原因：当你一不小心删掉了osd的auth信息时，重启osd服务，此时ceph -s查看发现osd down

如：

[root@ceph ~]# ceph osd tree
ID WEIGHT  TYPE NAME       UP/DOWN REWEIGHT PRIMARY-AFFINITY 
-1 0.02719 root default                                      
-2 0.01849     host ceph58                                   0 0.01849         osd.0        up  1.00000          1.00000 
-3 0.00870     host ceph28                                   1 0.00870         osd.1        up  1.00000          1.00000#集群OSD全部up[root@ceph ~]# ceph auth list
installed auth entries:osd.0key: AQDZ7T5ZmLx3MBAAR8Vhqt1UvreMUwSSmdfeSw==caps: [mon] allow profile osdcaps: [osd] allow *
osd.1key: AQDAFkRZEHhnGxAAjfbGRNNNT5kWvGl4jpKjYg==caps: [mon] allow profile osdcaps: [osd] allow *
client.adminkey: AQBA7T5ZAAAAABAAlJhtiG0oJVOeXlBc0Mzokw==caps: [mds] allow *caps: [mgr] allow *caps: [mon] allow *caps: [osd] allow *
client.bootstrap-osdkey: AQDA7T5ZBzemGhAAwQgt7wU3kVJps7IoLAg0TA==caps: [mon] allow profile bootstrap-osd#此时查看auth表中，osd.0和osd.1的auth值都正常[root@ceph ~]# ceph auth del osd.1    #将osd.1 auth值从auth表中删除
updated
[root@ceph ~]# ceph auth list     
installed auth entries:osd.0key: AQDZ7T5ZmLx3MBAAR8Vhqt1UvreMUwSSmdfeSw==caps: [mon] allow profile osdcaps: [osd] allow *
client.adminkey: AQBA7T5ZAAAAABAAlJhtiG0oJVOeXlBc0Mzokw==caps: [mds] allow *caps: [mgr] allow *caps: [mon] allow *caps: [osd] allow *
client.bootstrap-osdkey: AQDA7T5ZBzemGhAAwQgt7wU3kVJps7IoLAg0TA==caps: [mon] allow profile bootstrap-osd#此时发现osd.1的auth值已从auth表中清除[root@ceph ~]# systemctl restart ceph-osd@1  #重启osd
[root@ceph ~]# ceph osd tree
ID WEIGHT  TYPE NAME       UP/DOWN REWEIGHT PRIMARY-AFFINITY 
-1 0.02719 root default                                      
-2 0.01849     host ceph58                                   0 0.01849         osd.0        up  1.00000          1.00000 
-3 0.00870     host ceph28                                   1 0.00870         osd.1      down  1.00000          1.00000#此时发现osd.1已经down了

 

如何恢复？

步骤如下：

• 进入/var/lib/ceph/osd/ceph-*

[root@ceph ~]# cd /var/lib/ceph/osd/ceph-1
[root@ceph ceph-1]# ls
total 60
-rw-r--r--. 1 root root 202 Jun 17 01:34 activate.monmap
-rw-r--r--. 1 ceph ceph   3 Jun 17 01:44 active
lrwxrwxrwx. 1 ceph ceph  58 Jun 17 01:34 block -> /dev/disk/by-partuuid/87f73ff4-7add-4e83-94e9-29869c7c0123
lrwxrwxrwx. 1 ceph ceph  58 Jun 17 01:34 block.db -> /dev/disk/by-partuuid/f70d731a-666b-4828-8cb7-59c4aa498a91
-rw-r--r--. 1 ceph ceph  37 Jun 17 01:34 block.db_uuid
-rw-r--r--. 1 ceph ceph  37 Jun 17 01:34 block_uuid
lrwxrwxrwx. 1 ceph ceph  58 Jun 17 01:34 block.wal -> /dev/disk/by-partuuid/fbe8751d-c2ae-4db4-8a35-7ab699401b58
-rw-r--r--. 1 ceph ceph  37 Jun 17 01:34 block.wal_uuid
-rw-r--r--. 1 ceph ceph   2 Jun 17 01:34 bluefs
-rw-r--r--. 1 ceph ceph  37 Jun 17 01:34 ceph_fsid
-rw-r--r--. 1 ceph ceph  37 Jun 17 01:34 fsid
-rw-------. 1 ceph ceph 124 Jun 17 18:10 keyring
-rw-r--r--. 1 ceph ceph   8 Jun 17 01:34 kv_backend
-rw-r--r--. 1 ceph ceph  21 Jun 17 01:34 magic
-rw-r--r--. 1 ceph ceph   4 Jun 17 01:34 mkfs_done
-rw-r--r--. 1 ceph ceph   6 Jun 17 01:34 ready
-rw-r--r--. 1 ceph ceph   0 Jun 17 01:44 systemd
-rw-r--r--. 1 ceph ceph  10 Jun 17 01:34 type
-rw-r--r--. 1 ceph ceph   2 Jun 17 01:34 whoami#其中keying文件记录的就是该osd部分auth值

• 修改keying文件

[root@ceph ceph-1]# cat keyring 
[osd.1]key = AQDAFkRZEHhnGxAAjfbGRNNNT5kWvGl4jpKjYg==              #默认[root@ceph ceph-1]# cat keyring 
[osd.1]key = AQDAFkRZEHhnGxAAjfbGRNNNT5kWvGl4jpKjYg==caps mon = "allow profile osd"                              #增加caps osd = "allow *"                                        #增加

• 将keying文件的内容上传到mon

[root@ceph ceph-1]# ceph auth import -i keyring 
imported keyring[root@ceph ceph-1]# systemctl restart ceph-osd@1[root@ceph ceph-1]# ceph osd tree
ID WEIGHT  TYPE NAME       UP/DOWN REWEIGHT PRIMARY-AFFINITY 
-1 0.02719 root default                                      
-2 0.01849     host ceph58                                   0 0.01849         osd.0        up  1.00000          1.00000 
-3 0.00870     host ceph28                                   1 0.00870         osd.1        up  1.00000          1.00000