DevOpsArgoCD+ArgoRollouts Assignment Process

Task 1:

I created this repository to host both the source code of the application and also the kubernetes manifests.

First created a simple web application using Flask and python
Built a docker image using 'docker build' command (Find configuration in Dockerfile)
Pushed it to the repo tejdocker32/argocd on DockerHub
Included the docker repo in kubernetes manifests
Completed ArgoCD setup and used decoded credentials to login
Added necessary YAML files in the 'deployment' directory and added this path in application.yaml
Monitored the application deployment status on the cluster through the dashboard

Changed application's deployment to use the files in 'rollout' directory, specifying canary release strategy with 8 steps
Initial docker image is 'v1'.
To trigger a rollout changed the deployment image to 'v2'.
Monitored the rollout as they followed all the steps mentioned in the mainfests.

I had trouble with running the Flask application as there is a conflict with the versions being used.
Solution: Had use the flask version 2.2.2 + werkzeug version 2.3.7 to solve the problem
I had no idea the workflow of Kubernetes, minikube, ArgoCD
Solution: Watch several youtube videos to grasp the concepts and install necessary components
Had trouble deciding between NodePort and LoadBalancer for argoCD server
Solution: Decided to go with NodePort as it is less complicated
Had trouble working with argo rollouts as there are limited resources about it.
Solution: Had to workout the solution though a punishing trial-and-error series.

docker prune/ docker rmi commands help with removing any untagged or older versions of images
If we use the automatic syncing option for argoCD, we can opt the 'Prune resources' option to remove any unwanted resources in the cluster.