How does Quarkus support Kubernetes-native deployments?

Quarkus treats Kubernetes as a first-class deployment target through the quarkus-kubernetes extension (and related extensions for OpenShift, Knative, and Docker), which generates deployment manifests directly from the application's own configuration rather than requiring them to be hand-written and kept in sync separately.

./mvnw package -Dquarkus.kubernetes.deploy=true

Adding the extension and building the project automatically produces a kubernetes.yml (and equivalents for other targets) under target/kubernetes, including a Deployment and Service already wired to the application's configured port and container image; setting quarkus.kubernetes.deploy=true at build time goes a step further and applies those generated manifests directly to the currently configured cluster.

Beyond manifest generation, Quarkus also integrates health checks (via SmallRye Health, exposing /q/health/live and /q/health/ready endpoints that Kubernetes probes can use directly) and container-image building extensions (Jib, Docker, or Buildpacks) into the same build process, so a single Maven or Gradle build can take source code all the way to a running pod on a cluster, with the generated manifests already referencing the correct readiness and liveness probe paths without manual wiring.