mathewjustin / gkube-seeder Goto Github PK

"# kube-seeder"

This project is based on

• https://spring.io/guides/gs/spring-boot-docker/
• https://cloud.google.com/kubernetes-engine/docs/tutorials/hello-app

Create project and activate cloud shell in the https://console.cloud.google.com/appengine/

Open shell

• Step 1: Build the container image

  git clone https://github.com/mathewjustin/gkube-seeder.git
  

• Step 2: Set up environment variables, which are going to be reference in the future steps

   export PROJECT_ID="$(gcloud config get-value project -q)"
  

  PROJECT_ID Will be useful while publishing to the container registry

• Step 3: Build the container image

  On Step 1 we have cloned the repo to the project, next step is to build an image of this app and push to container registry.

   docker build -t gcr.io/${PROJECT_ID}/kube-seeder:v1 .
   or
   ./mvnw install dockerfile:build : for this particular project
  

  if this step throwing any error, switch to root user, connect to

  After this step you will be able to see the newly built image , by listing the local images "docker images"

• Step 4: Upload the container image

  Inorder for GKE to download and run the image, we need to push the image to container registry

  first authenticate with container registry.

  gcloud auth configure-docker
  

  Now run this command.

  docker push gcr.io/${PROJECT_ID}/kube-seeder:v1
  

• Step 5: Test it by running locally.

  docker run --rm -p 8080:8080 gcr.io/${PROJECT_ID}/kube-seeder:v1
  curl http://localhost:8080
  

• Step 6: Create a container cluster

  Create a new cluster for your app and configure nodes using UI, Or use this command

  gcloud container clusters create kube-cluster --num-nodes=3
  
  gcloud compute instances list
  

  If you have already created the cluster then run the following to retrieve the cluster credentials. gcloud container clusters get-credentials kube-cluster

• Step 7: Deploy the application

  The kubectl run command causes Kubernetes to create a Deployment named kube-web on your cluster. The Deployment manages multiple copies of your application, called replicas, and schedules them to run on the individual nodes in your cluster. In this case, the Deployment will be running only one Pod of your application.

  kubectl run kube-web --image=gcr.io/${PROJECT_ID}/kube-seeder:v1 --port 8080
  

  To see the Pod created by the Deployment, run the following command:

   kubectl get pods
  

• Step 8: Expose your application to the Internet

   kubectl expose deployment kube-web --type=LoadBalancer --port 80 --target-port 8080
  

  use the following command to see the details of service

    kubectl get service
  

• Step 9: Scale up your application - Manual

    kubectl scale deployment kube-web --replicas=3
    kubectl get deployment kube-web =>To see instance details
  

• Step 10: Updating versions.

Build new image with updated version, push to registry and run the following to do rolling update

 kubectl set image deployment/kube-web kube-web=gcr.io/${PROJECT_ID}/kube-seeder:v2

• Verify the app working as expected deployed.

  App will expose a simple /names end point

   <External IP>/names
  

  This end point will give a list of names.

  eg: /names ["Aaberg","Aalst","Aara","Aaren","Aarika","Aaron","Aaronson","Ab","Aba","Abad"]

//modifying