Skip to content

Fix LeaderElector waiting a full extra lease duration before acquiring an already-expired lease on startup - #4914

Draft
brendandburns with Copilot wants to merge 2 commits into
masterfrom
copilot/fix-leader-elector-expired-lease
Draft

brendandburns with Copilot wants to merge 2 commits into
masterfrom
copilot/fix-leader-elector-expired-lease

Conversation

Copilot AI commented Sep 18, 2026

Copy link
Copy Markdown
Contributor

On first observing an existing lock record, LeaderElector stamped its internal observedTimeMilliSeconds with the current wall-clock time rather than the record's actual renewTime. This made a freshly-started candidate treat an already-expired lease as still valid for one more full leaseDuration, delaying takeover instead of acquiring it immediately.

  • LeaderElector.tryAcquireOrRenew: when the observed record is being set for the very first time (no prior observedRecord), initialize observedTimeMilliSeconds from the record's own renewTime instead of System.currentTimeMillis(). Subsequent observations during normal operation still use the current time, preserving the existing clock-skew tolerance.
  • Test: added acquiresAlreadyExpiredLeaseImmediately in LeaderElectorTest, which seeds a lock with a record whose renewTime is well past leaseDuration and asserts a new candidate acquires leadership promptly rather than waiting out an extra lease period.
if (!oldLeaderElectionRecord.equals(this.observedRecord)) {
  boolean firstObservation = this.observedRecord == null;
  this.observedRecord = oldLeaderElectionRecord;
  this.observedTimeMilliSeconds =
      firstObservation
          ? oldLeaderElectionRecord.getRenewTime().getTime()
          : System.currentTimeMillis();
}

@kubernetes-prow kubernetes-prow Bot added the do-not-merge/work-in-progress Indicates that a PR should not merge because it is a work in progress. label Sep 18, 2026
@linux-foundation-easycla

linux-foundation-easycla Bot commented Sep 18, 2026

Copy link
Copy Markdown

CLA Not Signed

@kubernetes-prow kubernetes-prow Bot added the cncf-cla: no Indicates the PR's author has not signed the CNCF CLA. label Sep 18, 2026
@kubernetes-prow kubernetes-prow Bot added the size/XS Denotes a PR that changes 0-9 lines, ignoring generated files. label Sep 18, 2026
…tion

Co-authored-by: brendandburns <5751682+brendandburns@users.noreply.github.com>
@kubernetes-prow

Copy link
Copy Markdown
Contributor

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: Copilot
Once this PR has been reviewed and has the lgtm label, please ask for approval from brendandburns. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details Needs approval from an approver in each of these files:

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

@kubernetes-prow kubernetes-prow Bot added size/M Denotes a PR that changes 30-99 lines, ignoring generated files. and removed size/XS Denotes a PR that changes 0-9 lines, ignoring generated files. labels Sep 18, 2026
Copilot AI changed the title [WIP] Fix LeaderElector fails to acquire expired lease on initial run Fix LeaderElector waiting a full extra lease duration before acquiring an already-expired lease on startup Sep 18, 2026
Copilot AI requested a review from brendandburns September 18, 2026 14:32
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

cncf-cla: no Indicates the PR's author has not signed the CNCF CLA. do-not-merge/work-in-progress Indicates that a PR should not merge because it is a work in progress. size/M Denotes a PR that changes 30-99 lines, ignoring generated files.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

LeaderElector fails to acquire expired lease on initial run due to missing expiration check

2 participants