磐维分布式数据库日常维护命令

原创飞天 2025-03-09

磐维数据库官网

查看数据库集群整体运行状态

语法：ha_ctl monitor all  [-H] [-I] [-c clustername] -l dcslist
说明：如果创建集群的时候，集群名是panwei，则在命令行中可以省略-c clustername，否则需要加入-c clustername。
示例：ha_ctl monitor all -l http://192.*.*.60:2379 -HI
复制

示例如下：

[omm@node1 ~]$ ha_ctl monitor all -l http://192.*.*.60:2379 -HI
+----+------------+----------------+-------+---------+--------+
| No |    name    |      host      |  port |  state  | leader |
+----+------------+----------------+-------+---------+--------+
| 0  | ha_server1 | 192.*.*.60 | 20001 | running |  True  |
+----+------------+----------------+-------+---------+--------+
+----+------+----------------+------+----------------+---------+---------+
| No | name |      host      | port |    work_dir    |  state  |   role  |
+----+------+----------------+------+----------------+---------+---------+
| 0  | gtm1 | 192.*.*.60 | 6666 | /data/gtm/gtm1 | running | primary |
| 1  | gtm2 | 192.*.*.62 | 6666 | /data/gtm/gtm2 | running | standby |
+----+------+----------------+------+----------------+---------+---------+
+----+------+----------------+------+-----------------+---------+---------+
| No | name |      host      | port |     work_dir    |  state  |   role  |
+----+------+----------------+------+-----------------+---------+---------+
| 0  | cn1  | 192.*.*.62 | 5432 | /data/coord/cn1 | running | primary |
| 1  | cn2  | 192.*.*.64 | 5432 | /data/coord/cn2 | running | primary |
+----+------+----------------+------+-----------------+---------+---------+
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| No | group |  name |      host      |  port |     work_dir    |  state  |   role  |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| 0  |  dn1  | dn1_1 | 192.*.*.62 | 15432 | /data/dn1/dn1_1 | running | primary |
| 1  |  dn1  | dn1_2 | 192.*.*.64 | 15432 | /data/dn1/dn1_2 | running | standby |
| 2  |  dn1  | dn1_3 | 192.*.*.60 | 15432 | /data/dn1/dn1_3 | running | standby |
| 3  |  dn2  | dn2_1 | 192.*.*.64 | 20010 | /data/dn2/dn2_1 | running | primary |
| 4  |  dn2  | dn2_2 | 192.*.*.60 | 20010 | /data/dn2/dn2_2 | running | standby |
| 5  |  dn2  | dn2_3 | 192.*.*.62 | 20010 | /data/dn2/dn2_3 | running | standby |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
+----+----------------------------+--------+---------+----------+
| No |            url             |  name  |  state  | isLeader |
+----+----------------------------+--------+---------+----------+
| 0  | http://192.*.*.64:2379 | node_2 | healthy |  False   |
| 1  | http://192.*.*.62:2379 | node_1 | healthy |  False   |
| 2  | http://192.*.*.60:2379 | node_0 | healthy |   True   |
+----+----------------------------+--------+---------+----------+
复制

查看各组件的状态

查看ha server的运行状态

ha_ctl monitor server -l http://192.*.*.60:2379 -HI
复制

示例如下：

[omm@node1 ~]$ ha_ctl monitor server -l http://192.*.*.60:2379 -HI
+----+------------+----------------+-------+---------+--------+
| No |    name    |      host      |  port |  state  | leader |
+----+------------+----------------+-------+---------+--------+
| 0  | ha_server1 | 192.*.*.60 | 20001 | running |  True  |
+----+------------+----------------+-------+---------+--------+
复制

