Troubleshooting
Common symptoms, their causes, and how to fix them.
Container fails to start
Check docker logs <container-id> to see the actual error. Verify the manifest’s spec.image exists locally or is pullable from your configured registries. If the image doesn’t exist, update spec.image and re-run shrine apply.
Traefik does not route to my app
Verify the gateway plugin is active, the app’s spec.routing.domain matches the incoming request’s Host header, and the per-team network is connected to the platform bridge. Run docker network ls and docker network inspect <network> to confirm connectivity.
State drift after manual docker rm
Shrine treats Docker as the source of truth and reconciles on every apply. If you manually remove a container, running shrine apply <dir> or shrine deploy will redeploy it. This is expected behavior.
shrine apply reports validation errors
Fix every error listed in the multi-error report. Shrine surfaces all issues in a single pass, not one at a time. Re-run after fixing to ensure all errors are resolved.
--dry-run shows the right plan but apply fails
The plan succeeded but the actual Docker operation failed. Check Docker’s stderr for errors like missing images, port conflicts, or network issues. Verify the image exists, ports are available, and your registries are configured correctly.
See also
shrine apply— Deploy manifestsshrine status— Check running workloads