MPP主备集群搭建

部署环境

机器名	IP地址	初始状态	操作系统
A	8.142.141.128 172.21.215.74 (内网IP)	主库 GRP1_MPP_EP01 备库 GRP2_MPP_EP22	Debian 9
B	39.103.139.80 172.17.44.74 (内网IP)	主库 GRP1_MPP_EP02 备库 GRP2_MPP_EP11	Debian 9
C	本机	监视器	windows 11

部署规划

主库

实例名	PORT_NUM	MAL_INST_DW_PORT	MAL_HOST	MAL_PORT	MAL_DW_PORT	MPP_SEQNO
GRP1_MPP_EP01	5237	5243	172.21.215.74	5337	5253	0
GRP2_MPP_EP02	5237	5243	172.17.44.74	5337	5253	1

备库

实例名	PORT_NUM	MAL_INST_DW_PORT	MAL_HOST	MAL_PORT	MAL_DW_PORT	MPP_SEQNO
GRP1_MPP_EP11	5238	5244	172.17.44.74	5338	5254	0
GRP2_MPP_EP22	5238	5244	172.21.215.74	5338	5254	1

守护进程规划

组名	实例名	所在机器
GRP1	GRP1_MPP_EP01	172.21.215.74
	GRP1_MPP_EP11	172.17.44.74
GRP2	GRP2_MPP_EP02	172.17.44.74
	GRP2_MPP_EP22	172.21.215.74

数据准备

初始化实例

主库

--A机器
dminit path=/dm8/data/ PAGE_SIZE=32 EXTENT_SIZE=32 CASE_SENSITIVE=y CHARSET=1 DB_NAME=EP01 INSTANCE_NAME=GRP1_MPP_EP01 PORT_NUM=5237

--B机器
dminit path=/dm8/data/ PAGE_SIZE=32 EXTENT_SIZE=32 CASE_SENSITIVE=y CHARSET=1 DB_NAME=EP02 INSTANCE_NAME=GRP2_MPP_EP02 PORT_NUM=5237

备库

--A机器
dminit path=/dm8/data/ PAGE_SIZE=32 EXTENT_SIZE=32 CASE_SENSITIVE=y CHARSET=1 DB_NAME=EP22 INSTANCE_NAME=GRP2_MPP_EP22 PORT_NUM=5238

--B机器
dminit path=/dm8/data/ PAGE_SIZE=32 EXTENT_SIZE=32 CASE_SENSITIVE=y CHARSET=1 DB_NAME=EP11 INSTANCE_NAME=GRP1_MPP_EP11 PORT_NUM=5238

注册服务

主库

./dm_service_installer.sh -t dmserver -dm_ini /dm8/data/EP01/dm.ini -p EP01
./dm_service_installer.sh -t dmserver -dm_ini /dm8/data/EP02/dm.ini -p EP02

备库

./dm_service_installer.sh -t dmserver -dm_ini /dm8/data/EP22/dm.ini -p EP22
./dm_service_installer.sh -t dmserver -dm_ini /dm8/data/EP11/dm.ini -p EP11

启动主库

root@lxm2:/home/dmdba/dmdbms/script/root# systemctl start DmServiceEP01.service 
root@lxm:/home/dmdba/dmdbms/script/root# systemctl start DmServiceEP02.service

配置归档

--A机器
dmdba@lxm2:~$ rlwrap disql SYSDBA/SYSDBA@localhost:5237

Server[localhost:5237]:mode is normal, state is open
login used time : 2.574(ms)
disql V8
SQL> ALTER DATABASE MOUNT;
executed successfully
used time: 0.643(ms). Execute id is 0.
SQL> ALTER DATABASE ADD ARCHIVELOG 'dest=/dm8/data/EP01/dm_arch,type=local,file_size=64,space_limit=20480';
executed successfully
used time: 3.270(ms). Execute id is 0.
SQL> alter database archivelog;
executed successfully
used time: 7.166(ms). Execute id is 0.
SQL> alter database open;
executed successfully
used time: 18.316(ms). Execute id is 0.
SQL> SELECT ARCH_MODE FROM V$DATABASE;

LINEID     ARCH_MODE
---------- ---------
1          Y

used time: 2.610(ms). Execute id is 55400.


