主备守护集群搭建学习流程

本指南基于单台服务器单网卡（eth0, IP: 192.168.200.31）环境，通过端口错开的方式模拟双节点集群。

一、集群规划

配置项	主库 (GRP1_RT_01)	备库 (GRP1_RT_02)
业务/MAL IP	192.168.200.31	192.168.200.31
实例名	GRP1_RT_01	GRP1_RT_02
实例端口 (PORT_NUM)	5236	5237
MAL 端口 (MAL_PORT)	5336	5337
守护进程端口 (MAL_DW_PORT)	5436	5437
实例守护端口 (MAL_INST_DW_PORT)	5536	5537
OGUID	45331	45331
安装目录	/home/dmdba/dmdbms	/home/dmdba/dmdbms
数据目录	/dmdata/data/DAMENG	/dmdata/data1/DAMENG
归档目录	/dmdata/arch	/dmdata/arch1

二、初始化主库并备份数据

mkdir -p /dmdata/data /dmdata/arch /dmdata/dmbak
chown -R dmdba:dmdba /dmdata

初始化主库实例

su - dmdba
/home/dmdba/dmdbms/bin/dminit PATH=/dmdata/data INSTANCE_NAME=GRP1_RT_01 PAGE_SIZE=32 EXTENT_SIZE=32 LOG_SIZE=2048 SYSDBA_PWD=SYSDBA001 SYSAUDITOR_PWD=SYSDBA001

前台启动主库并开启归档

/home/dmdba/dmdbms/bin/dmserver /dmdata/data/DAMENG/dm.ini

在另开窗口执行disql登录：

/home/dmdba/dmdbms/bin/disql SYSDBA/SYSDBA001@192.168.200.31:5236

ALTER DATABASE MOUNT;
ALTER DATABASE ARCHIVELOG;
ALTER DATABASE ADD ARCHIVELOG 'DEST=/dmdata/arch, TYPE=LOCAL, FILE_SIZE=1024, SPACE_LIMIT=51200';
ALTER DATABASE OPEN;

备份主库数据

BACKUP DATABASE BACKUPSET '/dmdata/dmbak/BACKUP_FILE';
exit;

修改主库 dm.ini 参数
在 disql 中执行，或直接修改文件：

SP_SET_PARA_VALUE (2,'PORT_NUM',5236);
SP_SET_PARA_VALUE (2,'DW_INACTIVE_INTERVAL',60);
SP_SET_PARA_VALUE (2,'ALTER_MODE_STATUS',0);
SP_SET_PARA_VALUE (2,'ENABLE_OFFLINE_TS',2);
SP_SET_PARA_VALUE (2,'MAL_INI',1);
SP_SET_PARA_VALUE (2,'RLOG_SEND_APPLY_MON',64);
exit;

关闭前台启动的 dmserver 进程 (Ctrl+C)。

三、配置主库 INI 文件

配置 dmarch.ini

vi /dmdata/data/DAMENG/dmarch.ini

ARCH_WAIT_APPLY = 0
[ARCHIVE_LOCAL]
ARCH_TYPE = LOCAL
ARCH_DEST = /dmdata/arch/
ARCH_FILE_SIZE = 1024
ARCH_SPACE_LIMIT = 51200
[ARCHIVE_REALTIME1]
ARCH_TYPE = REALTIME
ARCH_DEST = GRP1_RT_02

配置 dmmal.ini

vi /dmdata/data/DAMENG/dmmal.ini

MAL_CHECK_INTERVAL = 10
MAL_CONN_FAIL_INTERVAL = 10
MAL_TEMP_PATH = /dmdata/data/malpath/
MAL_BUF_SIZE = 512
MAL_SYS_BUF_SIZE = 2048
MAL_COMPRESS_LEVEL = 0
[MAL_INST1]
MAL_INST_NAME = GRP1_RT_01
MAL_HOST = 192.168.200.31
MAL_PORT = 5336
MAL_INST_HOST = 192.168.200.31
MAL_INST_PORT = 5236
MAL_DW_PORT = 5436
MAL_INST_DW_PORT = 5536
[MAL_INST2]
MAL_INST_NAME = GRP1_RT_02
MAL_HOST = 192.168.200.31
MAL_PORT = 5337
MAL_INST_HOST = 192.168.200.31
MAL_INST_PORT = 5237
MAL_DW_PORT = 5437
MAL_INST_DW_PORT = 5537

