Code review comment for lp:~seyeongkim/charms/trusty/ceph/lp1411652

Hey Billy,

Some of this may change in the near future as we're planning to deprecate the current ceph charm in favor of a ceph-mon charm + ceph-osd charm to clarify what deployments should look like. At that point, we can cleanly separate the removal of services from osd and monitor specific deployments.

Chris
> On Feb 3, 2016, at 5:14 AM, Billy Olsen <email address hidden> wrote:
>
> fwiw, there's a few things at play here... the data on the OSD is intentionally not touched so that it can be reused later. Charm defaults to not reformatting the disk in an effort to prevent data loss so a user can bring the OSD back into the cluster. We'd need to check if this is handled automagically upon adding the device. In a MAAS environment, I believe the disks are wiped clean when the node is reclaimed by MAAS (an option at least). Thus, a juju destroy-unit ceph will instead be more catastrophic to the OSD themselves.
> --
> https://code.launchpad.net/~xtrusia/charms/trusty/ceph/lp1411652/+merge/281732
> You are reviewing the proposed merge of lp:~xtrusia/charms/trusty/ceph/lp1411652 into lp:~openstack-charmers/charms/trusty/ceph/next.