Conversation
Performance metrics 🚀
|
| on: | ||
| push: | ||
| branches: [main] | ||
| pull_request: |
There was a problem hiding this comment.
To be honest I would still run CodeQL for PRs against main, as it will block the PR from being merged if there are CodeQL issues.
There was a problem hiding this comment.
I think it's probably fine to run it on main only, because I don't remember a single time where it failed for a valid reason 😅 We'll still get the notification if it's failed on main and see the results here: https://github.com/getsentry/sentry-java/security/code-scanning
The reason we don't want to run it on PRs because without build-cache it'd be the longest one to run. Alternatively we could skip the run altogether, if there were no code changes to prevent the job from failing, but I think it's a good middle-ground for now.
romtsn
left a comment
There was a problem hiding this comment.
LGTM, let's see how it goes and we add a skip-check later, if we see it necessary 👍
📜 Description
This runs CodeQL with build caching disabled post-merge.
#skip-changelog
💡 Motivation and Context
CodeQL is slow and requires the compilation tasks to re-run. By disabling build caching we can ensure that CodeQL is run correctly. However since this is one of the longest checks, it doesn’t make sense to run it during PRs.
💚 How did you test it?
📝 Checklist
sendDefaultPIIis enabled.🔮 Next steps