配置 dmwatcher.ini

vi /dmdata/data/DAMENG/dmwatcher.ini

[GRP1]
DW_TYPE = GLOBAL
DW_MODE = AUTO
DW_ERROR_TIME = 20
INST_ERROR_TIME = 20
INST_RECOVER_TIME = 60
INST_OGUID = 45331
INST_INI = /dmdata/data/DAMENG/dm.ini
INST_AUTO_RESTART = 1
INST_STARTUP_CMD = /home/dmdba/dmdbms/bin/dmserver
RLOG_SEND_THRESHOLD = 0
RLOG_APPLY_THRESHOLD = 0

四、配置备库数据与 INI 文件

mkdir -p /dmdata/data1 /dmdata/arch1
chown -R dmdba:dmdba /dmdata/data1 /dmdata/arch1
# 使用 dmrman 恢复数据到备库目录
/home/dmdba/dmdbms/bin/dmrman CTLSTMT="RESTORE DATABASE '/dmdata/data1/DAMENG/dm.ini' FROM BACKUPSET '/dmdata/dmbak/BACKUP_FILE'"
# 注意：如果dm.ini尚未生成，需先初始化一个空实例或手工拷贝主库dm.ini后修改PORT_NUM等参数再执行恢复。
# 标准做法：初始化备库再覆盖恢复
/home/dmdba/dmdbms/bin/dminit PATH=/dmdata/data1 INSTANCE_NAME=GRP1_RT_02 PAGE_SIZE=32 EXTENT_SIZE=32 LOG_SIZE=2048 SYSDBA_PWD=SYSDBA001 SYSAUDITOR_PWD=SYSDBA001
/home/dmdba/dmdbms/bin/dmrman CTLSTMT="RESTORE DATABASE '/dmdata/data1/DAMENG/dm.ini' FROM BACKUPSET '/dmdata/dmbak/BACKUP_FILE'"
/home/dmdba/dmdbms/bin/dmrman CTLSTMT="RECOVER DATABASE '/dmdata/data1/DAMENG/dm.ini' FROM BACKUPSET '/dmdata/dmbak/BACKUP_FILE'"
/home/dmdba/dmdbms/bin/dmrman CTLSTMT="RECOVER DATABASE '/dmdata/data1/DAMENG/dm.ini' UPDATE DB_MAGIC"

配置备库 dm.ini
修改备库 /dmdata/data1/DAMENG/dm.ini，确保以下参数：

INSTANCE_NAME = GRP1_RT_02
PORT_NUM = 5237
DW_INACTIVE_INTERVAL = 60
ALTER_MODE_STATUS = 0
ENABLE_OFFLINE_TS = 2
MAL_INI = 1
ARCH_INI = 1
RLOG_SEND_APPLY_MON = 64

配置备库 dmarch.ini

vi /dmdata/data1/DAMENG/dmarch.ini

ARCH_WAIT_APPLY = 0
[ARCHIVE_LOCAL]
ARCH_TYPE = LOCAL
ARCH_DEST = /dmdata/arch1/
ARCH_FILE_SIZE = 1024
ARCH_SPACE_LIMIT = 51200
[ARCHIVE_REALTIME1]
ARCH_TYPE = REALTIME
ARCH_DEST = GRP1_RT_01

配置备库 dmmal.ini 与 dmwatcher.ini
备库的 dmmal.ini 和 dmwatcher.ini 内容与主库完全一致，直接拷贝并修改 dmwatcher.ini 中的 INST_INI 路径即可。

cp /dmdata/data/DAMENG/dmmal.ini /dmdata/data1/DAMENG/dmmal.ini
cp /dmdata/data/DAMENG/dmwatcher.ini /dmdata/data1/DAMENG/dmwatcher.ini

修改 /dmdata/data1/DAMENG/dmwatcher.ini 中的路径：

INST_INI = /dmdata/data1/DAMENG/dm.ini

五、注册服务与启动集群

注册后台服务 (root 用户执行)

