pwe3@conference.ietf.jabber.com - 2002/11/21
[11:09] %% swb has arrived.
[11:10] <swb> Twas brillig, and the slithey toves
[11:10] <swb> Did gyre and gymbal etc.
[11:12] %% JavierA has arrived.
[11:14] %% hanzzz has arrived.
[11:22] %% amalis has arrived.
[11:23] %% jaym has arrived.
[11:24] %% jaym has left.
[11:24] %% jaym has arrived.
[11:27] %% paul.knight has arrived.
[11:28] <paul.knight> in the wabe and the mimroths outgabe.
[11:30] <paul.knight> Eric Rosen: generalize signaling for PW provisioning
[11:32] <paul.knight> (same presentation as in PPVPN)
[11:34] <paul.knight> Distributed VPLS - can't be provisioned with simple VPN-ID or VCid
[11:35] <swb> You should find a record of your reporting from yesterday
[11:36] <paul.knight> at http://www.jabber.com/chatbot/logs/conference.ietf.jabber.com/ppvpn/2002-11-20.html
[11:40] <paul.knight> Hamid Ould-Brahim: need a Globally unique ID
[11:42] <paul.knight> Q: what is process: does PWE3 take PPVPN rqmts into account?
[11:42] <paul.knight> Danny: to some extent
[11:43] <paul.knight> Luca Martini: PWE3 is chartered to make a maintenance protocol ... some extensions to support PPVPN requirements are reasonable
[11:44] <paul.knight> Dinesh Mohan: Some extensions proposed fit in context of GVPLS
[11:45] <paul.knight> Danny: we're defining extensible mappings, and key definitions, not protocols
[11:45] <swb> Marco Carugi
[11:46] <paul.knight> Marco Carugi: same people are in two groups (PWE3 and PPVPN) we want same signaling used in point-to-point should be compatible for VPNs
[11:47] <paul.knight> David Zelig: PWE3 MIBs
[11:47] * paul.knight has changed the subject to: PWE3 MIBs
[11:49] %% paul.knight has left.
[11:49] %% paul.knight has arrived.
[11:51] <paul.knight> goo timing for MIB doctor review
[11:54] * paul.knight has changed the subject to: TDM
[11:54] <paul.knight> Shahram Davari
[11:54] <paul.knight> "Why PWE3 should adopt TDMoIP"
[11:55] <paul.knight> Reasons why design team has adopted this.... (see presentation)
[11:57] <paul.knight> reuse existing technologies like AAL1, interworking with ATM (used by many cariers for circuit/loop emulation)
[11:58] <paul.knight> Easy interoperation, simple SAR functionality
[11:59] <paul.knight> ASAP reuse - (Any Service Any Port)
[12:00] <paul.knight> Conclusion: no risk, available HW/SW, interworking with ATM
[12:01] <paul.knight> Question of adoption by WG ... Danny: move this to mailing list
[12:01] * paul.knight has changed the subject to: Issues with Control Word & Packet Loss
[12:02] <paul.knight> Yaakov Stein
[12:02] * paul.knight has changed the subject to: Unstructured TDM over PSN
[12:03] <paul.knight> Shasha Vainshtein
[12:06] %% swb has left.
[12:11] <paul.knight> Q: Steve Casmer (?) Does the 47 bytes include AAL1 framing? Does the proposal meet the requirements of the Design Team?
[12:11] %% paul.knight has left.
[12:12] %% paul.knight has arrived.
[12:13] <paul.knight> Dave McDyson: not clear wht is being proposed, not related to published drafts
[12:13] <paul.knight> A: Yaakov will address this in his talk, next
[12:14] <paul.knight> Yaakov Stein:
[12:15] <paul.knight> Two drafts: unstructured and structured TDM - this is basic approach of the Design Team
[12:15] <paul.knight> Each will have a main mode and one or two optional modes.
[12:17] %% paul.knight has left.
[12:17] %% paul.knight has arrived.
[12:17] * paul.knight has changed the subject to: PWE3 Control Word
[12:17] <paul.knight> Yaakov Stein
[12:19] <paul.knight> Three different formats currently in use.
[12:20] <paul.knight> PWE3 shoould adopt a single control word structure.
