Beginner

Exam Tips

Your final preparation guide — kubectl cheat sheet, exam day strategy, time management tips, and frequently asked questions to ensure you pass the CKA exam.

kubectl Cheat Sheet

Memorize these commands. Speed is critical on the CKA — you have about 6 minutes per task.

Essential Shortcuts

# Set up aliases (do this first on exam day)
alias k=kubectl
export do="--dry-run=client -o yaml"

# Generate YAML templates quickly
k run nginx --image=nginx $do > pod.yaml
k create deploy myapp --image=nginx --replicas=3 $do > deploy.yaml
k create job myjob --image=busybox $do > job.yaml
k create cronjob mycron --image=busybox --schedule="*/5 * * * *" $do > cron.yaml
k create service clusterip mysvc --tcp=80:8080 $do > svc.yaml

# Quick resource inspection
k get pods -A                    # All namespaces
k get pods -o wide               # Show node and IP
k describe pod <name>            # Detailed info
k logs <pod> -c <container>      # Container logs
k logs <pod> --previous          # Previous container logs
k exec -it <pod> -- /bin/sh      # Shell into Pod

# Context and namespace switching
k config use-context <context>   # Switch cluster
k config set-context --current --namespace=<ns>  # Set default namespace

GPU and ML Specific Commands

# Check GPU resources on nodes
k describe node <name> | grep -A 10 "Allocated resources"
k get nodes -o json | jq '.items[].status.allocatable'

# Check GPU Pods
k get pods -l nvidia.com/gpu --all-namespaces
k top pods --sort-by=memory -n ml-training

# Debug scheduling issues
k describe pod <pending-pod>    # Check Events section
k get events --sort-by='.lastTimestamp' -n <ns>

Exam Day Strategy

Before the Exam

  • Environment check: 30 minutes before, run the system check tool. Close all apps. Clear your desk.
  • Bookmark the docs: You can access kubernetes.io/docs during the exam. Bookmark key pages in advance: PVC, Ingress, NetworkPolicy, RBAC, etcd backup.
  • Practice the terminal: The exam uses a browser-based terminal. Practice copy-paste shortcuts (Ctrl+Shift+C/V in Linux terminal).

During the Exam

💡
The Three-Pass Strategy (120 minutes for 15-20 tasks):

Pass 1 (0-80 min): Do easy tasks first. If a task looks hard, flag it and move on. Aim to complete 12-15 tasks.

Pass 2 (80-110 min): Return to flagged tasks. Use the Kubernetes docs to look up YAML syntax you do not remember.

Pass 3 (110-120 min): Verify your work. Run kubectl get on each resource you created. Make sure Pods are Running, Services have endpoints, and PVCs are Bound.

Common Mistakes to Avoid

  • Wrong namespace: Always check which namespace the task requires. Set it with -n or change default namespace.
  • Wrong context: Some tasks switch clusters. Always verify with kubectl config current-context.
  • YAML indentation: Use 2 spaces, never tabs. A single indentation error can cost you the entire task.
  • Not verifying: Always kubectl get after creating resources. A resource that errors on creation scores zero.
  • Spending too long: If a 4% task takes more than 8 minutes, move on. A 7% task is worth coming back to.

Frequently Asked Questions

How hard is the CKA exam?

The CKA has a pass rate of approximately 50-60%. It is considered moderately difficult because it is performance-based (you must actually solve tasks, not just pick answers). With 2-4 weeks of dedicated practice, most candidates pass. Hands-on practice is essential — reading alone is not sufficient.

Can I use the Kubernetes documentation during the exam?

Yes. You can access kubernetes.io/docs, kubernetes.io/blog, and the GitHub Kubernetes repository during the exam. You CANNOT access other websites, Stack Overflow, or personal notes. Bookmark useful docs pages before the exam for quick access.

What happens if I fail the CKA?

The $395 exam fee includes one free retake. You can schedule the retake after a 24-hour waiting period. If you fail the retake, you must purchase a new exam. Your score report identifies weak areas to focus on.

How long is the CKA certification valid?

The CKA certification is valid for 3 years. To recertify, you must retake the current version of the exam. There is no recertification discount.

Is the CKA relevant for AI/ML engineers?

Absolutely. Kubernetes is the standard platform for ML infrastructure at scale. Companies like Google, Meta, and OpenAI run ML workloads on Kubernetes. A CKA demonstrates you can manage the infrastructure that AI systems depend on, making you valuable to any team building ML platforms.

Should I take the CKA or CKAD first?

For AI/ML engineers, take the CKA first. The CKA covers cluster administration (node management, storage, networking, troubleshooting) which is critical for GPU clusters and ML infrastructure. The CKAD focuses on application development, which is more relevant for developers deploying applications.

Do I need to know how to install a Kubernetes cluster from scratch?

You should know how to use kubeadm to initialize a cluster and join nodes. The exam may include tasks like upgrading a cluster or restoring etcd from a backup. You do not need to install Kubernetes entirely from binaries.

What tools are available in the exam environment?

The exam terminal includes: kubectl, vim, nano, jq, tmux, curl, and standard Linux tools. You can use vim or nano to edit YAML files. Practice with whichever editor you prefer before exam day.

Additional Study Resources

  • Kubernetes Official Documentation — kubernetes.io/docs is the primary reference (and available during the exam)
  • killer.sh — The CKA exam voucher includes two free killer.sh simulator sessions. These simulate the real exam environment and are harder than the actual exam.
  • minikube / kind — Set up a local cluster for hands-on practice. Practice creating every resource type from memory.
  • CNCF Curriculum — The official exam curriculum document lists every topic that can appear on the exam.
💡
You have completed this course! If you have worked through all 7 lessons, taken the practice exam, and practiced in a real cluster, you are well-prepared for the CKA exam. Remember: speed matters. Practice kubectl commands until they are muscle memory. Good luck!