CKA Cheat Sheet.
The night-before summary, built like the exam.

CKA Cheat Sheet: Certified Kubernetes Administrator

A condensed reference for exam-week review. The CKA tests hands-on cluster administration skills under time pressure, not multiple-choice recall.

Exam Logistics at a Glance

Content Domains (Quick List)

• Troubleshooting — 30% (heaviest weight; cluster, node, application, and networking failures)
• Cluster Architecture, Installation & Configuration — 25%
• Services & Networking — 20%
• Workloads & Scheduling — 15%
• Storage — 10%

Five domains total. Troubleshooting alone is nearly a third of the exam, so allocate practice time accordingly rather than splitting evenly.

Key Terms and Concepts to Memorize

Cluster Architecture

• Control plane components: kube-apiserver, etcd, kube-scheduler, kube-controller-manager, cloud-controller-manager
• Node components: kubelet, kube-proxy, container runtime (containerd)
• kubeadm workflow: kubeadm init, kubeadm join, certificate rotation, and static pod manifests under /etc/kubernetes/manifests
• etcd backup and restore using etcdctl snapshot save / snapshot restore
• RBAC objects: Role, ClusterRole, RoleBinding, ClusterRoleBinding

Workloads & Scheduling

• Deployments, ReplicaSets, DaemonSets, StatefulSets, and their update strategies
• Scheduling controls: nodeSelector, node affinity/anti-affinity, taints and tolerations
• Resource requests/limits and how the scheduler uses them
• ConfigMaps and Secrets injected as env vars or mounted volumes

Services & Networking

• Service types: ClusterIP, NodePort, LoadBalancer, ExternalName
• Ingress resources and IngressClass vs. legacy Ingress controllers
• NetworkPolicy objects for pod-to-pod traffic control
• CoreDNS as the in-cluster DNS provider
• CNI plugin basics — pods need a CNI installed before they leave Pending

Storage

• PersistentVolume (PV) vs PersistentVolumeClaim (PVC) binding lifecycle
• StorageClass and dynamic provisioning
• Access modes: ReadWriteOnce, ReadOnlyMany, ReadWriteMany
• Volume reclaim policies: Retain, Delete, Recycle

Troubleshooting

• kubectl describe, kubectl logs, and kubectl get events as first response tools
• Diagnosing CrashLoopBackOff, ImagePullBackOff, and Pending pod states
• Checking kubelet status and syslog/journalctl on nodes
• Diagnosing control plane failures via static pod logs and container runtime status

Common Gotchas and Traps

• Every task specifies a target cluster context — running a command against the wrong context is a top point-loss cause; always run kubectl config use-context first
• Time management matters more than perfect answers: skip a stuck task, flag it, and return later rather than burning minutes
• Imperative kubectl run/kubectl create commands with --dry-run=client -o yaml are faster than hand-writing YAML under time pressure
• Remember to save and exit editors correctly (vim gotchas trip up candidates unfamiliar with the exam terminal)
• Namespace defaults silently change results — many tasks require an explicit -n <namespace> flag
• etcdctl commands require explicit TLS cert/key/CA flags in most exam environments — missing them causes connection failures that look like etcd is down
• The exam allows only the official Kubernetes documentation as a reference; practice navigating docs quickly instead of memorizing every flag

Night-Before Checklist

• Confirm ID matches registration name exactly and test the PSI Secure Browser system check
• Clear your desk and test environment per proctoring rules to avoid a pre-exam delay
• Practice the vim/nano workflow you'll use for quick YAML edits
• Rehearse kubectl config use-context and namespace-flag habits until they're automatic
• Review etcd backup/restore commands and static pod manifest locations one more time
• Skim the weighted domain list above and mentally re-rank study priority toward Troubleshooting and Cluster Architecture
• Get sleep — a 2-hour performance exam rewards steady focus over last-minute cramming