Agenda

• Move to stable API and Kubernetes 1.18: @Grzegorz Lis (Nokia) @Daniel Milaszkiewicz @Magdalena Biernacka

  • OOM side: Everything is ready (use of stable API)

  • DCAE side:

    • biggest blueprints are ready

  • Installation: almost done on Azure Kubernetes Service (hope for first gating tomorrow) (@Sylvain Desbureaux )

• Seamless upgrade: @Tri Nguyen

  • Reviewed comments from @Krzysztof Opasiak done

  • @Krzysztof Opasiak will be back next Monday

• DCAE deployment issue:  @Vijay Kumar

  • DCAE struggles to deploy correctly when on "slow" Kubernetes

    • Azure deployments are not impacted for example

    • all deployments on Orange OpenStack are failing

    • some deployments are failing on other premices

  • happening since the last 10 days

  • maybe some memory, knowing that no limits are set into Azure

  • limits/requests seems to be too low

• Post Guilin start talk: @Vijay Kumar (will be started on email and discussed next week)

• IPv6/IPv4 dual stack: @Magdalena Biernacka @Grzegorz Lis (Nokia) @Daniel Milaszkiewicz

  • target working version: Honolulu

  • on SDNC side: SDNC-895 (SDNR DB a.k.a Elasticsearch)

  • DCAE:

    • update plugins to generate IPv6 version of services automatically

  • portal cassandra db is also failing

• Shared DBs: @Rajendra Mishra
  
  

  • VID move to shared db under review

• Move to helm v3 @Jakub Latusek

  • all charts seems to be OK on helm v3

  • now we have to change the deployment / compilation

• SO issue @Seshu Kumar Mudiganti @Sylvain Desbureaux

  • 1.7.1 is not working

  • merge request is more than just changing 1.7.1

    • decided to split the merge request

      • moving to "components" + keystore automatic retrieval → merge as soon as possible this one

      • bump the version (to 1.7.3-SNAPSHOT) (but seems still buggy)

• consul:  @Tri Nguyen

  • some scripts inside are pretty old (the one for healtchecks)

    • SDC, SDNC are not working anymore