Time budget: ~2 hours
Two scenarios. The first tests fault-finding; the second tests your understanding of production readiness.
You will need the following tools installed before starting:
- Docker — docker.com
- kind — kind.sigs.k8s.io
- kubectl — kubernetes.io
- Helm v3 — helm.sh
Verify everything is working before running the setup script:
docker info
kind version
kubectl version --client
helm versionRun the setup script. It creates a local Kubernetes cluster and installs both releases.
chmod +x setup.sh teardown.sh
./setup.shSetup takes about 90 seconds. When it finishes:
helm list -A # shows both releases across namespaces
kubectl get pods -A # expect some pods in scenario1 to be unhealthy — that's intentionalThe fault-app release is a multi-component application: an API and a background worker. It has been recently deployed and nothing is working correctly.
Your job is to find and fix every fault.
kubectl get pods -n scenario1
kubectl get svc -n scenario1Fix each issue via helm upgrade fault-app ./charts/scenario1 --namespace scenario1.
The prod-app release is a web application backed by a Redis cache. It deploys and runs without errors.
Your job is to review the chart through a production readiness lens — covering both resilience, security and maintainability. This is deliberately open-ended. Improve whatever you think matters most and be ready to talk through your reasoning. You can make whatever changes/additions to the manifest themselves as well as any templating improvements.
kubectl get pods -n scenario2
kubectl describe pod <pod-name> -n scenario2The app doesn't need to be externally accessible for this exercise, but if you want to verify it's serving traffic you can port-forward the service:
kubectl port-forward svc/prod-app-web 8080:80 -n scenario2then openhttp://localhost:8080.
When you're done, commit your changes and send us a Git bundle:
git add -A && git commit -m "exercise submission"
git bundle create k8s-exercise.bundle HEAD mainEmail k8s-exercise.bundle back to us. We'll review your changes and discuss your approach together in the follow-up session.
./teardown.sh