[SOLVED] Backup to remote data centre that's part of the same PVE cluster

[eugenevdm]

Hi there, I have a newbie question.

We have two data centres, with four VM hosts, Data Center A and Data Center B.

I have an installation Proxmox Backup Server in Data Centre B. I tried to backup a VM in Data Centre A to the PBS server in Data Centre B but it fails with an obvious error message about the disk not found:

INFO: starting new backup job: vzdump 124 --storage backup05.example.com --notes-template '{{guestname}}' --mode snapshot --remove 0 --node hv7
INFO: Starting Backup of VM 124 (qemu)
INFO: Backup started at 2023-10-23 12:53:38
INFO: status = stopped
INFO: backup mode: stop
INFO: ionice priority: 7
INFO: VM Name: host.example.com
INFO: include disk 'scsi0' 'raid1-2:vm-124-disk-0' 20G
ERROR: Backup of VM 124 failed - storage 'raid1-2' is not available on node 'hv7'
INFO: Failed at 2023-10-23 12:53:38
INFO: Backup job finished with errors
TASK ERROR: job errors

It seems PBS is trying to access the VM's disk to do the remote backup.

I guess my question is how would I backup using PBS if a VM is in a different data centre as the PBS server.

 

[Chris]

Hi,
please share your storage configuration cat /etc/pve/storage.cfg and the VM config qm config <VMID> --current, both from the node and the VM you are trying to backup up from as well as the node and VM with PBS installed.

I assume that the VM you are trying to backup is located on node hv7 but there is no storage raid1-2 on that node. Which leaves me with the question, is your VM actually running? And if so: How did you start the VM on that node without the disks for it being present? You can only perform the backup of the VM including this disk from the node where the disk is located.

 

[eugenevdm]

Hi @Chris,

I made some progress due to your on-point questions:

Which leaves me with the question, is your VM actually running?

The VM wasn't running. I've now instead chosen a running VM and backups are working! It took around 3 minutes to backup the 20 GB VM from the one data centre to the other and considering bandwidth limitations I'm actually very happy with this.

It appears I had another issue as well. I had to tie down `raid1-2` to the correct node using the Data Center / Storage / `Nodes` configuration,