Skip Daily Tests for Version changes

[sarthakaggarwal97]

Currently, we run all the daily tests whenever a PR is raised to one of the released branches.

We have started to automate the release process with automated release notes PR like #4247.

Daily tests are time consuming and often flaky. I think we do not need to run daily tests on these PRs, and CI could still be a good sanity check for version.h file changes.

This can help the release notes PR to be ready faster.

[coderabbitai]

📝 Walkthrough

Walkthrough

The daily GitHub Actions workflow now ignores pull requests that only change src/version.h.

Changes

Daily workflow filtering

Layer / File(s)	Summary
Exclude generated version header changes .github/workflows/daily.yml	Adds src/version.h to the pull request paths-ignore list.

Estimated code review effort: 1 (Trivial) | ~2 minutes

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Title check	✅ Passed	The title accurately summarizes the main change: skipping daily tests for version-related PRs.
Description check	✅ Passed	The description matches the change and objective, explaining why daily tests should be skipped for version.h-only PRs.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[valkey-review-bot]

The DCO check is failing: commit 3a1117c (ci: skip daily tests for version-only changes) has no Signed-off-by: trailer. Since this branch has a single commit, git rebase HEAD~1 --signoff and a force-push of ci/skip-daily-for-version-header will fix it.

[valkey-review-bot]

The paths-ignore addition is correct for the stated goal: GitHub skips the workflow only when every changed file matches an ignore pattern, so version-bump/release-notes PRs (src/version.h + 00-RELEASENOTES/markdown) stop triggering the daily suite, while any PR touching version.h alongside code still runs it. ci.yml and the other workflows do not ignore src/version.h, so regular CI still covers version-only PRs, and the schedule/workflow_dispatch/workflow_call triggers are unaffected by path filters. No issues with the change itself; the only blocker is the missing DCO sign-off noted separately.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 76.79%. Comparing base (f4dc3ca) to head (71e9e2a).

Additional details and impacted files

@@             Coverage Diff              @@
##           unstable    #4255      +/-   ##
============================================
- Coverage     76.89%   76.79%   -0.11%     
============================================
  Files           162      162              
  Lines         81487    81487              
============================================
- Hits          62663    62581      -82     
- Misses        18824    18906      +82     

see 20 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[roshkhatri]

LGTM!

[roshkhatri]

But we need to implement a signoff strategy on the versions with the failing tests, even if they are flaky.
Right now we are quite hand wavy about how we treat the flaky tests for daily failures on release notes PRs.

Now that we will skip the daily run, we should have the release owner add a signoff of the tests that failed on the last daily run on the release branch, and why we are moving forward with the release.

[sarthakaggarwal97]

Now that we will skip the daily run, we should have the release owner add a signoff of the tests that failed on the last daily run on the release branch

I think that's a good point. There could be scenarios where the release owner does not have the visibility about daily test failures on that branch. I don't feel too strongly about this anymore. I will close this for now. This seems like a good gate to have.