查看gtm的运行状态

ha_ctl monitor gtm -l http://192.*.*.60:2379 -HI
复制

示例如下：

[omm@node1 ~]$ ha_ctl monitor gtm -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+----------------+---------+---------+
| No | name |      host      | port |    work_dir    |  state  |   role  |
+----+------+----------------+------+----------------+---------+---------+
| 0  | gtm1 | 192.*.*.60 | 6666 | /data/gtm/gtm1 | running | primary |
| 1  | gtm2 | 192.*.*.62 | 6666 | /data/gtm/gtm2 | running | standby |
+----+------+----------------+------+----------------+---------+---------+
[omm@node1 ~]$ 
复制

查看dcs的运行状态

ha_ctl monitor dcs -l http://192.*.*.60:2379 -HI
复制

示例如下：

[omm@node1 ~]$ ha_ctl monitor dcs -l http://192.*.*.60:2379 -HI
+----+----------------------------+--------+---------+----------+
| No |            url             |  name  |  state  | isLeader |
+----+----------------------------+--------+---------+----------+
| 0  | http://192.*.*.64:2379 | node_2 | healthy |  False   |
| 1  | http://192.*.*.62:2379 | node_1 | healthy |  False   |
| 2  | http://192.*.*.60:2379 | node_0 | healthy |   True   |
+----+----------------------------+--------+---------+----------+
复制

查看cn节点的运行状态

ha_ctl monitor coordinator -l http://192.*.*.60:2379 -HI
复制

示例如下：

[omm@node1 ~]$ ha_ctl monitor coordinator -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+-----------------+---------+---------+
| No | name |      host      | port |     work_dir    |  state  |   role  |
+----+------+----------------+------+-----------------+---------+---------+
| 0  | cn1  | 192.*.*.62 | 5432 | /data/coord/cn1 | running | primary |
| 1  | cn2  | 192.*.*.64 | 5432 | /data/coord/cn2 | running | primary |
+----+------+----------------+------+-----------------+---------+---------+
复制

查看dn节点的运行状态

ha_ctl monitor datanode -l http://192.*.*.60:2379 -HI
复制

示例如下：

[omm@node1 ~]$ ha_ctl monitor datanode -l http://192.*.*.60:2379 -HI
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| No | group |  name |      host      |  port |     work_dir    |  state  |   role  |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| 0  |  dn1  | dn1_1 | 192.*.*.62 | 15432 | /data/dn1/dn1_1 | running | primary |
| 1  |  dn1  | dn1_2 | 192.*.*.64 | 15432 | /data/dn1/dn1_2 | running | standby |
| 2  |  dn1  | dn1_3 | 192.*.*.60 | 15432 | /data/dn1/dn1_3 | running | standby |
| 3  |  dn2  | dn2_1 | 192.*.*.64 | 20010 | /data/dn2/dn2_1 | running | primary |
| 4  |  dn2  | dn2_2 | 192.*.*.60 | 20010 | /data/dn2/dn2_2 | running | standby |
| 5  |  dn2  | dn2_3 | 192.*.*.62 | 20010 | /data/dn2/dn2_3 | running | standby |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
复制

