Hi everyone! I'm hoping to get some help with a queue issue that's been driving me crazy for days.

I recently deployed n8n on Railway (queue mode with Primary, Worker, Redis, and Postgres). Everything was working, but now all my executions are permanently stuck on "Queued / Starting soon" and the workflows don't trigger.

The Symptoms:

Earlier, I had a workflow that caused a bottleneck, and my Primary logs showed: (node:1) MaxListenersExceededWarning: Possible EventEmitter memory leak detected. 11 global:completed listeners added to [Queue].

Now, the primary logs just show Enqueued execution 232 (job 200) but the worker never seems to pick them up.

What I've already tried (without success):

Fixed the problematic workflow JSON that caused the initial hang.

Canceled all stuck/running executions manually from the UI.

Restarted ALL services in Railway (Primary n8n, n8n-worker, Postgres, and Redis).

Flushed the Redis cache / cleared the volume to remove deadlocks.

Confirmed that the N8N_ENCRYPTION_KEY is exactly the same on both the Primary and Worker containers.

I've only had this trial account running for 3 days. Is there a specific configuration for the workers on Railway I might have missed, or another way to forcefully unclog the Redis queue?

I'm attaching some screenshots of the logs and the UI. Any guidance would be greatly appreciated!