Attendees

 Luis Gomez  Andrew Hsia Rahul Sharma Anil Belur

Agenda

Minutes

Future works on Helm charts:

  1. ODL runtime stats/metrics: pod/container, module, process or cluster-specific metrics that could be exported through K8s monitoring, using Prometheus (JMX Exporter).
  2. Documentation for setting up Helm charts, jobs needs to be updated.
    1. Ensure Docker setup works with MiniKube for developers who want to bring up and test a single instance.
    2. Committers/Moving to a separate repo?
  3. Migrate a few CSIT (1node/2node) jobs onto K8s setup, document and triage/fix issues
  4. Pending release to be done for docker. 
  5. ODL Scalilibity: (ODL instances on separate clusters.)
    1. POC using concepts from (Strimzi Kafka CRD), KOPF (Operator framework) to ODL instance and Netconf devices is something the telco operators would be interested,
    2. Here is the Strimzi operator concept.
      https://strimzi.io/docs/operators/latest/overview.html#overview-components_str

      Also the Kubernetes Operator Framework in Python for quick POC
      https://kopf.readthedocs.io/en/stable/

  6. Infra, Nexus3 works with Charts museum for helm charts deployment and docker release jobs.

Action items