I have 2 of my nodes running on hardware I already had and was already on 24/7. A 12-bay Synology DS3617xs and a 4-bay Drobo 2nd gen.
I’m using spare space on those two devices. Their normal use is mostly backup and all with low performance requirements. The only difference to me is whether the disks are busy or not, which is probably fairly negligible.
I have one external USB enclosure with a 2TB disk running as a third node as well. That does pull power just for running a node, but it’s probably fairly cheap.
Both the Drobo and the external USB drive are connected to the Synology, which runs all three nodes on the same CPU. So that’s about as efficient you can get I guess.
The other side of it being economical is that I started with spare HDD space I already had. But I’ve since bought 2x 16TB HDD’s from Storj income alone. Since my node is only using about 13TB atm, it already paid for it’s own storage twice over. At this point I’m actually paying for the expansion I need for personal uses with Storj as well. Because of that I can make sure I always have enough disk space, no matter how much Storj wants to use.
The nodes are currently the following sizes:
- Syno: 24TB
- Drobo: 1.5TB
- Ext.USB: 1.9TB
But both the Syno and Drobo can be expanded and both will be when Storj starts using more space.
The external USB node is actually running on an HDD that was marked as failed by my Drobo. However it seems to be working fine. The Drobo registers serials of HDD’s it rejected and refuses to accept it back. I’m guessing it had a URE or something and the Drobo is just very strict about that. I ran a ton of SMART tests on it and surface scans and it seems to be completely fine. Since it’s too small to adopt in the array in my Synology, I figured I might as well put it to some good use as a separate node. It’s been running without a single audit failure for 3 months now. I don’t expect it to fail soon. But if it does, it doesn’t matter too much. It’s a small node anyway and I have 2 more to take over from it.