Digged further. ceph osd dump --format json |jq '.osds[]| select (.osd==4)' { "osd": 4, "uuid": "OSD UID", "up": 0, "in": 0, "weight": 0, "primary_affinity": 1, "last_clean_begin": 0, "last_clean_end": 0, "up_from": 0, "up_thru": 0, "down_at": 0, "lost_at": 0, "public_addrs": { "addrvec": [] }, "cluster_addrs": { "addrvec": [] }, "heartbeat_back_addrs": { "addrvec": [] }, "heartbeat_front_addrs": { "addrvec": [] }, "public_addr": "(unrecognized address family 0)/0", "cluster_addr": "(unrecognized address family 0)/0", "heartbeat_back_addr": "(unrecognized address family 0)/0", "heartbeat_front_addr": "(unrecognized address family 0)/0", "state": [ "autoout", "exists", "new" ] } Every other healthy osds fields are filled with proper values.. Clues match, but i keep on looking for the root cause. Best regards Philippe 21 sept. 2024, 21:29 de infolist@schwarz-fr.net:
Hi, on a healthy cluster, every osd creation made orphans osd.
# ceph osd tree ID CLASS WEIGHT TYPE NAME STATUS REWEIGHT PRI-AFF -1 26.38051 root default -7 3.63869 host A 1 hdd 3.63869 osd.1 up 1.00000 1.00000 -11 1.81940 host F 2 hdd 1.81940 osd.2 up 1.00000 1.00000 -5 7.27737 host J 3 hdd 3.63869 osd.3 up 1.00000 1.00000 5 hdd 3.63869 osd.5 up 1.00000 1.00000 -3 7.27737 host K 0 hdd 3.63869 osd.0 up 1.00000 1.00000 8 hdd 3.63869 osd.8 up 1.00000 1.00000 -9 6.36768 host S 6 hdd 3.63869 osd.6 up 1.00000 1.00000 7 hdd 2.72899 osd.7 up 1.00000 1.00000 4 0 osd.4 down 0 1.00000 9 0 osd.9 down 0 1.00000
neither 4 nor 9 osd appear in the crushmap (decompiled) Destroying and recreating osd recreate orphan osd.
Any hint ? Thanks for all Best regards Philippe