2 months ago
Our production system is currently down. This is a critical business application used during working hours, and this level of downtime is unacceptable for our operation.
Project/environment: TJRadar / production
Affected services: Backend and Postgres
Backend public URL: https://tjradar-backend.up.railway.app
The backend is failing to respond and returns:
502 Application failed to respond
In the backend logs, we are repeatedly seeing Postgres connection timeouts through Railway private networking
2 Replies
2 months ago
This thread has been opened as a bounty so the community can help solve it.
Status changed to Open Railway • about 2 months ago
2 months ago
This smells like the app is booting, trying to connect to Postgres over the private network, timing out, and then Railway reports it as a 502.
I’d check the exact DATABASE_URL the backend is using first. Make sure it points to the Postgres service in the same Railway environment and not an old/public TCP proxy URL.
Quick test from the backend shell:
pg_isready "$DATABASE_URL"
If that fails, check whether the internal host resolves at all:
getent hosts .railway.internal
If it eventually connects but your app exits first, add DB retry/backoff during startup. Then redeploy Postgres, wait until it’s healthy, and redeploy the backend.
2 months ago
Issue was solved, but the problem was a hardware Railway had will failure for a couple of hours.
Status changed to Solved iapolinario • about 2 months ago