-
Notifications
You must be signed in to change notification settings - Fork 40.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
NodeController sets NodeTaints instead of deleting Pods #41133
Conversation
5e70927
to
44bdb08
Compare
/approve |
@derekwaynecarr @kubernetes/rh-cluster-infra |
Please don't merge without my LGTM |
Yes, I'm going to write it. |
@k8s-bot gci gke e2e test this |
@davidopp I added another commit fixing problem introduced by @aveshagarwal PR. Also @aveshagarwal - you were supposed to update the implementation of helper functions to just use fields instead of annotations, not delete them, because they're small. If you did my rebase would be no-op, because you didn't it wasn't. This is the reason why we had additional level of abstraction in first place. I want those few hours back. |
@k8s-bot kops aws e2e test this |
1 similar comment
@k8s-bot kops aws e2e test this |
@davidopp PTAL |
BTW I noticed this PR doesn't use the "standard" mechanism for enabling alpha features, e.g. see #30468. Maybe you can fix in a followup PR. |
Sorry, that PR implemented the flag gate, doesn't demonstrate using it. But this PR does: #41617 |
/lgtm |
/lgtm |
[APPROVALNOTIFIER] This PR is APPROVED The following people have approved this PR: davidopp, eparis, gmarek, k8s-merge-robot, liggitt, mikedanese Needs approval from an approver in each of these OWNERS Files:
You can indicate your approval by writing |
Automatic merge from submit-queue |
…fault-toleration-seconds Automatic merge from submit-queue (batch tested with PRs 40932, 41896, 41815, 41309, 41628) Make DaemonSets survive taint-based evictions when nodes turn unreachable/notReady **What this PR does / why we need it**: DaemonPods shouldn't be deleted by NodeController in case of Node problems. This PR is to add infinite tolerations for Unreachable/NotReady NoExecute Taints, so that they won't be deleted by NodeController when a node goes unreachable/notReady. **Which issue this PR fixes** : fixes #41738 Related PR: #41133 **Special notes for your reviewer**: **Release note**: ```release-note Make DaemonSets survive taint-based evictions when nodes turn unreachable/notReady. ```
…ionseconds-admission-controller Automatic merge from submit-queue (batch tested with PRs 40932, 41896, 41815, 41309, 41628) enable DefaultTolerationSeconds admission controller by default **What this PR does / why we need it**: Continuation of PR #41414, enable DefaultTolerationSeconds admission controller by default. **Which issue this PR fixes**: fixes: #41860 related Issue: #1574, #25320 related PRs: #34825, #41133, #41414 **Special notes for your reviewer**: **Release note**: ```release-note enable DefaultTolerationSeconds admission controller by default ```
cc @timothysc @wojtek-t @davidopp
@aveshagarwal - this PR just uses library functions from previous one.
@kevin-wangzefeng - the only thing that's left is to write an admission controller. I don't remember what was the agreements. Are you going to write it, or should I?