停止整个数据库集群

ha_ctl stop all -l http://192.*.*.60:2379 
复制

示例如下：

[omm@node1 ~]$ ha_ctl stop all -l http://192.*.*.60:2379 
{
    "ret":0,
    "msg":"Success"
}
[omm@node1 ~]$  ha_ctl monitor all -l http://192.*.*.60:2379 -HI
+----+------------+----------------+-------+---------+--------+
| No |    name    |      host      |  port |  state  | leader |
+----+------------+----------------+-------+---------+--------+
| 0  | ha_server1 | 192.*.*.60 | 20001 | stopped | False  |
+----+------------+----------------+-------+---------+--------+
+----+------+----------------+------+----------------+---------+---------+
| No | name |      host      | port |    work_dir    |  state  |   role  |
+----+------+----------------+------+----------------+---------+---------+
| 0  | gtm1 | 192.*.*.60 | 6666 | /data/gtm/gtm1 | stopped | standby |
| 1  | gtm2 | 192.*.*.62 | 6666 | /data/gtm/gtm2 | stopped | primary |
+----+------+----------------+------+----------------+---------+---------+
+----+------+----------------+------+-----------------+---------+---------+
| No | name |      host      | port |     work_dir    |  state  |   role  |
+----+------+----------------+------+-----------------+---------+---------+
| 0  | cn1  | 192.*.*.62 | 5432 | /data/coord/cn1 | stopped | primary |
| 1  | cn2  | 192.*.*.64 | 5432 | /data/coord/cn2 | stopped | primary |
+----+------+----------------+------+-----------------+---------+---------+
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| No | group |  name |      host      |  port |     work_dir    |  state  |   role  |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| 0  |  dn1  | dn1_1 | 192.*.*.62 | 15432 | /data/dn1/dn1_1 | stopped | primary |
| 1  |  dn1  | dn1_2 | 192.*.*.64 | 15432 | /data/dn1/dn1_2 | stopped | standby |
| 2  |  dn1  | dn1_3 | 192.*.*.60 | 15432 | /data/dn1/dn1_3 | stopped | standby |
| 3  |  dn2  | dn2_1 | 192.*.*.64 | 20010 | /data/dn2/dn2_1 | stopped | primary |
| 4  |  dn2  | dn2_2 | 192.*.*.60 | 20010 | /data/dn2/dn2_2 | stopped | standby |
| 5  |  dn2  | dn2_3 | 192.*.*.62 | 20010 | /data/dn2/dn2_3 | stopped | standby |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
+----+----------------------------+--------+---------+----------+
| No |            url             |  name  |  state  | isLeader |
+----+----------------------------+--------+---------+----------+
| 0  | http://192.*.*.64:2379 | node_2 | healthy |   True   |
| 1  | http://192.*.*.62:2379 | node_1 | healthy |  False   |
| 2  | http://192.*.*.60:2379 | node_0 | healthy |  False   |
+----+----------------------------+--------+---------+----------+
复制

启动整个数据库集群

ha_ctl start all -l http://192.*.*.60:2379 
复制

示例如下：

[omm@node1 ~]$ ha_ctl start all -l http://192.*.*.60:2379 
{
    "ret":0,
    "msg":"Success"
}
[omm@node1 ~]$  ha_ctl monitor all -l http://192.*.*.60:2379 -HI
+----+------------+----------------+-------+---------+--------+
| No |    name    |      host      |  port |  state  | leader |
+----+------------+----------------+-------+---------+--------+
| 0  | ha_server1 | 192.*.*.60 | 20001 | running |  True  |
+----+------------+----------------+-------+---------+--------+
+----+------+----------------+------+----------------+---------+---------+
| No | name |      host      | port |    work_dir    |  state  |   role  |
+----+------+----------------+------+----------------+---------+---------+
| 0  | gtm1 | 192.*.*.60 | 6666 | /data/gtm/gtm1 | running | standby |
| 1  | gtm2 | 192.*.*.62 | 6666 | /data/gtm/gtm2 | running | primary |
+----+------+----------------+------+----------------+---------+---------+
+----+------+----------------+------+-----------------+---------+---------+
| No | name |      host      | port |     work_dir    |  state  |   role  |
+----+------+----------------+------+-----------------+---------+---------+
| 0  | cn1  | 192.*.*.62 | 5432 | /data/coord/cn1 | running | primary |
| 1  | cn2  | 192.*.*.64 | 5432 | /data/coord/cn2 | running | primary |
+----+------+----------------+------+-----------------+---------+---------+
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| No | group |  name |      host      |  port |     work_dir    |  state  |   role  |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| 0  |  dn1  | dn1_1 | 192.*.*.62 | 15432 | /data/dn1/dn1_1 | running | primary |
| 1  |  dn1  | dn1_2 | 192.*.*.64 | 15432 | /data/dn1/dn1_2 | running | standby |
| 2  |  dn1  | dn1_3 | 192.*.*.60 | 15432 | /data/dn1/dn1_3 | running | standby |
| 3  |  dn2  | dn2_1 | 192.*.*.64 | 20010 | /data/dn2/dn2_1 | running | primary |
| 4  |  dn2  | dn2_2 | 192.*.*.60 | 20010 | /data/dn2/dn2_2 | running | standby |
| 5  |  dn2  | dn2_3 | 192.*.*.62 | 20010 | /data/dn2/dn2_3 | running | standby |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
+----+----------------------------+--------+---------+----------+
| No |            url             |  name  |  state  | isLeader |
+----+----------------------------+--------+---------+----------+
| 0  | http://192.*.*.64:2379 | node_2 | healthy |  False   |
| 1  | http://192.*.*.62:2379 | node_1 | healthy |   True   |
| 2  | http://192.*.*.60:2379 | node_0 | healthy |  False   |
+----+----------------------------+--------+---------+----------+
[omm@node1 ~]$ 
复制

