ppvpn@conference.ietf.jabber.com - 2003/03/19
[11:16] %% paolo has arrived.
[11:16] %% paolo has left.
[11:45] %% brunod has arrived.
[12:10] %% brunod has left.
[14:12] %% paul.knight has arrived.
[14:12] <paul.knight> Test PPVPN jabber session. One person pls respond if u see this.
[14:20] <paul.knight> The sounds of Windows startup echoes through the room....
[14:23] <paul.knight> The projector screen is illuminated....
[14:25] <paul.knight> Marco speaks, and the session begins!
[14:26] <paul.knight> Presentation by Tissa S. will not be given.
[14:26] <paul.knight> Presentation by Michael Behringer will be added to agenda.
[14:27] <paul.knight> Anyone seeing this yet? I don't see anyone else.
[14:27] %% amalis has arrived.
[14:28] <amalis> Hi Paul.
[14:28] <paul.knight> Good, not totally isolated... Hi Andy.
[14:28] <paul.knight> Updated milestones being discussed.
[14:29] <paul.knight> Status of WG documents being discussed:
[14:29] <amalis> Marco is in his usual firehose mode.
[14:29] <paul.knight> framework.07
[14:29] <paul.knight> send to IESG for info RFC soon.
[14:30] <paul.knight> requirement.05: send to IESG in a few days
[14:31] <paul.knight> generic-reqts-o2.txt also to IESG soon.
[14:31] <paul.knight> ce-based-03 integrated into l3 framework, new ce-based-sol may cover it.
[14:33] <paul.knight> andersson-ppvpn-terminology-03.txt cleaned up, will be moved to WG soon
[14:33] <paul.knight> L2 Solutions to be addressed
[14:33] <paul.knight> metrics to be dropped
[14:33] <paul.knight> L3 solution space:
[14:34] <paul.knight> rfc2547bis to PS
[14:34] <paul.knight> vpn-vr to PS
[14:35] <paul.knight> some dependencies listed in the "protocol-dependencies" drafts
[14:36] <paul.knight> WG last call will follow for other 2547bis docs, VR docs if needed
[14:36] <paul.knight> L3 solution-specific MIBs
[14:36] <paul.knight> L2 space
[14:37] <paul.knight> (Please view the presentation file for the details, can't cover it all here)
[14:38] <paul.knight> L2 reqts ID target for submission May ?
[14:38] <paul.knight> L2 Framework - target late April or early May
[14:39] <paul.knight> Design teams - basically concluded; L2 DT will continue until completion of L2 reqts and L2 framework (soon).
[14:40] <paul.knight> L2 DT meetings synopsis (see presentation)
[14:42] <paul.knight> solutions documents should include section describing compliance to requirements as listed in the requirements docs.
[14:43] <paul.knight> Slide : status today (of L2 design team work)
[14:43] %% brunod has arrived.
[14:44] %% TonyLi has arrived.
[14:44] <paul.knight> some questions on clarity of reqts docs... hard to write solution docs to clearly state compliance with the reqts
[14:45] <paul.knight> WG discussion on mail list needed
[14:46] <paul.knight> Set of candidate L2 solutions - 4 on agenda today, but these are not the only ones; these are the ones which requested time to speaki today.
[14:47] <paul.knight> Experimental vs. Standards track: L2 design team somewhat in favor of moving some to experimental; need discussion on mail list.
[14:48] <paul.knight> Protocol extensions in PPVPN - need to have clear relationships with other WGs.
[14:49] <paul.knight> Alexs Zenin: PPVPN should not take on protocol work from other WGs
[14:49] <paul.knight> (Alex Zinin)
[14:50] <paul.knight> Marco: PPVPN frameworks for QOS, Security, ....(missed one)
[14:52] <paul.knight> Other items to be progressed - L2 Interworking (requirements); Service OAM - proposal to start from the Op and Mgt framework
[14:52] <paul.knight> Other items:
[14:52] <paul.knight> Multicast framework ?
[14:52] <paul.knight> L1/Optical VPNs (charter review, other bodies outside IETF)
[14:53] <paul.knight> Alex Z: WG has a lot to do now, hard to think about extending charter
[14:53] <paul.knight> Marco: agree
[14:54] <paul.knight> Presentation by Jeremy DeClerq
[14:54] <paul.knight> CE-based L3 PPVPNs progress
[14:55] <paul.knight> history recap... ce-based framework doc still alive; ce-based solutions is new
[14:56] <paul.knight> management responsibility - defined by SLA
[14:56] <paul.knight> routing clarification - CE operates in two routing realms
[14:57] <paul.knight> "door" is tunnel or firewall model
[14:57] <paul.knight> Internet access - addressed also
[14:59] <paul.knight> auto-discovery - configuration discovery; HTTP based; other protocols possible
[14:59] <paul.knight> Future steps: WG feedback needed - make it a WG doc?
[15:00] <paul.knight> Joe Touch, ISI: We have running implementation similar to what is described in document; it should be cited in the document.
[15:01] <paul.knight> It is similar to X-Bone system by USC-ISI.
