Node troubleshooting in Kubernetes - Practice Problems & Coding Challenges

Challenge - 5 Problems

💻 Command Output

intermediate

Check node status with kubectl

You run the command kubectl get nodes on your cluster. What output indicates that a node is NotReady?

Kubernetes

kubectl get nodes

A

NAME       STATUS     ROLES    AGE   VERSION
node1      Ready      worker   10d   v1.26.0
node2      NotReady   worker   10d   v1.26.0

B

NAME       STATUS     ROLES    AGE   VERSION
node1      Ready      master   10d   v1.26.0
node2      Ready      worker   10d   v1.26.0

C

NAME       STATUS     ROLES    AGE   VERSION
node1      Unknown    worker   10d   v1.26.0
node2      Ready      worker   10d   v1.26.0

D

NAME       STATUS     ROLES    AGE   VERSION
node1      Ready      worker   10d   v1.26.0
node2      SchedulingDisabled   worker   10d   v1.26.0

💡 Hint

Look for the STATUS column showing node health.

The NotReady status means the node is not healthy and cannot run pods. Other statuses like Ready or SchedulingDisabled mean different states.

❓ Troubleshoot

intermediate

Diagnose node unreachable error

You notice pods stuck in Pending state and suspect a node is unreachable. Which command helps you check node connectivity issues?

A kubectl logs <pod-name>

B kubectl get pods --all-namespaces

C kubectl exec <pod-name> -- ping 8.8.8.8

D kubectl describe node <node-name>

💡 Hint

You want detailed info about the node itself.

The kubectl describe node command shows node conditions, including network or disk pressure, which helps diagnose connectivity problems.

❓ Configuration

advanced

Identify correct kubelet config for node readiness

Which kubelet configuration option ensures the node reports Ready status only when the node is healthy?

A --fail-swap-on=false

B --read-only-port=10255

C --eviction-hard=memory.available<500Mi,nodefs.available<10%

D --node-status-update-frequency=10s

💡 Hint

Look for settings that control node health eviction.

The --eviction-hard flag sets thresholds for resource pressure. If these are exceeded, the node becomes NotReady to protect stability.

🔀 Workflow

advanced

Sequence to safely drain a node

What is the correct order of commands to safely drain a node before maintenance?

A 2,1,3,4

B 1,2,3,4

C 1,3,2,4

D 4,1,2,3

💡 Hint

First prevent new pods, then move existing pods, then allow scheduling again.

You first mark the node unschedulable with cordon, then safely evict pods with drain. After maintenance, use uncordon to allow pods again.

✅ Best Practice

expert

Identify best practice for node failure alerting

Which approach is best to get immediate alerts when a Kubernetes node becomes NotReady?

A Set up Prometheus to monitor node status and alert on <code>NotReady</code> condition

B Restart kubelet service on all nodes daily

C Use <code>kubectl logs</code> on pods to check node health

D Manually run <code>kubectl get nodes</code> every hour

💡 Hint

Automated monitoring and alerting is key for fast response.

Prometheus can scrape node metrics and alert immediately when a node status changes to NotReady, enabling quick action.