gtm相关操作

停止gtm

停止所有gtm：
ha_ctl stop  gtm  all  -l http://192.*.*.60:2379
停止某个gtm：
ha_ctl stop  gtm  gtm1 -l http://192.*.*.60:2379
ha_ctl stop  gtm  gtm2 -l http://192.*.*.60:2379
复制

示例如下：

[omm@node1 ~]$ ha_ctl monitor gtm -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+----------------+---------+---------+
| No | name |      host      | port |    work_dir    |  state  |   role  |
+----+------+----------------+------+----------------+---------+---------+
| 0  | gtm1 | 192.*.*.60 | 6666 | /data/gtm/gtm1 | running | primary |
| 1  | gtm2 | 192.*.*.62 | 6666 | /data/gtm/gtm2 | running | standby |
+----+------+----------------+------+----------------+---------+---------+
可以看到，当前gtm主节点为gtm1，下面停止gtm1：
[omm@node1 ~]$ ha_ctl stop  gtm  gtm1 -l http://192.*.*.60:2379
{
    "ret":0,
    "msg":"Success"
}
stop 命令执行成功。
[omm@node1 ~]$ ha_ctl monitor gtm -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+----------------+---------+---------+
| No | name |      host      | port |    work_dir    |  state  |   role  |
+----+------+----------------+------+----------------+---------+---------+
| 0  | gtm1 | 192.*.*.60 | 6666 | /data/gtm/gtm1 | stopped | standby |
| 1  | gtm2 | 192.*.*.62 | 6666 | /data/gtm/gtm2 | running | primary |
+----+------+----------------+------+----------------+---------+---------+
可以看到，gtm1被停止，状态为stopped，gtm2被提升成gtm主节点。
复制

启动gtm

启动所有gtm：
ha_ctl start  gtm  all  -l http://192.*.*.60:2379
启动某个gtm：
ha_ctl start  gtm  gtm1 -l http://192.*.*.60:2379
ha_ctl start  gtm  gtm2 -l http://192.*.*.60:2379
复制

示例如下：

[omm@node1 ~]$ ha_ctl monitor gtm -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+----------------+---------+---------+
| No | name |      host      | port |    work_dir    |  state  |   role  |
+----+------+----------------+------+----------------+---------+---------+
| 0  | gtm1 | 192.*.*.60 | 6666 | /data/gtm/gtm1 | stopped | standby |
| 1  | gtm2 | 192.*.*.62 | 6666 | /data/gtm/gtm2 | running | primary |
+----+------+----------------+------+----------------+---------+---------+
可以看到，当前gtm1节点状态为stopped，下面启动gtm1：
[omm@node1 ~]$ ha_ctl start  gtm  gtm1 -l http://192.*.*.60:2379
{
    "ret":0,
    "msg":"Success"
}
start 命令执行成功。
[omm@node1 ~]$ ha_ctl monitor gtm -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+----------------+---------+---------+
| No | name |      host      | port |    work_dir    |  state  |   role  |
+----+------+----------------+------+----------------+---------+---------+
| 0  | gtm1 | 192.*.*.60 | 6666 | /data/gtm/gtm1 | running | standby |
| 1  | gtm2 | 192.*.*.62 | 6666 | /data/gtm/gtm2 | running | primary |
+----+------+----------------+------+----------------+---------+---------+
可以看到，gtm1被启动，状态为running。
复制

