[pve-devel] [PATCH qemu-server] add discard_granularity to 4M for rbd storage

[Alwin Antreich]

On Fri, Jun 29, 2018 at 08:12:13AM +0200, Alexandre DERUMIER wrote:
> and for zfs , we can retrieve block size with:
> 
> zfs get volblocksize rpool/data/vm-105-disk-1
> NAME                      PROPERTY      VALUE     SOURCE
> rpool/data/vm-105-disk-1  volblocksize  8K        default
> 
> 
> 
> Maybe can we change current discard option to:
> 
> discard : 0 : disable
>           1 : enable (with defaut qemu value)
>           2 : enable : auto best value (zfs/rbd autodect and qemu for others storage)
>           1024-... : custom user value
> 
> ?
IMHO: I would be up for this approach, as a first measure, so the discard
granularity can be overwritten. A second step may be that Qemu should be
able to work with the proper granularity by itself.