sisimi的点点滴滴

你若盛开,清风自来

导航

Error EBUSY: osd.0 is still up; must be down before removal的解决办法

标签(空格分隔):ceph,ceph运维,osd故障


集群环境:

[root@node3 ~]# cat /etc/redhat-release 
CentOS Linux release 7.3.1611 (Core) 
[root@node3 ~]# ceph -v
ceph version 12.2.1 (3e7492b9ada8bdc9a5cd0feafd42fbca27f9c38e) luminous (stable)

问题描述:

在使用批量删除osd的shell脚本删除所有的osd时,提示:Error EBUSY: osd.0 is still up; must be down before removal.查看集群架构:

[root@node1 ceph]# ceph osd tree
ID CLASS WEIGHT TYPE NAME      STATUS REWEIGHT PRI-AFF 
-1            0 root default                           
-3            0     host node1                         
-5            0     host node2                         
-7            0     host node3                         
 0            0 osd.0              up        0 1.00000 

很明显就是需要把osd.0先down掉,然后我使用命令:systemctl stop ceph-osd@0 发现osd的状态并未改变,然后通过查看osd.0的进程,发现并没有运行。

解决方法:

如此简单,通过ceph命令强行标记为down,之后删除即可。下面先标记为down:

[root@node1 ceph]# ceph osd down osd.0
marked down osd.0. 

此时查看集群布局:

[root@node1 ceph]# ceph osd tree
ID CLASS WEIGHT TYPE NAME      STATUS REWEIGHT PRI-AFF 
-1            0 root default                           
-3            0     host node1                         
-5            0     host node2                         
-7            0     host node3                         
 0            0 osd.0            down        0 1.00000 

最后删除该osd.0:

[root@node1 ceph]# ceph osd rm 0
removed osd.0

确认是否删除:

[root@node1 ceph]# ceph osd tree
ID CLASS WEIGHT TYPE NAME      STATUS REWEIGHT PRI-AFF 
-1            0 root default                           
-3            0     host node1                         
-5            0     host node2                         
-7            0     host node3                         

参考文献:

How to delete a invalid osd in ceph cluster?

posted on 2017-11-10 11:14  sisimi_2017  阅读(3176)  评论(0编辑  收藏  举报