A DevOps engineer notices that a Kubernetes Deployment's Pods are CrashLoopBackOff. The application writes structured logs to stdout. Which command sequence should the engineer use to quickly identify the root cause?

kubectl logs --tail=50, then kubectl rollout restart deployment/

kubectl get events --sort-by=.lastTimestamp, then kubectl logs --previous

kubectl describe pod , then kubectl exec -it -- /bin/sh

kubectl get pod -o yaml, then kubectl delete pod

A Pod is stuck in `Pending` state. The team checks `kubectl describe pod` and sees the event: `0/3 nodes are available: 3 Insufficient cpu.` The Deployment requests 2 CPUs per replica, but the nodes only have 1 CPU available. What is the most appropriate action?

Remove the CPU request entirely from the container spec.

Change the Service type to NodePort to bypass scheduling.

Increase the number of replicas to spread the load.

Reduce the CPU request in the Deployment to fit the node capacity.

A team wants to automatically detect when a Kubernetes Deployment is not making progress during a rollout (e.g., new Pods are not becoming ready). Which Kubernetes feature should they use?

Use a HorizontalPodAutoscaler to scale the Deployment.

Monitor the Deployment's `status.conditions` with a custom script.

Configure a `progressDeadlineSeconds` on the Deployment.

Set up a liveness probe on the new Pods.

A team runs a stateful application in Kubernetes. They want to detect when the application is not able to serve requests due to a full disk on the node. They already have Prometheus monitoring node-level metrics. Which metric should they alert on?

`container_memory_usage_bytes` for the application's container.

`node_filesystem_avail_bytes` for the mount point used by the application's data.

`kube_pod_container_status_restarts_total` for the application's Pod.

`rate(container_cpu_usage_seconds_total[5m])` for the application's container.

A team deploys a new version of a microservice. The rollout is successful, but users report that some requests fail with `Connection reset by peer`. The team checks the logs and sees that the new Pods are receiving traffic, but the old Pods are still receiving traffic as well. The Deployment strategy is RollingUpdate. What is the most likely cause?

The Service's `sessionAffinity` is set to `ClientIP`, causing requests to stick to old Pods.

The new Pods have a different container port than the Service's targetPort.

The new Pods are not ready, so the Service is still sending traffic to the old Pods, which are being terminated.

The Ingress controller is caching the old endpoints and not updating to the new Pods.

Kubernetes and Cloud Native Associate (KCNA)

Domain 3Objective 2

Debugging KCNA Practice Questions (Page 2)

Part of the Cloud Native Application Delivery domain, which accounts for 16% of the KCNA exam. Linux Foundation does not publish an official question count, but from its 90-minute exam (~35–60 total, ~6–10 in this domain), expect 2–3 from this objective — we provide 23 practice questions to prepare you well beyond it. (estimate)

23questions here

5free pages

4concepts

16%of the exam

Questions 6–10

6application · medium
A DevOps engineer notices that a Kubernetes Deployment's Pods are CrashLoopBackOff. The application writes structured logs to stdout. Which command sequence should the engineer use to quickly identify the root cause?
Select an answer first
7application · medium
A Pod is stuck in `Pending` state. The team checks `kubectl describe pod` and sees the event: `0/3 nodes are available: 3 Insufficient cpu.` The Deployment requests 2 CPUs per replica, but the nodes only have 1 CPU available. What is the most appropriate action?
Select an answer first
8application · medium
A team wants to automatically detect when a Kubernetes Deployment is not making progress during a rollout (e.g., new Pods are not becoming ready). Which Kubernetes feature should they use?
Select an answer first
9expert · medium
A team runs a stateful application in Kubernetes. They want to detect when the application is not able to serve requests due to a full disk on the node. They already have Prometheus monitoring node-level metrics. Which metric should they alert on?
Select an answer first
10expert · hard
A team deploys a new version of a microservice. The rollout is successful, but users report that some requests fail with `Connection reset by peer`. The team checks the logs and sees that the new Pods are receiving traffic, but the old Pods are still receiving traffic as well. The Deployment strategy is RollingUpdate. What is the most likely cause?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to KCNA

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Linux Foundation. “KCNA” is a trademark of its owner, used for identification only.