What are you non-obvious, maybe strange usecases of Syncthing?

For example syncing the media library with your friend or maybe your entire /home/user folder between your PC and laptop?

I’d love to hear your ideas!

  • zaphod@lemmy.ca
    link
    fedilink
    English
    arrow-up
    4
    ·
    edit-2
    1 year ago

    Echoing other comments, my backup strategy for all our devices is: Syncthing to replicate data to my NAS, restic to generate encrypted backups, and then cron+rclone to offsite those backups to Google Drive.

    I absolutely love this setup.

    1. It works anywhere. Syncthing takes care of firewall punching and all that so whether I’m at home or on the road, I know the data is being replicated correctly.

    2. It’s immediate. Syncthing doesn’t run on some schedule. It’s constantly replicating so I know at minimum there’s a copy of all my data if something catastrophic happens.

    3. It’s private, encrypted, and entirely in my control.

    4. The setup is built of composable parts that can each be understood, modified, and debugged easily.

    Normally I’m a little cautious about rolling my own infrastructure for something critical like backups, but this setup is so simple and robust that I just don’t worry about it.

    Other use cases I’ve come up with:

    1. I use Paperless as a DMS. It has a watch folder for automatically ingesting documents. I set up Genius Scan + Syncthing on my phone, syncing scans to the Paperless drop folder, so I can scan from my phone and automatically upload to Paperless without any additional app. Just scan and off it goes.

    2. For a while I was playing Subnautica on my Steam Deck and my gaming rig. Subnautica doesn’t support the Steam Cloud so I used Syncthing to replicate the save data across my gaming devices.

    Not particularly weird or outlandish, but of course I also use Syncthing to replicate my keepass database across devices as well.

    I also have a personal wiki of Markdown notes that I sync between my laptop and phone using Syncthing.

    Oh, and I use it to replicate my Calibre library between my laptop and my calibre-web server.

    Basically it’s my swiss army knife of “I have data over here, I need to get that data over there” and it’s amazing!

    • lemmyvore@feddit.nl
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 year ago

      How does syncthing know what to upload to paperless watch dir if paperless keeps deleting the files after ingesting them?

      • zaphod@lemmy.ca
        link
        fedilink
        English
        arrow-up
        2
        ·
        11 months ago

        That’d exactly what you want! When the file initially lands in the sync folder, Syncthing sends it to paperless. Paperless ingests it, deletes it, and it disappears from my phone, now stored in paperless. Exactly what I need.

        If I wanted the files to stay on the phone I’d set up the phone as Send Only and the paperless side as Recieved Only.

        • lemmyvore@feddit.nl
          link
          fedilink
          English
          arrow-up
          1
          ·
          edit-2
          11 months ago

          Oh you delete them from the phone too, ok that makes sense.

          I don’t delete them from the phone (because they’re also in a different app I used since before Paperless) so that’s a bit of a dilemma. I’ll have to think about it better.

          • zaphod@lemmy.ca
            link
            fedilink
            English
            arrow-up
            2
            ·
            11 months ago

            Again, Syncthing supports one-way sync so allowing paperless to delete them and having that delete sync back to the phone is entirely optional.

            • lemmyvore@feddit.nl
              link
              fedilink
              English
              arrow-up
              1
              ·
              11 months ago

              I mean, how does Syncthing know not to copy a file again if it copied it once and paperless deleted it?

              • zaphod@lemmy.ca
                link
                fedilink
                English
                arrow-up
                2
                ·
                edit-2
                11 months ago

                The client on the sender side (the phone) knows it sent the file. It doesn’t care if the receiver side changed or deleted it. It sent the file. Its job is done. That’s why the mode is called “Send Only”.

                Meanwhile the client on the receiver side (my NAS) never pushes changes back. It only responds to received sync instructions. That why the mode is called “Receive Only”.

                It’s… all pretty simple. Not sure where the confusion lies?

                • lemmyvore@feddit.nl
                  link
                  fedilink
                  English
                  arrow-up
                  2
                  ·
                  11 months ago

                  The phone sync client would have to remember all the files it ever sends, which could be thousands. I’ve never seen a sync client that works like that, they usually compare the files that are in the source and destination folders. If syncthing can do this that’s really interesting.

                  • zaphod@lemmy.ca
                    link
                    fedilink
                    English
                    arrow-up
                    2
                    ·
                    edit-2
                    11 months ago

                    No it doesn’t.

                    Syncthing only needs to remember the current state of the files/folders it’s syncing. Not everything it’s every sync’d.

                    It does that by either periodically scanning the filesystem to look for changes since it last scanned (based on the file creation and modification dates that are stored in the filesystem), or it registers with the operating system to receive events when files are created, modified, or deleted.

                    When Syncthing notices a create, update, or delete, it pushes those changes to the receiver and then updates it’s record of the filesystem state accordingly.

                    It also pushes whole files, not deltas. So it doesn’t care how the files changed, only that they did.

                    Even with hundreds of thousands of files to sync this is a relatively small amount of state as it’s just file paths and their create/modify dates.