Building Self-Healing Containers in Kubernetes

Introduction

Kubernetes offers several features that can be used together to create self-healing applications in a variety of scenarios. In this lab, you will be able to practice your skills at using features such as probes and restart policies to create a container application that is automatically healed when it stops working.

Solution

Log in to the provided lab server using the credentials provided:

ssh cloud_user@<PUBLIC_IP_ADDRESS>

Set a Restart Policy to Restart the Container When It Is Down

1. Find the pod that needs to be modified:

   kubectl get pods -o wide
   

2. Take note of the beebox-shipping-data pod’s IP address.

3. Use the busybox pod to make a request to the pod to see if it is working:

   kubectl exec busybox -- curl <beebox-shipping-data_ IP>:8080
   

   We will likely get an error message.

4. Get the pod’s YAML descriptor:

     kubectl get pod beebox-shipping-data -o yaml > beebox-shipping-data.yml
   

5. Open the file:

   vi beebox-shipping-data.yml
   

6. Set the restartPolicy to Always:

   spec:
     ...
     restartPolicy: Always
     ...
   

Create a Liveness Probe to Detect When the Application Has Crashed

1. Add a liveness probe:

   spec:
     containers:
     - ...
       name: shipping-data
       livenessProbe:
         httpGet:
           path: /
           port: 8080
         initialDelaySeconds: 5
         periodSeconds: 5
       ...
   

2. Save and exit the file by pressing Escape followed by :wq.

3. Delete the pod:

   kubectl delete pod beebox-shipping-data
   

4. Re-create the pod to apply the changes:

   kubectl apply -f beebox-shipping-data.yml
   

5. Check the pod status

   kubectl get pods -o wide
   

6. If you wait a minute or so and check again, you should see the pod is being restarted whenever the application crashes.

7. Check the http response from the pod again (it will have a new IP address since we re-created it):

   kubectl exec busybox -- curl <beebox-shipping-data_IP>:8080
   

   If you wish, you can explore and see what happens as the application crashes and the pod is restarted automatically.