As a developer I want to re-run only the job that has failed
Currently on ansible-pulp, if one job fails, you can not restart only this job, you have to restart all jobs for the current workflow. I want to be able to restart only the job that has failed
Currently ansible-pulp has 2 workflows:
- Pulp CI: 3 jobs
- Cron Jobs: 9 jobs
It seems, we only can start entire workflows: https://github.community/t5/GitHub-Actions/Manually-restart-actions-and-entire-workflows/m-p/35907/highlight/true#M2365 Which in this case we would end up with at least 9 workflows
- Status changed from ASSIGNED to POST
As maintaining 9 workflows doesn't seem good, I propose adding an automatic retry when job fails.
Also refactored, so we have one workflow just for pull requests, and another workflow for master branch health
On NPM, I was able to put 2 workflows in 1, but I could not split workflows by cases due to deploy depending on more than one case, see: https://github.com/pulp/pulp_npm/pull/13/files#r399558331
- Status changed from POST to MODIFIED
- % Done changed from 0 to 100
Applied in changeset ansible-pulp|9f29d5aa2666febe88c4c93fe472dba171286d00.
Retry does not work, it keeps the status as failed, and it is not possible to change job status https://github.community/t5/GitHub-Actions/Set-job-status-to-success-even-when-a-step-fails/td-p/44234
Please register to edit this issue