Kubernetes is powerful… but also tricky.
If you’ve worked with it in production, you’ve probably seen pods restarting endlessly, images failing to pull, or volumes stuck in Pending.

In this guide, I’ll share 15 of the most common Kubernetes errors, why they happen, and how to fix them. I’ve added YAML snippets, commands, and diagrams so you can debug faster.

---

🔄 Pod Lifecycle (Quick Reference)

flowchart LR
  A[Pending] --> B[ContainerCreating]
  B --> C[Running]
  C -->|Crash| D[CrashLoopBackOff]
  C -->|Exit 0| E[Succeeded]
  C -->|Exit 1| F[Failed]

---

1. ❌ CrashLoopBackOff

• Meaning: Pod starts, crashes, and restarts in a loop.

• Fix Steps:

    kubectl logs <pod-name>
    kubectl describe pod <pod-name>
  

• Example YAML (fix missing env vars & probes):

apiVersion: v1
kind: Pod
metadata:
  name: myapp
spec:
  containers:
  - name: app
    image: myapp:v1
    env:
    - name: DB_HOST
      value: "db-service"
    livenessProbe:
      httpGet:
        path: /health
        port: 8080
      initialDelaySeconds: 5
      periodSeconds: 10

---

2. ❌ ImagePullBackOff

• Meaning: Kubernetes cannot pull the container image.

• Fix:

  • Check image name & tag

  • Add secret for private registry

spec:
  containers:
  - name: app
    image: myregistry.com/team/app:v2
  imagePullSecrets:
  - name: regcred

kubectl create secret docker-registry regcred \
  --docker-server=myregistry.com \
  --docker-username=user \
  --docker-password=pass

---

3. ❌ ImagePullPolicy = Never

• Meaning: Pod tries to use only local images.

• Fix:

imagePullPolicy: IfNotPresent

---

4. ❌ Node NotReady

• Meaning: Worker node unhealthy.

• Fix:

kubectl describe node <node-name>
systemctl restart kubelet

---

5. ❌ Pod Stuck in Pending

• Meaning: Pod can’t find a node.

• Fix:

  • Check resources, taints, tolerations.

kubectl describe pod <pod>

---

6. ❌ OOMKilled

• Meaning: Pod killed due to memory overuse.

• Fix (set limits):

resources:
  requests:
    memory: "256Mi"
    cpu: "250m"
  limits:
    memory: "512Mi"
    cpu: "500m"

---

7. ❌ Pod Stuck in Terminating

• Meaning: Pod won’t delete.

• Fix:

kubectl delete pod <pod> --grace-period=0 --force

---

8. ❌ Service Not Accessible

• Meaning: Service misconfigured or no endpoints.

• Fix:

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

Diagram:

flowchart LR
  Client --> Service --> Endpoints --> Pod

---

9. ❌ DNS Resolution Failure

• Meaning: CoreDNS not resolving service names.

• Fix:

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

---

10. ❌ RBAC Forbidden

• Meaning: User doesn’t have permission.

• Fix (RoleBinding):

apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: dev-access
  namespace: dev
subjects:
- kind: User
  name: dev-user
roleRef:
  kind: Role
  name: pod-reader
  apiGroup: rbac.authorization.k8s.io

---

11. ❌ Unauthorized (Kubeconfig Issue)

• Meaning: Wrong kubeconfig credentials.

• Fix: Update ~/.kube/config or re-download from cluster.

---

12. ❌ PVC Pending

• Meaning: PVC not binding.

• Fix:

apiVersion: v1
kind: PersistentVolume
metadata:
  name: pv-volume
spec:
  storageClassName: manual
  capacity:
    storage: 1Gi
  accessModes:
    - ReadWriteOnce
  hostPath:
    path: "/data/pv"

---

13. ❌ ImagePullSecret Missing

• Meaning: Secret not found.

• Fix: Ensure correct reference under imagePullSecrets.

---

14. ❌ Read-Only Filesystem

• Meaning: Volume mounted as read-only.

• Fix:

volumeMounts:
- name: data
  mountPath: /app/data
  readOnly: false

---

15. ❌ Kube-API Server Down

• Meaning: Control plane unresponsive.

• Fix:

systemctl restart kube-apiserver

---

⚡ Key Takeaways

• Always start with:

    kubectl describe pod <pod>
    kubectl logs <pod>
  

• Define resource limits, probes, and secrets properly.

• Use monitoring (Prometheus, Grafana, CloudWatch) to catch issues early.

---

💡 Have you faced any of these errors in production? Share your experience in the comments — let’s debug together!