8 days ago
Hi Railway Support,
I have a production service that is failing during provisioning after the image has successfully built and been pushed.
Project: energetic-courtesy
Environment: production
Service: stark
Original deployment:
dbfa083c-ee64-4a18-9737-0e6aa193a59d
This deployment remained DEPLOYING / CREATED for more than 8 hours and eventually changed to FAILED.
Same-artifact redeployment:
03b13594-ed55-4b08-a9d8-a94823216810
This deployment also remained at DEPLOYING / CREATED. The service reports stopped=true.
In both cases:
- Build completed successfully.
- Image export and push completed successfully.
- No runtime/startup logs were generated.
- There is no evidence the container was scheduled.
- The application process never started.
- The production volume /app/data reports READY.
- CPU, memory, network, and volume usage are well below limits.
- Other services in the same production environment are running successfully, including Maker API, PostgreSQL, Redis, worker, and scheduler.
- There is no application crash evidence because execution never reached application startup.
The previous successful STARK deployment is still retained by the control plane.
Could you please check the Railway-side provisioning/scheduling state for these deployments and determine why a container is not being scheduled after successful image creation?
Please let me know if you need any additional deployment/project identifiers or logs.
2 Replies
8 days ago
This thread has been opened as a bounty so the community can help solve it.
Status changed to Open Railway • 8 days ago
8 days ago
Can you share your Deploy Logs? If there's no application startup logs, Try set a Custom Start Command in the Service Settings
mayori
Can you share your Deploy Logs? If there's no application startup logs, Try set a Custom Start Command in the Service Settings
8 days ago
Thanks. The Deploy Logs contain no application startup/runtime logs.
The build itself completes successfully, including image export and push. After that, the deployment remains at DEPLOYING, with the instance at CREATED.
Original deployment: dbfa083c-ee64-4a18-9737-0e6aa193a59d — eventually changed to FAILED.
Exact same-artifact redeployment: 03b13594-ed55-4b08-a9d8-a94823216810 — remains DEPLOYING / CREATED.
In both cases there are no startup/runtime logs at all, so it appears the application process never begins.
Other services in the same environment are running normally, and the volume reports READY.
Before I change the Custom Start Command, could you confirm whether Railway actually scheduled/started a container for either deployment? If the container never started, I want to rule out a provisioning/scheduling issue first.
I can also share the build/deploy logs if needed.