Cannot Cancel Workflow

A job that that was triggered during last nights GitHub Actions outage is stuck in the “Queued” status and when attempting to cancel errors with: “Failed to cancel workflow.”

Is there anyway to force or work around this issue?


I am also having the same issue. Won’t be able to cancel the workflow. It’s been running for 19 mins, still running. How do I cancel the workflow. It doesn’t allow me to cancel.

1 Like

Hi @khanveerr, there’s currently a major incident that’s causing new Actions workflow runs to remain queued and preventing currently running workflows from being cancelled. The engineering team are aware and investigating so I’d recommend keeping an eye on our incident page here for updates:

Hi again, thanks for your patience during this incident this morning, and sorry for any disruption caused.

The engineering team have deployed a fix so your Actions workflows should now queue and run correctly again. The team believe that most queued jobs will recover and run correctly but this may not happen immediately.

Secondly, search on the Actions page and via the API may return incomplete results. If you have an application, integration, or workflow that relies on this then be aware that these results may be incomplete for the next few days.

@thomasshaped has there been any update on this?

I’m still experiencing issues where new actions are not firing, as well as having a single action queued with no way to remove it (when I try and cancel it I get an error).


We still have a job that’s been queued since yesterday and can’t be cancelled. Will this eventually resolve itself or should we do something?

1 Like

Hi everyone, I missed this thread when I posted a second update yesterday, sorry! Any successfully queued runs should have completed not long after the green status yesterday.

If you’re still seeing queued runs that means that these runs were never actually queued successfully and are now in an invalid state in the UI. These runs will never complete and will need to be triggered again via whichever event triggered them in the first place.

I’ve been told there will be a background job that should clean these up and resolve their state in the UI, but this can take up to 72 hours.

We can also attempt to clean these up manually so if you’re seeing any “stuck” runs and you want rid of them then please open a ticket to the support team with links to any of the stuck runs here:

@airwalk225 can you post a link to your repository so I can take a look? Runs should now be triggering successfully again so if you’re still having trouble there may be something else going on.

If you can’t post publicly then I’d suggest opening a ticket with support over here with an explanation of what’s going on, including which recent events miss triggering runs if you can:

here’s ours: Actions · oracle/opengrok · GitHub

I have this issue too. I can’t stop actions.
Please check here.