Additional developers (5+) are being made available by Ericsson - to work mainly on NONRTRIC topics, and related topics in other projects (OAM, near-RealTime RIC, etc). Some will be properly introduced at this week's TOC meeting and next week's SCRUM meeting.
Public holiday in US, so reduced attendance. We decided to use the meeting slot to brainstorm on a number of diverse topics:
Where to put/host OAM/NONRTRIC use-case/applications in ONAP ... Where should the cross-function coordination services exist?
A model-driven self-serve approach would be a better approach, so we need to assist in requirements gathering for future ONAP releases to help realise this.
We need a simple use-case (Release B) that demonstrates/exercises a closed-loop combination of O1, A1, DCAE, Policy, etc.
One example might be (simple) PNF on-boarding, then operation
We discussed the onboarding process, including how SDN-R functionality can be triggered etc.
When/how does a new PNF register/integrate with it's near-RealTime RIC? How will the NONRTRIC locate the appropriate near-RealTime RIC for the PNF?
Where will the near-RealTime RIC coordinates be stored? A&AI, ConfigDB, SDNC, ... ?
Discussed the origins and possible roles for ConfigDB or something similar.
Committers on Repos ...
Committer policy does not exist for either project yet.
Suggest 2+ (from different companies) should review submissions initially
Want to avoid any one organisation or individual having unfair influence, and maintain a rigorous quality control process,
Being a committer comes with code-review responsibilities, so as development ramps up this may lead to unreasonable workloads ... will review later
Moving beyond initial work of terminating A1/O1 interfaces toward more crosscutting usecase development, greater coordination between OAM and NONRTRIC projects will be required.
Suggest John&Martin propose to TOC to act as initial committers on both OAM and NONRTRIC projects
Committers will be added/removed as required when the project scopes become more concrete.