Backing up QCOW2 VM - SR_BACKEND_FAILURE_82

[Andrw0830]

Are you using XOA or XO from the sources?

XO from the sources

Which release channel?

stable

Provide your commit number

cc7ba

Describe the bug

For the past month since I've spun up a VM with 1- 300GB VHD and 1 - 6TB QCOW2, I'm having issues with backups where sometimes the weekly backup of this VM to a block base backup repo will succeed or fail and cause the base copy of the disks to remain on the NFS SR. Last week I had to run the rescan disk manually to get the garbage collection to run to coalesce the disks, but now that's not working and I have numerous base copies of my VMs listed. Look at below error on latest failure which seemed to be when the base copies for all other VMs that tried to copy showed up.

Error message

SR_BACKEND_FAILURE_82(, Failed to snapshot VDI [opterr=Command ['/usr/bin/qemu-img', 'info', '--backing-chain', '--output=json', '/var/run/sr-mount/cf29ab2b-ca03-e55e-38ed-451512514dac/9889ca88-c17f-4e77-9b8f-77c6217b6786.qcow2'] failed (): Operation not permitted], )

To reproduce

1. Start a XOA backup for the VM
2. Error message appears above
3. Because it doesn't backup, it keeps the base copy of the disk on the SR and causes other VM backups on that SR to skip to protect the Chain.

Expected behavior

Backups should be able to run on running QCOW2 VM and garbage collection should be able to run properly to remove base copies like it does for VDI VMs.

Screenshots

No response

Node

22.23.0

Hypervisor

XCP-NG 8.3

Additional context

No response

[MrGrymReaper]

The situation with S3 Deletion on Google Cloud Storage has changed, recently it’s now possible to delete stored files and directories. But using the API and SDK documented at this link (https://docs.cloud.google.com/storage/docs/deleting-objects#client-libraries_1).