Self-hosted runners - offline

We’re using Actions with 2 self-hosted Windows runners, 2 Mac runners, 2 Linux runners.
Linux and Windows are rock-solid: never had to manually restart or otherwise troubleshoot them going offline.
Mac runners, every so often (about every 1-3 months) will show as Offline, where the solution would be to either manually stop and start the service (last happened today), or, worse, log in via UI (didn’t have to do that in about 4 months).


  1. Why does it happen, specifically on Mac, but not on the other OSs? All of our runners take about the same amount of load/jobs.
  2. More importantly, is there a way to trigger an alert (email/Slack/other webhook) when a provisioned self-hosted runner changes its status to or from Offline?

Hey and welcome @w3sip ! How has it been in the last weeks? I’m afraid I can’t answer your questions straight away, but to improve chances of someone stumbling onto something I would be interested if the service on Mac logs anything or if you have any idea what the root cause of the crash there is.

It hasn’t happen in awhile, so the memory is fuzzy – but my recollection is that there wasn’t anything useful in the logs. Previously (when log in via UI was required), this was directly related for self-upgrade of the runner, though I’m not sure if the last occasion or two were.
I’ll make sure to preserve the logs next time it happens.

Regardless, I think a notification feature for runner’s status change would be unrelated to this, don’t you agree?