GitHub action still waiting for a runner after near 23h

I’m trying to replace Travis CI and this is my first attempt to use GitHub Actions. My first test worked perfectly but this second test has been in queue waiting for a runner for more than 22h and still waiting. Is this normal? What is the expected average delay for running an Action? Is there some problem with my Action?

That sounds unusual. If there was an error in your workflow you should get an error message, not an endless wait. The only where I would expect a practically endless wait is if you’re using self-hosted runners and those are offline. :sweat_smile:

Otherwise, you can check if there’s a known issue at https://www.githubstatus.com/, and possibly file a bug report: Bug report - GitHub Support

Thank you. I think I do not use a self-hosted runner (it is a simple action to trigger a build after each PR, but … I’m new to this). The action is still waiting (1d 43m) so I will open a bug report. Thank you again.

1 Like

Occasionally things get lost. You should treat things as dead if they aren’t resolved w/in 3 days. At that point, you can file a ticket to support.github.com w/ a link to the run and they can clean it out. (Typically these are because of some GHA outage for which githubstatus.com would have reported an item, although that isn’t always the case.)

Recently I found a pair that got stuck ~8 months ago.

Overall, for the fairly large number of runs we’ve had, the handful of items that were lost.

Thank you. I didn’t see your post until today.

The problem was mine. I that I coded:

runs-on: ${{matrix.os}}

instead of

runs-on: ${{matrix.os}}-latest

And therefore there were not runners available with that name, and the job was endlessly waiting. Perhaps GitHub should detect requests for non existing runners and abort with a meaningful message error.