Crash everyday
Anonymous
PROOP

18 days ago

Hello Railway Support,

I’m experiencing a recurring issue with my Railway deployment.

Problem description:
My service goes down almost every day. BetterStack monitoring detects incidents where the readiness endpoint returns HTTP 503. The issue resolves automatically after a few minutes, but it happens regularly and affects stability.

Service details:

  • Platform: Railway

  • Application: n8n (production instance)

  • Endpoint checked: /healthz/readiness

  • Error: Status 503

  • Typical duration: about 2 minutes

  • Incident example:

    • Started: Feb 24, 2026 at 8:37 PM EET

    • Resolved automatically: Feb 24, 2026 at 8:43 PM EET

What I observe:

  • The service becomes temporarily unavailable.

  • BetterStack sends “New incident started” alerts.

  • After a short time, Railway recovers automatically without manual restart.

  • This behavior happens daily.

What I need help with:

  • Understanding why the service temporarily fails readiness checks.

  • Identifying whether this is caused by Railway infrastructure, resource limits, or container restarts.

  • Recommendations on how to prevent these daily outages.

If needed, I can provide additional logs or configuration details.

Thank you.

$20 Bounty

1 Replies

xmrafonso
FREETop 5% Contributor

17 days ago

Hey,

Any ideas of exit codes when it crashes? I recall that n8n sometimes suffers from memory spikes and if your service does not have enough memory allocated, it can cause it to run out of memory and make railway force restart the container.
Any type of informations on this regard would help.

Loading...