删除ceph集群mds

ceph集群新搭建以后是只有一个默认的存储池rbd的池
## 创建文件接口集群

1.创建一个元数据池

[root@mytest ~]# ceph osd pool create metadata  20 20
pool 'metadata' created

2.创建一个数据池

[root@mytest ~]# ceph osd pool create data   20 20
pool 'data' created

3.创建一个文件系统

[root@mytest ~]# ceph fs new ceph  metadata data
new fs with metadata pool 4 and data pool 5

4.创建一个mds

[root@mytest ~]# ceph-deploy mds create mytest

5.部署完检查状态

[root@mytest ceph]# ceph -s
    cluster 7e5469ac-ae1f-494f-9913-901f60c0a76b
     health HEALTH_OK
     monmap e1: 1 mons at {mytest=192.168.0.76:6789/0}
            election epoch 1, quorum 0 mytest
     mdsmap e60: 1/1/1 up {0=mytest=up:active}
     osdmap e70: 2 osds: 2 up, 2 in
      pgmap v252: 104 pgs, 3 pools, 1962 bytes data, 20 objects
            75144 kB used, 30624 MB / 30697 MB avail
                 104 active+clean
  client io 108 B/s wr, 0 op/s

删除文件接口集群（删除mds）

1.停止mds进程

[root@mytest ceph]# /etc/init.d/ceph stop mds
=== mds.mytest ===
Stopping Ceph mds.mytest on mytest...kill 9638...done

2.将mds状态标记为失效

[root@mytest ceph]# ceph mds fail 0
failed mds.0

3.删除ceph文件系统

[root@mytest ceph]# ceph fs rm ceph --yes-i-really-mean-it

4.删除完了检查状态

[root@mytest ceph]# ceph -s
    cluster 7e5469ac-ae1f-494f-9913-901f60c0a76b
     health HEALTH_OK
     monmap e1: 1 mons at {mytest=192.168.0.76:6789/0}
            election epoch 1, quorum 0 mytest
     osdmap e71: 2 osds: 2 up, 2 in
      pgmap v253: 104 pgs, 3 pools, 1962 bytes data, 20 objects
            75144 kB used, 30624 MB / 30697 MB avail
                 104 active+clean

可以看到已经没有了mds的那一条了