A bit more detail: simply being powered on won’t necessarily stop that.
You want something checksumming the data and making sure it’s not silently rotting off the disk.
ZFS does this, something like snapraid can do it too, and there’s various other methods of making checksums you can validate data integrity with and be able to repair minor corruption. (PAR files, for example.)
A real-world example of this kind of oops is everyone’s favorite Youtube Tech Personality™ LTT who lost a fuck-ton of data due to not scrubbing data on a ZFS array and had to go through months of restoration to get most of it back, so uh, yeah, make sure you’ve taken steps to detect and correct the bitrot that’s going to happen anyways.
Yeah, that wasn’t meant to be remotely comprehensive: there’s a lot of ways you can do this ranging from the filesystem to what kind of archives you’re storing, to programs that make parity data for validation.
…also, since I haven’t started a flamewar yet today, I don’t think I’d personally use BTRFS. It’s still too new, has had data consistency issues too recently, and just plain doesn’t have the kind of historical performance record for something I’d want to use for archival purposes.
Come back in another decade and we’ll see how it’s been going.
I know SSDs need to be powered on occasionally due to how they store data, which is why (also due to cost lol) I have most of my stuff on HDDs, though I know those can have issues if you leave them out unprotected.
Not much more I can do without spending a lot more money than I have already. But so far I’ve never needed to get my second backup.
Which reminds me I need to re-backup the second drive…
Beware of bit rot, hard drives are meant to be powered occasionally to hold data. Using a recycled computer as a NAS is a great low cost solution.
A bit more detail: simply being powered on won’t necessarily stop that.
You want something checksumming the data and making sure it’s not silently rotting off the disk.
ZFS does this, something like snapraid can do it too, and there’s various other methods of making checksums you can validate data integrity with and be able to repair minor corruption. (PAR files, for example.)
A real-world example of this kind of oops is everyone’s favorite Youtube Tech Personality™ LTT who lost a fuck-ton of data due to not scrubbing data on a ZFS array and had to go through months of restoration to get most of it back, so uh, yeah, make sure you’ve taken steps to detect and correct the bitrot that’s going to happen anyways.
Just as an addition, nothing against ZFS it is a great FS, but BTRFS can this too.
Yeah, that wasn’t meant to be remotely comprehensive: there’s a lot of ways you can do this ranging from the filesystem to what kind of archives you’re storing, to programs that make parity data for validation.
…also, since I haven’t started a flamewar yet today, I don’t think I’d personally use BTRFS. It’s still too new, has had data consistency issues too recently, and just plain doesn’t have the kind of historical performance record for something I’d want to use for archival purposes.
Come back in another decade and we’ll see how it’s been going.
All fair and good arguments, so no flamewar today. Sorry 😜
I know SSDs need to be powered on occasionally due to how they store data, which is why (also due to cost lol) I have most of my stuff on HDDs, though I know those can have issues if you leave them out unprotected.
Not much more I can do without spending a lot more money than I have already. But so far I’ve never needed to get my second backup.
Which reminds me I need to re-backup the second drive…