nxd010/k3s-infrastructure

This project is a complete, documented example of building a production style Kubernetes infrastructure entirely on your own local hardware, rather than in the cloud. It uses Hyper-V, the virtualization feature built into Windows, to create three virtual machines that form a small Kubernetes cluster using K3s, a lightweight version of Kubernetes, with one control plane machine and two worker machines. On top of that cluster, the project sets up a GitOps deployment pipeline with ArgoCD, meaning that pushing new code to a Git repository automatically triggers a deployment to the cluster with no manual steps, plus a full monitoring stack using Prometheus to collect metrics and Grafana to display them as dashboards. The project also demonstrates automatic scaling. Under a load test of fifty thousand requests, the cluster handled a peak of around three thousand four hundred requests per second with zero failed requests and a mean response time under sixty milliseconds, automatically scaling the running copies of the application up to ten pods as demand increased. Rather than just describing the setup, the README is written as a full step by step guide to reproduce it yourself starting from nothing. It walks through enabling Hyper-V and creating a private virtual network, provisioning three Ubuntu Server virtual machines with a PowerShell script, installing K3s on each machine to form the cluster, setting up a local Docker image registry, building and deploying a sample application, installing ArgoCD for automated Git based deployments, and finally installing the Prometheus and Grafana monitoring stack. Each step includes the exact commands needed, from PowerShell scripts on the Windows host to shell commands run inside each Ubuntu virtual machine. To follow this guide yourself you would need a Windows 11 Pro machine with at least sixteen gigabytes of RAM, one hundred gigabytes of free storage, and a processor with virtualization support enabled, since the whole cluster runs as virtual machines on that single computer. This is best suited for someone who already has some familiarity with servers, networking, and command line tools, since the guide assumes comfort working in PowerShell and Linux terminals.