We could also just look at setting rbd_flatten_volume_from_snapshot=True in cinder.conf for the ceph job, since if I understand that code correctly, it should disassociate the backing volume from the snapshot in the ceph backend which should speed this up.
We could also just look at setting rbd_flatten_ volume_ from_snapshot= True in cinder.conf for the ceph job, since if I understand that code correctly, it should disassociate the backing volume from the snapshot in the ceph backend which should speed this up.