Hello team,

I’ve been encountering frequent issues with ghost executions and infinitely running workflows in my n8n service hosted on Railway.

My setup consists of four interconnected sub-workflows. The first workflow handles job queuing and data forwarding to the subsequent ones. However, certain executions are behaving abnormally—showing as “running” for several hours without any actual execution data. When checked in n8n, these executions appear to have errors, but no data or node execution details are visible.

For example:

• On October 4, Workflow 1 had an execution with ID #2543 that appeared to run for 36 hours. However, no data reached the final node that triggers the next sub-workflow, indicating that no payload was passed forward.

• Subsequently, the next two workflows showed failed executions (#2544 and #2545), but opening them revealed no execution data either.

Railway logs for the same period show:

Worker started execution 2543 (job 2491)
Enqueued execution 2543 (job 2491)
Task (fKCLDEBq) rejected by Runner with reason "Offer expired - not accepted within validity window"

Could yo please investigate or provide guidance on how to resolve this