切换gtm

切换命令为：gtm gtm1 gtm2将当前主节点从gtm1切换到gtm2上
ha_ctl switchover gtm gtm1 gtm2 -l http://192.*.*.60:2379
复制

示例如下：

# 查看gtm
[omm@node1 ~]$ ha_ctl monitor gtm -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+----------------+---------+---------+
| No | name |      host      | port |    work_dir    |  state  |   role  |
+----+------+----------------+------+----------------+---------+---------+
| 0  | gtm1 | 192.*.*.60 | 6666 | /data/gtm/gtm1 | running | primary |
| 1  | gtm2 | 192.*.*.62 | 6666 | /data/gtm/gtm2 | running | standby |
+----+------+----------------+------+----------------+---------+---------+
可以看到，目前gtm主运行在了gtm1（192.*.*.60）节点。下面执行switchover命令把gtm主节点从gtm1切换到gtm2：
[omm@node1 ~]$ ha_ctl switchover gtm gtm1 gtm2 -l http://192.*.*.60:2379
{
    "ret":0,
    "msg":"Success"
}
switchover命令执行成功。
[omm@node1 ~]$ ha_ctl monitor gtm -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+----------------+---------+---------+
| No | name |      host      | port |    work_dir    |  state  |   role  |
+----+------+----------------+------+----------------+---------+---------+
| 0  | gtm1 | 192.*.*.60 | 6666 | /data/gtm/gtm1 | running | standby |
| 1  | gtm2 | 192.*.*.62 | 6666 | /data/gtm/gtm2 | running | primary |
+----+------+----------------+------+----------------+---------+---------+
可以看到，switchover后，gtm主运行在了gtm2 （192.*.*.62）节点。
复制

cn相关操作

停止cn

停止所有cn：
ha_ctl stop coordinator all  -l http://192.*.*.60:2379
停止某个cn：
ha_ctl stop coordinator cn1  -l http://192.*.*.60:2379
ha_ctl stop coordinator cn2  -l http://192.*.*.60:2379
复制

示例如下：

[omm@node1 ~]$  ha_ctl monitor coordinator -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+-----------------+---------+---------+
| No | name |      host      | port |     work_dir    |  state  |   role  |
+----+------+----------------+------+-----------------+---------+---------+
| 0  | cn1  | 192.*.*.62 | 5432 | /data/coord/cn1 | running | primary |
| 1  | cn2  | 192.*.*.64 | 5432 | /data/coord/cn2 | running | primary |
+----+------+----------------+------+-----------------+---------+---------+
可以看到，cn1和cn2的状态都为running。
[omm@node1 ~]$ ha_ctl stop coordinator cn1  -l http://192.*.*.60:2379
{
    "ret":0,
    "msg":"Success"
}
stop命令执行成功。
[omm@node1 ~]$  ha_ctl monitor coordinator -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+-----------------+---------+---------+
| No | name |      host      | port |     work_dir    |  state  |   role  |
+----+------+----------------+------+-----------------+---------+---------+
| 0  | cn1  | 192.*.*.62 | 5432 | /data/coord/cn1 | stopped | primary |
| 1  | cn2  | 192.*.*.64 | 5432 | /data/coord/cn2 | running | primary |
+----+------+----------------+------+-----------------+---------+---------+
可以看到，stop执行后，cn1的运行状态为stopped。
复制

启动cn

启动所有cn：
ha_ctl start coordinator all  -l http://192.*.*.60:2379
启动某个cn：
ha_ctl start coordinator cn1  -l http://192.*.*.60:2379
复制

