The home lab started as a learning environment (a Lenovo TS140 in a closet running ESXi) and grew into a multi-node platform for experimentation. These essays cover the builds, the creative projects, and the philosophy of keeping hobbies unmonetized. Alongside the infrastructure: bartop arcade cabinets, 3D printing experiments, and whatever else my brain latches onto when the kids are asleep and the work laptop is closed.
Infrastructure teams build tools because their blast radius is enormous. When your work touches every application in the organization, the leverage of a good tool (or the damage of a bad one) scales differently than it does for an application team.
I didn't build my home office all at once. It evolved piece by piece over a few years, driven by actual pain points rather than aspirational Pinterest boards. Here's what I'm running, what I'd buy again, and what I'm still evaluating.
The homelab isn't about saving money on cloud hosting. It's about having a space where I can build, break, and iterate without a ticket, a timeline, or a stakeholder.
Palworld hit 1.0 this week and I wanted back in. A Proxmox VM, an Ubuntu 26 template, Docker Compose, and about an hour of troubleshooting DNS quirks later, I had a dedicated server running in the basement.
The moment you start thinking about ROI on a hobby, you've turned something you love into something you owe. I've watched it happen with photography, copywriting, and smoking meat, and the pattern is always the same.
My kid builds redstone contraptions with no spec and no deadline. My engineers build their best prototypes under the same conditions. The pattern isn't a coincidence.