| Time | Item | Who | Notes |
|---|
| IM-D | Administrative | Kam | - Future meetings
- 2018 December 3 - 7: ONF OIMT & OTCC Meeting in Santa Clara (co-located with Dec. 4 - 6 ONF Connect 2018)
- Jan. 21-25: Q12+Q14 in South China (may invite OIMT & OTCC)
- 1Q 2019 (February or March 18-22) in Darmstadt hosted by DT (Bernd)
- April 8-12: Q14 in Europe (may invite OIMT & OTCC)
- 2Q 2019 in Asia (China, Japan), need host
- Late May - Early July: SG15 in Geneva
- 3Q 2019 in Australia hosted by Cisco (Chris), Chris would like to nail down the date of the meeting
- 4Q 2019 in California, co-located with ONF "CONNECT 2019" workday
|
| | | Chris | - Proposal from Chris:
- Have a UML, model and CIM introduction ‘course’ that we could run at each F2F for any newcomers (as required)
- We could record it and post the URL and slides publically
- Need more people to engage in preparation
- Input material: The TR-512 suites and the guidelines documents
- Overview: Why, what, where (links to further details)
- There is general support from the group
- Chris will take a start
|
| | | | - Briefly discussed post v1.4 tasks and focus
- Tme of next deliverable
- Cleaning up of document
- When for v2.0
- Should forcus on engagement in the December F2F meeting
|
| | | | - Q14/15 Liaison (Defer to IM-E)
|
| | Q3 deliverable | Nigel, Kam | - TR-512 v1.4 review feedback
- Discussed TR-512.A.7 (Control/Interaction examples) comments from Malcolm (TR-512.A.7_OnfCoreIm-Appendix-ControlAndInteractionExamples-CHV2-clean-ND-MB-Updates)
- Section 3.2 The SDN Controller
- The "Client Context and Server Context" is key in the ONF SDN architecture.
- Should not move too far away from this architecture
- This archticture is also the base of the ITU-T SDN architecture (which has further details on the control components).
- Peer context and mesh of peer
- Master and slave
- Two styles of interaction between two controller:
- One controller sets the police & allocates resources for the other controller,
- They talk about thinks for control
- Decided to split this section (the description of SDN controller) into two part
- First the representation of SDN architecture (client/server context)
- Then the representation of more general Peer context. This needs more further work
|
| IM-E | Q14/15 Liaison | | |
| | TAPI | Karthik | - Photonic & OTSi Model:
- TAPI association:
- Karthik: Coming up from MEF and OIF demo, the TAPI associations need to be bidirectional
- Nigel: Cost (not cheap) vs benefit (powerful)
- Nigel: Navigation direction: most (if not all) from bottom up;
- Nigel: Bidirectional only make sense in the database
- Nigel: Reverse direction can be inferred
- Nigel: Need to look individually
- Karthik: look at the individual association
 - LinkTerminatesOnNEP: Link 0..1 -- 2..* NodeEdgePoint (FcPort + LTP)
- Core: Link (LinkPort) --> LTP (in TAPI is NEP)
- In implementation, NEP exists first, then the link
- TPAI wants Link <----> NEP
- CEPRelatesToCSEP: (FcPort) ConnectivityServiceEndPoint 0..1 – * ConnectionEndPoint (FcPort + LTP)
- CEPAgregateCEP: ConnectionEndPoint – ConnectionEndPoint; this is assembly relationship, for example NMCA aggregate NMCs. Not inverse multiplexing.
- CEPGroupsCEP
- Don't use the same verb for assembling (grouping) and inverse multiplexing
- ConnectionSupportsClientLinks: Connection – Link
 - In Core, FCSupportsLink (in core): Link <-- FC;
- Link --> FC is more for YANG
- Nigel: Need to be very careful to have multiple ways for doing the same thing.
|
| | Wireless transport | Martin | - Nothing to discuss.
- Facing some unexptected YANG issues. Dynamic pointer to read a table.
|
| | Administrative | | - No TAPI, OT-IM, and OIMT calls next week.
|
| IM-F | Q3 deliverable | Nigel, Kam | - TR-512 v1.4 review feedback
|