Hi all! My ceph version is 10.2.11 and I use rgw and EC(7+3). When I use muliti clients to read and write on one rgw bucket .The bucket have 200 shards. There look like dead look on osd? Thanks for all! The osd block ops all have the similiar below logs and the block osds not always on the same osd. "description": "osd_op(client.3147989.0:2782587465 13.36f21cde .dir.517af746-28f1-454c-ba41-0c4fd51af270.896917.11.99 [call rgw.bucket_complete_op] snapc 0=[] ack+ondisk+write+known_if_redirected e26026)", "initiated_at": "2019-09-04 13:38:48.089766", "age": 3520.459786, "duration": 3521.432087, "type_data": [ "delayed", { "client": "client.3147989", "tid": 2782587465 }, [ { "time": "2019-09-04 13:38:48.089766", "event": "initiated" }, { "time": "2019-09-04 13:38:48.089794", "event": "queued_for_pg" }, { "time": "2019-09-04 13:38:48.180941", "event": "reached_pg" }, { "time": "2019-09-04 13:38:48.180962", "event": "waiting for rw locks" }, { "time": "2019-09-04 13:38:48.430598", "event": "reached_pg" }, { "time": "2019-09-04 13:38:48.430616", "event": "waiting for rw locks" }, { "time": "2019-09-04 13:38:49.150673", "event": "reached_pg" }, { "time": "2019-09-04 13:38:49.150691", "event": "waiting for rw locks" }, { "time": "2019-09-04 13:38:51.421887", "event": "reached_pg" }, { "time": "2019-09-04 13:38:51.421904", "event": "waiting for rw locks" }, { "time": "2019-09-04 13:38:51.990943", "event": "reached_pg" }, { "time": "2019-09-04 13:38:51.990961", "event": "waiting for rw locks" }, { "time": "2019-09-04 13:38:54.343921", "event": "reached_pg" }, { "time": "2019-09-04 13:38:54.343938", "event": "waiting for rw locks" }