Hey! Stuck executions piling up in the queue is a known scenario, and there are a few things you can check.
Quick context on the limits: Automate executions are supposed to time out after 45 minutes in the execution queue, with a 4-hour cap on event runtime (docs). If something’s been sitting > 12 hours, that timeout isn’t firing for some reason, which usually points to one of these:
-
“Queue effect executions” is enabled on the Condition tab. With this on, events run strictly serially so one slow event will block everything behind it. If you don’t actually need ordering guarantees, turning it off avoids this whole pile-up. (docs)
-
Internal parallelism cap hit by a burst of triggers, same symptom (everything queues behind the slow ones), but resolves on its own once the head-of-line drains.
-
The execution itself is stuck, in which case the 45-min queue timeout never trips. That’s the case where you actually need manual intervention.
You can cancel individual executions, but there is no bulk-cancel, unfortunately:
-
Open the automation → History view.
-
Click the stuck execution to open its sidebar.
-
Hit the red Cancel button.
That only cancels the one execution you’ve selected, not the whole automation. The button is disabled if the execution is already cancelled/completed, or if it’s still computing the input set.