`microcloud add` shows already used disks when adding new systems

[roosterfish]

When running microcloud add, MicroCloud loads the disks from the initiator and all the new joining systems.
However it still displays disks which are already used on the initiator (e.g. for MicroCeph OSDs) and allows to select them for either local or remote storage.

Instead MicroCloud should filter out already used disks before presenting them to the user.

[roosterfish]

This can be narrowed down a bit more.

In case the disk on the initiator is already used for local storage, it will be skipped because it's not anymore pristine and contains a partition.
The problematic disks are only the ones configured as OSD.

When re-selecting the disk already used as OSD the following errors occur:

• When configuring distributed storage: The already used disk isn't displayed
• When configuring local storage: This already used disk is displayed and the following error occurs when trying the create the cluster wide local storage pool: ⨯ Error: Failed to run: zpool create -f -m none -O compression=on local /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_lxd_disk1: exit status 1 (/dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_lxd_disk1 is in use and contains a unknown filesystem.)

So this bug right now doesn't harm the environment, but it still causes bad user experience as you have to remove the cluster member from MicroCloud, reset all the snaps and start all over in case you have selected a disk on the initiator which is already used as OSD.