Talk:Rook Storage for Kubernetes

clearing/reusing drives in rook/ceph:

lsblk /dev/xxxx

if you see a ceph ID, you need to remove it:

dmsetup remove ceph--(whatever you see in lsblk)

you also need to clean out the partition table:

wipefs -a /dev/xxxx

if that device had been in ceph before, you need to remove the old OSD (that should be marked as down/out). open the rook-ceph-tools container:

ceph osd purge <number> --yes-i-really-mean-it

You need to make sure that the osd is truly dead ...

Removing error messages that make no sense:

- warning about low host disk space

ceph config set mon mon_data_avail_warn 10

- warning about pool objects per pg

ceph config set mgr mon_pg_warn_max_object_skew 0.0