Signed-off-by: Brad Hubbard <bhubbard@redhat.com>
Ceph detects that a placement group is missing information about
writes that may have occurred, or does not have any healthy
copies. If you see this state, try to start any failed OSDs that may
- contain the needed information.
+ contain the needed information. In the case of an erasure coded pool
+ temporarily reducing min_size may allow recovery.
*Stale*
The placement group is in an unknown state - the monitors have not received