示例如下：

[omm@node1 ~]$  ha_ctl monitor coordinator -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+-----------------+---------+---------+
| No | name |      host      | port |     work_dir    |  state  |   role  |
+----+------+----------------+------+-----------------+---------+---------+
| 0  | cn1  | 192.*.*.62 | 5432 | /data/coord/cn1 | stopped | primary |
| 1  | cn2  | 192.*.*.64 | 5432 | /data/coord/cn2 | running | primary |
+----+------+----------------+------+-----------------+---------+---------+
可以看到，cn1的运行状态为stopped。
[omm@node1 ~]$ ha_ctl start coordinator cn1  -l http://192.*.*.60:2379
{
    "ret":0,
    "msg":"Success"
}
start命令执行成功。
[omm@node1 ~]$  ha_ctl monitor coordinator -l http://192.*.*.60:2379 -HI
+----+------+----------------+------+-----------------+---------+---------+
| No | name |      host      | port |     work_dir    |  state  |   role  |
+----+------+----------------+------+-----------------+---------+---------+
| 0  | cn1  | 192.*.*.62 | 5432 | /data/coord/cn1 | running | primary |
| 1  | cn2  | 192.*.*.64 | 5432 | /data/coord/cn2 | running | primary |
+----+------+----------------+------+-----------------+---------+---------+
[omm@node1 ~]$ 
可以看到，cn1的运行状态为running。
复制

dn相关操作

停止dn

ha_ctl stop  datanode GROUP_NAME NAME  -l http://192.*.*.60:2379
示例：
ha_ctl stop  datanode dn1 dn1_1  -l http://192.*.*.60:2379
复制

示例如下：

[omm@node1 ~]$ ha_ctl monitor datanode -l http://192.*.*.60:2379 -HI
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| No | group |  name |      host      |  port |     work_dir    |  state  |   role  |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| 0  |  dn1  | dn1_1 | 192.*.*.62 | 15432 | /data/dn1/dn1_1 | running | primary |
| 1  |  dn1  | dn1_2 | 192.*.*.64 | 15432 | /data/dn1/dn1_2 | running | standby |
| 2  |  dn1  | dn1_3 | 192.*.*.60 | 15432 | /data/dn1/dn1_3 | running | standby |
| 3  |  dn2  | dn2_1 | 192.*.*.64 | 20010 | /data/dn2/dn2_1 | running | primary |
| 4  |  dn2  | dn2_2 | 192.*.*.60 | 20010 | /data/dn2/dn2_2 | running | standby |
| 5  |  dn2  | dn2_3 | 192.*.*.62 | 20010 | /data/dn2/dn2_3 | running | standby |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
可以看到，所有dn的运行状态都为running。
[omm@node1 ~]$ ha_ctl stop  datanode dn1 dn1_1  -l http://192.*.*.60:2379
{
    "ret":0,
    "msg":"Success"
}
stop命令执行成功。
[omm@node1 ~]$ ha_ctl monitor datanode -l http://192.*.*.60:2379 -HI
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| No | group |  name |      host      |  port |     work_dir    |  state  |   role  |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| 0  |  dn1  | dn1_1 | 192.*.*.62 | 15432 | /data/dn1/dn1_1 | stopped | standby |
| 1  |  dn1  | dn1_2 | 192.*.*.64 | 15432 | /data/dn1/dn1_2 | running | primary |
| 2  |  dn1  | dn1_3 | 192.*.*.60 | 15432 | /data/dn1/dn1_3 | running | standby |
| 3  |  dn2  | dn2_1 | 192.*.*.64 | 20010 | /data/dn2/dn2_1 | running | primary |
| 4  |  dn2  | dn2_2 | 192.*.*.60 | 20010 | /data/dn2/dn2_2 | running | standby |
| 5  |  dn2  | dn2_3 | 192.*.*.62 | 20010 | /data/dn2/dn2_3 | running | standby |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
可以看到，dn1_1的运行状态为stopped，同时它的role切换为standby，dn1_2的role提升为primary。
复制

