My PostgreSQL service repeatedly fails at PRE_DEPLOY_COMMAND even though the current Railway service configuration explicitly shows preDeployCommand: [].

Two fresh deployment attempts have failed identically. SNAPSHOT_CODE completes successfully, then PRE_DEPLOY_COMMAND fails with “Pre-deploy command failed.” The configured PostgreSQL start command is never reached.

The service uses the official postgres:16 image. Historically, there was a faulty pre-deploy command, but it has been removed. Railway's deployment diagnostics also report that the command has been cleared and suggest a fresh deployment should succeed. Fresh deployments nevertheless continue to fail at the same PRE_DEPLOY_COMMAND stage.

Could Railway staff please check whether an old pre-deploy command remains cached or persisted somewhere not exposed by the current service configuration?

There is an existing persistent PostgreSQL volume containing data, so I do not want to delete, detach, replace or reformat the volume while troubleshooting.

I can provide project, service and deployment IDs privately if Railway staff need them.

Also remove Project: VanWise-TC1, the Project ID, Service ID, and deployment IDs that are currently visible in your draft.

Keep your existing title. Then you can safely press Create Public Thread.

One other thing: the page says “5 similar threads found.” Before submitting, it may be worth tapping Show—there's a chance Railway already has a known fix for this exact problem.