19 Common Kubernetes Failures and How to Fix Them

1. Service Access Failure (Certificate Issue)

Cause: The cluster cannot recognize the TLS certificate because it is custom, expired, or otherwise invalid.

Solution: Replace or renew the certificate.

2. Service Access Failure (Connection Refused)

Cause: Incorrect port mapping; the service is running but the port is not exposed correctly.

Solution: Delete the existing Service and recreate it with the proper port mapping.

kubectl delete svc nginx-deployment

3. Service Exposure Failure (AlreadyExists)

Cause: The service has already been created for the container.

Solution: Delete the existing Service and recreate it with the correct configuration.

kubectl delete svc nginx-deployment

4. External Access Failure (ClusterIP)

Cause: The Service type is ClusterIP, which does not expose the service outside the cluster.

Solution: Change the Service type to NodePort so it can be accessed via any node IP.

kubectl edit svc nginx-deployment

5. Pod ErrImagePull

Cause: The container image cannot be pulled (e.g., missing, private, or corrupted).

Solution: Replace the image reference with a valid one and redeploy.

kubectl set image deployment/myapp myapp-container=new-image:tag

6. Init Container Stuck (PodInitializing)

Cause: The init container never finishes, often because of DNS resolution failures or missing resources.

Solution: Create the required Service (e.g., myservice) and ensure CoreDNS can resolve its name.

kubectl apply -f myservice.yaml

7. CrashLoopBackOff

Cause: The container image is faulty, causing repeated start‑up failures.

Solution: Replace the problematic image with a stable version.

kubectl set image deployment/myapp myapp-container=stable-image:tag

8. Pod Pending (Image Pull Issues)

Cause: The image name is incorrect or the registry cannot be reached.

Solution: Correct the image name and pull the image manually if needed.

kubectl delete pod readiness-httpget-pod
kubectl run test-nginx --image=10.0.0.81:5000/nginx:alpine

9. Pod Ready State Not Reached

Cause: The pod’s command fails, preventing the container from becoming ready.

Solution: Enter the container, fix the command or resource definition, and redeploy.

10. Pod Creation Failure (YAML Errors)

Cause: The YAML file contains duplicate fields or uses non‑ASCII characters.

Solution: Clean up the YAML (remove duplicate containers entries, fix encoding) and apply again.

11. Flannel DaemonSet Pod Init Failure

Cause: Image pull failure on the node (e.g., network or registry issue).

Solution: Restart Docker on the affected node, manually pull the image, or reinstall the Flannel plugin.

kubectl create -f kube-flannel.yml
kubectl get nodes

12. Service ErrImagePull

Cause: Incorrect image name specified in the pod spec.

Solution: Delete the faulty pod and redeploy with the correct image.

kubectl delete pod test-nginx
kubectl run test-nginx --image=10.0.0.81:5000/nginx:alpine

13. Cannot Enter Specified Container

Cause: The YAML defines duplicate containers fields, so the intended container does not exist.

Solution: Remove the extra containers entry and recreate the pod.

14. PersistentVolume Creation Failure

Cause: Duplicate name fields in PV definitions.

Solution: Change the PV name to a unique value.

15. PVC Mount Failure

Cause: Mismatch between accessModes of the PVC and the available PVs (e.g., PVC requests RWO >1Gi but only a different PV exists).

Solution: Adjust the PVC or PV accessModes to match.

16. PV Content Not Accessible

Cause: The NFS volume is empty or permissions are incorrect.

Solution: Create the required files in the NFS share and set appropriate permissions.

17. Node Status Query Failure

Cause: Missing Heapster monitoring service.

Solution: Install Prometheus (or another monitoring stack) to provide node metrics.

18. Pod Stuck in Pending

Cause: No schedulable nodes available because existing pods already consume all resources for the requested image.

Solution: Delete existing pods or scale down workloads, then redeploy.

19. Helm Install Failure

Cause: Incorrect chart file naming; Chart.yaml is missing or misnamed.

Solution: Rename the file to Chart.yaml and retry the install.

mv chart.yaml Chart.yaml