At work, there happened to be a Seagate BlackArmor 220 NAS device floating around uncommitted. I suggested that I might bring it home and test it a bit. That happened.
The primary driver for looking into this kind of appliance is that I’m currently running a mid-tower Linux box all the time … and only personally using it rarely. It does, however, also function as the house DNS server, the home SMB server, and the backup central host. It fires off scripts to connect to remote systems (like the virtual host for this place) and pull down backups, too.
That system draws the same as a hundred watt bulb, full time. By way of contrast, that BlackArmor device draws between 17 and 20 watts, depending on disk activity. That’s what I measured with my Kill-a-Watt meter.
One of the other things I want from a home storage system is the ability for the Apple gear to use it properly, as an iTunes server as well as a Time Machine backup disk. The tricky bit about that latter is that Apple changed the AFP and discovery protocols a bit, so that gear (networked drives) that worked as Time Machine destinations under Leopard and earlier OS X releases stopped working when Lion tries to connect. I did a bunch of reading, and found that the BlackArmor appliance was a suitable Time Machine target.
Turns out that’s true with fresh firmware for the 220 from just a month or so ago, which I found on the Seagate site. There are similar firmware updates for other BlackArmor models. Once updated, I was able to get Time Machine up and working with Agog, the new Mac Air. I also tested files on and off using CIFS – that worked fine, too. So it would be a suitable target for Marcia’s backups. What doesn’t work for me is that I can’t trivially script activities for the system to pull backups, which is a crucial part of what I want such a system to do.
Most of these appliance-style NAS devices aren’t configurable or controllable to that level, although I’ve found that there’s ways to expose the underlying Linux in the BlackArmor. The upside of an appliance is that it’s already working properly, and it’s relatively inexpensive.
I could buy components, and build a low-power system. I’m thinking that maybe the next release of FreeNAS might do the trick with that. And I can do a lot more with a fully-accessible FreeBSD-based OS than with a hacked Linux image on an underpowered appliance. Figure that the Linux box is costing me on the order of a hundred bucks a year for electricity. If it costs me $400 to build a 20 watt replacement system, the build pays for itself in … 5 years. Which is about the life of the system. That ain’t bad. I wonder if I can build what I want for $400?
* * *
[[Warning: Pedantry]] You see that I used the Oxford comma in the title. I still do most of the time, although not always. I’m not even sure if consistency is called for. Ah, well.