> > "We should be selective about the kind of traffic that flows > over PTP LSPs > as these will be handled as a special case by the LSR. The > only LSP user > plane traffic MUST be PTP, but the LSP MAY also carry essential > MPLS/MPLS-TP control plane traffic such as BFD and LSP-Ping." > > Your suggestion above is fine as well, I don't have any major > objections > to it.
And neither do I with your text. Will update the next revision with the above. > > > I am still not sure if I understood your point. Are you > suggesting that > > all such packets be marked as NC? > > I'm not talking marking at all. > > When the LSR wants to send out the PTP message, should it go into the > normal output queuing mechanism or should it just be sent out > first thing, > regardless of what the queuing configuration is (in that > case, it'll beat > EF when it comes to "importance" for output queuing). Ok, I think I understand what youre alluding to. You will like implementations to send out the PTP message using the highest priority queue that's available at the egress. Is this correct? If this is what you want, then I would disagree. While I understand the urgency to send out these packets, its way to less than all the other control traffic that needs to be sent (like BFD, LACP, EFM, etc) out. Cheers, Manav > > -- > Mikael Abrahamsson email: [email protected] > _______________________________________________ TICTOC mailing list [email protected] https://www.ietf.org/mailman/listinfo/tictoc
