[pve-devel] [PATCH storage] Automatically round up to the next valid size when resizing a ZFS volume

[Fabian Grünbichler]

On December 5, 2019 11:47 am, Fabian Ebner wrote:
> so ZFS won't complain when we do things like 'qm resize 102 scsi1 +0.01G'
> 
> Signed-off-by: Fabian Ebner <f.ebner at proxmox.com>
> ---
>  PVE/Storage/ZFSPoolPlugin.pm | 11 +++++++++++
>  1 file changed, 11 insertions(+)
> 
> diff --git a/PVE/Storage/ZFSPoolPlugin.pm b/PVE/Storage/ZFSPoolPlugin.pm
> index 456fb40..9bc3fcf 100644
> --- a/PVE/Storage/ZFSPoolPlugin.pm
> +++ b/PVE/Storage/ZFSPoolPlugin.pm
> @@ -662,6 +662,17 @@ sub volume_resize {
>  
>      my $attr = $format eq 'subvol' ? 'refquota' : 'volsize';
>  
> +    # for ZFS volumes, size has to be a multiple of the volume block size,
> +    # so we round up to the next valid size
> +    if ($format eq 'raw') {
> +	my $volblocksize = int($class->zfs_get_properties($scfg, 'volblocksize', "$scfg->{pool}/$vname"));
> +	if ($volblocksize >= 2048) {
> +	    $volblocksize = int($volblocksize / 1024);
> +	    my $remainder = $new_size % $volblocksize;
> +	    $new_size = $new_size + $volblocksize - $remainder if $remainder;
> +	}

we already align/pad to 1M on allocation (zfs_create_zvol), maybe we 
could do the same here to increase consistency, and avoid a round-trip 
via zfs_get_properties?

> +    }
> +
>      $class->zfs_request($scfg, undef, 'set', "$attr=${new_size}k", "$scfg->{pool}/$vname");
>  
>      return $new_size;
> -- 
> 2.20.1
> 
> 
> _______________________________________________
> pve-devel mailing list
> pve-devel at pve.proxmox.com
> https://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
> 
>