Hello, yesterday i did: ceph osd purge 32 --yes-i-really-mean-it I also started to upgrade: ceph orch upgrade start --ceph-version 15.2.4 It seems its really gone: ceph osd crush remove osd.32 => device 'osd.32' does not appear in the crush map ceph orch ps: osd.32 ceph01 error 4h ago 2h <unknown> docker.io/ceph/ceph:v15 <unknown> <unknown> root@ceph01:~# ceph orch osd rm status NAME HOST PGS STARTED_AT osd.32 ceph01 n/a 2020-07-29 09:50:11.026643 (some hours ago) root@ceph01:~# ceph health detail HEALTH_ERR Module 'cephadm' has failed: auth get failed: failed to find osd.32 in keyring retval: -2; Low space hindering backfill (add storage if this doesn't resolve itself): 105 pgs backfill_toofull; Degraded data redundancy: 9882/33167949 objects degraded (0.030%), 1 pg degraded, 1 pg undersized [ERR] MGR_MODULE_ERROR: Module 'cephadm' has failed: auth get failed: failed to find osd.32 in keyring retval: -2 Module 'cephadm' has failed: auth get failed: failed to find osd.32 in keyring retval: -2 Any idea what i would try to remove that OSD? Thanks, Michael