]> git.apps.os.sepia.ceph.com Git - ceph-ci.git/commitdiff
docs: Clarify the relationship of min_size to EC pool recovery
authorBrad Hubbard <bhubbard@redhat.com>
Mon, 10 Apr 2017 05:14:20 +0000 (15:14 +1000)
committerBrad Hubbard <bhubbard@redhat.com>
Mon, 10 Apr 2017 05:14:20 +0000 (15:14 +1000)
Signed-off-by: Brad Hubbard <bhubbard@redhat.com>
doc/rados/operations/pg-states.rst

index 03a92426de5f923f4105d1064fe6436829d8532e..233042945046aacde62501dbf805960843847612 100644 (file)
@@ -54,7 +54,8 @@ map is ``active + clean``.
   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