CKA Certified Kubernetes Administrator: Exam Format and Study Plan

The Certified Kubernetes Administrator (CKA) exam is a two-hour, online-proctored, performance-based test in which you solve 15-20 tasks on live Kubernetes clusters from a command line. You need 66% or higher to pass. The most reliable way to prepare is to build and repair clusters yourself, practise with kubectl until the common commands are automatic, and learn where things live in the Kubernetes documentation, because you may use it during the exam.

This guide is for administrators, DevOps engineers and developers who already use containers and want a structured plan. By the end you will know the exam format and rules, the weight of each domain, which skills to practise for each, and how to spread the work over eight weeks. Figures here were checked against the Linux Foundation pages listed in References on 28 October 2026; the exam page and curriculum change with each Kubernetes release, so confirm them before you book.

CKA exam format at a glance

The price is quoted from the Linux Foundation certification page and can change, so check it there. Registrations for CKA include access to the killer.sh exam simulator (two attempts, each with 36 hours of access from activation), but the "single" registrations do not.

Exam domains and what to practise

The Linux Foundation lists five domains with these weights.

The detailed competencies sit in the CNCF curriculum PDF whose version matches the Kubernetes version (the curriculum repository names it CKA_Curriculum_v1.37.pdf). Download the one matching your exam date and treat it as your checklist. The sections below describe how to turn each domain into practice; they are study suggestions, not a restatement of the curriculum.

Troubleshooting (30%)

This is the biggest slice, so it deserves the most lab time. Practise diagnosing problems rather than only creating objects:

• Pods stuck in Pending, CrashLoopBackOff or ImagePullBackOff, using kubectl describe and kubectl logs (including --previous).
• A worker node showing NotReady: check the kubelet with systemctl status kubelet and journalctl -xeu kubelet, both of which the kubeadm upgrade documentation uses for kubelet health.
• Control plane components that fail to start, by reading their static Pod manifests and logs.
• Services with no endpoints, caused by selector or port mismatches.

A good exercise is to ask a friend, or script yourself, to introduce a fault into a practice cluster and then diagnose it with a time limit.

Cluster Architecture, Installation and Configuration (25%)

Expect cluster lifecycle work. Practise these on a cluster you built with kubeadm:

• Installing a cluster with kubeadm and joining worker nodes.
• Upgrading a cluster. The Kubernetes documentation describes the order: upgrade kubeadm, run kubeadm upgrade plan and kubeadm upgrade apply on the first control plane node, run kubeadm upgrade node on additional control plane nodes and workers, drain each node, upgrade kubelet and kubectl, restart the kubelet and uncordon the node.
• Backing up and restoring etcd. The Kubernetes documentation covers etcdctl snapshot save and restoring with etcdutl; read the current page rather than relying on memory, because the tooling has changed over the years.
• Role-based access control: Roles, ClusterRoles, bindings and kubectl auth can-i.
• Working with Helm and Kustomize where the curriculum lists them.

Services and Networking (20%)

Practise exposing workloads and controlling traffic: ClusterIP, NodePort and LoadBalancer Services, NetworkPolicies, CoreDNS behaviour, and the Ingress and Gateway API resources. The Gateway API documentation is explicitly allowed for CKA, so learn where its task pages are.

Workloads and Scheduling (15%)

Deployments, rolling updates and rollbacks, ConfigMaps and Secrets, resource requests and limits, autoscaling, and scheduling controls such as node selectors, affinity, taints and tolerations. Practise creating manifests quickly with imperative commands and editing the generated YAML.

Storage (10%)

PersistentVolumes, PersistentVolumeClaims, StorageClasses, access modes, reclaim policies and mounting volumes into Pods. This is the smallest domain, but tasks are often quick marks if you have practised the manifests.

