XO reports wrong number of backups

codesmaker

Hi,

First of all, Thanks a lot for the nice product and for all your contributions to the community!

I'm using Xen Orchestra, complied from source, version 5.103.1. I added a remote NFS storage which I use for my VM backups.

I created a backup job for one of my VMs that runs once a week with backup retention set to 2. When I try to restore, XO shows all the backups taken. Initially, I thought that the backup retention wasn't working as it should but when I checked the files on disk, I could see that there were only 2 backups which means that the set retention in the backup job is working correctly.

Any idea why XO is reporting that all the backups are available for restore?

Thanks!

julien-f

@codesmaker Can you try to delete the cache.json.gz files to see if that helps?

If it does, that may be due to an old bug, tell me in case the issue reappears.

olivierlambert

Hi,

Please provide the commit number instead the version, it's more relevant (as per https://xen-orchestra.com/docs/community.html#current-version)
Double check you are on latest commit after rebuild to confirm you have the same behavior
If it's the case, please provide some screenshots so you we can easily understand

Thanks!

codesmaker

Hi Olivier,

Thanks for the quick reply!

I'm currently on the latest commit (81abc091d) and I updated from (c52e0a553). I can still see the same issue.

As you can see in the second screenshot, there are 6 backups while when I check the available backups on disk, there are only 2 dated November 5 and October 29.

olivierlambert

Can you show us what do you have on the disk? (eg a simple tree -h on the Xo backup folder will do the trick.)

codesmaker

Here you are:

olivierlambert

Pinging @julien-f

julien-f

@codesmaker Can you try to delete the cache.json.gz files to see if that helps?

If it does, that may be due to an old bug, tell me in case the issue reappears.

vmpr

@julien-f that also fixed my issue, cheers!

codesmaker

@julien-f After two months of testing, I can confirm that this issue persists.

julien-f

@codesmaker I'm really skeptical that this issue still exists, please update to latest version and delete cache.json.gz to force refresh for VMs without a recent backup run.

The commits that fix this issue:

0

julien-f committed to vatesfr/xen-orchestra

feat(backups/cleanVm): detect and fix cache inconsistencies (#6575)

0

julien-f committed to vatesfr/xen-orchestra

fix(backups): wait for cache to be updated before running cleanVm (#6580)

Introduced by 191c12413

codesmaker

@julien-f Thanks a lot for the fast answer. I'll update to the latest version and test.