Effective CKA exam preparation starts with one decision: allocate your study hours to the domain weights instead of treating every topic as equal. The five CKA domains are weighted Troubleshooting at 30%, Cluster Architecture, Installation and Configuration at 25%, Services and Networking at 20%, Workloads and Scheduling at 15%, and Storage at 10%, according to the official Linux Foundation exam page. Because the weights are uneven, the top two domains alone control 55% of the score, and a candidate who masters troubleshooting under time pressure can absorb a weaker Storage performance. The plan below turns those weights into a six-week schedule with format rules, timed drills and simulator sessions that match how the exam is actually scored.
What the CKA Exam Tests
The CKA is an online, proctored, performance-based exam that gives candidates 2 hours of command-line tasks on an environment based on Kubernetes v1.35. There are no multiple-choice questions: every point comes from changing or repairing live cluster state with kubectl and Linux tools, which is why terminal practice beats passive video courses. The exam environment tracks the most recent Kubernetes minor release within approximately 4 to 8 weeks of the upstream release date, so confirm the current version before buying lab bundles that may already be outdated.
Scoring is automatic and results arrive by email within 24 hours. A score of 66% or above must be earned to pass the CKA exam. The listing bundles two exam attempts at the $445 exam-only price, so a first attempt can double as a diagnostic, and registration stays valid for 12 months, which gives you room to schedule around Kubernetes release cycles.
Domain Weights and Study Time
The official domain list is the backbone of any study plan. This table pairs each weight with the tasks that appear in performance-based questions, so you can convert percentages into practice hours.
| Domain | Weight | Priority practice |
|---|---|---|
| Troubleshooting | 30% | Failed pods, NotReady nodes, control-plane component logs, service reachability, resource monitoring |
| Cluster Architecture, Installation & Configuration | 25% | kubeadm init and join, RBAC, cluster lifecycle, HA control plane, Helm, Kustomize, CRDs and operators, CNI/CSI/CRI extension interfaces |
| Services & Networking | 20% | Pod connectivity, NetworkPolicy, ClusterIP, NodePort, LoadBalancer, Ingress, Gateway API, CoreDNS |
| Workloads & Scheduling | 15% | Deployments with rolling updates and rollbacks, ConfigMaps, Secrets, autoscaling, pod admission and scheduling |
| Storage | 10% | Storage classes, persistent volumes and claims, volume access modes, reclaim policies, dynamic provisioning |
One change deserves special attention. The Services and Networking domain now includes the competency to use the Gateway API to manage Ingress traffic, alongside Ingress controllers and CoreDNS. CNCF flagged this addition when it refreshed the CKA competencies, and older study guides that stop at classic Ingress resources leave gaps in a domain worth a fifth of the score. If your materials predate the Gateway API competency, replace them.
Six-Week Study Plan
- Weeks 1 to 2, architecture and install: build and break kubeadm clusters repeatedly, configure RBAC bindings, and install components with Helm. For splitting hours per domain, follow the CKA prep plan with study time by domain weight.
- Weeks 2 to 3, workloads: deployments, rolling updates and rollbacks, ConfigMaps and Secrets, autoscaling, resource limits and node affinity.
- Weeks 3 to 4, services and networking: NetworkPolicy enforcement, all three service types, endpoints, CoreDNS queries and Gateway API routing until each takes under ten minutes.
- Weeks 4 to 5, storage: storage classes and persistent volume claims, access modes, reclaim policies and dynamic provisioning.
- Weeks 5 to 6, troubleshooting on the clock: diagnose broken clusters, nodes and components against a timer, and use the format rules and timed drills guide to rehearse two-hour pacing.
- Final week, simulator: run both Killer.sh sessions at full length and log every domain that leaks points.
Simulator and Exam-Day Tactics
Each CKA registration includes two Killer.sh simulation attempts, with 36 hours of access per attempt and 17 questions per session. Treat each simulation as a dress rehearsal: same two-hour limit, same terminal-only workflow, no notes outside the allowed documentation. Graded results show exactly which domains lose points, which is more actionable than any third-party question bank.
On exam day, small mechanics save minutes:
- Run the PSI system check on the exact machine and network you will use; dual monitors are not supported.
- Practice with the k alias and bash completion configured the way the exam environment provides them.
- Read each task to the end, work on the designated cluster, and verify your work with kubectl before moving on.
- Flag hard tasks and return later; unanswered tasks earn nothing.
- Use only the allowed documentation, such as the Kubernetes docs, never external search results.
The trade-off is simple: breadth-only study burns hours on the lightest domain while the heaviest one stays weak. Anchor the calendar to the weights, rehearse in the terminal under the clock, and treat the bundled retake as a safety net rather than the plan.
Sources
- Linux Foundation – Certified Kubernetes Administrator (CKA) exam page
- Linux Foundation – CKA, CKAD and CKS certification FAQ
- CNCF Blog – Kubernetes is evolving, the CKA exam too