I have a workflow that processes multiple YouTube videos and sends them to a local transcription service via HTTP POST. The response from the transcription service (studio_transcript_m) contains 3 valid items, all with status: 0.
These 3 items are passed to an IF node (If_no_err) that checks:
{{ $json.status }} is equal to 0
However, in the execution trace:
- Only 2 items are processed by the
IFnode (both correctly) - The execution ends with the message:
“Workflow execution had an error — The execution was cancelled”
But:
- I did not manually cancel the execution (it happened during the night)
- All inputs and outputs are valid JSON
- No node shows an error
- The transcription step for all 3 items completed successfully (confirmed by output logs)
It seems n8n cancelled the execution after 2 of 3 items were passed to the IF node, before the last one could be evaluated.
What is the error message (if any)?
Workflow execution had an error – The execution was cancelled
(No node failed, and no error stack is shown)
Share the output returned by the last node
One of the successful transcription results:
{
"status": 0,
"id": "LJnchPGJ_os",
"transcription": {...
}
}
All 3 items look like this, and all have status: 0.
Information on your n8n setup
- n8n version: 1.99.1
- Database: PostgreSQL
- n8n EXECUTIONS_PROCESS setting: own
- Running n8n via: Docker
- Operating system: macOS
My question:
Why would n8n cancel an execution after 2 of 3 items were already processed by an IF node?
Could this be:
- A timeout or memory limit reached during evaluation?
- A problem in the execution engine when handling long-running requests?
- Something else I can debug or catch earlier?
Happy to provide logs, workflow JSON, or screenshots if helpful.
Thanks in advance!
