A readiness check reports whether an application can receive traffic now. Kubernetes uses the result through a readiness probe: when the check fails, Kubernetes stops sending traffic to the pod until the condition is resolved. This differs from a liveness check, which signals an unrecoverable problem and can trigger a restart. section 10.2
MicroProfile Health exposes health information through HTTP endpoints. In Quarkus, SmallRye Health implements the feature, and the readiness results are available at the health endpoints, including `/q/health/ready`; `/q/health` returns all health checks grouped by type. section 10.2 section 10.2.1
A readiness check is useful when a service should remain running but temporarily stop receiving requests:
- Quarkus automatically adds a readiness check for each JDBC data source and verifies that the database connection works. section 10.2.1
- A failed database check reports that the application cannot process requests, allowing traffic to be withheld without treating the condition as a restart-worthy failure. section 10.2 section 10.2.1
- Readiness is evaluated periodically by monitoring tools or Kubernetes at configured probe intervals. section 10.2
Checking the Inventory service
Add the SmallRye Health extension with `quarkus extension add smallrye-health`, then run the Inventory service in Dev mode. Quarkus starts the MySQL instance through Dev Services and exposes a database readiness check. The Dev UI can show its status, while `http://localhost:8083/q/health` returns JSON. With the database available, the check is `UP`; after stopping the MySQL container, the overall status and database check become `DOWN`, including the connection error. section 10.2.1