启动dn

ha_ctl start  datanode GROUP_NAME NAME  -l http://192.*.*.60:2379
示例：
ha_ctl start  datanode dn1 dn1_1  -l http://192.*.*.60:2379
复制

示例如下：

[omm@node1 ~]$ ha_ctl monitor datanode -l http://192.*.*.60:2379 -HI
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| No | group |  name |      host      |  port |     work_dir    |  state  |   role  |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| 0  |  dn1  | dn1_1 | 192.*.*.62 | 15432 | /data/dn1/dn1_1 | stopped | standby |
| 1  |  dn1  | dn1_2 | 192.*.*.64 | 15432 | /data/dn1/dn1_2 | running | primary |
| 2  |  dn1  | dn1_3 | 192.*.*.60 | 15432 | /data/dn1/dn1_3 | running | standby |
| 3  |  dn2  | dn2_1 | 192.*.*.64 | 20010 | /data/dn2/dn2_1 | running | primary |
| 4  |  dn2  | dn2_2 | 192.*.*.60 | 20010 | /data/dn2/dn2_2 | running | standby |
| 5  |  dn2  | dn2_3 | 192.*.*.62 | 20010 | /data/dn2/dn2_3 | running | standby |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
可以看到，dn1_1的运行状态为stopped。
[omm@node1 ~]$ ha_ctl start  datanode dn1 dn1_1  -l http://192.*.*.60:2379
{
    "ret":0,
    "msg":"Success"
}
start命令执行成功。
[omm@node1 ~]$ ha_ctl monitor datanode -l http://192.*.*.60:2379 -HI
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| No | group |  name |      host      |  port |     work_dir    |  state  |   role  |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| 0  |  dn1  | dn1_1 | 192.*.*.62 | 15432 | /data/dn1/dn1_1 | running | standby |
| 1  |  dn1  | dn1_2 | 192.*.*.64 | 15432 | /data/dn1/dn1_2 | running | primary |
| 2  |  dn1  | dn1_3 | 192.*.*.60 | 15432 | /data/dn1/dn1_3 | running | standby |
| 3  |  dn2  | dn2_1 | 192.*.*.64 | 20010 | /data/dn2/dn2_1 | running | primary |
| 4  |  dn2  | dn2_2 | 192.*.*.60 | 20010 | /data/dn2/dn2_2 | running | standby |
| 5  |  dn2  | dn2_3 | 192.*.*.62 | 20010 | /data/dn2/dn2_3 | running | standby |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
可以看到，dn1_1的运行状态为running。
复制

切换dn

切换命令为：datanode dn1 dn1_2  dn1_1将当前dn1的主分片从dn1_2切换到dn1_1上
ha_ctl switchover  datanode dn1 dn1_2  dn1_1 -l http://192.*.*.60:2379
复制

示例如下：

[omm@node1 ~]$  ha_ctl monitor datanode -l http://192.*.*.60:2379 -HI
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| No | group |  name |      host      |  port |     work_dir    |  state  |   role  |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| 0  |  dn1  | dn1_1 | 192.*.*.62 | 15432 | /data/dn1/dn1_1 | running | standby |
| 1  |  dn1  | dn1_2 | 192.*.*.64 | 15432 | /data/dn1/dn1_2 | running | primary |
| 2  |  dn1  | dn1_3 | 192.*.*.60 | 15432 | /data/dn1/dn1_3 | running | standby |
| 3  |  dn2  | dn2_1 | 192.*.*.64 | 20010 | /data/dn2/dn2_1 | running | primary |
| 4  |  dn2  | dn2_2 | 192.*.*.60 | 20010 | /data/dn2/dn2_2 | running | standby |
| 5  |  dn2  | dn2_3 | 192.*.*.62 | 20010 | /data/dn2/dn2_3 | running | standby |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
可以看到，当前dn1_1分片的role为standby，dn1_2分片的role为primary。下面将dn1的主分片切换到 dn1_1。
[omm@node1 ~]$ ha_ctl switchover  datanode dn1 dn1_2  dn1_1 -l http://192.*.*.60:2379
{
    "ret":0,
    "msg":"Success"
}
switchover命令执行成功。

