Kubecenter: a K8s Cluster Management Web App

This is a K8S Cluster management web app built with Go. It enables the developers to manage the K8S clusters without creating .yaml files or even runing the commands on the nodes on the servers. Visit Kubecenter to see the Swagger API doc and test the APIs. Prometheus is also integrated for monitoring the cluster CPU/memory usages. Visit the Prometheus UI to see the graphs. Type custer_cpu or cluster_mem to see the graphs. My K8s cluster consists of one master node and two worker nodes.

Language/Framework/Tools/IDE

• Deployment: Docker, Docker Compose, Gogs, Drone, AWS EC2
• Language: Go 1.22.5 darwin/amd64
• API Documentation: Swagger
• Backend Framework: Gin v1.10.0
• IDE: GoLand
• Tools/Middlewares/Libraries:
  • client-go: Go clients for talking to K8S cluster
  • K8s: for creating cluster and recource management
  • Gogs + Drone: for createing an automated CI/CD pipeline
  • harbor: for creating private image registry
  • Go Testing: for unit tests
  • Prometheus: for monitoring cluster CPU/memory usages
  • dotenv: for reading in env variables from .env

Highlights:

• Created a Docker image registry using Harbor and established a CI/CD pipeline using Gogs and Drone, enabling private image storage and automated integration/deployment
• Integrated Prometheus to monitor Cluster CPU/Memory usages, enhancing operational observability
• Built a customized IoC (Inverse of Control) module for dependency injection and decoupling of application components
  • IoC module could be a open-source project. I used it in all my Go projects

Architecture

Not all APIs are included below, visit Kubecenter to see the Swagger API doc.

CI/CD Pipeline: disgram to be updated

Below is just presenting the UI of harbor, gogs, and drone, techinical details will be shared soon.

harbor, gogs, drone

What I'm working on now

• cache: use Redis to cache up the cluster info
• audit service: create/update/delete operations should be audited first, this should be a microservice