XCP-ng
    • Categories
    • Recent
    • Tags
    • Popular
    • Users
    • Groups
    • Register
    • Login

    scheldued snapshots and manual snapshots

    Scheduled Pinned Locked Moved Xen Orchestra
    6 Posts 3 Posters 954 Views 1 Watching
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • akurzawaA Offline
      akurzawa
      last edited by

      hi

      is this safe to have sheldued snapshots enabled (7 automatic dialy snaps) and some old manual snaps? how does xcp-ng manages the old snap?

      if the manual snap would be like three months old could it be dangerous? inconsistency problems and so on?

      f2773467-b8eb-48b0-bf62-70522f451b68-image.png

      1 Reply Last reply Reply Quote 0
      • olivierlambertO Offline
        olivierlambert Vates 🪐 Co-Founder CEO
        last edited by

        Scheduled snaps won't have any interference with manuals snapshots (XO is using a specific tracking for automated snapshots).

        1 Reply Last reply Reply Quote 0
        • akurzawaA Offline
          akurzawa
          last edited by

          OK.

          If I want to delete all snapshots, in which direction should I start deleting snaps? from oldest to newest? or from newest to oldest?

          1 Reply Last reply Reply Quote 0
          • olivierlambertO Offline
            olivierlambert Vates 🪐 Co-Founder CEO
            last edited by

            Just delete them all, no order is needed.

            1 Reply Last reply Reply Quote 1
            • C Offline
              cg
              last edited by

              Keep in mind that too much active snapshots may degrade your disk performance enormous, as for every transation every snapshot has to be considered for changes.
              (At least it was like that, a while a go with XenServer, Oliver may correct me on that - also ZFS should bring a big boost for snapshots)

              1 Reply Last reply Reply Quote 0
              • olivierlambertO Offline
                olivierlambert Vates 🪐 Co-Founder CEO
                last edited by

                You'll end up with a chain, yes (so every read will need to make "hops" on each member of the chain).

                Longer the chain, slower the operation. But it shouldn't be too catastrophic either (for a heavily IOPS loaded VM, you'll notice the difference, but not in other less IOPS hungry scenarios)

                1 Reply Last reply Reply Quote 1
                • First post
                  Last post