Test your ability to design, deploy, configure, observe, and network Kubernetes applications through 20 independent CKAD-style challenges. Each task gives you a prepared application scenario and acceptance criteria, then asks you to choose the commands or manifests that produce the required live state and pass automated checks.
The exam covers image delivery, scheduled and multi-container workloads, release strategies, health and debugging, application identity and security, configuration, storage, Services, Ingress, and NetworkPolicy. It is a readiness check for learners who already know Kubernetes application fundamentals, not a guided introduction or a substitute for the official CKAD exam.
What You Will Learn
- Build a local application image, make it available to Kubernetes, and deploy it without publishing to an external registry
- Create CronJobs, init and sidecar containers, and workloads that combine
emptyDirand PVC-backed storage - Perform constrained rolling updates, run a stable/canary release behind one Service, and deploy staged Helm and Kustomize packages
- Configure startup, readiness, and liveness probes and repair CrashLooping applications from status, Events, and logs
- Migrate deprecated CronJob and PodDisruptionBudget API versions while preserving working application behavior
- Inject ConfigMaps and Secrets, enforce ResourceQuota and container resources, and assign a least-privilege ServiceAccount
- Apply Restricted-compatible security contexts and create a schema-valid custom resource from a staged CRD
- Expose Deployments with Services, route host and path traffic through Ingress, restrict ingress with NetworkPolicy, and repair EndpointSlices
Who This Course Is For
This practice exam is for application developers, platform engineers, and CKAD candidates who want to test whether they can turn requirements into working Kubernetes resources without step-by-step commands. It is most useful after guided CKAD study or comparable hands-on experience.
Prerequisites: You should be comfortable with Linux CLI work, Kubernetes YAML, kubectl, Pods, Deployments, Jobs and CronJobs, ConfigMaps, Secrets, volumes, Services, labels, selectors, and basic troubleshooting. Familiarity with container image builds, probes, RBAC, resource controls, Helm, Kustomize, Ingress, NetworkPolicy, and security contexts is recommended.
Learning environment: The course contains 20 separate challenge labs in an Ubuntu 22.04 CKAD terminal environment with the single-node labex-v135 Kubernetes profile. Seventeen challenges are labeled Intermediate and three Advanced. Each starts in its own prepared VM, provides a scenario, scope, acceptance criteria, and limited hints, and uses two automated checks to validate the required application and cluster state.
Frequently Asked Questions
Is this an official CKAD exam or official question set?
No. This is an independent LabEx learning resource and is not affiliated with, endorsed by, or sponsored by The Linux Foundation or CNCF. Its coverage is informed by public CKAD domains, but the tasks and their distribution do not represent the real exam’s questions or exact task mix.
How are time and scoring handled?
The recommended full attempt is 120 minutes. Each of the 20 challenges is worth 5 simulated points for a total of 100, and 66 points is the LabEx simulated practice threshold. These values support self-assessment only and neither predict nor constitute an official CKAD result.
How is this different from CKA practice?
The tasks stay centered on applications: workload design, release configuration, probes, logs, configuration, security contexts, application RBAC, Services, and traffic policy. They do not assess CKA-style node maintenance, kubeadm lifecycle, control-plane repair, or etcd backup.
Do I need an external image registry or remote chart repository?
No. The image challenge builds locally and makes the image available to the prepared cluster without a public registry. The Helm chart and Kustomize base are staged in the workspace. Other workloads use local, staged, imported, or cached images so the tasks do not depend on publishing your own artifacts externally.
