Date:  

Participants

Swaminathan Seetharaman

LIN MENG

Ahila P

Aleem Raja

Alessandro Gerardo D'Alessandro

Alex Borovitzky

Aniello Paolo Malinconico

Avi Chapnick

Borislav Glozman

Byung-Woo Jun

Deepika S

Fei Zhang

Fernando Oliveira

Kamel Idir

krishna moorthy

Naveen Sankad

PRAVEEN VELUGUBANTLA

Pravin Kumar Damodaran

Sanchita Pathak

Sandeep Shah

Seshu Kumar Mudiganti

Sowmya Narayanasetty

Agenda

  1. CSIT/Robot tests for SO (Seshu)
  2. Alessandro’s queries on the setup
  3. Incorporating ETSI Network Service with E2E Network Slicing (Fred)
  4. CPS updates and approach
  5. M3 readiness & Integration Testing

Notes and Actions

  1. CSIT/Robot tests for SO (Seshu)
    1. Seshu Kumar Mudiganti  provided a quick overview of the SO refactoring and modularization work, making it a plug-n-play architecture with reduced footprint based on the deployment needs and reduced dependencies across non-overlapping use case functionalities and sub-components.
    2. As far as E2E Network Slicing use case is concerned, NSSMF Adaptor, OOF Adaptor and CNF Adaptor have now been separated out, and will be part of the Adaptors as sub-components.
    3. The E2E Network Slicing team is requested to contribute to the CSIT for NSSMF Adaptor, OOF Adaptor and CNF Adaptor, as well as the Robot tests for E2E and Macro flows in SO.
    4. Swaminathan Seetharamanand LIN MENG to discuss within the use case team and provide an update to Seshu Kumar Mudiganti . The discussion will continue in the weekly SO call on Wed, and as well offline with Seshu as needed to take it forward in the Honolulu timeframe. The exact scope, timelines, etc. will be agreed in the coming days.
  2. Alessandro's queries on the setup
    1. Alessandro Gerardo D'Alessandro raised a couple of blocking issues faced during NSI/NSSI Selection during interaction with OOF. Alessandro Gerardo D'Alessandroto invite for a call with krishna moorthyto address the issues.
    2. Most of the earlier issues are now addressed, except a few - these will be taken up via mails.
  3. Incorporating ETSI Network Service with E2E Network Slicing (Fred)
    1. Fernando Oliveira gave a short presentation on E2E Network Slicing making use of ETSI Network Services. This was also presented at the LFN DTF earlier this month, and the slides are available at https://wiki.lfnetworking.org/download/attachments/50529096/LFN-Feb2020DDF-ETSI5GSlicing.pdf?version=1&modificationDate=1612302881000&api=v2
    2. Based on the proposal, we agreed that we can start with Core Slicing - the Core NSSMF can invoke the NFVO over SOL5 (instead of it calling Macro workflow via NBI). Subsequently RAN Slicing can be considered.
    3. This requires collaboration between the 2 teams, in particular, the folks working on Core Slicing and Modeling. Follow-on discussions and meetings will be scheduled to take this forward for Istanbul release. Swaminathan Seetharamanto initiate a mail thread including all the relevant folks.
  4. Slack channel link for our use case: https://onapproject.slack.com/archives/C01P1EXREJG
  5. CPS approach will be discussed in next week's call.

Recording

  • No labels

2 Comments

  1. Milind Jalwadi

    Hi Fernando Oliveira Swaminathan Seetharaman ,


    Thank you for putting up the new requirements for NSSMF.


    We had a brief discussion on this during our slicing call on 8th March, accordingly would like to put across few points as below –


    • Since ONAP has an “non ETSI” way of orchestration, by the means of NBI / SO Macro flow, we should retain this and could add to any new implementation
    • The NST or NSSTs I think have references only to the Network Service. So I believe all the CBAs / policies etc. should be bundled with the SDC service. Not entirely sure if there are any new models suggested in any recent specs
    • NSSMF role when it comes to NS instantiation or configuration as part of slice scenarios, should always be that of the consumer of ONAP NFVO functionality which could include the non ETSI and ETSI interfaces.


    We could discuss more during the next call on 15th March.


    Thanks,

    Milind

    1. Fernando Oliveira

      Milind Jalwadi

      • I agree that we should retain the current mechanism and add a new mechanism as a branch, perhaps based on the existence of an NSD reference in the NSST.
      • I don't think that any new model is required.  Perhaps a slight change to the NSST model to optionlly have a reference to an NSD.
      • I think that we need to discuss how the non-ETSI and ETSI interfaces are implemented.

      I will join the next slicing call on 15th March to discuss more.