[15:02] <paul.knight> Alexs Zinin: running code is not enough, need WG consensus.
[15:02] <paul.knight> JOe Touch: need to read literature, cite prior work.
[15:03] <paul.knight> Marco: please continue discussion on mailing list.
[15:04] <paul.knight> Jeremy: back to presentation: future steps needed on AS document
[15:04] <paul.knight> dmitri: parts of earlier document are in new document
[15:05] <paul.knight> End of Jeremy's presentation.
[15:05] <paul.knight> Hierarchy of Provider Edge...
[15:05] <paul.knight> Huawei (Lee not available for presentation)
[15:06] <paul.knight> Graphic presentation, will not try to capture much of it here...
[15:08] <paul.knight> Li Defeng not here - anyone know name of presenter?
[15:11] <paul.knight> "01.txt" version not yet on system; would like to move this to a WG doc if possible
[15:11] <paul.knight> Eric Rosen: This is simply a deployment model; it is not in the scope of the WG
[15:12] <paul.knight> Eric: there is no interoperability problem identified with the 2547bis model, so this is not really in the charter of the WG.
[15:13] <paul.knight> Marco: agree there is not an impact on the protocol of 2547bis .. any other comments besides from Huawei and Cisco?
[15:13] <paul.knight> Especially service providers?
[15:14] <paul.knight> Speaker: agree, not bring up into WG
[15:14] <paul.knight> Marco: discuss on list; see if there is service provider interest.
[15:15] <amalis> The speaker was Robert Jaksa.
[15:15] <paul.knight> Michael Behringer: MPLS VPN Import/Export Verification presentation.
[15:17] <paul.knight> draft-behringer-mpls-vpn-auth-01.txt
[15:17] <paul.knight> link the MD5 authentication keys from CE-PE-PE-CE
[15:18] <paul.knight> Quick comparison table - with Bonica draft
[15:20] <paul.knight> Looking to get feedback on whether this is of interest, and should become a WG item...
[15:20] <paul.knight> Mark Duffy: this does not provide security beyond the Service provider; like having the fox watch the henhouse.
[15:21] <paul.knight> Alex Zinin: does this change the MD5 between the PEs? Does it affect per-route authentication?
[15:22] <paul.knight> A.Z. there is a disconnect of authentication - need to see if this is addressed.
[15:23] <paul.knight> Marco - see if this can be worked in the security framework; also see how it meets all requirements met by Bonica draft.
[15:24] <paul.knight> Speaker: this is attractive because it needs no changes on CE; this can be deployed soon.
[15:24] <paul.knight> Vach Kompella: Hierarchical VPLS presentation
[15:25] <paul.knight> - Evolution of draft...
[15:26] <paul.knight> Future: we have some "experience with the service"; interoperability between 6 vendors
[15:26] <paul.knight> Request this become a WG document.
[15:27] %% gih has arrived.
[15:28] <paul.knight> Robert Jaksa: why change VC type?
[15:28] <paul.knight> Vach: behavior of pseudowire is like Ethernet
[15:30] <paul.knight> Eric Rosen: Doc is pretty good: some issues - this is written as an MPLS-specific solution; iot doesn't have to be that way; lets change it before it becomes a WG doc.
[15:31] <paul.knight> Vach: use a VPN-ID as a generic field; it could contain anything; avoid tying it to any protocol.
[15:31] <paul.knight> Vach: need a wider audience looking at this.
[15:32] <paul.knight> Marco: not prolong things, but let's have a mailing list discussion.
[15:32] <paul.knight> A.Z.: if it is of interest, it can become a WG document; doesn't have to be perfect at the beginning.
[15:34] <paul.knight> Robert Jaksa: it is not general enough to be a WG document
[15:34] <paul.knight> Marco: we need people to work on it, not just say it is interesting.
[15:35] <paul.knight> Alex Zinin: a WG document should be able to incorporate changes better.
[15:36] <paul.knight> HUMMMM: There is interest in making this a WG document.
[15:36] <paul.knight> Marco: There is enough interest that we should get the mailing list consensus to move this to a WG doc.
[15:38] <paul.knight> ISOCORE interoperability test: Please look at the presentation on the WG web site. (Rajiv ??)
[15:38] <paul.knight> Vasille Radoaca: GVPLS presentation
[15:39] <paul.knight> Characteristics of GVPLS ; reference model
[15:39] <paul.knight> List of key elements for -01 version
[15:41] <paul.knight> additional requirements addressed by -01 version...
[15:42] <paul.knight> future: convergence of signaling with draft-rosen-ppvpn-signaling-02.txt
[15:43] <paul.knight> Robert Jaksa: relation of this to other work on hierarchical VPNs?
[15:43] %% TonyLi has left.
[15:44] <paul.knight> Speaker: packet sequencing problem? Need to define this as an issue.
[15:45] <paul.knight> Marco: debate this draft on mailing list. Other drafts are not being presented here; may not be fair to progress drafts to WG status based on whether they were presented here.
[15:46] <paul.knight> Cheng-Yin Lee:
[15:46] <paul.knight> Hybrid Virtual Private LAN
[15:47] %% gih has left.
