Migrating to Unraid

In 2013, the home server problem looked simple from the outside: build a box, add disks, put files on it, and stop thinking about storage. In practice, the decision came down to what kind of system I wanted to live with for the next several years.

FreeNAS was the obvious serious option. It had ZFS, a strong reputation, and a storage model that appealed to anyone who cared about data integrity. The downside was that ZFS wanted planning discipline up front. That is not a bad thing in an enterprise storage design, but it was awkward for a home lab that was going to grow one disk at a time.

Unraid fit the way the system was actually going to evolve.

Mixed Disks Mattered

The biggest factor was disk flexibility. With FreeNAS and ZFS, capacity expansion worked best when the pool was designed around matched disks and upgraded in groups. If the pool was built from smaller drives, replacing one disk with a larger disk did not immediately give the pool more usable space. The practical path was to upgrade every disk in the vdev before the larger size really paid off.

That made sense technically, but it was not how I bought drives. I was not going to replace a full set of disks every time prices dropped or a larger drive became affordable.

Unraid let me mix disk sizes and upgrade incrementally. Add a disk, replace a disk, grow over time. As long as the parity disk was large enough, the array could evolve without treating every storage upgrade like a full rebuild plan. For a home server, that was the deciding point.

The UI Was Easier To Live In

FreeNAS felt powerful, but Unraid felt approachable. The web UI made the common tasks obvious: shares, disks, parity, users, and system status were all easy to find. That mattered because the server was not supposed to become a second job.

The goal was not to build the most academically correct storage appliance. The goal was to build something dependable that I could maintain quickly, understand at a glance, and hand off to future me without needing to rediscover every design decision.

Docker Changed The Direction

The other thing that pushed Unraid forward was application hosting. At the time, the home server was already becoming more than a file server. Media tools, download automation, web utilities, databases, and small internal apps all wanted a place to run.

Unraid’s Docker support made that model feel natural. Instead of treating applications as awkward add-ons to a storage platform, containers became part of the system’s everyday workflow. Services could be added, removed, updated, and isolated without turning the base operating system into a junk drawer.

That was the beginning of the lab becoming a lab, not just a NAS.

The Tradeoff

There were tradeoffs. ZFS had stronger native data integrity features, and FreeNAS was a better fit for some storage-first designs. If I had been building a fixed disk set with a more formal pool layout, the choice may have gone the other way.

But the requirements were different: mixed disks, gradual upgrades, simple management, and a path toward running more services on the same hardware. For that shape of problem, Unraid was the better fit.

Looking back, the choice held up. The storage array grew over time instead of being replaced in big jumps, Docker became central to how I tested services, and the server gradually turned into the infrastructure workbench that Node804 now documents.