While day 1 (startup) is described, don’t see anything on maintenance or in-service upgrades at the optical layer. A couple of notes (covered in some of the multi-layer RFCs done previously):
Avoid disruption in the network – minimal impacts on both optical and packet. Do not want to trigger a network-wide router software update when performing operations on an optical module.
Recommend module software – software upgrades/maintenance on the packet level or optical level are not disruptive on the other level.
Need to manage properly, including security levels for the different entities.
I noted it is said “no protection” – maybe want to soften to say “protection/restoration is for further study”?
Measures to include network wide multi-layer re-optimization strategy to prevent fragmentation and optimal use of capacity (e.g., RFC5557 provides guidance when using a PCE).
Measures to protect against network destabilization
Measures are needed to support graceful deletion at all layers
Fault escalation strategy to coordinate across the layers
Mechanisms for aggregating, summarizing, enabling and disabling of inter-layer notifications so as not to overwhelm the controller.
A couple of wording nits:
Section 2 “which results in non-optimal and inefficient networking”/s/”which may result in non-optimal and inefficient networking” (marketing/opinion)
Section 5.6 “to avoid proven protection races”/s/”to avoid possible protection races” (proven? May happen if the multi-layers are not managed correctly, a vendor fails to implement proper detection/escalation).
While day 1 (startup) is described, don’t see anything on maintenance or in-service upgrades at the optical layer. A couple of notes (covered in some of the multi-layer RFCs done previously):
A couple of wording nits: