Practice Test For CNCF Kubernetes Exam (2026 Edition)

It is Monday morning, and the cluster is misbehaving. A deployment that worked on Friday now shows pods stuck in Pending, one node reports NotReady, a service that should be reachable inside the cluster returns no response, and a persistent volume claim waits for storage that never binds. That Monday morning scene is the starting point for this practice set. It starts with symptoms, partial output, and a need to narrow the problem step by step. This set of Kubernetes practice tests contains 560 CNCF practice questions built around that kind of work, so you can rehearse the decisions a cluster administrator makes under pressure, in a calm, self-paced study setting, before you sit a real exam. It offers CKA-style practice for Kubernetes administrators, DevOps engineers, and platform engineers preparing for a CNCF Kubernetes exam, and cluster administration exam practice that covers both building a cluster and fixing one.

Think like a cluster troubleshooter

A strong cluster troubleshooter does not guess. The troubleshooter checks what is running, what changed, and what the cluster itself is reporting, then follows the chain from symptom to cause. A pod in CrashLoopBackOff may point to a bad image, a missing ConfigMap, a failed probe, a resource limit, or a security context that blocks the process from starting. A service with no endpoints may point to a label mismatch or pods that never became ready. A node problem may point to the kubelet, the container runtime, disk pressure, or a network route that disappeared.

Many questions present a short scenario, a piece of output, or a configuration snippet, and ask you to choose the most useful next check, the most likely cause, or the correct fix. Other questions test the building blocks behind the symptom, so you understand why the fix works. Over 560 questions, you build a repeatable way of working: describe the symptom, locate the layer where it lives, test one idea at a time, and confirm the result after the change.

The six practice areas

The 560 questions are divided across six practice areas:

1. Cluster Architecture, Installation and Configuration: 102 questions

2. Workloads and Scheduling: 89 questions

3. Services and Networking: 89 questions

4. Storage: 99 questions

5. Troubleshooting Control Plane, Node, and Networking, Part 1: 92 questions

6. Troubleshooting Workloads, Storage, and Security Contexts, Part 2: 89 questions

Cluster Architecture, Installation and Configuration covers the kube-apiserver, etcd, kube-scheduler, kube-controller-manager, and cloud-controller-manager, plus the kubelet, kube-proxy, and the container runtime. Questions cover kubeadm installation, cluster upgrades, kubeconfig files, role-based access control, service accounts, namespaces, labels, resource quotas, etcd backup and restore, certificate handling, and high availability, with attention to what can go wrong during an upgrade or a restore.

Workloads and Scheduling covers pods, including multi-container pods, init containers, probes, and lifecycle behaviour, and controllers such as Deployments, ReplicaSets, DaemonSets, StatefulSets, Jobs, and CronJobs. Scheduling questions cover node selectors, affinity and anti-affinity, taints and tolerations, resource requests and limits, rolling updates, rollbacks, scaling, and the reasons a pod stays Pending when no node can accept it.

Services and Networking covers ClusterIP, NodePort, and LoadBalancer services, service discovery through DNS and CoreDNS, endpoints and EndpointSlices, ingress resources and ingress controllers, and the path a request follows from a client to a pod. Questions also cover container network interface plugins, pod networking, network policies, and faults such as a service with no endpoints, a DNS name that does not resolve, or a policy that blocks expected traffic.

Storage covers persistent volumes and persistent volume claims, storage classes, provisioners, access modes, reclaim policies, and volume expansion. Questions follow the chain from a claim to a class to a provisioner to a volume, and cover the reasons a claim stays Pending, a pod stays in ContainerCreating, or data does not survive a pod restart. Stateful workloads, volume mounts, ConfigMaps, and Secrets are covered alongside the storage concepts that support them.

The two troubleshooting areas make up 181 questions. Part 1 focuses on the control plane, nodes, and networking: a failed API server, a scheduler that stops placing pods, an etcd problem, a kubelet that stops reporting, a node under disk or memory pressure, and a network path that breaks between nodes or between pods. Part 2 focuses on workloads, storage, and security contexts: image pull failures, probe failures, resource exhaustion, storage binding and mount failures, permission errors, and security context settings such as runAsUser, fsGroup, capabilities, seccomp profiles, and read-only file systems that prevent a container from starting.

Learning from each explanation

Every question includes a detailed explanation, and that explanation is where much of the learning happens. A correct answer is explained in plain terms, with the concept behind it and the reason the other options do not fit. An incorrect option is not simply marked wrong. The explanation shows the trap in it, whether it is a plausible command with the wrong target, a setting that belongs to a different resource, a step that comes too early, or a fix that treats a symptom while leaving the cause in place.

Read the full explanation even when you answered correctly, because a correct guess can still hide a gap in understanding. Note the topics where you hesitate, repeat those sections after a few days, and compare your second attempt with your first. If you miss several questions on etcd recovery, scheduling, ingress routing, or security contexts, that pattern tells you where to spend your next study session. The goal is not to memorise 560 answers. The goal is to build the mental model that lets you answer a new question on the same topic, with different output and a different fault. This set supports self-paced study with no fixed schedule, so you can focus on your weakest area first or mix areas to keep your review balanced.

Who this practice set serves

This practice set serves Kubernetes administrators who run clusters in production or staging and want to check their readiness for a CNCF Kubernetes exam. It serves DevOps engineers who deploy, scale, and support container workloads, and platform engineers who build shared clusters and need to handle access control, scheduling, quotas, networking, and storage requests with confidence. It also serves system administrators and cloud engineers moving into Kubernetes work who want practice questions that connect configuration choices to running behaviour. Prior hands-on experience helps, because many questions assume you have seen kubectl output, edited a manifest, or inspected a pod at least once.

Independent practice disclosure

This practice set is an independent educational resource. It is not affiliated with, endorsed by, or sponsored by the Cloud Native Computing Foundation (CNCF). Kubernetes and CNCF names are used only to identify the subject area of this practice material and for educational purposes. No practice score in this set promises or guarantees a pass, a particular score, a job, or a salary outcome. Use this set to find gaps, strengthen weak areas, and build confidence through repeated, careful practice, combined with hands-on cluster work and official documentation.