim,
Its probably best if you now update your draft with a better
description
of what scenario you are looking at, details about the customers
requirements, justification of why new work is needed, and an analysis
of why existing solutions are inappropriate or undesirable in your
scenario. This has already been a long discussion but in my opinion
it has not been all that informative yet with regards to these issues.
Similar thoughts on my part.
Further, I would like get a better understanding as to why RA was
chosen for this. From my viewpoint the RA/RS is a very nice
mechanism for a router to tell all the hosts on a link common
information (e.g., prefixes, link MTU, lifetimes, etc.). ND does
have the ability to run in a unicast mode where different answers
could be provided depending on the address of the node sending the
solicitation, but this not something we have thought very much about
nor specified. I think it could be made to work, but it would need a
lot more thought and discussion.
Also, RAs were, of course, designed to allow routers to send
information to hosts. The proposal here is, if I understand it
correctly, is to use RAs to provision other routers. Not the
original intent.
It would be great if these issues could be discussed in a revised draft.
Thanks,
Bob
Also, at least I personally want to know what problem we are
solving before spending a lot of effort in analysing a solution --
we know there is an infinite number of ways to design a protocol
but the real question is whether it solves a problem.
Looking forward to your new draft.
--Jari
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------