Ad-hoc Meeting on Peer-to-peer SIP

Friday, July 14, 2006, 1230-1500, Room 519A
Minutes edited by Gonzalo Camarillo and Mary Barnes
Based on notes by Spencer Dawkins and Dean Willis
Meeting chaired by Mary Barnes and Gonzalo Camarillo

Agenda Bash

Led by Chairs

Slides

The chairs highlighted that this was an official SIPPING WG adhoc meeting, thus the Note Well statement does apply. The objective of the session is to get agreement on the basic concepts and terminology and ensure that the problem to be solved in the proposed BoF/WG is well scoped. The current plan is to have a BoF in San Diego (IETF-67) with the intention of forming a new RAI WG.

Concepts and Terminology

Led by Dean Willis

Slides

Relevant documents: draft-willis-p2psip-concepts-00.txt

Dean presented a basic overview of the model proposed for P2P SIP. Clarifications of some key terms was provided such as the difference between a P2P client and Peers, as well the difference between Enrollment and Insertion. It was clarified that the "Network" should be removed from the document.

Charter Discussions

Led by David Bryan

Slides

David presented 4 key decision points for the group in terms of scoping the work.

1. What gets standardized? The proposal was for 2wo protocols - an overlay client protocol and an overlay peer protocol. Could be the same protocol, but don't know that yet. At least one could be vanilla SIP, but don't know that yet, either.

   Poll: Do folks agree with the proposal? Lots of hands are OK, 80 percent. Disagree? none in this room.

2. Are topologies with NATs in scope? This question was divided into 3 subquestions: Do NATs need to be addressed at all? Is locating media relays for NAT traversal required ? If so, should existing IETF work be reused? Proposal was "Yes" for all the subquestions.

   Poll: "yes" before we asked the question - provided there's wording in the charter on reusing ICE, etc. Almost unanimous agreement on the chart. Disagree? No one spoke up.

3. Can we limit the scope to name lookups? Proposed answer: Yes. Very analogous to SIP registrar functionality.

   Poll: All yeses for limiting scope to name lookup.

4. Anonymization service?

   Poll: Interested at all in solving this problem? 15 against, 30 for.

   Poll: Should this be excluded completely? 8 for.

   Poll - Dean's compromise: This is very important, but delivery could be handled separately - "nothing that precludes" would be sufficient. Thus, proposing to not put in initial charter. 40 support this compromise proposal.

   Poll: Include in charter? None for.

With some final comments made that the anonymization is fundamentally bound up in SIP, and you can't fix it without changing SIP. Gonzalo emphasized that any SIP changes required for this work will go through SIP change process.

The importance of starting the chartered work with a small number of milestones was emphasized, with a threat analysis document being suggested as one of the initial documents, which lead to a discussion of the fundamental security requirements for this work. The suggestion was made to charter with explicit security tradeoffs, but without requiring extensions to the state of the art.

A suggestion was made since there was time left after David's presentation to work through the detailed charter changes. However, the majority believed that could be resolved outside the meeting based on the feedback received.

Conclusion: Proposed charter needs updating to reflect the tentative consensus, with discussion to continue on the mailing list.