-
Notifications
You must be signed in to change notification settings - Fork 1.9k
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
Inconsistent silent failure at the end of action on ubuntu-24.04-arm
#2064
Comments
We're seeing the same thing, except during the initial checkout, not during cleanup. About 10% of the time it will just fail after 1 second, with no error message. Only happens on arm runners. It's not the same step either, random different steps may fail on different runs - but its always the checkout action. |
@brandond yes, it happens during initial checkout, sorry if my wording was unclear. "Cleanup" refers to the line from logs of that step. |
We are seeing the same thing over at https://github.com/electricitymaps/electricitymaps-contrib after we switched to the new arm runners. All random and all seems to work if you restart the workflow again. |
Also independently reported to the runner announcement, https://github.com/orgs/community/discussions/148648#discussioncomment-11866202, there's one more affected workflow example. |
also cc @lkfortuna |
This is not an action/checkout issue but a hardware issue with the runners themselves, we are tracking in the community discussion, I am going to close this one out here in favor of keeping things in one issue. |
actions/checkout@V4
occasionally fails silently at the very end of the action, preventing the rest of the job from running. I never encountered this on other runners, this seems specific to ARM (ubuntu-24.04-arm
) runners.May be relevant to #2053, but that one mentions sparse checkout and lacks debugging info.
Example run: https://github.com/python/mypy/actions/runs/13183337167/job/36799380042?pr=18620
Relevant step:
Failed execution logs
Successful execution logs
And here's the diff:
"Post job cleanup" looks relevant: it only happened mid-job (before "syncing repository") in the fail case for some reason.
The text was updated successfully, but these errors were encountered: