Hi Railway Support,

I’m having a persistent deployment issue with one of my services. The build completes successfully, but the deployment fails immediately at the “Deploy / Creating containers” stage. There are no runtime/deploy logs produced, only build logs.

Service / Environment

• Service name: “Maximus Daemon Paper DRY-RUN”

• Builder: Nixpacks

• Region: us-west2

• Public Networking enabled

• Healthcheck path configured: /api/health

• Start command is controlled via config-as-code (railway.toml), not the UI

What happens

• Build phase completes with: “Successfully Built!”

• Deployment then fails at Deploy (00:00) / Creating containers

• “Network” and “Post-deploy” steps show Not started

• “Deploy Logs” tab shows No logs in this time range

• The UI start command can revert because it is overridden by railway.toml

What we verified / changed

• We ensured the start command is deterministic and should keep the container alive:

  • railway.toml includes:

    • deploy.startCommand = "bash railway_start.sh"

    • deploy.healthcheckPath = "/api/health"

    • deploy.healthcheckTimeout = 120

• We modified railway_start.sh so that:

  • PID1 runs a lightweight health server responding 200 on /api/health (and binds to $PORT, default 8080)

  • The daemon runs in the background with a restart loop (exponential backoff)

  • The container should not exit even if the daemon crashes

• Even with this, deployment still fails at “Creating containers” and there are no runtime logs visible.

Suspicion / question

Historically this same repo/service deployed successfully (healthcheck passed and the API was reachable). Now the deployment fails before any runtime output is shown, which suggests:

• The container might not be starting at all (image pull / container creation issue), or

• The platform might be failing before stdout/stderr is attached, or

• Something changed in the deployment pipeline that prevents launching the container despite a successful build.

What I need from you

Could you please check, on the Railway side, the exact reason why the container cannot be created/launched for this deployment? Specifically:

1. Is there an internal error during container creation, image pull, network attach, or healthcheck init that is not exposed in logs?

2. Can you confirm whether the built image is being pushed/pulled correctly for this service?

3. If possible, can you provide the internal failure reason / event message for the “Deploy / Creating containers” failure (e.g., image not found, permission, quota/limits, runtime crash before logs attach, etc.)?

I can provide the service ID, deployment ID, and the relevant railway.toml / start script if needed.

Thanks for your help.