--B机器
ALTER DATABASE MOUNT;
ALTER DATABASE ADD ARCHIVELOG 'dest=/dm8/data/EP02/dm_arch,type=local,file_size=64,space_limit=20480';
alter database archivelog;
alter database open;
SELECT ARCH_MODE FROM V$DATABASE;

备份还原

联机备份EP01并脱机还原到EP11

--备份
SQL> BACKUP DATABASE BACKUPSET '/dm8/data/BACKUP_FILE_01';
executed successfully
used time: 00:00:03.753. Execute id is 55501.
SQL> exit

--拷贝
dmdba@lxm2:/dm8/data$ scp -r BACKUP_FILE_01/ dmdba@39.103.139.80:/dm8/data
dmdba@39.103.139.80's password: 
BACKUP_FILE_01.bak                                                                                                                                       100%   89MB  13.4MB/s   00:06    
BACKUP_FILE_01.meta                                                                                                                                      100%   89KB   8.5MB/s   00:00    
BACKUP_FILE_01_1.bak                                                                                                                                     100%  485KB  15.0MB/s   00:00    
--备机还原
dmdba@lxm:~$ rlwrap dmrman
dmrman V8
RESTORE DATABASE '/dm8/data/EP11/dm.ini' FROM BACKUPSET '/dm8/data/BACKUP_FILE_01';
RESTORE DATABASE '/dm8/data/EP11/dm.ini' FROM BACKUPSET '/dm8/data/BACKUP_FILE_01';
file dm.key not found, use default license!
RECOVER DATABASE '/dm8/data/EP11/dm.ini' FROM BACKUPSET '/dm8/data/BACKUP_FILE_01';
Normal of FAST
Normal of DEFAULT
Normal of RECYCLE
Normal of KEEP
Normal of ROLL
[Percent:100.00%][Speed:0.00M/s][Cost:00:00:02][Remaining:00:00:00]                                 
restore successfully.
time used: 00:00:02.765
RMAN> RECOVER DATABASE '/dm8/data/EP11/dm.ini' FROM BACKUPSET '/dm8/data/BACKUP_FILE_01';
Database mode = 0, oguid = 0
Normal of FAST
Normal of DEFAULT
Normal of RECYCLE
Normal of KEEP
Normal of ROLL
EP[0]'s cur_lsn[37302], file_lsn[37302]
[Percent:100.00%][Speed:0.00PKG/s][Cost:00:00:00][Remaining:00:00:00]                               
recover successfully!
time used: 00:00:02.540
RMAN> RECOVER DATABASE '/dm8/data/EP11/dm.ini' UPDATE DB_MAGIC;
RECOVER DATABASE '/dm8/data/EP11/dm.ini' UPDATE DB_MAGIC;
Database mode = 0, oguid = 0
Normal of FAST
Normal of DEFAULT
Normal of RECYCLE
Normal of KEEP
Normal of ROLL
EP[0]'s cur_lsn[38498], file_lsn[38498]
recover successfully!
time used: 00:00:01.055

联机备份EP02并脱机还原到EP22

--备份
SQL> BACKUP DATABASE BACKUPSET '/dm8/data/BACKUP_FILE_02';
--拷贝
dmdba@lxm2:/dm8/data$ scp -r BACKUP_FILE_02/ dmdba@8.142.141.128:/dm8/data
--备库还原
RESTORE DATABASE '/dm8/data/EP22/dm.ini' FROM BACKUPSET '/dm8/data/BACKUP_FILE_02';
RECOVER DATABASE '/dm8/data/EP22/dm.ini' FROM BACKUPSET '/dm8/data/BACKUP_FILE_02';
RECOVER DATABASE '/dm8/data/EP22/dm.ini' UPDATE DB_MAGIC;

