Ceph is a beast and will require at least 2-3 technicians with intricate Ceph knowledge to run multiple (!) Ceph clusters in a business continuity responsible m

Ceph is a beast and will require at least 2-3 technicians with intricate Ceph knowledge to run multiple (!) Ceph clusters in a business continuity responsible manner. Because you must be able to deal with Ceph quirks. If you can shard your data over multiple independent stand-alone ZFS boxes, that would be much simpler and more robust. But it might not scale like Ceph.