initial 'auto' scenario support
ceph-volume now has an `auto` sub-command that should allow automatic OSD creation given any number of drives. The initial support for this mechanism
should allow to create bluestore OSDs from a number of drives.
This support should include:
- 1:1 drive to OSD if all detected/available devices are HDD (spinning)
- as large as possible block.db devices if there is a mix of HDD/solid (data on hdd, block.wal on solid)
- logic to identify incorrect/unsafe sizing of block.db or data LVs
- sizing and device type (solid/spinning) detection for usage, via ceph.conf or ENV vars