BESS Working Group Agenda, IETF102 Montreal ------------------------------------------- 9:30-11:30 Friday Morning session I, Viger Chair: Matthew Bocci Minutes: Mankamana Misha 1. Working Group Status Chairs, 15 min · 3 RFC published · 2 in queue to RFC editor · 3 Drafts in AD review · One document ready for submission io IESG · 3 drafts ready for WGLC · 3 new WG document since last IETF (IETF 101) · Mvpn-yang no reply during adoption call. · Pim-proxy draft expired. . Jorge – received feedback, before next IETF plan to update. Need another adoption call after updating the document · Evpn-irb-mcast . New draft uploaded recently. Still updating (Z Zhang) · Bum-procedure ready for adotion call · Msdp-interop ready for adoption call · L2vpn/evpn yang ready for last call (Himanshu) · Evpn-internet-subnet-forwarding . Ali: significant update in document, need to have last call again as new changes are significant. 2. draft-ietf-bess-mvpn-fast-failover-03 Greg Mirsky, 10 min (Remote) · Updated local procedure to define bootstrap in this version of draft. · Update includes mpBFD operations · Ready for working group last call. · Ali – Are we having 2 p2mp BFD sessions? Switchover should happen only when secondary tunnel is up. · Greg: Each PE would originate its own p2mp sessions. It need to be documented, that before switching over to secondary tunnel policy has to make sure it is ready. . Ali: I meant more higher level. It was mor emaking sure that the other tunnel is healthy before switching over. . Greg: Yes, that is good observation, but is outside of the policy. But does need to be explicitly documented in the draft 3. draft-ietf-bess-evpn-igmp-mld-proxy-02 Ali Sajassi, 10 min · In queue for WGLC · New version contain change in ES-import RT area. Multiple options were considered. . After discussion mechanism “Define multiple EVI-RT ECs” was taken. New revision defines 4 types of EVI-RT EC’s. · Susan: IDR has some discussion going on extra EC or flexible EC, in case it need to be discussed for this draft. · Ali : It has already been thought through · Mankamana: Can we take out IGMP V1 · Ali: If nobody has objection IGMPv1 will be taken out of the draft and will mention only v2 and v3 . Ali: Jorge suggested to remove the IGMP from wrt sync, leave and join and generalize it and reflect it through text in the draft . Jorge: Removing is a good thing, but do we need to change the bits in the draft? . Ali: no 4. draft-sajassi-bess-evpn-mvpn-seamless-interop-01 Ali Sajassi, 10 min · Significant change from first version, which covers comment from WG. · Z-Zhang, Andrew: If Source PE is sending on Inclusive tunnel of MPVN, multi-home PE would get multiple copy of traffic · Ali : To take care in next version · Ashutosh: Avi network is planning to implement it. · Jorge: We should not adopt the document, as there is one other document which does solves the same requirement. · Ali : This document is useful in different segment of network. · Andrew: TTL decrementing would have security issue · Ali: It is as per MVPN procedure. · 15 people think it's ready for adoption call · 8 people think it is not ready for adoption call 5. draft-jain-bess-evpn-lsp-ping-07 Parag Jain, 10 min · It was published 3 years ago and presenting 2 years back · Jorge: Is there any advantage over CFM? · Ali: CFM and this draft is doing different check. CFM is at ethernet layer and this draft does Service layer. 6. draft-sajassi-bess-evpn-per-mcast-flow-df-election-01 Mankamana Misha, 5 min · Asks for WG adoption, there is already implementation . Ali: This is how you going to loadbalance the traffic for the fat vlan over the same networks 7. draft-malhotra-bess-evpn-unequal-lb-04 Neeraj Malhotra/Ali Sajassi, 10 min · Need to add DF type 5 · Ready for adoption call. . Chair: Who read the fd=draft: 9 . Chair: Who think it is ready draft it is ready for adopton: room seems to indicate supportive · Jeff: BW EC, can we make transitive?· . Ali: Its BW extended community we are talking about and is defined in other draft. It is not-transitive, and we want it to be transitive. AUthors are contexted, but not responding 8. draft-malhotra-bess-evpn-irb-extended-mobility-04 Neeraj Malhotra/Ali Sajassi, 10 min · It does not have major changes. Just some clarification added. · Ready for WG adoption · Wen : mac is associated with many IP address, and IP address moves to different MAC. It is proposed using MAC sequence, it need to address the case if multiple IP move happens at same time. · Ali: sequence number is attached to MAC, so it should not create any issue. . Ali: you pick the seq# of the parent. Lets take this offline 9. draft-rabadan-sajassi-bess-evpn-ipvpn-interworking-01 Jorge Rabadan, 10 min · All feedback incorporated in this version . Ready for WG adoption call. . Chair: We just have 1 minute for question. How many read this version of the draft? . Chair: I count only few people 10. draft-salam-bess-evpn-oam-req-frmwk-00 Donald Eastlake, 5 min · 4th version of draft and authors think it is ready to adoption. 11. draft-gmsm-bess-evpn-bfd-00 Donald Eastlake, 5 min . Jorge: We use BFD for fast fail-over. I was wondering why here thiscan be discriminators are distributed through LSP ping and not through BGP routes . Ali: This is the old draft and that was the thinking of few years ago. If now the approach is with BGP for this approach, then that is ok and this work can go that path . Donald: Yes, i am happy to change the direction of the draft in that suggesting and go with resurection forward with the newer approach 12. draft-rosen-bess-secure-l3vpn-01 Ron Bonica, 10 min . Ron: Can this draft be adopted? · Linda: it is useful for SD-WAN application. · Susan: security consideration need to be updated. It should not be out of scope. · Jeff: can we use it to distribute public key or not . It would be good to be clear in the document. · Jim: can I use any encap tunnel? · Ron : Yes · Around 7 people read the draft. · Ali: Before calling adoption, should we run it by IPsec working group ? 13. draft-bonica-6man-vp-dest-opt-00 Ron Bonica, 10 min · It would be called as 6man WG item. · Need feedback from WG to see if proposal breaks anything. · Ali: what is advantage of doing it ? we are trying to solve hardware issue here. Geneve draft already does it, why do we need new document? why new encapsulations is needed. · Ron: there would be some device , some device does not support MPLS. This would be useful when we do not want to MPLS. · Ali: this is new encapsulation, and we done that in NVO3, so why is the what we developed not enough? . Ron: I take this as an action item to document why GENEVE is not suficiant . Adrian Farrel: This is with IP running and adding the required context. It is not a new encapsulation.