[15:48] <paul.knight> Proposal intended to meet majority of end customer requirements now; not wait for most other VPLS proposals to be standardized
[15:51] <paul.knight> customer-located devices (CLE) managed by providers; connect to CEs on customer site and to PE
[15:53] <paul.knight> speeding up... many slides are covered in the draft
[15:54] <paul.knight> Other CE routers are visible at IP level.
[15:55] <paul.knight> Speaker: (Ali ?) full mesh needed between CLEs
[15:56] <paul.knight> CY Lee: spanning tree is run between CLEs over the SP; one per VPN
[15:57] <paul.knight> SPeaker: spanning tree needs to run across many metro networks across a wider geography; some scalability if crossing many metro regions.
[15:57] <paul.knight> CY Lee: BPDUs are tunneled, not a big problem with scalability
[15:58] <paul.knight> Eric Rosen: CLE here is really the CE in other discussions. This is really not what we need to address in the WG; although it may be a good service.
[15:59] <paul.knight> Juha Heinanen - presentation on RADIUS-based VPN discovery
[15:59] <paul.knight> There is a need for BGP-free operation of VPN discovery.
[16:00] <paul.knight> BGP cannot be cost-justified in all PEs, example in a multi-tenant unit.
[16:00] <paul.knight> RADIUS has needed attributes and scalability
[16:02] <paul.knight> description of message formats
[16:02] <paul.knight> graphic example of operation
[16:06] <paul.knight> Is there support for this?
[16:07] <paul.knight> Marco: interest? Many hands raised.
[16:09] <paul.knight> Marco: Another questions:support for making the L2 Solutions documents whould be progressed as Experimental rather than Standards track?
[16:09] <paul.knight> Vach Kompella: there was some question at the DT meeting.
[16:10] <paul.knight> Alex Zinin: info from Design teams does not have high weight.
[16:10] <paul.knight> Kireeti Kompella: I have different memory from my brother...
[16:11] <paul.knight> Eric Rosen: people remember their position as what was decided or agreed to. There is not a good consensus on this; or even good understanding of the experimental/standards track tradeoff.
[16:12] <paul.knight> A.Z. Experimental requirements are well defined (2026). Let's decide soon on this. We may need an interim meeting of the WG.
[16:13] <paul.knight> Kireeti: I agree with Eric's comment on memory of the group decisions. The DT meeting is not binding anyway.
[16:13] <paul.knight> AZ: design team is not binding; but it is of interest.
[16:14] <paul.knight> KIreeti: make these experimental, work on them; then pick the better ones based on experience.
[16:14] <paul.knight> Vach: don't just move a bunch of things to experimental because we can't make a decision.
[16:15] <paul.knight> Marco: Vach's comment is reasonable.
[16:16] <paul.knight> More presentations: security framework
[16:19] <paul.knight> Fabio Chiussi - PPVPN QOS
[16:20] <paul.knight> Framework document
[16:21] <paul.knight> new version -01.txt
[16:22] <paul.knight> future changes: need to make more readable, merge with draft-declerq-ppvpn-l3vpn-qos-00.txt
[16:23] <paul.knight> This will merge L2 and L3 approach and adopt Jeremy's more clear approach.
[16:25] <paul.knight> An other presentation on the DeClerq draft just mentioned (by Fabio).
[16:26] <paul.knight> Identified issues listed..
[16:27] <paul.knight> Goal: QOS framework draft as the reference doc. for other Qos drafts defining related solutions/signaling/extensions/MIBs, etc.
[16:29] <paul.knight> Speaker: Gustov Ausaul (?) Alcatel: issues were supposed to be addressed as I understood, but this is merging things, not addressing the issues
[16:31] <paul.knight> Speaker: not sure if this is in the scope of the WG -- it is deployment guidelines, etc. In the second one, there are some good items which could be in 2547bis. Not clear this adds to results of the WG, just more documents.
[16:33] <paul.knight> Marco: the point is to have this abstracted from the other documents so they can progress, not to include them in the documents in each of the solutions. We will not kill this at this time we will put some constraints and dependencies on this process.
[16:33] <paul.knight> Speaker was David (?)
[16:34] <paul.knight> New presentation:
[16:34] <paul.knight> Luyuan Fang: Security Framework for PPVPN
[16:36] <paul.knight> document to be used to analyze properties of the PPVPN solutions
[16:37] <paul.knight> try to get WG consensus on what should be in/out of this doc.
[16:37] <paul.knight> OUtline discussed....
[16:38] <paul.knight> Provider and User security requirements are related, but different.
[16:40] <paul.knight> Avoid slowing down existing work.
[16:41] <paul.knight> Ananth Nagaranjan: coordinate with Applicability statements?
[16:42] <paul.knight> Marco: we are out of time, we can't have the presentation on Management (OAM, MIB, customer network mgmt)
[16:42] <paul.knight> Signing off!
[16:42] %% paul.knight has left.
[16:44] %% amalis has left.
[16:48] %% brunod has left.
[17:25] %% paul.knight has arrived.
[17:25] %% paul.knight has left.