Automated-restart-carshed-pods

Creating an automated system to detect and restart Kubernetes pods stuck in a CrashLoopBackOff state.

Folder structure

Test steps

Deploy the Crashy Application

1. Apply the crashy-app.yaml manifest::

   kubectl apply -f manifests/crashy-app.yaml
   
   

2. Verify that the pod enters a CrashLoopBackOff state:

   kubectl get pods
   

3. Run the CrashLoop Detection Script

4. Make the script executable::

   chmod +x scripts/watch_crashloop.sh
   

5. Run the script:.

   ./scripts/watch_crashloop.sh
   
   **Observe the script detecting and restarting the pod.**
   

6. Deploy the CronJob (Optional)

   Apply the restart_crashloop_cronjob.yaml manifest:

   kubectl apply -f scripts/restart_crashloop_cronjob.yaml
   

7. Verify the CronJob:

   kubectl get cronjobs
   
   **Deploy the Kubernetes Operator (Optional)**
   

8. Build the operator Docker image:

   docker build -t crashloop-operator:latest manifests/operator/
   

9. Push the image to a container registry (e.g., Docker Hub):

   docker tag crashloop-operator:latest your-dockerhub-username/crashloop-operator:latest
   docker push your-dockerhub-username/crashloop-operator:latest
   

10. Deploy the operator to your Kubernetes cluster:

    apiVersion: apps/v1
    kind: Deployment
    metadata:
      name: crashloop-operator
    spec:
      replicas: 1
      selector:
        matchLabels:
          app: crashloop-operator
      template:
        metadata:
          labels:
            app: crashloop-operator
        spec:
          containers:
          - name: operator
            image: your-dockerhub-username/crashloop-operator:latest
    

11. Apply the operator deployment:

    kubectl apply -f operator-deployment.yaml
    
    

Expected Output

The crashy-app pod enters a CrashLoopBackOff state.

The script, CronJob, or operator detects the CrashLoopBackOff and restarts the pod.

The pod is recreated and continues to crash, but the system ensures it is restarted automatically.