# 注册主库服务
/home/dmdba/dmdbms/script/root/dm_service_installer.sh -t dmserver -p GRP1_RT_01 -dm_ini /dmdata/data/DAMENG/dm.ini -m mount
# 注册备库服务
/home/dmdba/dmdbms/script/root/dm_service_installer.sh -t dmserver -p GRP1_RT_02 -dm_ini /dmdata/data1/DAMENG/dm.ini -m mount
# 注册主库守护进程
/home/dmdba/dmdbms/script/root/dm_service_installer.sh -t dmwatcher -p Watcher01 -watcher_ini /dmdata/data/DAMENG/dmwatcher.ini
# 注册备库守护进程
/home/dmdba/dmdbms/script/root/dm_service_installer.sh -t dmwatcher -p Watcher02 -watcher_ini /dmdata/data1/DAMENG/dmwatcher.ini

启动数据库并设置主备模式 (dmdba 用户执行)

# 启动主库
/home/dmdba/dmdbms/bin/DmServiceGRP1_RT_01 start
# 启动备库
/home/dmdba/dmdbms/bin/DmServiceGRP1_RT_02 start
# 登录主库设置 OGUID 和模式
/home/dmdba/dmdbms/bin/disql SYSDBA/SYSDBA001@192.168.200.31:5236

SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 1);
SP_SET_OGUID(45331);
ALTER DATABASE PRIMARY;
SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 0);
exit;

# 登录备库设置 OGUID 和模式
/home/dmdba/dmdbms/bin/disql SYSDBA/SYSDBA001@192.168.200.31:5237

SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 1);
SP_SET_OGUID(45331);
ALTER DATABASE STANDBY;
SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 0);
exit;

启动守护进程

/home/dmdba/dmdbms/bin/DmWatcherServiceWatcher01 start
/home/dmdba/dmdbms/bin/DmWatcherServiceWatcher02 start

六、配置监视器

创建确认监视器配置文件

vi /home/dmdba/dmdbms/bin/dmmonitor.ini

MON_DW_CONFIRM = 1
MON_LOG_PATH = /home/dmdba/dmdbms/log
MON_LOG_INTERVAL = 60
MON_LOG_FILE_SIZE = 512
MON_LOG_SPACE_LIMIT = 2048
[GRP1]
MON_INST_OGUID = 45331
MON_DW_IP = 192.168.200.31:5436, 192.168.200.31:5437

创建非确认监视器配置文件 (用于日常交互)

vi /home/dmdba/dmdbms/bin/dmmonitor_manual.ini

MON_DW_CONFIRM = 0
MON_LOG_PATH = /home/dmdba/dmdbms/log
MON_LOG_INTERVAL = 60
MON_LOG_FILE_SIZE = 512
MON_LOG_SPACE_LIMIT = 2048
[GRP1]
MON_INST_OGUID = 45331
MON_DW_IP = 192.168.200.31:5436, 192.168.200.31:5437

注册并启动确认监视器后台服务 (root 用户执行)

/home/dmdba/dmdbms/script/root/dm_service_installer.sh -t dmmonitor -p Monitor -monitor_ini /home/dmdba/dmdbms/bin/dmmonitor.ini
su - dmdba -c "/home/dmdba/dmdbms/bin/DmMonitorServiceMonitor start"

使用非确认监视器查看集群状态

/home/dmdba/dmdbms/bin/dmmonitor /home/dmdba/dmdbms/bin/dmmonitor_manual.ini

输入 show 命令验证主备库状态是否均为 OPEN 且模式分别为 PRIMARY 和 STANDBY。

七、集群日常启停流程

停止集群流程：

停止确认监视器：/home/dmdba/dmdbms/bin/DmMonitorServiceMonitor stop
停止备库守护进程：/home/dmdba/dmdbms/bin/DmWatcherServiceWatcher02 stop
停止主库守护进程：/home/dmdba/dmdbms/bin/DmWatcherServiceWatcher01 stop
停止备库实例：/home/dmdba/dmdbms/bin/DmServiceGRP1_RT_02 stop
停止主库实例：/home/dmdba/dmdbms/bin/DmServiceGRP1_RT_01 stop
启动集群流程：
启动主库实例：/home/dmdba/dmdbms/bin/DmServiceGRP1_RT_01 start
启动备库实例：/home/dmdba/dmdbms/bin/DmServiceGRP1_RT_02 start
启动主库守护进程：/home/dmdba/dmdbms/bin/DmWatcherServiceWatcher01 start
启动备库守护进程：/home/dmdba/dmdbms/bin/DmWatcherServiceWatcher02 start
启动确认监视器：/home/dmdba/dmdbms/bin/DmMonitorServiceMonitor start