Several of my homelab projects want a local GPU for inference — running a vision LLM for OCR, offloading work that would otherwise hit a cloud API, keeping a small model resident and fast. So I built a dedicated box for it: a Dell Precision 3640 (i3-10105F, 16 GB) with a Gigabyte RTX 3060 12 GB, running Ubuntu Server and serving a vision model through Ollama.

Two things made this worth writing up. First, the actual GPU-serving setup has a few non-obvious choices (which driver, no CUDA Toolkit, exposing the port safely). Second — the fun part — I made the whole install zero-touch: plug in a USB stick, walk away, and the box comes up on a static IP with SSH pubkey auth ready, no monitor or keyboard ever attached. This post covers both, and both live as reusable docs/scripts in my infra-config repo.

Two things happened on the same day that looked like problems and weren’t — at least not the problems they appeared to be. My NAS was reporting almost all its RAM in use, which sounds like a leak or an imminent crash. And a TV show I’d grabbed through LunaSea on my phone had downloaded fine but then “wouldn’t import” into the library, which sounds like a broken automation pipeline.

Both are among the most common false alarms in a self-hosted media stack. Neither was what it looked like. Here’s what was actually going on, and the real fixes.