Choosing a storage server requires deciding on the operating system...
Choosing a storage server requires deciding on the operating system...
I've decided to repurpose my old tower as a storage server. I own three relatively new 6TB drives housed in external enclosures, but I've grown tired of the need for separate drives that sometimes hold duplicates. Now I prefer a single JBOD with everything consolidated. I also have five 2TB disks (one likely non-functional, let's say four remain), plus several 1TB or smaller drives. My goal is to find a setup that offers solid storage, prioritizing capacity over speed, and includes redundancy. These aren't brand-new units, but I need something easy to expand. I'm open to swapping in newer, larger drives if space allows, though I have a limit on how many can fit. I started with TrueNAS, which doesn’t seem to support the flexibility I require, possibly because of configuration issues. I downloaded TrueNAS Scale and thought it might work, but I’m concerned about managing the drive array effectively. Ideally, a system would allow bad sectors to be detected and automatically removed from the JBOD, with capacity adjusting dynamically—adding or removing drives should seamlessly update the total storage. Since rebuilding the entire server isn’t an option, I’m unsure if such a feature exists or how to implement it. If you know of any solutions or guides that address this need, I’d really appreciate your help. Thanks!
I believe the focus should be on "wizardry" rather than just technology for a storage system that operates this way. Even RAID0-like setups will fail if a drive goes down, so there are various methods to link more drives together to form a larger unit. Each approach has its own constraints (or limitations from today’s view). Also... even in a massive tower configuration, a single 1TB hard drive consumes unnecessary power. It’s not the storage quality that matters, but rather the size of the drives themselves. If you’re limited to one 1TB drive, make the most of it; otherwise, skip investing in such small units.
Consider exploring unraid since it seems to be the best option for managing numerous drives of varying sizes. Still, be aware that data redundancy can be affected if you don’t monitor closely for mistakes. Older storage devices are more likely to fail, so keep this in mind—always safeguard critical information elsewhere. Also note that your motherboard might lack sufficient SATA/EIDE ports for all the drives, which could require a quality host bus controller. You might find it worthwhile to purchase a suitable 4 or 6 TB drive instead of multiple smaller ones. I maintain backups myself using cold storage in an HDD case (currently holding eight old drives ranging from 320 GB to 1 TB). I’ve compiled a list of their contents. The drawback here is limited access and the need to connect them periodically to preserve data quality. For any NAS system, aim for identical disk sizes from the same brand or at least matching rotational speeds. If feasible, opt for datacenter-grade drives, particularly if the system will run continuously. For occasional backups, your HDDs could suffice, but most NAS operating systems struggle with mixed-sized HDDs, especially regarding data reliability.