I’ve recently embarked on a new project: building a K3s cluster to learn Kubernetes from the ground up and to begin self-hosting various services. This post documents the initial setup and my progress so far.

The Hardware and Virtualization Stack

To keep the physical footprint and power consumption low, I opted for a single, efficient Mini-PC as the foundation for the entire cluster.

  • Host Machine: Mini-PC with an Intel N100 CPU and 16GB of RAM
  • Hypervisor: Proxmox
  • Virtual Machines: 3 nodes running Ubuntu Server
  • Kubernetes: A 3-node K3s server cluster

While running a multi-node cluster on a single physical machine isn’t ideal, it’s a fantastic and cost-effective setup for a learning environment.

Progress and Source Code

The cluster is up and running, and I’ve started documenting the process. You can find the shell scripts for provisioning and the Kubernetes YAML manifests I’m developing in my K3s GitHub Repository.

Next Steps

The project is still evolving. My immediate goals are to stabilize the setup and begin deploying my first applications. I’ll be sharing updates as I work through challenges and reach new milestones.

edited by Gemini :)