BuildButler

Flaky Tests

Configure thresholds for flaky test detection in BuildButler.

Configure thresholds for flaky test detection. Tests that exceed these thresholds will be flagged on the Flaky Tests dashboard.

Flaky Tests settings

Transition Count

Mark a test as flaky after a set number of pass/fail flips within a rolling window of runs.

  • Flips — number of status changes (default: 5)
  • Runs — the rolling window to look back over (default: 30)

Reliability Threshold

Flag tests with a pass rate below a given percentage (default: 80%).

Remove flaky tests from calculations

When enabled, known flaky tests are excluded from pass/fail counts and pass rate calculations across dashboards and build details.

Click Save to apply, or Reset to Defaults to restore original values.

On this page