I’m running my media server with a 36tb raid5 array with 3 disks, so I do have some resilience to drives failing. But currently can only afford to loose a single drive at a time, which got me thinking about backups. Normally I’d just do a backup to my NAS, but that quickly gets ridiculous for me with the size of my library, which is significantly larger than my NAS storage of only a few tb. And buying cloud storage is much too expensive for my liking with these amounts of storage.

Do you backup only the most valuable parts of your library?

  • Imprint9816@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    0
    ·
    edit-2
    2 months ago

    I admit this is not a helpful answer but…

    If you want to have hundreds of gigabytes or more of media storage plus backups, its going to be expensive. There is no secret cheap way.

    This is what makes debrid options so appealing. You can amass terabytes of media data for a cheap monthly cost.

    You can then supplement that with a small nas or drive of rare or hard to find media / offline selection in which case you could probably run raid 10 with the small amount that you would actually need to backup.

  • hydrogen@lemmy.ml
    link
    fedilink
    English
    arrow-up
    0
    ·
    2 months ago

    “Cloud” storage is indeed more expensive. But depending where you live, you count the electricity cost in, and you use the storage ‘only’ for backups. Maybe it makes more sense to pay for remote storage in a datacenter. Check out Hetzner Storage Box, it’s what I use.

    If it’s still too expensive, maybe ask a friend or family member (maybe someone that uses your media) to setup a nas at their home for backup purpose. (I use this for my media)

    Make sure you encrypt your backups if you use a remote location for your backups.

    You have to decide what’s valuable for you. For me my media is, I can just download everything again, but the time I put in to have every movie the correct subtitles without ads, the correct posters, metadata etc. I value my time, I don’t want to do it again if I loose everything.

  • jabjoe@feddit.uk
    link
    fedilink
    English
    arrow-up
    0
    ·
    2 months ago

    Remote storage (Pi at parents house with a big disk) and cron’ed btrfs send over ssh.

    • jay@mbin.zerojay.com
      link
      fedilink
      arrow-up
      0
      ·
      2 months ago

      You really shouldn’t trust anything important to a pi. I hope that you at the very least have that pi on a UPS if you’re going to risk your data this way.

      • jabjoe@feddit.uk
        link
        fedilink
        English
        arrow-up
        0
        ·
        2 months ago

        It’s a backup. On the main machine there are two disks (fast & big and slow & smaller) not in raid, with a btrfs copy.

        It would be quite an event to lose all three copies.

  • Azzu@lemm.ee
    link
    fedilink
    English
    arrow-up
    0
    ·
    2 months ago

    I feel like other people seeding is enough backup for me. I don’t backup my library at all.

    • ShortN0te@lemmy.ml
      link
      fedilink
      English
      arrow-up
      0
      ·
      2 months ago

      Have you read upon the actual recovery experience with cloud backup personal?

      It is very impractical to do a recovery from this service.

      • XNX@slrpnk.net
        link
        fedilink
        English
        arrow-up
        0
        ·
        2 months ago

        How so? You can download it all or download specific files or folders

        • ShortN0te@lemmy.ml
          link
          fedilink
          English
          arrow-up
          0
          ·
          2 months ago

          So this is just from the top of my head, there are several reddit threads about this.

          You can only bulk download in 5gb (maybe 50gb) chunks. With limited speed and unreliable connections (aborted downloads etc)

          Have you ever tried downloading a big size of your data? Never used it myself.

          • XNX@slrpnk.net
            link
            fedilink
            English
            arrow-up
            0
            ·
            2 months ago

            Oh i haven’t heard of that. I haven’t downloaded huge chunks no. I think you can connect with clients that allow pausing and continuing can’t you?

      • lightnsfw@reddthat.com
        link
        fedilink
        English
        arrow-up
        0
        ·
        2 months ago

        I’d have to double check but I believe that only works on windows and I know for sure you can’t use it for network shares. I have it on my windows desktop but wasn’t able to backup my media server with it.

        • Arn_Thor@feddit.uk
          link
          fedilink
          English
          arrow-up
          0
          ·
          2 months ago

          Doesn’t work for network shares as far as I can tell. At least not without some funky magic

      • ancoraunamoka@lemmy.dbzer0.com
        link
        fedilink
        English
        arrow-up
        0
        ·
        2 months ago

        I mean… do the math and you can figure out by yourself that it’s a fair price but in no way some sort of very convebient situation for the users. A 20tb hard drive goes for about 450€ and then you can consider the advantages that they have buying hdd at scale.

  • randomaside@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    0
    ·
    2 months ago

    Data hoarding is a truly unique experience. Just my two cents

    • raid is not a backup. Don’t use raid5 unless you’re using a filesystem like zfs that checksums your data. Raid5 is vulnerable to scenarios with a “write hole” that leads to bit rot.

    • split up your dataset into smaller more manageable datasets so you can more easily back it up in different ways like external drives, cloud storage, etc. You can then limit the dataset size to never exceed the same of your backup target.

    • snapshots, use them. Snapshots in your filesystem can make your backups more manageable by only sending the differential data as opposed to something like Rsync which may need to rsync an entire file.

    I use ZFS and have found that compression with ZSTD works pretty well for getting extra use out of your disks but unless you have a lot of RAM and some special metadata NVME disks, don’t use reduplication as it will be a serious performance impact.

    Now if you aren’t using a FOSS system like truenas and instead you’re using a system like a qnap off the shelf, the qnap hybrid backup and sync manager has a really elegant solution for doing policy based differential backups to back blaze b2 storage. Not only does this give you a copy of your data, you also get immutable points in time archives of your data.

    Good luck in your data hoarding endeavors!

    • WhyJiffie@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      0
      ·
      2 months ago

      Important clarification: snapshots only make backups faster if you do backups with zfs send, or with other filesystems they have the appropriate command too.

  • figjam@midwest.social
    link
    fedilink
    English
    arrow-up
    0
    ·
    2 months ago

    My library is not nearly as large as yours but I’ve got the things that i can get back over time like movies and shows and stls and I’ve got the things that i can’t get back cached elsewhere.

  • Boomkop3@reddthat.com
    link
    fedilink
    English
    arrow-up
    0
    ·
    2 months ago
    • automated scheduled copies
    • syncing data to multiple machines
    • scheduled copy to the cloud (not particularly frequent tho)
  • ChonkaLoo@lemmy.zip
    link
    fedilink
    English
    arrow-up
    0
    ·
    2 months ago

    I have a couple of external USB drives I bought on sale I backup my NAS to once a week. It’ll protect against drive failure at least. Almost got hit by ransomware a while back so I don’t keep anything on there without some backup.

  • cmnybo@discuss.tchncs.de
    link
    fedilink
    English
    arrow-up
    0
    ·
    2 months ago

    You can put a big hard drive in an external enclosure and use it for offline backups. There is no point in paying for cloud storage for something you can just download again if needed. Save the cloud storage for backing up non-replaceable data.

  • LiPoly@lemmynsfw.com
    link
    fedilink
    English
    arrow-up
    0
    ·
    2 months ago

    Personally, I back up everything on my NAS except my movie library, because that is something I can relatively easily restore by just downloading buying it again, and because it’s of course the biggest chunk of data. For the other data, I’m using a very affordable Hetzner server auction system with a lot of disks in a striped array. This gives me the maximum amount of storage, and given that I can just create the backup again should the stripe fail, I’m not worried about redundancy on the backup itself.

  • Chadus_Maximus@lemm.ee
    link
    fedilink
    English
    arrow-up
    0
    ·
    edit-2
    2 months ago

    I save the documents in a usb and format all my drives. Can’t have backup troubles if you never back anything up.