[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Discussion of draft-ietf-v6ops-ent-analysis-02.txt



I'm not at all clear that DSTM is being recommended by the working group - or any specific transition mechanism. Our charter specifically excludes discussion of transition mechanisms, and I note that the IESG recently approved Teredo as a Proposed Standard, which I think means that the IETF recommends that. And personally, I'm not convinced that any of the proposed transition mechanisms are sufficient, as I don't see documented anywhere a single transition mechanism simultaneously solves all four problems:

- an IPv4-only system finding and communicating with another IPv4-only system across an IPv6-only infrastructure
- an IPv6-only system finding and communicating with another IPv6-only system across an IPv4-only infrastructure
- an IPv4-only system finding and communicating with an IPv6-only system
- an IPv6-only system finding and communicating with an IPv4-only system


My history with this document and with the DSTM proposal is that I took over the WG last March. In the WG meeting in march, DSTM was not discussed, and the discussion of the enterprise analysis identified a number of items to be corrected. We have just gotten the updated document, and it needs thorough review.

On May 20, 2005, at 9:14 AM, Margaret Wasserman wrote:


Does the WG really have consensus on this recommendation?

 Later in the transition process, after the enterprise has
 transitioned to a predominately IPv6 infrastructure, the architect
 should implement the Dual-IP Transition Mechanism [DSTM, DSTM+]. Or
 in the case of early deployment of IPv6-dominant networks DSTM can
 be used too.

This whole document comes down to a brief set of recommendations, and I think that it is important that we only make recommendations that have the clear consensus of the WG.

I am not personally convinced that we should recommend DSTM as a solution for this stage of network transition (I haven't even read DSTM+, so I have no opinion on that), but perhaps I am part of the rough? Fred, was there an explicit consensus call on each of these recommendations that I missed while I was busy somewhere else?

If the WG does have consensus that, at some point in an enterprise transition, administrators "should implement DSTM", what is our plan to standardize it?

Margaret