Rules of the exam environment

The Linux Foundation's exam tips page describes the environment. Read it fully before exam day; the points that surprise people are these:

• Each task names a host. You SSH to it, work there, then run exit to return to the base system. Nested SSH is not supported.
• On the SSH hosts, kubectl already has the alias k and Bash autocompletion, and yq, curl, wget and man are available. The base system does not have these tools, so do the work on the designated hosts.
• Use sudo -i for elevated privileges, or prefix individual commands with sudo.
• Do not reboot the base node; it will not restart your exam environment.
• The terminal is Linux, so copy and paste with Ctrl+Shift+C and Ctrl+Shift+V, or the right-click menu. Use Ctrl+Alt+W instead of Ctrl+W, which would close the browser tab. In vi, the Insert key is disabled, so press i to insert and Esc to leave insert mode.
• You use the PSI Secure Browser, a single monitor, and no virtual machine or other applications. Run the PSI system check beforehand, and use a reliable, preferably wired, connection.

Allowed documentation is limited to kubernetes.io/docs, the Kubernetes blog, Helm documentation, task-specific pages linked in the exam's quick reference box and, for CKA only, the Gateway API documentation. Opening external search results is not allowed. Use the search box on kubernetes.io/docs, and spend practice time finding pages such as the kubectl quick reference, the kubeadm upgrade guide and the etcd operations guide so that you can reach them in seconds.

Shell setup that saves time

Because the alias k and completion exist on exam hosts, set up the same thing in your lab. The kubectl quick reference gives these Bash lines:

source <(kubectl completion bash)
echo "source <(kubectl completion bash)" >> ~/.bashrc
alias k=kubectl
complete -o default -F __start_kubectl k

Beyond that, rely on imperative commands to generate manifests, then edit them. For example:

kubectl create deployment web --image=nginx --dry-run=client -o yaml > web.yaml

The --dry-run=client -o yaml pair prints the object without creating it. Practise this for Pods, Deployments, Services, Jobs and CronJobs until you do not need to look up the syntax. Use kubectl explain (for example kubectl explain pod.spec.containers) to check field names offline within the cluster.

An eight-week study plan

Adapt the pacing to your experience. The plan assumes about six to eight hours a week.

Keep a running list of commands you had to look up. In weeks 7 and 8, turn that list into short drills.

Practice environments

A multi-node cluster you build yourself teaches more than a single-node tool, because the exam includes node-level work. Two or three small virtual machines running a supported Linux distribution are enough. Use the Kubernetes version that matches the exam where you can, since upgrade steps and defaults change between releases. The killer.sh simulator included with a standard CKA registration is useful for timing and for a realistic interface, and it is worth saving until you have completed the earlier weeks.

Exam day checklist

• Confirm the Kubernetes version and curriculum for your exam date.
• Run the PSI system check, clear your desk and use a single monitor.
• Plan your time: with 15-20 tasks in 120 minutes, you average six to eight minutes a task. Skip a task that stalls and return to it.
• Read each task for the host name and the namespace before you type anything.
• Verify your work after each task, for example with kubectl get and kubectl describe.
• Remember to exit back to base between tasks.

Where to go next

For another certification study plan, see AWS Solutions Architect Associate Study Guide: Domains and Labs.

References

• Linux Foundation, Certified Kubernetes Administrator (CKA): https://training.linuxfoundation.org/certification/certified-kubernetes-administrator-cka/
• Linux Foundation, CKA and CKAD exam tips: https://docs.linuxfoundation.org/tc-docs/certification/tips-cka-and-ckad
• Linux Foundation, CKA, CKAD and CKS FAQ: https://docs.linuxfoundation.org/tc-docs/certification/faq-cka-ckad-cks
• Linux Foundation, certification resources allowed: https://docs.linuxfoundation.org/tc-docs/certification/certification-resources-allowed
• CNCF curriculum repository: https://github.com/cncf/curriculum
• Kubernetes documentation, kubectl quick reference: https://kubernetes.io/docs/reference/kubectl/quick-reference/
• Kubernetes documentation, upgrading kubeadm clusters: https://kubernetes.io/docs/tasks/administer-cluster/kubeadm/kubeadm-upgrade/
• Kubernetes documentation, operating etcd clusters: https://kubernetes.io/docs/tasks/administer-cluster/configure-upgrade-etcd/