-
Notifications
You must be signed in to change notification settings - Fork 0
Minutes 2016 11 2
Meeting Materials | Attending |
---|---|
WebEx Recording password: cSy4T9un | Andy Bierman, Alexander Clemm, Walker, Einar Nilsen-Nygaard, Eric Voit, Tim Jenkins, Peipei Guo |
Agenda:
- Review the IETF97 slides.
- Discuss requests for WG feedback which are to be made in Seoul.
- Draft slides are here
for evaluation by the larger NETCONF team in Seoul
- Feedback Request #1: Streams and Filter Types: if/when OpState adopted, should we augment with a corresponding new filter-type allowing the OpState datastore to be selected? (believe answer is yes.)
- Feedback Request #2: Topic Filtering: Assess WG interest in classifying Model, Subtree, & Leaf via independent categorizations such as Event-type & severity. (Andy to edit Eric's first attempt at slide which is linked above.)
- Feedback Request #3: Metadata Filter: This is obviously a need well past subscriptions, and tied to #1. We should look to charter some YANG Metadata filter draft. (in NETCONF or NETMOD?) Who wants to help here?
- Feedback Request #4: Do we include an Optional Update-Number to help with the Detection of Loss/Duplication independent of TCP?
- Feedback Request #5: HTTP2 compatibility with GRPC: we have proposed messages/exchanges for seamless transport over GRPC implementations. Do people agree with this, or see issues here with the mechanism proposed?
Other thoughts:
-
Metadata filter + Dampening Period
-
What if there is a move of an object to Intended or Applied Datastore before it hits the Operational-State datastore.
-
A dampening period might slow the information that something has hit operational-state.
-
Based on this, we might need to have the metadata filter type apply for only certain values, construct different subscriptions for different datastores, and/or
-
Should propose a meeting / get-together in Korea on Metadata filtering.
-
Alberto will have recording for call-home with subscriptions with Open Daylight for embedded in the slides
-
Will cancel next week's meeting based on our getting the slides mostly done already.
-
On the reduced attendance today. It is likely due to this being a slide review call rather than an issues closure call.