Hi, all. I installed a Home Assistant VM via the community scripts repo several months ago. I just set up a new PBS instance, and noticed when I was backing everything up for the first time that the HA VM backup failed:
The rest of the workloads I'm running are LXC's, and they're backing up successfully. I'm able to get a backup of the VM manually via "Stop", so the overall backup process is working.
It seemed to be complaining about the qemu-guest-agent, so I've checked the Proxmox settings; the guest agent is enabled and freeze/thaw is also checked. I've tried running the agent in VirtIO and ISA modes (reboots in between) but that didn't make a difference. Looking on the VM itself, I don't see any processes with "qemu" in them, so it doesn't look like the agent is actually installed. And of course, HAOS doesn't seem to allow any package managers to install the agent manually. Before I go down the path of trying to build a new VM and move the configuration over, I wanted to poll the crowd about:
1) I'm assuming that the community script should be installing the right HAOS image with the qemu guest agent in it, right?
2) I'm also assuming the root cause here is the agent not being available on the VM to freeze things to get a successful backup.
3) Anything I can do other than going through the whole build-new-and-move-config process?
Proxmox is running at v 8.4.1, PBS is v 3.4.1, and HA is v 2025.5.3 (Core/Supervisor) / 15.2 (OS) / 20250516.0 (Frontend).
Thanks!
INFO: starting new backup job: vzdump 100 --all 0 --notes-template '{{guestname}}' --storage pbs --mode suspend --node zox1 --fleecing 0 --prune-backups 'keep-all=1' --notification-mode autoINFO: Starting Backup of VM 100 (qemu)INFO: Backup started at 2025-06-01 20:26:56INFO: status = runningINFO: backup mode: suspendINFO: ionice priority: 7INFO: VM Name: zomeINFO: include disk 'scsi0' 'ZFS_Data:vm-100-disk-1' 32GINFO: include disk 'efidisk0' 'ZFS_Data:vm-100-disk-0' 4MINFO: suspending guestINFO: creating Proxmox Backup Server archive 'vm/100/2025-06-02T01:26:56Z'INFO: skipping guest-agent 'fs-freeze', agent configured but not running?ERROR: VM 100 qmp command 'backup' failed - backup register image failed: command error: EINVAL: Invalid argumentINFO: aborting backup jobINFO: resuming VM again after 4 secondsINFO: resume vmERROR: Backup of VM 100 failed - VM 100 qmp command 'backup' failed - backup register image failed: command error: EINVAL: Invalid argumentINFO: Failed at 2025-06-01 20:27:03INFO: Backup job finished with errorsINFO: notified via target `mail-to-root`INFO: notified via target `HA_Notify_Webhook`TASK ERROR: job errorsThe rest of the workloads I'm running are LXC's, and they're backing up successfully. I'm able to get a backup of the VM manually via "Stop", so the overall backup process is working.
It seemed to be complaining about the qemu-guest-agent, so I've checked the Proxmox settings; the guest agent is enabled and freeze/thaw is also checked. I've tried running the agent in VirtIO and ISA modes (reboots in between) but that didn't make a difference. Looking on the VM itself, I don't see any processes with "qemu" in them, so it doesn't look like the agent is actually installed. And of course, HAOS doesn't seem to allow any package managers to install the agent manually. Before I go down the path of trying to build a new VM and move the configuration over, I wanted to poll the crowd about:
1) I'm assuming that the community script should be installing the right HAOS image with the qemu guest agent in it, right?
2) I'm also assuming the root cause here is the agent not being available on the VM to freeze things to get a successful backup.
3) Anything I can do other than going through the whole build-new-and-move-config process?
Proxmox is running at v 8.4.1, PBS is v 3.4.1, and HA is v 2025.5.3 (Core/Supervisor) / 15.2 (OS) / 20250516.0 (Frontend).
Thanks!
