Date:
Participants
Alessandro Gerardo D'Alessandro
Sendil Kumar J
Sanyam Dubey
Agenda
- Intent Based Closed loop - DCAE Questions
- Open Questions in TN NSSMF
- Closed loop setup issues raised by Alessandro Gerardo D'Alessandro
- ??
Notes and Actions
- Lukasz Rajewski and Alessandro Gerardo D'Alessandro suggested to enhance the use case documentation - Action Item: Ahila P and team to consolidate the use case documentation and produce and smart guide. Distinguish the ONAP and use case configurations.
- DCAE queries and impacts:
| Questions | Comments | Action items |
| How to set configuration policy to check if bandwidth usage is continuously above threshold for x reported times? Is there any DCAE code changes required? | Attached a sample policy. threshold_nsi.json | |
| For step 2, what is the interface (REST?) for SDNC to send VES data to DCAE? Is there an example? | VES Provider available at DeviceManager bundle Reference: https://gerrit.onap.org/r/c/ccsdk/features/+/111374 | |
| For step 4, is there any DCAE code changes required to generate the DMaaP event which is consumed by Policy? | Exisitng MS handles the PM data generated from RAN NFs. A new API needs to be written in slice Analysis MS to collect the PM metrics of cloud leased lines and to analyze it. Slice Analysis MS: https://gerrit.onap.org/r/gitweb?p=dcaegen2/services.git;a=tree;f=components/slice-analysis-ms;h=36f2fda6d284acd3f1729a2162da20d0635c6294;hb=HEAD | 1. identify the counters 2. Send a VES message to DCAE 3. Expose a new API in SLA MS 4. New Policy and changes in Policy 5. SO, SDNR changes |
| AAI APIs to retrieve the intent instances | Existing APIs should be sufficient | |
| DCAE MS impacts to handle the intent based closed loop | 1. Feasibility of using ML MS, as it is not officially exposed 2. sNSSAI needs to be used as an intent id. So that the corrected configuration from the analysis can be evaluated against the intent before triggering the closed loop. | 1. Sendil and Ahila will work on the ML MS enhancements and optimizations. 2. Ahila will check the procedure to expose the MS to open source 3. Schedule a call and discuss futher on the open items on the need basis |
3. TN NSSMF clarifications: (a) Henry Yu - to check the issue reported in mail (b)Ahila P & Deepika S - to analyze the feasibility of using existing topology for the FH and MH simulators.
4. Option1 and Option2 sequences - Lukasz Rajewski refer this page NSMF and Common NSSMF. In case of option1, internal NSSMFs will be invoked and the respective flows are available in the same space. For option2, the NSSMFs are external to ONAP, returns stubbed response and furthermore no logic and sequences are involved.
