Repository navigation
test(sdk): cut SDK test time and skip build output in docs:check - #369
Merged
Merged
Conversation
The Hub enrollment and Compute Optimizer region-failure cases waited through the SDK's real 0.5+1+2+4+8s retry backoff, about 16s each. Run them on fake timers like the status probe retry tests: waitFor advances the admission waits before the first request, then the test advances past the backoff. Assertions, including all six attempts, are unchanged. Locally 15.5s -> 0.3s and 16.6s -> 1.4s.
docs:check walked every directory except .git and node_modules, so after a VS Code build it failed on packages/vscode/dist/vscode/media/README.md, a staged copy whose links resolve only in the published extension. Skip the Git-ignored dist, .turbo and coverage directories as well.
dannysteenman
marked this pull request as ready for review
October 30, 2026 13:12
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
Vitest defaults to one worker fewer than the CPU count, three on the 4-vCPU CI runners. The discovery HTTP suites spend most of their time waiting on admission timers, so the CPU-bound files queue behind them. Use 1.5x the CPU count; Vitest caps percentage values at the CPU count, so compute it.
roborev: Combined Review (
|
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
The SDK test task on CI drops from 92 to 100 seconds after #368 to 59 to 67 seconds, and
pnpm docs:checkno longer fails after a local VS Code build. These are the follow-ups listed in #368.What changes:
docs:checkskips the Git-ignoreddist,.turbo, andcoveragefolders. Before, a VS Code build left a staged README inpackages/vscode/distwhose links only resolve in the published extension, and the check failed on it.One risk to watch: more workers than CPUs could stretch the timer-paced tests, which is why Turbo already runs at
--concurrency=2. Two CI runs showed no timeouts, and per-file times stayed within about 2 seconds.Diagram
n/a, test and checker changes only.
Scope
@cloudburn/sdk(tests and test config only)Release Notes
Verification
pnpm verifypasses locally at the head commit. The newdocs:testcase failed before the checker fix withlocal link target does not existand passes after it.--affectedfrom ci: run only affected package checks and parallelize the SDK suite #368 now works on real PRs: the first run here selected 9 of 30 tasks and Validation took 105s.Boundary Checks
@cloudburn/rulesDocumentation
docs/TESTING.mdwith the skipped output.docs/README.mdanddocs/reference/commands.mdstay true: they list whatdocs:checkvalidates, which is unchanged.Related Issues
Refs #368