Adjust osd-purge state and shell script to check for cluster health prior to execution
https://git.cybbh.space/vta/saltstack/blob/master/apps/ceph/files/purge-osd.sh and https://git.cybbh.space/vta/saltstack/blob/master/apps/ceph/osd/purge-osd.sls currently do not check for cluster health nor perform a re-balance prior to being dropped from the cluster.
There are two key issues that need to be addressed:
- Check that ceph is in a state of 'HEALTHY' prior to any osd actions being taken
- Ensure that all placement groups are active+clean prior to fully removing OSDs and hosts from the crush map.
This issue should serve as a tracking issue as these two items are addressed.