ceph部署时错误分析

(1)、执行ceph-deploy install ceph01 ceph02 ceph03时报错

解决办法：
经过查询方知道是由于网络慢的原因导致的
可现在ceph02、ceph03节点安装上ceph

apt install -y ceph

再执行上述步骤
问题解决

(2)、HEALTH_WARN too few PGs per OSD (21 < min 30)

root@ceph01:/mnt/mycluster# ceph -s    cluster 1838d346-9ecc-4b06-a5f9-cc086a7e7eec     health HEALTH_WARN            too few PGs per OSD (21 < min 30)     monmap e1: 3 mons at {ceph01=192.168.93.14:6789/0,ceph02=192.168.93.15:6789/0,ceph03=192.168.93.11:6789/0}            election epoch 4, quorum 0,1,2 ceph03,ceph01,ceph02     osdmap e43: 9 osds: 9 up, 9 in            flags sortbitwise,require_jewel_osds      pgmap v109: 64 pgs, 1 pools, 0 bytes data, 0 objects            46380 MB used, 134 GB / 179 GB avail                  64 active+clean

由于是新配置的集群，只有一个pool

root@ceph01:/mnt/mycluster# ceph osd lspools0 rbd,

查看rbd pool的PGS

root@ceph01:/mnt/mycluster# ceph osd pool get rbd pg_numpg_num: 64

pgs为64，因为是3副本的配置，所以当有9个osd的时候，每个osd上均分了64/9 *3=21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

root@ceph01:/mnt/mycluster# ceph osd pool set rbd pg_num 128set pool 0 pg_num to 128root@ceph01:/mnt/mycluster# ceph -s    cluster 1838d346-9ecc-4b06-a5f9-cc086a7e7eec     health HEALTH_WARN            pool rbd pg_num 128 > pgp_num 64     monmap e1: 3 mons at {ceph01=192.168.93.14:6789/0,ceph02=192.168.93.15:6789/0,ceph03=192.168.93.11:6789/0}            election epoch 4, quorum 0,1,2 ceph03,ceph01,ceph02     osdmap e45: 9 osds: 9 up, 9 in            flags sortbitwise,require_jewel_osds      pgmap v127: 128 pgs, 1 pools, 0 bytes data, 0 objects            46382 MB used, 134 GB / 179 GB avail                 128 active+clean

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

root@ceph01:/mnt/mycluster# ceph osd pool get rbd pgp_numpgp_num: 64root@ceph01:/mnt/mycluster# ceph osd pool set rbd pgp_num 128set pool 0 pgp_num to 128

最后查看集群状态，显示为OK，错误解决：

root@ceph01:/mnt/mycluster# ceph -s    cluster 1838d346-9ecc-4b06-a5f9-cc086a7e7eec     health HEALTH_OK     monmap e1: 3 mons at {ceph01=192.168.93.14:6789/0,ceph02=192.168.93.15:6789/0,ceph03=192.168.93.11:6789/0}            election epoch 4, quorum 0,1,2 ceph03,ceph01,ceph02     osdmap e47: 9 osds: 9 up, 9 in            flags sortbitwise,require_jewel_osds      pgmap v142: 128 pgs, 1 pools, 0 bytes data, 0 objects            46385 MB used, 134 GB / 179 GB avail                 128 active+clean

(3)、HEALTH_WARN clock skew detected on mon.ceph02

root@ceph01:/var/lib/ceph/osd/ceph-2/current# ceph -s    cluster 1838d346-9ecc-4b06-a5f9-cc086a7e7eec     health HEALTH_WARN            clock skew detected on mon.ceph02            Monitor clock skew detected      monmap e1: 3 mons at {ceph01=192.168.93.14:6789/0,ceph02=192.168.93.15:6789/0,ceph03=192.168.93.11:6789/0}            election epoch 4, quorum 0,1,2 ceph03,ceph01,ceph02     osdmap e47: 9 osds: 9 up, 9 in            flags sortbitwise,require_jewel_osds      pgmap v151: 128 pgs, 1 pools, 0 bytes data, 0 objects            46384 MB used, 134 GB / 179 GB avail                 128 active+clean

这是由于集群的时间不一致，同步一下时间ok

root@ceph01:/var/lib/ceph/osd/ceph-2/current# ntpq -p     remote           refid      st t when poll reach   delay   offset  jitter============================================================================== 0.ubuntu.pool.n .POOL.          16 p    -   64    0    0.000    0.000   0.000 1.ubuntu.pool.n .POOL.          16 p    -   64    0    0.000    0.000   0.000 2.ubuntu.pool.n .POOL.          16 p    -   64    0    0.000    0.000   0.000 3.ubuntu.pool.n .POOL.          16 p    -   64    0    0.000    0.000   0.000 ntp.ubuntu.com  .POOL.          16 p    -   64    0    0.000    0.000   0.000+ns.ustc.edu.cn  202.118.1.81     3 u  526  256  376   31.628  -10.231  24.584*dns1.synet.edu. 202.118.1.47     2 u    3  256  177   38.342  -19.102  13.452-85.199.214.100  .GPS.            1 u    1   64  377  225.205   24.569   3.503+202.118.1.130   202.118.1.46     2 u  246  256  377   37.415  -20.863  17.250-101.236.4.18    202.118.1.46     2 u  230  256  377    3.861    1.357   7.409