# Runbook

## Site shows fallback content

1. Check `docker compose ps` and the backend health status.
2. Request `http://localhost:8000/api/site/home/` directly.
3. Inspect `docker compose logs backend postgres`.
4. Verify `PUBLIC_CMS_API_URL` resolves from the frontend container.
5. Restart the frontend after recovery so its static page is rebuilt from CMS data.

## Wagtail does not start

1. Confirm PostgreSQL is healthy.
2. Check `DATABASE_URL` and allowed hosts without printing secret values in shared logs.
3. Run `docker compose exec backend python manage.py migrate`.
4. Review backend logs for a specific migration or configuration error.

## Images are missing

1. Verify the `media_data` volume is mounted at `/app/media`.
2. Check `WAGTAILADMIN_BASE_URL` uses a URL reachable by the browser.
3. Confirm the image still exists in Wagtail and has not been removed from the volume.

## A service is unhealthy

1. Run `docker compose ps` and inspect `docker compose logs --tail=200 SERVICE`.
2. For PostgreSQL, check volume capacity and database/user configuration.
3. For the backend, check `/health/`, migrations, and database connectivity.
4. For the frontend, request port `4321`, then check the API separately because
   fallback content can mask a CMS outage.

## Rollback

Revert the application commit and rebuild containers. Preserve database/media volumes.
Before reversing migrations or deleting volumes, make and validate coordinated backups.