You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Pushing to an internal PR naturally triggers all checks that run on push and pull_request events.
Like the other jobs, the check job gets skipped on the pull_request trigger for internal PRs, since the push event already runs the same stuff.
Since the check (push) job depends on other jobs and doesn't run (or even get queued) until they're finished, and the check (pull_request) job gets skipped, there's a window where the requirements set in the branch protection rules are fulfilled (as skipped checks are valid) and the PR gets merged even though some jobs aren't finished yet.
For example, this pytest job finished at 13:32:34, while the PR had already been merged at 13:32:16.
The text was updated successfully, but these errors were encountered:
It looks like if we change some of our workflows and tighten them up a bit as shown here we should solve the issue, since it would cancel the workflow instead of skipping it.
Pushing to an internal PR naturally triggers all checks that run on
push
andpull_request
events.Like the other jobs, the check job gets skipped on the
pull_request
trigger for internal PRs, since thepush
event already runs the same stuff.Since the
check (push)
job depends on other jobs and doesn't run (or even get queued) until they're finished, and thecheck (pull_request)
job gets skipped, there's a window where the requirements set in the branch protection rules are fulfilled (as skipped checks are valid) and the PR gets merged even though some jobs aren't finished yet.For example, this pytest job finished at 13:32:34, while the PR had already been merged at 13:32:16.
The text was updated successfully, but these errors were encountered: