RE: RDDP status of drafts

Caitlin,
   Even though for the consensus I will say that it is OK to remove the
TCP SCTP adaptation for the Active-Active requirements, I still think it
is worth the effort to probably have a separate effort in defining those
elements for Active-Active model and also a mix of an active-active and
an active-passive for a good failover solution that can be coined in
this area.
   The more I think about the more it makes sense to have some sort of a
standardization in this area for a "near-zero packet loss architecture"
to be defined under this arena. This may become a generic mechanism so
that can it can applied to a non-rdma space too.

Thanks
SG


-----Original Message-----
From: Caitlin Bestler [mailto:[email protected]] 
Sent: Tuesday, March 02, 2004 8:46 AM
To: [email protected]
Cc: [email protected]
Subject: Re: [rddp] RDDP status of drafts


On Mar 2, 2004, at 2:12 AM, [email protected] wrote:

> -- SCTP Mapping
>
> draft-ietf-rddp-sctp-00.txt
>
> To become a Proposed Standard RFC.
>
> Despite the -00 version number, this is the fourth version of this 
> draft,
> and it is relatively mature.

Earlier I raised the question as to whether the TCP and
SCTP adaptations should be brought into alignment on
support for Active-Active session initiation.

I believe the consensus from reflector discussion was that
it was not worthwhile to add this functionality to MPA/TCP.

I suggest a quick poll at the WG meeting and/or reflector as
to whether the Active-Active support should be removed from
the SCTP mapping. As that legal sessions are defined by an
enumeration of legal patterns, removing one is a very simple
edit should that be the consensus.

I personally remain ambivalent on the issue, seeing both the
merits of flexibility and of eliminating test cases.



_______________________________________________
rddp mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/rddp