< 10000 bdi class="js-issue-title markdown-title">WIP: kube-controller-manager: Release leader election lease on exit #132620
Conversation
Make sure the leader election lease is released on exit. This is currently not the case and KCM replicas must wait on lease expiration, which is too long and unnecessary.
|
This issue is currently awaiting triage. If a SIG or subproject determines this is a relevant issue, they will accept it by applying the The Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. |
|
Hi @tchap. Thanks for your PR. I'm waiting for a kubernetes member to verify that this patch is reasonable to test. If it is, they should reply with Once the patch is verified, the new status will be reflected by the I understand the commands that are listed here. Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. |
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: tchap The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
03c0159 to
2b9124f
Compare
2b9124f to
8fc918e
Compare
|
There is already a KEP for this IIUIC https://github.com/kubernetes/enhancements/tree/master/keps/sig-api-machinery/5366-graceful-leader-transition , can you sync with @Jefftree |
What type of PR is this?
/kind bug
What this PR does / why we need it:
kube-controller-managerdoes not release leader election lease on exit. This causes a prolonged period of inactivity when in a HA setup. So this PR makes sure the lease is actually released on exit.Which issue(s) this PR is related to:
Fixes #132599
Special notes for your reviewer:
This is a WIP. The approach taken so far is to change the termination sequence to actually not exit in leader election callbacks, but rather return from the Run function. This path was taken to ensure both the main and the migration lease is released if possible. This would not happen if one of the leader election threads was to exit immediately, even on success.
I actually implemented a helper object to manage the synchronization, but there are no tests so far. This is a PoC.
Does this PR introduce a user-facing change?
Additional documentation e.g., KEPs (Kubernetes Enhancement Proposals), usage docs, etc.:
N/A