I’m experiencing an issue with my execution queue being stuck.
What happened: I had 5 workflows scheduled at 30-minute intervals. After they started running, I changed each to 1 hour 10 minute intervals, which caused scheduling overlaps. This triggered the concurrency limit (5 concurrent executions), and all workflows went to “Queued” status.
I then cancelled all executions and deleted those 5 workflows. However, my other scheduled workflows are now permanently stuck in “Queued” status, even though Current Active Executions shows 0 out of 5.
What I’ve tried:
Cancelled all pending executions
Deleted the problematic workflows
Restarted the workspace
Manual execution works fine
Only scheduled triggers show “Queued”
Request: Could you please manually reset/clear the execution queue on my instance? It seems the queue state didn’t properly clear after I cancelled and deleted those workflows.
I’ve seen this exact thing a couple of times – when you cancel executions during a concurrency overload, the internal queue counter sometimes gets “stuck” and doesn’t reset to zero, even though the dashboard shows 0/5 active.
Manual runs work because they bypass the scheduler queue completely.
Unfortunately we can’t manually clear the queue from the community forum (only the n8n team can do that on hosted/cloud instances), but since you’re self-hosting, here’s what almost always fixes it:
(or if you have direct DB access, truncate the execution table carefully).
Restart the container.
This forces a clean state and the scheduler queue usually resets itself.
If that doesn’t do it, try adding --concurrency=1 temporarily when starting n8n, let one scheduled workflow run through, then remove the flag and restart again.
Let me know how it goes – it’s annoying when it gets stuck like this! Good luck