IETF 100: BABEL Working Group Raffles City Convention Center, Singapore Thursday, 16 November 2017. 18:10 - 19:10 Bras Basah Room Chairs: Russ White (LinkedIn) Donald Eastlake (Huawei) Minutes: Barbara Stark Jabber: Ted Lemon [times below are as scheduled, not as actually taken] 5 min. Administrativia (scribes), Agenda Bashing, Chairs 5 min. Status, Review of Milestones, Chairs 10 min. Update on RFC6126bis and Source Specific drafts. David Schinazi draft-ietf-babel-rfc6126bis-04 draft-ietf-babel-source-specific-01 15 min. Security. Juliusz Chroboczek 10 min. BIER in IPv6 Zheng (Sandy) Zhang draft-zhang-bier-bierin6-00 5 min. Wrap-Up, Chairs ------------------ Agenda and Status https://datatracker.ietf.org/meeting/100/materials/slides-100-babel-agenda-and-status/ Donald presented. The Babel Routing Protocol (draft-ietf-babel-rfc6126bis) is in WG LC. That ends tomorrow. ------------------ Update on RFC6126bis and Source Specific drafts https://datatracker.ietf.org/meeting/100/materials/slides-100-babel-update-on-rfc6126bis-and-source-specific-drafts/ David Schinazi presented. Tony Przygienda: With mandatory bits you have proven your intelligence. There were no other comments or questions on this presentation. ------------------ Security https://datatracker.ietf.org/meeting/100/materials/slides-100-babel-security/ Juliusz Chroboczek presented remotely. David: I really prefer DTLS secuirty to RFC 7298 security. You can bootstrap with symmetric keys or whatever you want. It eases the trust model, which is needed by the HOMENET WG. Regarding MTI [Mandatory To Implement], I would be mostly against that. When running over IPsec, it doesn't make sense. But could make MTI in homenet. Ted Lemon: One thing that came up in a discussion was that you're using multicast for link quality detection. Juliusz: Yes. Ted: Have you looked at how things have changed in DTLS 1.3? Juliusz: No. Ted: If significant differences between DTLS 1.2 and 1.3, then maybe standardize on 1.3. Stephen Farrell: Standardizing on 1.3 would probably be cleaner. On MTI, there is a BCP that says you must have MTI security recommendation. But you would probably get pushback. Juliusz: First we need a solid proposal and solid implementation. We can discuss MTI after that. David: Responding to Ted's point. As far as I know, changes from 1.2 to 1.3 would not make big difference. Ted: What I remember is that 1.3 may be a little easier. Stephen: A profile for 1.3 might be smoother because it's better designed. Barbara: My understanding is that 1.3 deprecated many mechanisms. It may be good to consider. Most of the underlying mechanisms -- handshakes, etc. -- didn't change. Barbara: Would you suggest holding off on finalizing the homenet profile until we have a security recommendation? Juliusz: Never. ------------------ BIER in IPv6 https://datatracker.ietf.org/meeting/100/materials/slides-100-babel-bier-in-ipv6/ Zheng (Sandy) Zhang presented. Juliusz: I read the draft and think it's good. Juliusz: One thing. We now have multiple encapsulations of BIER. How do I know which encapsulation to use? E.g., IPv6 or BABEL? Tony P: It is architecturally unspecified. If there are multiple, I can run per link whichever I choose. We can easily migrate from this sort of hack to a solution with an Ethertype. It's a solution without a distinction. Juliusz: We will take this offline. Tony P: I know on both sides the binding exists and you can choose which encapsulation to use. But, yes, we can take it offline. ------------------- Donald: That is the end of the agenda items. We will see people on the list and at the next meeting in London in March.