fix(test): raise pipeline-e2e timeout to 30s (#1169)#1170
Merged
Conversation
…ntion flake (#1169) The `eval pipeline e2e > runs full input → grade → bench pipeline` test spawns three sequential `bun apps/cli/src/cli.ts pipeline ...` subprocesses (input → grade → bench). In isolation it completes in ~7-12s, well under the 5000ms default per-test timeout, but under suite contention from `bun --filter agentv test` it routinely overshoots and times out. Bump the per-test timeout to 30s using `it(name, fn, timeout)` (Bun's test runner supports the numeric third-arg form). Other suite tests are not modified — only the test cited in the issue. Closes #1169
Deploying agentv with
|
| Latest commit: |
c50d372
|
| Status: | ✅ Deploy successful! |
| Preview URL: | https://e0da385c.agentv.pages.dev |
| Branch Preview URL: | https://fix-1169-pipeline-e2e-timeou.agentv.pages.dev |
This was referenced Apr 27, 2026
christso
added a commit
that referenced
this pull request
Apr 27, 2026
…1170) (#1176) The `pipeline input` describe block in `apps/cli/test/commands/eval/pipeline/input.test.ts` exhibits the same 5s-default timeout flake under suite contention as the e2e test fixed in #1170. Each test spawns a `bun apps/cli/src/cli.ts pipeline input ...` subprocess that completes in ~1-2s in isolation but routinely overshoots 5s under `bun --filter agentv test` contention. Apply the same per-test timeout bump (`it(name, fn, 30_000)`) to all 10 tests in the describe block. The user brief named three sibling tests (`writes code_graders/<name>.json for deterministic assertions`, `omits experiment from manifest...`, `falls back to eval file basename...`); the remaining seven exhibited identical SIGTERM-at-5s failures in the same suite run, so the bump is applied uniformly to prevent partial coverage churn. Same one-line treatment per the #1170 precedent.
2 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #1169
Problem
apps/cli/test/commands/eval/pipeline/pipeline-e2e.test.ts(eval pipeline e2e > runs full input → grade → bench pipeline) flakes under suite contention frombun --filter agentv test, timing out at the 5000 ms per-test default. The test passes in isolation in ~7-12s — it's a real end-to-end test that spawns three sequentialbun apps/cli/src/cli.ts pipeline ...subprocesses (input → grade → bench), so 5s is too tight when the suite is busy.Fix
Raise the per-test timeout to 30 s using Bun's
it(name, fn, timeout)numeric-third-arg form. One-line change, no scope creep. The pipeline plumbing assertion ("the wiring works") is preserved as-is — fixture not trimmed because the existing fixture is already minimal and the timeout bump alone fully solves the flake.Red / green evidence
(fail) eval pipeline e2e > runs full input → grade → bench pipeline ... ^ this test timed out after 5000ms.bun test apps/cli/test/commands/eval/pipeline/pipeline-e2e.test.ts→1 pass, 0 failin ~12 s.bun --filter agentv testran multiple times; the pipeline-e2e test no longer appears in any failure output. Pre-push hook (which runs the full suite) passed cleanly on the green push.Notes
pipeline inputtests inapps/cli/test/commands/eval/pipeline/input.test.tsexhibit the same 5s-timeout-under-contention pattern and flake intermittently. They are out of scope for test: pipeline-e2e flake at 5000ms default timeout #1169 and would benefit from the same treatment in a follow-up.