[bess] Re: Port-Based Service Interface in draft-ietf-bess-rfc7432bis

2024-06-20 Thread Alexander Vainshtein
Ali, Lots of thanks again for a prompt and very encouraging response. I defer to you and the rest of the authors of 7432bis whether Informative reference to MEF documents describing processing of L2 Control Protocols should be added. (IMHO it could be useful while not causing any procedural issu

[bess] Re: Port-Based Service Interface in draft-ietf-bess-rfc7432bis

2024-06-19 Thread Ali Sajassi (sajassi)
Hi Sasha, Jorge: I agree with you both that the clarification should be done wrt “data” traffic and the detailed handling of L2CP is outside of the scope of 7432bis as already covered in MEF documents. Wrt data traffic on a port, we have two uses cases: 1. Access ports where the data traf

[bess] Re: Port-Based Service Interface in draft-ietf-bess-rfc7432bis

2024-06-19 Thread Alexander Vainshtein
Jorge, Again, lots of thanks for a prompt response. I fully agree that my second bullet is effectively covered by the relevant MEF technical specifications (MEF45.1). IMHO an Informational reference to these documents would be useful. As for my first bullet: from my POV the text of Section 6.2.1

[bess] Re: Port-Based Service Interface in draft-ietf-bess-rfc7432bis

2024-06-18 Thread Jorge Rabadan (Nokia)
Hi Sasha, I don’t have a strong opinion on your first bullet, if there are no objections. Although it could be interpreted as if RFC7432 didn’t support untagged traffic on port-based service interfaces, which is not the case. About the second bullet, we are not defining L2CP behavior in any of

[bess] Re: Port-Based Service Interface in draft-ietf-bess-rfc7432bis

2024-06-18 Thread Alexander Vainshtein
Jorge, Lots of thanks for a prompt and most helpful response. IMHO the text is Section 6.2.1 should explicitly state your interpretation, i e.: * Untagged customer traffic is encapsulated and forwarded "as is" * Untagged Layer 2 control protocols traffic (identified by carrying well-kno

[bess] Re: Port-Based Service Interface in draft-ietf-bess-rfc7432bis

2024-06-18 Thread Jorge Rabadan (Nokia)
Hi Sasha, The implementations I know, all the traffic – tagged and untagged – is mapped to the EVPN broadcast domain, for that type of service. Since no pop/push is done of the vlan tags, untagged traffic would be encapsulated into the EVPN packet and forwarded as is. My interpretation in this