kubernetes crisp

Got it — you want crisp, 30–45 second interview-ready answers for each Kubernetes real-time scenario, so you can speak confidently without sounding like you’re reading a script.

I’ll make them in this format:

Scenario → Concise Answer (Why + How + Example in bullets, with imperative command if relevant)

---

1. Pods Stuck in CrashLoopBackOff

• Why: Container is repeatedly failing — bad config, missing env vars, wrong image.
• How: Describe how to check logs, events, and fix root cause.
• Example:

  kubectl describe pod <pod>
  kubectl logs <pod> -c <container>
  

• Interview line: “First, I’d check kubectl describe and logs to see the failure reason, fix the image/config, then redeploy.”

---

2. Pod Pending (Scheduling Issue)

• Why: No matching node (resource shortage, node selector mismatch, taints).
• How: Check events, resource requests, node taints.
• Example:

  kubectl describe pod <pod>
  kubectl get nodes --show-labels
  

• Interview line: “I’d inspect scheduling events, verify resource requests, and adjust node selectors or scale nodes.”

---

3. High CPU/Memory Usage

• Why: Container limits too low, memory leak, heavy workload.
• How: Use metrics, adjust resources, scale pods.
• Example:

  kubectl top pod
  kubectl top node
  kubectl scale deploy <name> --replicas=5
  

• Interview line: “I’d use kubectl top to confirm high usage, then tune requests/limits or scale the deployment.”

---

4. Service Not Accessible

• Why: Service type mismatch, wrong selector, missing endpoints.
• How: Check service endpoints and pod labels.
• Example:

  kubectl get svc
  kubectl describe svc <svc>
  kubectl get endpoints <svc>
  

• Interview line: “I’d verify the service has correct selectors and endpoints, then test connectivity from inside the cluster.”

---

5. Image Pull BackOff

• Why: Wrong image name, private repo without creds.
• How: Fix image reference or add imagePullSecret.
• Example:

  kubectl create secret docker-registry mysecret --docker-username=... --docker-password=...
  

• Interview line: “I’d confirm the image name, check registry access, and add an imagePullSecret if needed.”

---

If you want, I can prepare all 20 Lloyds scenarios in this exact 30–45 second bullet style with imperative commands so you have a single “speak-ready” document for the interview.

Do you want me to go ahead and prepare the full 20 in this short-answer style?

Alright, here’s a Kubernetes Troubleshooting Interview Quick-Answer Sheet — each scenario condensed into 30–45 seconds with key imperative commands you can rattle off in an interview without over-talking.

---

1. Pod Stuck in CrashLoopBackOff

Short Answer:

"Usually due to application crashes or bad configs. First, check logs and events to find the root cause, fix config, and redeploy."

Commands:

kubectl get pod <pod> -n <ns>

kubectl logs <pod> -n <ns> --previous

kubectl describe pod <pod> -n <ns>

kubectl rollout restart deploy <deploy> -n <ns>

---

2. Pod Pending

Short Answer:

"Pending means the scheduler can’t place the pod — usually no matching nodes, insufficient resources, or affinity rules blocking scheduling."

Commands:

kubectl describe pod <pod> -n <ns>

kubectl get nodes -o wide

kubectl describe node <node>

---

3. ImagePullBackOff

Short Answer:

"Registry auth issues, wrong image name, or tag mismatch. Check events and fix image config or secret."

Commands:

kubectl describe pod <pod> -n <ns>

kubectl create secret docker-registry <name> \

  --docker-server=<server> --docker-username=<user> \

  --docker-password=<pass> --docker-email=<email>

kubectl patch serviceaccount default \

  -p '{"imagePullSecrets":[{"name":"<name>"}]}'

---

4. Node Not Ready

Short Answer:

"Can be due to kubelet failure, disk pressure, network issue, or cloud node health problem. Check node status and kubelet logs."

Commands:

kubectl get nodes

kubectl describe node <node>

journalctl -u kubelet -f

---

5. High Pod Restarts

Short Answer:

"Often caused by memory/CPU limits, readiness probe failure, or app bugs. Analyze logs, events, and resource usage."

Commands:

kubectl logs <pod> -n <ns>

kubectl top pod <pod> -n <ns>

kubectl describe pod <pod> -n <ns>

---

6. OOMKilled

Short Answer:

"Container exceeded memory limit. Increase limits or optimize memory usage."

Commands:

kubectl describe pod <pod> -n <ns>

kubectl edit deploy <deploy> -n <ns>

---

7. Service Not Accessible

Short Answer:

"Check if service type and selector match pods, and ensure endpoints are created."

Commands:

kubectl get svc -n <ns>

kubectl describe svc <svc> -n <ns>

kubectl get endpoints <svc> -n <ns>

---

8. DNS Resolution Failing

Short Answer:

"Usually CoreDNS crash or misconfig. Check CoreDNS pods and logs."

Commands:

kubectl get pods -n kube-system -l k8s-app=kube-dns

kubectl logs -n kube-system <coredns-pod>

---

9. PVC Pending

Short Answer:

"StorageClass mismatch or no available PV. Check PVC events and storage backend."

Commands:

kubectl describe pvc <pvc> -n <ns>

kubectl get sc

kubectl get pv

---

10. Deployment Not Updating

Short Answer:

"Wrong selector, failing readiness probe, or paused rollout. Check rollout status."

Commands:

kubectl rollout status deploy <deploy> -n <ns>

kubectl rollout resume deploy <deploy> -n <ns>

kubectl describe deploy <deploy> -n <ns>

---

If you want, I can expand this to all 20 Lloyds interview scenarios in the same short-answer + imperative command style so you can literally answer in 40 seconds per question in the interview w

ithout going blank.

Do you want me to prepare all 20 like this in one printable PDF? That will make it a perfect rapid-fire prep sheet.