Your Postgres service is confirmed in a CRASHED state. The timing aligns with our major service disruption on May 19, which affected image registries and is now resolved. A fresh redeploy should re-pull the image cleanly - try redeploying from your service's settings. Your volume data is intact and persists independently of deployments, but it can only be accessed through the running service. If the redeploy still fails with the same error, you can change the service's source image in Settings (e.g., to the official postgres:18 image from Docker Hub) while keeping the same volume mount, which would let Postgres start and give you access to your data.

I noticed something that might be related: my account sometimes shows "trial maxed out" at the top, and other times shows "16 days or $4.50 left". I registered recently and I do have credits remaining, so I'm not sure why it's inconsistent.

Could this billing/trial state be blocking redeployments? When I try to redeploy via the dashboard (changing source image to postgres:18), the "commit and deploy" prompt appears but does nothing. Via CLI I get:

Failed to fetch: error sending request for url (https://backboard.railway.com/graphql/v2) — operation timed out

Railway status page shows all systems operational. What should I try next?

After multiple failed redeploy/restart attempts, I switched the source image to the official postgres:18 Docker Hub image which got the service running again. Once it was up and the data was confirmed intact, I switched back to the original ghcr.io/railwayapp-templates/postgres-ssl:18 image and redeployed — it's now running fine on the SSL image as well. Thanks for the help!

Resolved at 2026-05-20 12:06 UTC (14:06 EET) — after a few hours of downtime. One of my many retry attempts just happened to land at the right time.