[omm@node1 ~]$  ha_ctl monitor datanode -l http://192.*.*.60:2379 -HI
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| No | group |  name |      host      |  port |     work_dir    |  state  |   role  |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
| 0  |  dn1  | dn1_1 | 192.*.*.62 | 15432 | /data/dn1/dn1_1 | running | primary |
| 1  |  dn1  | dn1_2 | 192.*.*.64 | 15432 | /data/dn1/dn1_2 | running | standby |
| 2  |  dn1  | dn1_3 | 192.*.*.60 | 15432 | /data/dn1/dn1_3 | running | standby |
| 3  |  dn2  | dn2_1 | 192.*.*.64 | 20010 | /data/dn2/dn2_1 | running | primary |
| 4  |  dn2  | dn2_2 | 192.*.*.60 | 20010 | /data/dn2/dn2_2 | running | standby |
| 5  |  dn2  | dn2_3 | 192.*.*.62 | 20010 | /data/dn2/dn2_3 | running | standby |
+----+-------+-------+----------------+-------+-----------------+---------+---------+
可以看到，dn1_1分片的role由standby被切换为primary，dn1_2分片的role由primary被切换为standby。
复制

总结

在生产系统执行命令的时候务必要小心谨慎，清楚命令的真实用途，并再三确认执行的命令是否正确，否则可能会造成严重的后果。高危操作在申请的割接窗口执行，切记！

关于作者：
网名：飞天，墨天轮2024年度优秀原创作者，拥有 Oracle 10g OCM 认证、PGCE认证以及OBCA、KCP、ACP、磐维等众多国产数据库认证证书，目前从事Oracle、Mysql、PostgresSQL、磐维数据库管理运维工作，喜欢结交更多志同道合的朋友，热衷于研究、分享数据库技术。
微信公众号：飞天online
墨天轮：https://www.modb.pro/u/15197
如有任何疑问，欢迎大家留言，共同探讨~~~

墨力计划磐维数据库

最后修改时间：2025-03-10 17:43:46

「喜欢这篇文章，您的关注和赞赏是给作者最好的鼓励」

关注作者

文章被以下合辑收录

磐维数据库（共46篇）

磐维数据库，简称"PanWeiDB"。是中国移动信息技术中心首个基于中国本土开源数据库打造的面向ICT基础设施的自研数据库产品。其产品内核能力基于华为openGauss开源软件，并进一步提升了系统稳定性。磐维数据库具有高性能、高可靠、高安全、高兼容等特点，能够满足复杂多变的业务需求。磐维数据库提供了自动化、流程化的解决方案，实现了一键式数据迁移。这种高效的数据迁移方式不仅提高了迁移数据的效率，也降低了操作难度，为用户带来了极大的便利。

磐维数据库官网
查看数据库集群整体运行状态
查看各组件的状态
- 查看ha server的运行状态
- 查看gtm的运行状态
- 查看dcs的运行状态
- 查看cn节点的运行状态
- 查看dn节点的运行状态
停止整个数据库集群
启动整个数据库集群
gtm相关操作
- 停止gtm
- 启动gtm
- 切换gtm
cn相关操作
- 停止cn
- 启动cn
dn相关操作
- 停止dn
- 启动dn
- 切换dn
总结

磐维分布式数据库日常维护命令

磐维数据库官网

查看数据库集群整体运行状态

查看各组件的状态

查看ha server的运行状态

查看gtm的运行状态

查看dcs的运行状态

查看cn节点的运行状态

查看dn节点的运行状态

停止整个数据库集群

启动整个数据库集群

gtm相关操作

停止gtm

启动gtm

切换gtm

cn相关操作

停止cn

启动cn

dn相关操作

停止dn

启动dn

切换dn

总结

文章被以下合辑收录

评论