KaaS

KaaS (Kubernetes as a Service) is a self-service platform written with Golang that allows users to deploy, monitor, and manage their applications and databases effortlessly.

Technologies used

• Golang, Programming language.
• Echo, HTTP web framework.
• PostgreSQL, Database management system.
• Minikube, Tool for running Kubernetes locally.
• client-go, Go client for Kubernetes.
• Helm, Kubernetes package manager.
• Docker, Containerization platform.

Features

• Self-Service Deployment: Users can deploy their applications and databases with a simple request.

• Database Management: Automates the creation and management of PostgreSQL databases.

• Health Monitoring: Regularly checks the status of deployed applications and records their health metrics.

Getting Started

Prerequisites

• Golang
• Docker
• Kubernetes cluster

Installation

Build the Docker image:

docker build -t <image> -f build/kaas/Dockerfile .

Deploy to your Kubernetes cluster using Helm:

helm package kaas-api

helm install kaas-api-release ./kaas-api-0.1.0.tgz

Configuration

Modify the values.yaml file to set your desired configurations, such as replica counts, image repositories, and database settings.

Monitoring

To monitor the health of an application, I use a CronJob that regularly checks the application's health status every 5 minutes. You can still change this interval as per your requirements.

The health of each application is monitored using HTTP GET requests to the root endpoint (/). If the response status is 200, the application is considered healthy. Monitoring results are stored in the PostgreSQL database.

Examples

Deploying an Application

To deploy an application, send a POST request to the /application/create endpoint with the following JSON payload:

{
  "AppName": "example-application",
  "Replicas": 3,
  "ImageAddress": "your-docker-repo/your-app-image",
  "ImageTag": "latest",
  "ServicePort": 8080,
  "Resources": {
    "CPU": "500m",
    "RAM": "256Mi"
  },
  "Envs": [
    {
      "Key": "DB_USER",
      "Value": "db_user",
      "IsSecret": true
    },
    {
      "Key": "DB_NAME",
      "Value": "db_name",
      "IsSecret": false
    }
  ]
}

Check Application Status

To check the status of a specific application, send a GET request to the /application/status/{appName} endpoint, replacing {appName} with the name of your application. Example:

{
  "DeploymentName": "example-application",
  "Replicas": 3,
  "ReadyReplicas": 2,
  "PodStatuses": [
    {
      "Name": "example-application-pod-1",
      "Phase": "Running",
      "HostIP": "192.168.49.2",
      "PodIP": "10.244.1.4",
      "StartTime": "2024-06-07T12:30:00Z"
    },
    {
      "Name": "example-application-pod-2",
      "Phase": "Running",
      "HostIP": "192.168.49.2",
      "PodIP": "10.244.1.5",
      "StartTime": "2024-06-07T12:32:00Z"
    },
    {
      "Name": "example-application-pod-3",
      "Phase": "Pending",
      "HostIP": "",
      "PodIP": "",
      "StartTime": ""
    }
  ]
}

Deploying PostgreSQL as a Self-Service

To deploy a PostgreSQL database, send a POST request to the /service/postgres endpoint with the following JSON payload:

{
  "AppName": "example-postgres",
  "Resources": {
    "CPU": "500m",
    "RAM": "1Gi"
  },
  "External": true
}

Monitoring an Application

To monitor the health of an application, send a GET request to the /applicartion/health/{appName} endpoint, replacing {appName} with the name of your application. Example:

{
  "AppName": "example-application",
  "FailureCount": 0,
  "SuccessCount": 3,
  "LastSuccess": "2024-07-01T12:30:00Z",
  "LastFailure": ""
}