Multiple drives of a lower (1-2tb) capacity might be more expensive, but they’d technically be more resistant to a single failure over fewer larger (4tb+) drives when in a pool/array.

  • The Pantser@lemmy.world
    link
    fedilink
    English
    arrow-up
    0
    ·
    6 days ago

    More drives mean more heat and more power usage. Better to have two large with parity enabled. The chances of both failing is very rare. Should be using an online backup service like backblaze or having a plan to take a copy off-site.

    • merthyr1831@lemmy.mlOP
      link
      fedilink
      English
      arrow-up
      0
      ·
      6 days ago

      Yup, already keep an off-site copy of the important stuff via Hetzner but my Linux ISOs are the big data hogs

    • schizo@forum.uncomfortable.business
      link
      fedilink
      English
      arrow-up
      0
      ·
      6 days ago

      The chances of both failing is very rare.

      If they’re sequential off the manufacturing line and there’s a fault, they’re more likely to fail around the same time and in the same manner, since you put the surviving drive under a LOT of stress when you start a rebuild after replacing the dead drive.

      Like, that’s the most likely scenario to lose multiple drives and thus the whole array.

      I’ve seen far too many arrays that were built out of a box of drives lose one or two, and during rebuild lose another few and nuke the whole array, so uh, the thought they probably won’t both fail is maybe true, but I wouldn’t wager my data on that assumption.

      (If you care about your data, backups, test the backups, and then even more backups.)