ceph config help mon_max_pg_per_osd查看默认值

admin

[root@compute01 ~]# ceph config help mon_max_pg_per_osd
( i8 x  |+ F) }mon_max_pg_per_osd - Max number of PGs per OSD the cluster will allow
  (uint, advanced)
* H9 d. J. C! n  Default: 250
/ p& H5 T$ K8 k  Minimum: 1
/ I5 X2 O  r0 E! ^! \  Maximum:
  H, i; g! s5 N- L/ D5 e  Can update at runtime: true
  Services: [mgr]
) a6 k7 E7 u& y7 Z! i& V- U1 lIf the number of PGs per OSD exceeds this, a health warning will be visible in `ceph status`.  This is also used in automated PG management, as the threshold at which some pools' pg_num may be shrunk in order to enable increasing the pg_num of others.


查看ceph的mon_max_pg_per_osd默认值：
ceph daemon mgr.`hostname -s` config get mon_max_pg_per_osd
{
    "mon_max_pg_per_osd": "1000"
}
& h6 x0 J' n4 A( k/ L7 ~- U
; o$ B. W8 ]7 N. T' z: K2 V7 d. F在线实时调整配置参数(重启后失效)

9 Q1 E4 f! q  J. H0 g; I6 X, g! m```
( ?7 s; x* d) C7 z. j. S[root@ceph ~]# ceph daemon mgr.ceph1 config set mon_max_pg_per_osd 300
{
1 k$ K/ W, G) B  V! s1 ?    "success": "mon_max_pg_per_osd = '300' (not observed, change may require restart) "
3 i0 E, d( z4 C7 o& d+ u}
```

  w. Q3 p7 n8 f6 a3 @) N# r
. M6 G) h2 e" z" v; W
启动保存配置
: V, l; D; X- Y2 o需要在/etc/ceph/ceph.conf加上以下内容

```
: j4 s; n' f4 [" _6 ^[mgr.ceph]
mon_max_pg_per_osd = 300
( `7 g; w6 D) j7 y: z```
% L: H7 N# c; M% X
5 t, b. _$ B1 Z2 B
6 T6 `1 `# `5 ?8 G( X- _

admin

查看ceph mon_max_pg_per_osd  最大连接数
; r1 ^% E& \' q$ h) z' Uceph daemon mgr.`hostname -s` config get mon_max_pg_per_osd
- h4 G- S4 \) X% i{
    "mon_max_pg_per_osd": "1000"
}

admin

[root@compute01 ~]#  ceph daemon mgr.compute01 config set mon_max_pg_per_osd 1000
{
9 Y: _) ^% l7 h9 p    "success": "mon_max_pg_per_osd = '1000' (not observed, change may require restart) "
}