Hi Dhruv, Thanks for your suggestions! That really helps to progress the this work. Please see in-line with Quan.
[Pce] draft-xiong-pce-nrp-id-01 Dhruv Dhody <[email protected]> Fri, 22 July 2022 18:32 UTCShow header Hi, Consider adding capability exchange before using this NRP-ID TLV. Quan>Good point, NRP capability could be carried in Open message and we will update that in next version. We need to clarify what it means to take NRP into consideration during path computation - if the expectations between two implementations are not the same we would run into interoperability issues. Quan>Yes, we could clarify that our goal is to take NRP into consideration during path computation in slicing networks. A NRP is a collection of resources (bufferage, queuing,scheduling, etc.) in the underlay network to support the IETF Network Slice service. The PCE may compute paths constrained within a NRP domain. If that involve interoperability issues, we will update that in next version. There is another document - draft-dong-pce-pcep-vtn-01 (presented at 112) which also defines a new TLV for LSPA. Has there been any discussion with the authors of that I-D? Quan>We have no discussion with the authors of draft-dong-pce-pcep-vtn-01 so far. I am not sure the difference between VPN+ and NRP which is up to TEAS Working Group. We are glad to discuss with them to progress this work. The draft-xiong-pce-nrp-id-01 has replaced the draft-peng-pce-te-constraints-06 and the latter draft proposed the slice related constraint early in July 2019 with the terminology changing from AII to Slice-ID. Till this June, the TEAS WG adopted the [I-D.ietf-teas-ns-ip-mpls] which defined the NRP-ID as the identifier of slice resources, we updated the terminology to NRP-ID as the constraint for PCE path computation. Hope this helps! Thanks! Dhruv [Pce] draft-xiong-pce-nrp-id-01 Dhruv Dhody _______________________________________________ Pce mailing list [email protected] https://www.ietf.org/mailman/listinfo/pce
