|
Hi Stewart and Andy, I support this draft to become WG draft. Congratulation!! I also have some comments on the draft: 1) Since Packet PW requires using CW,
it is necessary to state how to use flow label and S/N for packet PW. 2) It is good to have the PW status
mapped to virtual interfaces. However, each protocol interface may have its own
link status protocol running, the draft should state the procedures to determine
interface status, or require each protocol interface to turn off its link
protocol when using Packet PW 3) Since Packet PW acts as a data
link between two clients LSRs, the draft should state link operational
procedures (for client networks) over a packet PW such as how to determine admin
cost, TE metrics, etc. 4) I assume that a Packet PW can obtain
transport resiliency from server MPLS network, it would be nice for the draft
to state them and considers that the consequence may change admin cost in
client networks. Cheers, Lucy |