Instructions and covered skills
Complete the test in a quiet setting and keep your attention on each scenario. Read configuration details carefully before selecting a response. Consider how Kubernetes controllers and kubelets act over time rather than focusing on one field in isolation. Do not rely on assumptions about application behavior that the scenario does not state. Turn off notifications and avoid switching between unrelated tasks. Review your choices for rollout consequences, probe timing, and availability effects before submitting.
Key Areas
This test focuses on operating Kubernetes Deployments through controlled application changes. It covers the relationship between desired replicas, available replicas, updated replicas, and rollout progress. Candidates should understand how the RollingUpdate strategy uses maxSurge and maxUnavailable to control replacement capacity and service continuity. They should also recognize when a rollout becomes stalled, how progressDeadlineSeconds is evaluated, and how pausing a Deployment affects reconciliation.
Health management is another central area. The test examines readiness, liveness, and startup probes, including the distinct actions triggered by each result. Readiness controls whether a Pod can receive Service traffic, while liveness can restart a running container and startup delays liveness and readiness evaluation for applications that require more initialization time. Candidates should be able to reason about probe paths, ports, timing thresholds, failure thresholds, and realistic application states.
Recovery and observability are also included. Candidates should know how to inspect rollout status, review Deployment revision history, undo a revision, and interpret why a rollout has not reached its expected state. Questions emphasize configuration consequences: replica calculations, availability guarantees, endpoint membership, and the behavior of Pods during a release.
Recommended Preparation
Review the Kubernetes Deployment specification, especially strategy, minReadySeconds, revisionHistoryLimit, progressDeadlineSeconds, and the status conditions reported by a Deployment. Practice using kubectl rollout status, kubectl rollout history, kubectl rollout pause, kubectl rollout resume, and kubectl rollout undo in a nonproduction cluster.
Study the container probe fields initialDelaySeconds, periodSeconds, timeoutSeconds, successThreshold, and failureThreshold. Work through release scenarios involving slow startup, failed dependency checks, unavailable replicas, and image corrections. Focus on selecting probe endpoints that reflect whether an application can serve traffic or should be restarted, rather than merely whether a process exists.