How to run a Kubernetes cluster on your laptop

Prerequisites for the setup:

1. Make sure you are not connected to any VPN

2. Hyper-V should be disabled in your system since minikube uses VirtualBox [Hyper-V is disabled by default]

Note:

Minikube package embeds both docker client and the daemon. You don’t need to install docker separately. But note that you can only execute docker commands from the minikube host, and not from your laptop.

Detailed Steps to setup kubernetes in local from scratch

 

1. Install Kubectl : curl -LO https://storage.googleapis.com/kubernetes-release/release/v1.16.0/bin/windows/amd64/kubectl.exe

 

2. Update the windows PATH to include kubectl.exe

 

3. Check if “kubectl version” runs properly

 

4. Install Minikube : https://github.com/kubernetes/minikube/releases/latest/download/minikube-installer.exe

 

5. Install Virtual Box

 

6. Start the minikube environment : minikube start

 

7. Check the dashboard : minikube dashboard

 

8. Create a tomcat deployment in the Kubernetes cluster with 3 replicas: kubectl run mytomcat –image=tomcat –replicas=3 [[This will create all the necessary pods for the deployment]]

 

9. List all PODs : kubectl get pods

 

10. You can login to any of the created PODs using : kubectl exec -it <pod_name> /bin/bash

 

11. Expose the deployment externally as a service : kubectl expose deployment mytomcat –port=8080 –target-port=8080 –type=NodePort

 

12. Check the exposed IP and Port : kubectl describe service mytomcat

 

13. Access the service : minikube service mytomcat

 

14. Check whether the requests are getting load balanced between all the replicas by checking access logs of each tomcat containers : kubectl exec -it mytomcat-75b679fc45-8t94j /bin/bash

 

15. Scale up the number of containers from 3 to 4 : kubectl scale –replicas=4 deployment mytomcat

 

16. Scale down the number of containers from 4 to 2 : kubectl scale –replicas=2 deployment mytomcat

 

17. Check the number of PODs Running : kubectl get pods

 

18. To perform a rolling update, first get the details of the deployment using : kubectl edit deployment mytomcat

 

19. Run the command to change the tomcat version : kubectl –record deployment/mytomcat set image deployment/mytomcat mytomcat=tomcat:7.0.96-jdk8-adoptopenjdk-hotspot

 

20. Look at the sequence of events below, how the new containers are spawned one by one and the old ones are terminated in a rolling fashion:

 

Poll on the URl to see if you get response code other than 200 : while true; do curl -I http://192.168.99.104:31906/; done

You can get a rollout history : kubectl rollout history deployment/mytomcat

deployment.apps/mytomcat REVISION  CHANGE-CAUSE 1          kubectl deployment/mytomcat set image deployment/mytomcat tomcat=tomcat:7.0.96-jdk8-adoptopenjdk-hotspot –record=true 2          kubectl deployment/mytomcat set image deployment/mytomcat mytomcat=tomcat:7.0.96-jdk8-adoptopenjdk-hotspot –record=true 3          kubectl deployment/mytomcat set image deployment/mytomcat mytomcat=tomcat:8-jdk11-adoptopenjdk-hotspot –record=true

Rolling back to the previous version : kubectl rollout undo deployment/mytomcat