-
Notifications
You must be signed in to change notification settings - Fork 28
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
fix: fix get failured checks #28
Conversation
Signed-off-by: Zixuan Liu <nodeces@gmail.com>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There's a misunderstanding. The new CI requires using the new API for re-running just failed jobs and not the complete workflow. Please report what is broken first and we can think of a solution.
I explained some of the changes in https://lists.apache.org/thread/n05xp17krb4f3cc0jjkh65ytqk1w6130 . You can also provide feedback on the mailing list. |
With the new CI there's another case where you see a red mark. It's when flaky tests fail, but later pass in a retry, the failed test result will get recorded to a test report. That doesn't block merging, but will show up so that the failures can be inspected. It's a bit confusing since GitHub Actions has a bug that the test reports get attached randomly to a workflow when there are multiple workflows executing. I'll write to the dev list explaining this since I have received a few questions about this. It might appear that re-running doesn't work because of the new test reports. Those failures won't clear off, but they don't block merging either. |
In your case, it's the flaky tests report: https://github.com/apache/pulsar/pull/15065/checks?check_run_id=5867398812 . |
@lhotari Thanks for your explanation!
I have a question about the
This is detailed.
This is confusing, I'm looking forward your solution. |
Yes. That happens when an actual build run hasn't failed. As explained in previous comments, the test reports added to the build are causing this. The benefit of seeing the failed test reports is that we can start paying more attention to the flaky test problem and fixing them. Flaky tests might be indicating real production code issues.
There's really no other solution than documenting this. We could remove the rendered test reports, but I think that integrating test reports to GitHub Actions UI is useful. |
There's more details in https://lists.apache.org/thread/60x7sqg2p4mlssj5jtow6zwq3jksf6w3 . Currently there's an issue with GitHub Actions that the workflow run search doesn't return all results and that's why rerunning failed jobs doesn't work in some cases. |
I have fixed the rerunning in cases where the PR's head sha doesn't change (for example closing and reopening the PR to trigger a new build): #33 |
Signed-off-by: Zixuan Liu nodeces@gmail.com
Motivation
I tried to rerun the failed test, but the failed test cannot be rerun. I checked the https://github.com/apache/pulsar/runs/5869256539?check_suite_focus=true#step:5:314, no tests were run.
Using https://docs.github.com/en/rest/reference/actions#list-workflow-runs-for-a-repository is not working, this API cannot find my actions, it seems to fetch all actions.
You can use this command to reproduce:
If we use the old way to get the failed checks, it is exactly: