AI Planet DevOps Demo

All Changes with commits present in repo - read commit messages

Initial steps:

Setup A kubernetes cluster on mac m2 Selected minikube
minikube start - using docker driver

Task 1: Setup and Configuration

1.1 Create a Github repository

Created Repository: https://github.com/leulad/ai-planet-devops

1.2 Install ArgoCD

• create argocd namespace
  kubectl create namespace argocd \
• Install ArgoCD using standard installation
  kubectl apply -n argocd -f https://raw.githubusercontent.com/argoproj/argo-cd/stable/manifests/install.yaml

1.3 Install Argo Rollouts

• create argo-rollouts namespace
  kubectl create namespace argo-rollouts
• Install using standard installation method
  kubectl apply -n argo-rollouts -f https://github.com/argoproj/argo-rollouts/releases/latest/download/install.yaml
• Install kubectl plugin
  brew install argoproj/tap/kubectl-argo-rollouts

Task 2: Creating the GitOps Pipeline

2.1 Dockerize the application

• Created sample node js webapp v1
• Create production grade Dockerfile
• Added github actions to build docker image and push to docker hub located at lulad7/ai-planet-demo
• github actions workflow file located at .github/workflows/main.yml

2.2 Deploy the application using ArgoCD

• Created namespace ai-planet-demo
  kubectl create ns ai-planet-demo
• Created Deployment.yaml and Service.yaml - NodePort inside deployment folder
• Change default admin password
  argocd account update-password
• Created a project on argocd named ai-planet-demo \
• Added https://github.com/leulad/ai-planet-devops.git as a scoped repository inside the project
• Created and connected the Repository
• Added the application ai-planet-demo to the project ai-planet-demo

  Challenge: Unable to run amd64 version on mac m2 requiring arm64 image,

  • Fix:- added platform: arm64 to github actions main.yml file on build and push docker image step
• Application ai-planet-demo with image version ai-planet-demo:1.0.2 running succesfully AI Planet version 1

Task 3: Implementing a Canary Release with Argo Rollouts

3.1 Define a rollout strategy

Modified deployment.yaml

First Rollout

• changed from Deployment type to Rollout
• added canary strategy to deployment file with 1 minute interval between each step

  strategy:
          canary:
          steps:
          - setWeight: 20
          - pause: {}
          - setWeight: 40
          - pause: {duration: 1}
          - setWeight: 60
          - pause: {duration: 1} 

• Using exisiting version for first deployment ai-planet-demo:1.2
• Push to repository to main branch and trigger github actions
• At the same time Wait for argocd to detect the change in manifest files and redeploy the rollout
• Run command to inspect rollout
  kubectl argo rollouts get rollout ai-planet-demo-rollout -n ai-planet-demo --watch
• Rollout successfully deployed for first time without waiting or approval from operator

3.2 Trigger a rollout

Second Rollout

• Make modification to we app change to AI Planet Version 2
• Change Github actions file to build image version ai-planet-demo:1.3
• Modify Rollout Deployment file to use image version ai-planet-demo:1.3
• Push repository to main branch and trigger github actions automatically
• At the same time Wait for argocd to detect the change in manifest files and deploy revision2 of the rollout automatically

3.3 Monitor the Rollout

Run command:
kubectl argo rollouts get rollout ai-planet-demo-rollout -n ai-planet-demo --watch

• First only set to 20% trafic and waits for operator approval
• Approve and promote the rollout
  kubectl argo rollouts promote ai-planet-demo-rollout -n ai-planet-demo
  then: 40% -> 1 minute pause -> 60% -> 1 minute pause -> 100% rollout

Task 4: Document and Cleanup

• Documentation Comleted alongside working on project
• Beautify Readme.md file

4.2 Clean Up

• Delete the application and project from ArgoCD
• Run command to uninstall argo rollouts and delete namespace
  kubectl delete -n argo-rollouts -f https://github.com/argoproj/argo-rollouts/releases/latest/download/install.yaml
kubectl delete ns argo-rollouts
• Run command to uninstall argo rollout plugin brew uninstall argoproj/tap/kubectl-argo-rollouts
• Run command to uninstall ArgoCD and delete namespace
  kubectl delete -n argocd -f https://raw.githubusercontent.com/argoproj/argo-cd/stable/manifests/install.yaml
kubectl delete ns argocd
• Run command to delete ai-planet-demo namespace
  kubectl delete ns ai-planet-demo
• Run command to delete minikube cluster profile
  minikube delete
• Follow recommended procedures to disable Github Actions and Delete repository if required