---
title: "The home server I finally stopped turning off"
date: 2026-09-03
description: "How Raspberry Pi experiments became a useful home server, and why learning when not to self-host mattered as much as Docker and DNS."
tags:
  - "self-hosting"
  - "linux"
  - "docker"
  - "home-server"
---

I use my home server all the time now, but for years I kept building one and then turning it off.

Around the start of 2020, I got a Raspberry Pi and repeatedly installed Raspbian or Debian on it. I would add Sonarr, Radarr, maybe Prowlarr, a torrent client and Plex. Sometimes I would run Pi-hole too. There was no reverse proxy, and I was not putting my own domains behind it. It was primitive, but I learnt something each time.

It never stuck, right? The pattern was always: “Oh, I’m going to play with this,” and then I would turn it off.

For me, that was not yet a home server I depended on. It was a computer I sometimes used as a server.

## Doing it properly

In summer 2024, I decided I wanted something that would stay running. I had an old i5 desktop lying around, I knew it worked, and it was easy to connect drives to. Why the hell not?

I installed [OpenMediaVault](https://www.openmediavault.org/) from the start. Over the next two or three months, the setup accumulated the things that had been missing from the Pi experiments: Docker-managed services, the fuller media stack, [Traefik](https://doc.traefik.io/traefik/) as a reverse proxy, [Tailscale](https://tailscale.com/docs/concepts/what-is-tailscale), proper DNS and network sharing.

The useful change was not any one service. I could deploy something new behind HTTPS and decide whether it should be reachable publicly or only inside my network. Two Pi-hole instances gave me two DNS servers. The machine gradually became infrastructure rather than another experiment waiting to be unplugged.

I also bought a separate OptiPlex with 4 GB of RAM and installed Debian on it. Its main job was Plex Pass transcoding. Plex read the media over the network from storage attached to the OpenMediaVault machine, so the little OptiPlex could act as a transcoding helper rather than another full copy of the server. That forced me to work out network shares and deployment across two Linux machines. It also gave me a second place to run DNS.

None of this is exotic, but it joined together parts of computing that application-focused software-engineering jobs often keep separate. Docker, Linux, DNS, storage, reverse proxies and network failures stopped being someone else’s layer. They became problems I had to understand because I wanted my own services to work.

## Knowing when to stop

Not every learning exercise deserved to survive.

That same summer, I tried turning an old laptop into my router with [OpenWrt](https://openwrt.org/). It kept dropping out. While I was learning how the rest of the server worked, I was also repeatedly taking down my own internet.

Eventually I bought a router.

The useful lesson was not that I needed to become better at running OpenWrt. It was that you need to know when to just not do it yourself. I got to the point where I thought: “I just want something that works. I’m going to stop trying to overcomplicate it. I’m just going to make it work.”

That judgement is part of running a home server too. You do not have to turn every available layer into a personal infrastructure project.

## It is still a headache

A home server is not free. It uses energy. Plex Pass costs money, as can a VPN or another service you decide to depend on.

It also fails. Sometimes you want to watch a film and the thing is down, which probably would not have happened with a simpler setup. Self-hosting will cause a fucking headache.

What are you going to do, right? I still think it is worth it.

The value is not that every service is cheaper or more convenient than its hosted equivalent. It is that the server gives me real reasons to learn Docker, networking, Linux, storage and operations. The problems matter because I use the result all the time. It is just good stuff, right?

By 2026, I had a different desktop that I was no longer using because my MacBook had become my everyday computer. I did not have a neat business case for rebuilding the server. I simply had a more capable machine sitting there, with more room to run things, and thought: why not make that the server?

That is where this part of the story ends. The move to Proxmox is the next one.