flaky: Windows scripts job hits test timeouts (check-dco, i18n runner/translate, check-changed)

Ouverte
#550 1 commentaire 0 réactions 0 personnes assignées Voir sur GitHub

Les mainteneurs répondent en général sous 1 jour

Personne n'a encore pris cette issue.

Évaluation

Difficulté
3/5
Temps estimé
1-2 jours
Accessibilité débutants
56/100
Type d'issue
Bug
Clarté
Plutôt claire
Activité
Active
Stack technique
bun, github-actions, rust, typescript
Domaine
ci-cd, testing

Piste de recherche

Start by reading the four named tests: scripts/check-dco.test.ts, scripts/i18n/runner.test.ts, scripts/i18n/translate.test.ts, and scripts/check-changed.test.ts. Check how bun run release:test sets test timeouts on Windows and how the cargo metadata test handles its cache. Confirm the intermittent timeout locally or in CI; done means the Windows scripts job passes reliably without masking genuine test failures.

Rédigé par le modèle d'indexation à partir du texte de l'issue.

Description

area: ci os: windows type: bug

Repo scripts (windows-latest) fails intermittently on bun run release:test because tests hit bun's default 5 s timeout (and one its 60 s timeout) on the Windows runner. Everything else in the run passed; a rerun of the failed job passed.

Evidence

Run 44112407920 (PR #462, head 5a0796e0), job 134145061890, test summary:

  • scripts/check-dco.test.ts: a signed-off commit by a real identity passes timed out after 5000 ms.
  • scripts/i18n/runner.test.ts: bun scripts/i18n/check.ts > info findings are hidden unless --info, and never fail the run timed out after 5000 ms.
  • scripts/i18n/translate.test.ts: the prompt > includes at most 20 existing pairs of the same area, never the keys being translated timed out after 5000 ms.
  • scripts/check-changed.test.ts:129: --json --stdin (the atlas_affected plugin) > a crate under crates/ resolves through cargo metadata to its dependents timed out after 60000 ms.

Frequency: 1 occurrence in the 4 or so Windows scripts jobs I watched on 2026-11-17 (the runners were heavily queued). All four tests spawn a child process (git, bun, cargo metadata), which is slow on Windows when the machine is busy.

Suspected cause (not verified)

Spawning git/bun/cargo metadata per test on a contended Windows runner exceeds the default test timeouts. The cargo metadata test has no warm cache on that job. Raising the per-test timeout on Windows (or a CI-wide --timeout), or caching cargo metadata, would be the options.

No fix in the PR where it was seen.

Langage dominant
Rust
Étoiles
3
Forks
1
Merge moyen
9 h 51 min
PR mergées (30 j)
310

Préparer son environnement

Par où commencer

  1. Lisez l'issue en entier, puis le guide de contribution du projet.
  2. Signalez en commentaire que vous la prenez — cela évite que deux personnes fassent le même travail.
  3. Forkez le dépôt et travaillez sur une branche.
  4. Ouvrez une pull request qui référence le numéro de l'issue.

Autres issues de codeatlasdev/atlas

Toutes les issues de codeatlasdev/atlas

Issues similaires

Plus d'issues Rust

Recevez les nouvelles issues par e-mail

Un résumé court des issues GitHub adaptées aux débutants.