Notes: day of learning ~ learning how to create k8s go operators using operator-sdk, bi-weekly python-client PR/issues triage meeting #17

kubernetes

May 24, 2021

Quick notes from today:

day of learning ~ learning how to create k8s go operators using operator-sdk

• Last friday (May 21, 2021) was a Day of Learning (DOL) for APAC teams at Red Hat. Jose Cueto and I (as suggested by Jose) decided to explore how to write kubernetes go operators using the operator-sdk framework.
• Putting the following links (that Jose compiled) for myself, so, I could refer them quickly later as I require:
  • Set up kind cluster
    • https://kind.sigs.k8s.io/docs/user/quick-start/#installation
    • https://kind.sigs.k8s.io/docs/user/quick-start/#creating-a-cluster
  • Followed the following tutorial for building operators using operator-sdk
    • https://sdk.operatorframework.io/docs/building-operators/golang/tutorial/
  • Load local operator (container) image in a kind cluster
    • https://kind.sigs.k8s.io/docs/user/quick-start/#loading-an-image-into-your-cluster
  • Deploy reference addon in kind cluster.
    • https://sdk.operatorframework.io/docs/building-operators/golang/tutorial/#2-run-as-a-deployment-inside-the-cluster
  • Deploy olm in kind cluster
    • https://sdk.operatorframework.io/docs/building-operators/golang/tutorial/#3-deploy-your-operator-with-olm
  • I also learnt how to deploy/reconcile operators in multiple clusters from a single centralised Hive clusters using SelectorSyncSets (SSS).
  • Some more links, that are important:
    • Highly recommend these links for understanding a bunch of operator concepts.
      • https://olm.operatorframework.io/docs/tasks/

  Although, I’m still facing some issues while deploying my dummy go operator through OLM. Jose & me continued to troubleshoot the issues today as well. Not much success on that, seems like I’ve some different configuration on my machine than Jose’s. And these differences are what we are trying to figure out currently. Once I’m done with some tangible results on this part, I would aim to write a little better blog around the steps I followed, & what all things I learnt.

  But I thoroughly enjoyed thi time’s DOL. Totally worth it!

---

bi-weekly python-client PR/issues triage meeting

• Attended the bi-weekly kubernetes python-client PR/issues triage meeting
• Picked up the following issue. It’s a simple one, but the solution involves regenerating the python-client, so would be a very good learning exercise for me.
  • BatchV1Api documentation examples are mis-indented #1470
• Also, I’ve a pending PR to finish ~ [WIP] add documentation for the server & client side timeout #1467
• Some major discussions from today’s meeting:
  • Proposal to reduce the alpha -> beta, beta -> GA release soak time to 1 week from previously 2 weeks.
  • Although the new pattern would still ensure there will be a 2 week time for feedbacks between a beta to next alpha release.
• And, I have volunteered to participate in creating the upcoming python-client v18.0.0 beta release. I might pick up the release documentation or other small bits (I’m yet not aware what might be those) that might come up during the process.

---

That’s all for the day!

PS: Find here, the links to all the kubernetes-contributions for the month of May, 2021.