Title: Postgres service internal networking broken - TCP timeout on private domain
rocky-paimon
PROOP

a month ago

Description:

Service Postgres-rwa-user (ID: 8b2cca4f-fc7a-41b1-b5ac-9efe8446c0e1) in project paimon-prod (ID: 1882c87d-7d56-4130-9668-e7f2ed2a3661), production environment.

The PostgreSQL process is running normally (logs show "database system is ready to accept connections", and Railway Data UI can query it). However, the internal private domain

postgres-80hj.railway.internal:5432 is unreachable from other services in the same project — TCP connection times out.

- DNS resolves correctly to 10.169.22.44 / fd12:ded4:fe6f:1:5000:ae:52a9:162c

- TCP connection to port 5432 times out (both IPv4 and IPv6)

- Redis in the same project works fine via internal networking

- Public TCP proxy also fails (server closes connection unexpectedly)

- Multiple redeploy/restart/rollback attempts did not fix the networking

- ghcr.io image pulls also timeout (i/o timeout to 140.82.116.34:443)

Please investigate the container networking for this service.

$20 Bounty

1 Replies

Status changed to Awaiting Railway Response Railway about 1 month ago


Status changed to Open Railway about 1 month ago


quantumturbulence1
FREE

11 days ago

The ghcr.io timeout is a useful clue: this may be broader container networking/egress trouble, not just Postgres config.

I would prove the split from inside the affected service:

```bash

getent hosts "$PGHOST"

nc -vz "$PGHOST" "$PGPORT"

curl -I https://ghcr.io


Welcome!

Sign in to your Railway account to join the conversation.

Loading...