after backup, snapshots are not deleted



  • In storages -> Disks i see them like this: " 01-WEB-HOST 0 "
    Here, https://ibb.co/PT3853J is a screenshot of 2 disk that appear when i filter for "type:VDI-snapshot"
    Thanks



  • These VDI snapshots are attached to VM snapshots which is normal.

    XO keeps at least a snapshot per backup for each VM if:

    • you are using rolling snapshots
    • or you are using delta backups
    • or you are using continuous replication


  • I am not using none of the above
    I use: Backup only



  • XO should delete the backup then.

    We are currently completely rewriting the backup code, we'll test more once this is released.



  • until the latest updates it worked just fine, only now i realized that is not deleting them



  • This post is deleted!


  • Sorry to bring up an old topic, but is this still a thing? I recently implemented a continuous replication job (4 times a day) to go along with our nightly backups, and after a few days it starts erroring out saying the VDI chain for various VMs is too long. If I manually delete any snapshots that aren't associated with a VM then it'll start coalescing, but I can't keep doing manual cleanups every few days!

    Screenshot to show what I mean. I cleaned these out yesterday and it's back up to a chain of 4 unassigned snapshots.

    6c1f150c-67bf-4d0e-ba2a-12430926ad0e-image.png

    Cheers for help etc. 🙂



    1. VDI chain protection is a… protection so that's normal if it's displayed when the chain isn't clean
    2. This is more a symptom of something wrong on your SR.

    We have very very few info to be able to draw any conclusion:

    • what's your XCP-ng version?
    • is it up to date?
    • are you using XO from the sources or XO? If from the source, what's your xo-server and xo-web version?
    • check your XCP-ng logs (especially SMlog)


  • 🤔

    • This is happening on both XCP-ng 7.6 and 8.0 pools (both fully patched and rebooted).
    • We're using XOA version 5.38 (stable channel).
    • It's happening on both iSCSI and Hardware HBA storages.

    I'll check on the logs when I have more time next week, but I can't see any common factor other than it's all using thick provisioning. We've just got some FreeNAS based NFS storage up and running and I'm moving one of the problem VMs to that right now, so we'll see if it helps at all.

    Edit: It's also not a VDI protection thing but a VDI chain too long thing. It's simply not coalescing unless I manually delete the unassociated snapshots.



  • If you are using XOA (stable channel), I strongly suggest to use a support ticket 🙂



  • Knew I'd forgotten something this week! I was supposed to be getting my boss to set me up with a support account. 😕



  • Also XO forum is now transferred to https://xcp-ng.org/forum/category/12/xen-orchestra

    For community questions it's the place to go. But again, as you have XOA with support, creating a ticket is the best thing to do!


Log in to reply