frontend-test fails intermittently in CI and passes on a retry #404
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
frontend-testgoes red on unrelated pushes and is green when the run is restarted, so the failure is in the run, not in the diff.Measured on #403, a CHANGELOG-only change: the job failed after 220s, while the same suite locally is 1634 passed in 10.5s. Its neighbours on that push were just as slow (
backend-test254s,e2e254s), which points at runner contention rather than a bad test - we push 13 jobs at one runner. Related: the e2e flakiness under the same contention, and #402.Worth finding out before adding more retries: which tests fail, whether it is always the same ones, and whether it is timing out rather than asserting. The job's own log is the starting point - the Forgejo API does not serve it, so this needs the web interface.
Note a restart is not free: it hides how often this happens, and the number is what decides whether this is worth fixing at the runner or in the suite.