HEALTH_WARN too few PGs per OSD (16 < min 30）

摘要: ceph添加完osd，查看集群状态报错

health HEALTH_WARN too few PGs per OSD (16 < min 30)

执行ceph –s 发现集群状态并非ok，具体信息如下：

$ sudo ceph -s
cluster 257faba1-f259-4164-a0f9-1726bd70b05a
health HEALTH_WARN
too few PGs per OSD (16 < min 30)
monmap e1: 1 mons at {bdc217=192.168.13.217:6789/0}
election epoch 2, quorum 0 bdc217
osdmap e50: 8 osds: 8 up, 8 in
flags sortbitwise
pgmap v119: 64 pgs, 1 pools, 0 bytes data, 0 objects
715 MB used, 27550 GB / 29025 GB avail
64 active+clean

由于是新配置的集群，只有一个pool

$ sudo ceph osd lspools
0 rbd,

查看rbd pool的PGS

$ sudo ceph osd pool get rbd pg_num
pg_num: 64

pgs为64，因为是2副本的配置，所以当有8个osd的时候，每个osd上均分了64/8 *2=16个pgs,也就是出现了如上的错误 小于最小配置30个

解决办法：修改默认pool rbd的pgs

$ sudo ceph osd pool set rbd pg_num 128
set pool 0 pg_num to 128
$ sudo ceph -s
cluster 257faba1-f259-4164-a0f9-1726bd70b05a
health HEALTH_WARN
64 pgs stuck inactive
64 pgs stuck unclean
pool rbd pg_num 128 > pgp_num 64
monmap e1: 1 mons at {bdc217=192.168.13.217:6789/0}
election epoch 2, quorum 0 bdc217
osdmap e52: 8 osds: 8 up, 8 in
flags sortbitwise
pgmap v121: 128 pgs, 1 pools, 0 bytes data, 0 objects
715 MB used, 27550 GB / 29025 GB avail
64 active+clean
64 creating

发现需要把pgp_num也一并修改，默认两个pg_num和pgp_num一样大小均为64，此处也将两个的值都设为128

$ sudo ceph osd pool set rbd pgp_num 128
set pool 0 pgp_num to 128

最后查看集群状态，显示为OK，错误解决：

$ sudo ceph -s
cluster 257faba1-f259-4164-a0f9-1726bd70b05a
health HEALTH_OK
monmap e1: 1 mons at {bdc217=192.168.13.217:6789/0}
election epoch 2, quorum 0 bdc217
osdmap e54: 8 osds: 8 up, 8 in
flags sortbitwise
pgmap v125: 128 pgs, 1 pools, 0 bytes data, 0 objects
718 MB used, 27550 GB / 29025 GB avail
128 active+clean