I run nightly KVM backups to an offsite NFS share that I have configured directly as a Proxmox storage, but I unfortunately only have space for a limited number of backups, so naturally I have the storage "Max Backups" option set to 3. Checked back a few (5) days later to find the storage almost completely overflowing with 5 backups per VM.
As I'm writing this, it occurs to me that one of my hook scripts may be interfering the vzdump's "old backup removal" logic, because I rename the backups to include their VM hostname (follows the format "vzdump-qemu-${vmid}-${date}.${hostname}.vma.zst").
So I guess my question now is: does vzdump use the filenames to determine which backups to delete in a "max backup" situation? If not, then where could I start troubleshooting to find the source of this issue? The backup log itself didn't have any useful information.
(As of making this realization, I've edited my hook script to keep the original filenames, but I think having the hostname in the backups' filenames is a pretty handy way to tell what it is at a glance if you have 10+ VMs)
As I'm writing this, it occurs to me that one of my hook scripts may be interfering the vzdump's "old backup removal" logic, because I rename the backups to include their VM hostname (follows the format "vzdump-qemu-${vmid}-${date}.${hostname}.vma.zst").
So I guess my question now is: does vzdump use the filenames to determine which backups to delete in a "max backup" situation? If not, then where could I start troubleshooting to find the source of this issue? The backup log itself didn't have any useful information.
(As of making this realization, I've edited my hook script to keep the original filenames, but I think having the hostname in the backups' filenames is a pretty handy way to tell what it is at a glance if you have 10+ VMs)