Railway stops the container ~3 seconds later and restarts it in a loop

Question

Hey team.I'm having a frustrating issue with a Node.js/Express backend that runs perfectly locally but keeps crash-looping on Railway.

The problem:
My app starts (`Server running on port 8080`), but Railway stops the container ~3 seconds later and restarts it in a loop. No error is printed after startup — the process just dies silently.

What I've ruled out:
- App runs fine locally with `NODE_ENV=production node src/app.js` — no crashes
- DATABASE_URL` is set and Prisma connects successfully (migrations run fine in logs)
- No hardcoded PORT — I let Railway inject it
- trust proxy 1` is set (was getting ERR_ERL_UNEXPECTED_X_FORWARDED_FOR from express-rate-limit, now fixed)
- No placeholder env vars
- uncaughtException` and `unhandledRejection` handlers are in place but nothing is logged before the stop

Relevant log pattern:
```
Starting Container
Prisma: No pending migrations to apply
Stopping Container        ← happens ~3s after start, no error
Starting Container
Server running on port 8080
Environment: production
(dies again)
```

Stack: Node 22, Express, Prisma 5.22, express-rate-limit, Railway Postgres (internal URL)

Is Railway sending a health check signal that my app isn't responding to in time? Could there be a startup timeout I'm hitting? Any way to see what signal is killing the container?

Thanks