Hi,

I'm experiencing an issue where my backend container shuts down just a few seconds after deployment, even though the logs show that the Uvicorn server starts properly. Here’s what I observe:

• The deployment logs show:

  INFO:     Uvicorn running on http://0.0.0.0:<port>
  INFO:     Shutting down
  INFO:     Application shutdown complete.
  INFO:     Finished server process [1]
  

• Then the container stops and the service is marked as failed.

• I have removed --reload from my startCommand, and I’m correctly using the $PORT variable injected by Railway.

• My healthcheck path is set to /healthz, and I have verified that this endpoint returns a 200 OK when tested locally.

• The same command works perfectly on my local machine.

Here is my startCommand:

uvicorn backend.main:app --host 0.0.0.0 --port $PORT --log-level info

I’ve also tried wrapping it via start_backend.py with:

port = int(os.getenv("PORT", 8000))
uvicorn.run(app, host="0.0.0.0", port=port, log_level="info")

But the result is the same — the process starts, then exits within a few seconds.

Is there anything that could be triggering an early shutdown from Railway’s side? Or is there a recent runtime behavior that would cause the container to stop even if Uvicorn is active?

Thanks in advance — I’d really appreciate help figuring this out.

Best,