Bess Status PagesBGP Enabled ServiceS (Active WG)
Rtg Area: Alvaro Retana, Alia Atlas, Deborah Brungard | 2014-Oct-17 —Chairs:
IETF-99 bess minutes
Session 2017-07-20 1550-1750: Congress Hall I - Audio stream - bess chatroom
BESS @ IETF99 ------------- WG Status chairs Martin (as chair): status of draft-ietf-bess-evpn-inter-subnet-forwarding following WG LC? Ali: almost 70% comments incorporated, rest to come soon. MIBs update Hiroshi Jeffrey H.: do not add BGP rate information in your MIBs Tsuno: OK draft-drake-bess-datacenter-gateway-04 Adrian Wim H.: BGP SR-TE policy is another solution Adrian F.: but not the way we would do this Wim: but you have more options possible for TE in other draft Adrian: need to look at it in the context of new spring draft (draft-farrel-spring-sr-domain-interconnect) Jeffrey H.: SR-TE policy document is potentially usable in context presented here, but not fully Wim: if you use binding SID representing the DC on the other side, you acheive the same thing Wim: we need to look into these two documents to avoid competing solutions Ali: Should SPRING document progress before that one? Martin (as chair): how many have read this document? Approx 10 Yang (draft-ietf-rtgwg-ni-model and L2, L3, and EVPN yangs models) Lou/Acee, Patrice, Dhanendra, 15min Himanshu: There was a PW yang model presentation in PALS. Maybe there is a need for coordination/consolidation. draft-mohanty-bess-mutipath-interas-00 Satya no question/comment draft-boutros-bess-evpn-geneve-00 Sami Greg: control plane solution for NVO3 or for Geneve Sami: Geneve only as this is selected solution by NVO3 Ali: there is also a draft for VXLAN Jorge: if you propose specific/new bits here, they first need to be approved in NVO3 draft-snr-bess-evpn-loop-protect-00 Jorge Ali: draft is too verbose for what it proposes, can be condensed. Jorge: we felt 7432 wasn't clear enough Satya: Are AC 3 and 4 on same Ethernet Segment Jorge: different, because no loop if on the same. draft-skr-bess-evpn-pim-proxy-00 Jorge Stig V.: We have many options in PIM, why a fixed model here? Jorge: we only have selected those that we need / felt were important for the use-case. Stig V.: not sure you want to go into dense mode. draft-sajassi-bess-evpn-mvpn-seamless-interop-00 Niloofar / Ali Jorge: lot's of things not explained yet in the document Ali: not holes, but let's discuss these Jorge (on multi-homed CE): if a source is behind switch and received behind PE3, because of all-active MH, mcast traffic can be hashed to PE1 how to make sure PE3 picks the correct PE (PE1, not PE2) Ali: because we do it via source active discovery Jorge: how do you discover the source Ali: same way as EVPN Jeffrey Z.: caveats in the solution Ali: we'll describe what is not covered draft-sajassi-bess-evpn-ip-aliasing-00 Ali, 10min Jorge: PLan to incmlude RT5 for mass withdraw? Would be nice to have it. Ali: ok Satya: you assume the L2 and L3 RT are different? Ali: the EVI are different. draft-sajassi-bess-evpn-vpls-all-active-00 Ali no question/comment draft-malhotra-bess-evpn-irb-extended-mobility-00 Neeraj no time for question draft-kaliraj-bess-bgp-sig-private-mpls-labels-00 Kaliraj (remote) no question/comment draft-jain-bess-evpn-lsp-ping Parag presentation not done because of time consummed by other presentations.