Re: IPv6 WG Last Call:draft-ietf-ipv6-ndproxy-02.txt

[Erik Nordmark <erik.nordmark at sun.com>]

Margaret Wasserman wrote:

> Could the current chairs perhaps comment on how they believe that the ND 
> Proxy document, as it currently exists, fits within the IPv6 WG 
> charter?  Is there anyone else who has an opinion on this?
>
> This is not merely a legalistic concern.  I am concerned that the people 
> who should be involved in the standardization an RSTP-based L3 bridging 
> technology may not be involved in the IPv6 WG.   In fact, they may not 
> even be aware that this work is underway here. Therefore, this document 
> may not have been developed and reviewed by a group that has the balance 
> of experience needed to reach valid IETF consensus on this work.

FWIW I'm also concerned that the RSTP part of the protocol might not be 
as well understood. I've discovered that at least STP makes assumptions 
about the failure of the underlying "transport", such as the failures 
being visible to the STP. This is satisfied when the "transport" is IEEE 
802.3 etc, because the link layer provides some form of "link test", but 
wouldn't be satisfied when just encapsulating STP in IPv6.

It *might be* that RSTP (IEEE 802.1D-2004 with its addenda and 
corrigenda, and not necessarily the stuff that vendors claim being 
RSTP), does not make that assumption, but before putting out a 
specification that carries IEEE 802.1D-2004 over IPv6, I think we should 
verify this with the experts in IEEE 802.1.

Alternatively, we can go limit the draft to handling the simple cases 
(with the 'P' bit as a way to detect loops and shut them off).

  Erik

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