I wanted a sanity check on my current rsync flags. Posts on Reddit seem to highlight the use of rsync -avz for most use cases, for some instances even for when someone asks for mirroring a drive: https://www.reddit.com/r/DataHoarder/comments/rau071/rsync_command_to_mirror_drive/. This has not worked for me for the following case:

Drive #1:

file1.txt
file2.txt

Drive #2:

file1.txt
file2.txt
file3.txt

With -avz, file3 on the destination would not be deleted. With experimenting, I ended up now using rsync -havziP --delete-after --info=progress2 dir1/ dir2/, which actually ended up mirroring the drives for me. My question is: is this the best rsync approach for mirroring drives or was there a better option that works better?

Side note: it is interesting that rclone sync from rclone (https://rclone.org/commands/rclone_sync/) claims to delete by default, while with rsync it seems to be something you have to distinctly mention.

  • limelight79@lemmy.world
    link
    fedilink
    arrow-up
    1
    ·
    10 hours ago

    You have your answers, but I just want to note that what I do is run rsync from cron nightly, without the delete option.

    Basically this reduces the chances of losing something by accidentally deleting it on the primary drive. Every few weeks/months, I run the rsync command with the delete option to clean up.

    It’s not a perfect backup by any means, especially since I could erroneously delete something right before that second rsync command and lose it for good. But it does provide a measure of safety.

  • IsoKiero@sopuli.xyz
    link
    fedilink
    English
    arrow-up
    10
    ·
    2 days ago

    There’s also --delete-before which might help if your destination is tight on available space. And, as usual with ‘traditional’ tools, man-page is pretty good, there’s a ton of parameters which might be helpful. And, as @hades@feddit.uk already mentioned, compression (-z) may actually hurt performance if you have a lot of bandwidth or if you’re copying over already compressed data like JPGs.

    • Harisfromcyber@lemmy.mlOP
      link
      fedilink
      arrow-up
      2
      ·
      1 day ago

      I wasn’t aware of the full impact of -z. Thanks for clarifying that for me (thanks to @hades@feddit.uk as well!). I’ll definitely keep an eye on the speed with -z vs without it when I do my next dry run.

      • IsoKiero@sopuli.xyz
        link
        fedilink
        English
        arrow-up
        2
        ·
        20 hours ago

        If I’m not mistaken, dry-run (-n) doesn’t give you the results as it doesn’t actually transfer nor compress data. It’s only useful to verify that the command you’ll testing will actually do what you want. For performance measurement you can use --progress and --stats.

      • BCsven@lemmy.ca
        link
        fedilink
        arrow-up
        1
        ·
        edit-2
        1 day ago

        Graphic interface, you just click options. And it has drop downs for favourite folders you’ve configured before so you can switch between backup/sync locations.

        Once you apply, it runs rsync, and it it shows you the CLI command used as visual reference