[12:21] <paul.knight> Suggest: based on Martini control word
[12:21] <paul.knight> maybe use extended CW if needed (s by CEP)
[12:22] <paul.knight> Q: What is justification? Just a desire to have one format for one WG?
[12:23] <paul.knight> Yaakov: They all hold similar info.
[12:23] <paul.knight> Q: NO, they are different.
[12:24] <paul.knight> Mark Townsley: Yes, combine them. Martini approach of reserved field is easier to parse.
[12:25] <paul.knight> Ghassem Koleyni: Martini is best, it is used in ITU-T as well.
[12:26] <paul.knight> Shasha Vainshtein: Not sure we can do away with the three CWs.
[12:27] <paul.knight> Maybe have a layer 1 CW and a layer 2 CW.
[12:27] %% jaym has left.
[12:27] * paul.knight has changed the subject to: Packet Loss
[12:28] <paul.knight> Yaakov Stein
[12:29] <paul.knight> Effect of packet loss on voice quality
[12:29] <paul.knight> graph: results of using different algorithms:
[12:30] <paul.knight> previous sample, linear interpolation, statistical interpolation provide higher quality, in ascending order
[12:31] <paul.knight> Statistical interpolation requires computation, probably not appropriate
[12:31] <paul.knight> can keep toll quality service in the face of 2-3% packet loss
[12:32] <paul.knight> Read the draft draft-stein-pwe3-(.....) sorry - missed
[12:34] <paul.knight> Shasha V.: replacement of missing packet should be a local process, not have packets rebuilt by the network
[12:35] * paul.knight has changed the subject to: Bi-directional LSPs for classical MPLS
[12:35] <paul.knight> Rohit Dube
[12:36] <paul.knight> Scales better than unidirectional LSPs
[12:37] <paul.knight> Needs RSVP-TE extension
[12:37] <paul.knight> Other info already available.
[12:38] <paul.knight> Being extended to handle asymmetric bandwidth reservation
[12:38] * paul.knight has changed the subject to: Header Compression Drafts
[12:38] <paul.knight> Jerry Ash
[12:40] <paul.knight> End-to-End (VoIP | VoMPLS) Header Compression (two drafts)
[12:44] %% hanzzz has left.
[12:44] <paul.knight> http://www.ietf.org/internet-drafts/draft-ash-e2e-vompls-hdr-compress-00.txt
http://www.ietf.org/internet-drafts/draft-ash-e2e-crtp-hdr-compress-00.txt
[12:46] <paul.knight> tradeoffs: better compression with crtp, but needs to resynch under packet loss
[12:48] <paul.knight> Q: This is so different from other work here, it should be done elsewhere.
[12:49] <paul.knight> Danny: Lets take this to list to discuss its home.
[12:50] * paul.knight has changed the subject to: ITU-T SG13 Activities Regarding X-MPLS Interworking
[12:50] <paul.knight> Ghassem Koleyni
[12:51] <paul.knight> ATM-MPLS Network Interworking
[12:51] <paul.knight> cell mode, frame mode
[12:55] <paul.knight> Register with ITU-T to get access to the ITU-T documents at the URL in the presentation.
[12:56] * paul.knight has changed the subject to: Draft Updates
[12:56] <paul.knight> Luca Martini
[12:56] <paul.knight> ATM, Ethernet and contorl word drafts
[12:57] <paul.knight> (I'll never make a living as a typist.)
[12:58] <paul.knight> ATM draft:
[12:59] <paul.knight> will use ITU text for format
[13:00] <paul.knight> Ethernet draft
[13:00] <paul.knight> single mode for encaps
[13:00] <paul.knight> Control draft:
[13:01] <paul.knight> want comments from WG on draft-rosen-ppvpn-l2-signaling-02.txt
[13:02] <paul.knight> draft-ietf-pwe3-control-01.txt
[13:06] %% paul.knight has left.
[13:06] %% paul.knight has arrived.
[13:08] <paul.knight> Q: why include options unless they will be implemented? This will delay things.
[13:08] %% paul.knight has left.
[13:21] %% JavierA has left.
[13:25] %% amalis has left.