Kubernetes scenario practice

Make the right kubectl call under pressure.

Short incident scenarios with graded feedback — practice the decisions you make in real clusters, not isolated trivia.

850+ scenario casesNo account needed~5 min per session

scenario · pod_debuggingStandard

A Pod in namespace shop keeps restarting. Status: CrashLoopBackOff.

?What is your best next step?

1kubectl delete pod api-7f8b9c -n shop
2kubectl describe pod api-7f8b9c -n shop
3kubectl logs api-7f8b9c -n shop
4kubectl rollout restart deployment/api -n shop

✓ Best first step

Describe shows Events, exit codes, and probe failures — then logs --previous.

Five minutes, three beats — the same arc CKA and on-call practice both reward: orient, decide, then sequence.

  1. Warm up

    Two quick foundation prompts refresh the concepts you need before the incident hits.

  2. Decide

    Read realistic terminal output, then pick the best first kubectl step from graded options.

  3. Sequence

    When the fix takes more than one move, order the runbook — describe before delete.

Full topic catalog850 scenarios · 7 core paths
Common questionsFAQ
Do I need a Kubernetes cluster to practice here?

No. Every scenario includes realistic kubectl output and incident context. You decide the next command from what you see — the same reading you'd do on a real cluster, without provisioning nodes or installing a control plane in this app.

What's included in the free Pod Debugging path?

During early access, every topic is open — including Pod Debugging, all core paths, CKA/CKAD prep, platform/GitOps, and interview scenarios. Pod Debugging is still the best place to start.

What happens in a typical five-minute session?

Core topics: two foundation warm-ups, four incident scenarios (pick the best first kubectl step), then one runbook ordering exercise. CKA, CKAD, and platform topics add a third warm-up and may include a second runbook — still designed for a short sitting, not a multi-hour lab.

CKA or CKAD — which exam pack should I choose?

CKAD practice focuses on workloads — Deployments, probes, Services, ConfigMaps, Jobs, and manifest fixes. CKA practice focuses on cluster operations — nodes, networking, storage, and control-plane troubleshooting. The Linux Foundation does not require CKA before CKAD (or the reverse); pick the lens that matches your role.

Is this enough to pass the CKA or CKAD exam on its own?

No. Official CKA and CKAD exams are performance-based: you work in a live environment, create and edit resources, and verify fixes under a two-hour clock with roughly 15–20 tasks. Use Decision Trainer to drill diagnostic order under pressure; pair it with hands-on cluster practice (kind, minikube, or a cloud sandbox) and the official curriculum.

Is this official CNCF or Linux Foundation training?

No. Decision Trainer is independent, unofficial practice. Kubernetes®, CKA®, and CKAD® are registered trademarks of The Linux Foundation. This site is not affiliated with CNCF or The Linux Foundation.

How is this different from a kubectl cheat sheet?

Cheat sheets list commands. Here you choose the best next step from terminal output while a pod is CrashLooping or a Service has no Endpoints — closer to on-call triage and exam scenarios than memorizing syntax in isolation.

Do I need an account — and is there a subscription?

No account is required to start. Progress stays in your browser until you sign up. A free account syncs training history across devices. There is no subscription — during early access everything is open.