4 months ago
I’m currently running an n8n setup on Railway with a Primary/Worker architecture using Postgres and Redis. Recently, my Worker deployment (Inflearn_n8n) keeps crashing with the error message: “Deploy Ran Out of Memory!” in the production environment.
Here’s the situation:
• Platform: Railway (self-hosted n8n with queue mode)
• Affected Service: Worker (Primary still runs but sometimes fails healthchecks afterward)
• Error: “Deploy Ran Out of Memory” / Exit code 137
• Redis and Postgres are connected and seem stable
• The issue happens even when workflows are idle or processing small data volumes
2 Replies
I’m currently deploying n8n on Railway (Docker image: n8nio/n8n) and the deployment consistently gets stuck during the “Network: Performing healthchecks” phase. It never progresses to the post-deploy stage and eventually times out.
4 months ago
Use this template, it works best.
https://railway.com/deploy/n8n-with-workers
Make sure you are subscribed to use enough resources.
Atleast 1 GB of RAM. Check out here: https://docs.n8n.io/embed/prerequisites/