Troubleshooting

Diagnose from durable intent outward

Start with the desired revision and operation receipt, then inspect target deployment state, runtime acknowledgement, and finally a real network probe. This distinguishes control failure from runtime failure.

Start with the durable operation and desired revision, then move outward:

1. Check /api/ready and administrator component health.
2. Inspect the relevant operation, deployment, task, or certificate order.
3. Compare desired and acknowledged revisions.
4. Check the responsible Horizon lane and failed jobs.
5. Check the private runtime service and its last-valid state.
6. Reconcile only the smallest affected scope.

Choose a symptom:

• DNS
• Edge and origin
• TLS and cache
• Telemetry

Do not repair desired state in PowerAdmin or by editing generated edge files. Those are derived and will be overwritten.

Docker reports storage/logs: file exists

This message during make dev-up comes from Docker attempting concurrent copy-up into the shared Laravel storage volume, not from Laravel logging. The current Compose contract mounts core-storage with volume.nocopy: true, and the application entrypoint creates the required directories.

1. Confirm the checkout contains the current compose.dev.yml.
2. Run docker compose -f compose.dev.yml config and verify the /app/storage volume reports nocopy: true.
3. Run make dev-up again.

Preserve durable development state

Do not use docker compose down -v or remove core-storage as a workaround. If the current contract still fails, capture docker compose ps, the daemon error, and docker volume inspect cdnfoundry-dev_core-storage before changing anything.