Skip to content

Latest commit

 

History

History
22 lines (17 loc) · 2.68 KB

week3.md

File metadata and controls

22 lines (17 loc) · 2.68 KB

Weekly Note: 22 May - 28 May

Delivered this week

  • Created a PR #46200 for NameConflicts. Came across many problems while doing this and fixed a few bugs!
    • This initially caused a panic (stack trace). This occurred because now we deleted from storage instead of the client (after this PR). This was racing the storage and causing a nil dereference. PR with fix: #46430.
    • Did a rebase again but still did not work. Now I got: "reason": "InstanceDeletionFailed" and "message": "could not list instances: unable to find a custom resource client for noxus.mygroup.example.com". The problem turned out to be that the storage map was only filled if a CR (not the CRD!) was accessed via some CRUD request. So this worked only if you created CRs before the CRD deletion. PR with fix: #46501.
    • Fix naming controller so that when a resource is deleted, other resources in the same group are added to the queue. This is necessary since the deleted CRD can free names.
    • Added a requeueAllOtherGroupCRDs for the above. Had a little trouble making the synchronization was right. (Finally got it working!)
  • Added a test for checking self link of cluster scoped resources.
  • Wrote the schema for validation of Pod for the proposal.
  • Did more community bonding:
    • Talked to David about the project and CRDs.
    • Helped a few folks on Slack and answered questions about getting started with contributing: 1, 2, 3.
  • Did my first code review!
  • Some twitter fun:

@TheNikhita Thanks for your work so far @TheNikhita! You fixed a bug that we hit in Kubernetes and is getting backported to the next 1.6.x branch 👍👍👍

— Brandon Philips (@BrandonPhilips) May 25, 2017

Pitfalls

Really need to ramp up on the proposal. Want to get in the tests before the code freeze on June 1 and will then continue to work on the proposal completely.