Layer 1 VPN WG (l1vpn) ====================== TUESDAY, March 20, 2007 0900-1130 Morning Session I CHAIRS: Hamid Ould-Brahim Tomonori Takeda Adrian Farrel AGENDA: ================================================================ Agenda and admin (chairs) ================================================================ ================================================================ WG Status and milestones (chairs) ================================================================ Hamid went over the slides and showed WG status and milestones. Tomonori: For the basic mode applicability statement I-D, we are planning to add references to existing MIB modules and see whether new work is needed. ================================================================ Optimization of the auto-discovery information (Don) draft-ietf-l1vpn-bgp-auto-discovery-01.txt also draft-ietf-l1vpn-basic-mode-01.txt ================================================================ Don went over the slides, explaining the changes from 00 version. A new NLRI format was proposed, and feedback was solicited. Lou: What was nice in the previous version is the format was common. Are you going to continue to keep the format in the basic mode framework document? Don: I think the format should be common, but it does not need to be in a common document. Lou: Do we need multiple format definitions? Suggest to remove the format from the BGP document. Hamid: Do you agree with the change? Lou: I need to think about BGP semantics to decide how to map to OSPF context. I am not yet convinced by changes, but my main concern is commonality. Yakov: Having a common format is good, but sometimes we have to go separately. Lou: Well, better to have common definition, encoding is next. Hamid: Signaling document was defining encoding. The technical point is to be able to not change signaling when you change routing protocols. Lou: My reservation is that we should have the common definition. Yakov: If we can do it then OK. Nice, but not a MUST Tomonori: Not yet clear whether we need a change in signaling. Don: I don't think so. It is just BGP autodiscovery. Adrian: So is there no change in signaling? Don: I don't think so. But we should check it. Tomonori: How about in shuffling? Shuffling needs to carry PPIs. Don: I think so. That informaiton is still local. Hamid: In inter-domain and inter-provider, its advantage is scalability. But we can't use multi-domain as a reason, because it off charter. Yakov: Also it is exactly the same model as l2vpn and l3vpn. Another benefit is how much you have to advertise when there is a change in the customer configuration. So it allows changes to the PPI information and moving PPI from one port to another without requiring network re-advertisement of BGP auto-discovery information. Hamid: Let's discuss in the mailing list. Lou: On OSPF auto-discovery, we don't have any slide, and the draft is ready to go. The only issue is the format. Better to wait for the basic mode framework, keep the draft updating not to expire, and then issue a last call. Hamid: BGP-TE will be presented in IDR WG. Yakov: I don't think this is an IDR WG document. How about a L1VPN WG document? Adrian: Not really sure. Needs to see the interest at least. Yakov: Is it OK to poll the mailing list? Adrian: How many people read the draft? (six or seven hands) Lou: The question is where this belongs, then whether this is a WG document or not. Igor: What happened for OSPF auto-discovery is that we took the common function to the OSPF WG. It is the same. Adrian: We will wait for the outcome of the IDR WG first. Yakov: OK. ================================================================ OSPF Based L1VPN Auto-Discovery (Lou) draft-ietf-l1vpn-ospf-auto-discovery-01.txt ================================================================ Lou Berger presented a quick update without using slides. The I-D is stable but there is a new issue for encoding format as just discussed. If the base mode document is updated to catch the encoding change, then this I-D might not need any work. ================================================================ Enhanced mode analysis (Dan) draft-li-l1vpn-enhanced-mode-analysis-01.txt ================================================================ Dan went over the slides, explaining the analysis of four sub-models. He pointed out that more comments and feedback from WG is needed. Adrian: What is the answer? You have done very useful work. Have you reached some conclusion? Dan: We just listed four sub-models. We do not have any clear answer for this yet. Need to look at question of what the criteria are for making the decision. Need input on that question, then we make a decision. ================================================================ Plans and next steps (chairs) ================================================================ Hamid: We will discuss the impact on signaling in basic mode, and whether we should have a common format for BGP auto-discovery and OSPF auto-discovery. We invite proposal on the enhanced mode.