Re: [mpls] Working Group Adoption Poll for draft-andersson-mpls-mna-fwk

[Adrian Farrel <[e-mail]>]

Hey,

Just pouncing on one thing as it is an old, old discussion.

On 12/07/2022 18:00 Stewart Bryant <[e-mail]> wrote

   2. The payload type (for example, L2 or L3 packets) cannot be directly determined because there is no payload indication.

This is an MPLS design philosophy, but not intrinsic to the protocol. You could use one of a number of PW types including PW types that carry a generic payload identifier. Alternatively you could create a new FEC type that had the property that the last label was a payload type indicator, or that the payload started with a type indicator.

[AF] Stewart is right that this was a deliberate design decision in MPLS. Whether the decision was right or still holds today is a different matter.

[AF] Debate in the past has included using different LSPs for different payloads, different bottom of stack labels for different payloads, or (as Stewart says) different PW type for different payloads. There has been a perceived need for per-packet payload indicator so that packets of different payloads can share an LSP or PW. 

A question arises as to whether this information is needed at the egress only, or at other nodes. Further, whether global semantic is needed or whether the iidentifier can be negotiated between participating nodes (easiest is only ingress and egress need to know on an LSP)

A