[SOLVED] Backup Fleecing on pool-based backup job

[TheAbian]

Hi,

We currently have a Proxmox cluster with 10 nodes and a PBS.
Sometimes the VMs hang / run unperformant during the backup, so I have read up on fleecing.

Each node now has a local SSD, mounted as LVM-thin, in /etc/vzdump.conf I have activated fleecing and set the local storage.
A manuel backup works perfectly fine with fleecing.


´The backup job on the other hand overwrites the vzdump.conf and I have to tick the box under ‘advanced’ for ‘fleecing’, but I also have to select a specific storage for fleecing. However, this makes no sense, as the VMs for the pool-based job are scattered across the cluster.
If I leave ‘fleecing’ unticked, the job does not use fleecing at all.


What am I missing?

Thank's in advance!

 

[fiona]

Hi,
yes, a backup job created via the UI will explicitly set fleecing to enabled or disabled depending on whether the checkbox is ticked or not. What you would need is a fallback option that is selected by default, but that is not implemented in the UI right now. Feel free to open a feature request for this: https://bugzilla.proxmox.com/

If you create a backup job via CLI/API and don't specify a setting for fleecing at all, the node fallback from vzdump.conf will be used.

 

[TheAbian]

Hi,
yes, a backup job created via the UI will explicitly set fleecing to enabled or disabled depending on whether the checkbox is ticked or not. What you would need is a fallback option that is selected by default, but that is not implemented in the UI right now. Feel free to open a feature request for this: https://bugzilla.proxmox.com/

If you create a backup job via CLI/API and don't specify a setting for fleecing at all, the node fallback from vzdump.conf will be used.

Hi,
thank you very much, that was the information I was missing.
Now everything works as I expected it to.