I’m working on a minimal ceph deployment onto ubuntu 18.04 server VMs with loop devices. For my osds, the task “run ceph-volume lvm batch --report” is failing, apparently because the size key is never populated in ceph_volume/util/device.py. As far as I can tell, loop devices are skipped over while enumerating device information in get_devices().
In my osds.yml, I have specified the device files for my loop devices, but otherwise I am using default settings. Is there a proper way to configure Ceph such that it will accept and use loop devices?
I realize that I could add additional disks to each VM, but for my use case it is preferable to have access to the backing files used by the loop devices.