配置文件

配置主库 GRP1_MPP_EP01

dm.ini

INSTANCE_NAME = GRP1_MPP_EP01
PORT_NUM = 5237 #数据库实例监听端口
DW_INACTIVE_INTERVAL = 60 #接收守护进程消息超时时间
ALTER_MODE_STATUS = 0 #不允许手工方式修改实例模式/状态/OGUID
ENABLE_OFFLINE_TS = 2 #不允许备库 OFFLINE 表空间
MAL_INI = 1 #打开 MAL 系统
ARCH_INI = 1 #打开归档配置
MPP_INI = 1 #启用 MPP 配置
RLOG_SEND_APPLY_MON = 64 #统计最近 64 次的日志发送信息

dmmal.ini

MAL_CHECK_INTERVAL = 5 #MAL 链路检测时间间隔
MAL_CONN_FAIL_INTERVAL = 5 #判定 MAL 链路断开的时间
[MAL_INST1]
 MAL_INST_NAME = GRP1_MPP_EP01 #实例名，和 dm.ini 中的 INSTANCE_NAME 一致
 MAL_HOST = 172.21.215.74 #MAL 系统监听 TCP 连接的 IP 地址
 MAL_PORT = 5337 #MAL 系统监听 TCP 连接的端口
 MAL_INST_HOST = 8.142.141.128 #实例的对外服务 IP 地址
 MAL_INST_PORT = 5237 #实例的对外服务端口，和 dm.ini 中的 PORT_NUM 一致
 MAL_DW_PORT = 5253 #实例对应的守护进程监听 TCP 连接的端口
 MAL_INST_DW_PORT = 5243 #实例监听守护进程 TCP 连接的端口
[MAL_INST2]
 MAL_INST_NAME = GRP2_MPP_EP02
 MAL_HOST = 172.17.44.74 
 MAL_PORT = 5337
 MAL_INST_HOST = 39.103.139.80
 MAL_INST_PORT = 5237
 MAL_DW_PORT = 5253
MAL_INST_DW_PORT = 5243
[MAL_INST3]
 MAL_INST_NAME = GRP1_MPP_EP11
 MAL_HOST = 172.17.44.74 
 MAL_PORT = 5338
 MAL_INST_HOST = 39.103.139.80
 MAL_INST_PORT = 5238
 MAL_DW_PORT = 5254
MAL_INST_DW_PORT = 5244
[MAL_INST4]
 MAL_INST_NAME = GRP2_MPP_EP22
 MAL_HOST = 172.21.215.74
 MAL_PORT = 5338
 MAL_INST_HOST = 8.142.141.128
 MAL_INST_PORT = 5238
 MAL_DW_PORT = 5254
 MAL_INST_DW_PORT = 5244

dmarch.ini

[ARCHIVE_REALTIME1]
ARCH_TYPE = REALTIME #实时归档类型
ARCH_DEST = GRP1_MPP_EP11 #实时归档目标实例名
[ARCHIVE_LOCAL1]
ARCH_TYPE = LOCAL #本地归档类型
ARCH_DEST = /dm8/data/EP01/arch #本地归档文件存放路径
ARCH_FILE_SIZE = 128 #单位 Mb，本地单个归档文件最大值
ARCH_SPACE_LIMIT = 0 #单位 Mb，0 表示无限制，范围 1024~4294967294M

dmmpp.ctl

--新建dmmpp.ini
[service_name1]
mpp_seq_no = 0
mpp_inst_name = GRP1_MPP_EP01
[service_name2] 
mpp_seq_no = 1
mpp_inst_name = GRP2_MPP_EP02

--生成dmmpp.ctl
dmctlcvt TYPE=2 SRC=/dm8/data/EP01/dmmpp.ini DEST=/dm8/data/EP01/dmmpp.ctl

启动主库

dmserver /dm/data/EP01/dm.ini mount

设置OGUID

dmdba@lxm2:~$ rlwrap disql SYSDBA/SYSDBA@localhost:5237#"{mpp_type=local}"

Server[localhost:5237]:mode is normal, state is mount
login used time : 1.708(ms)
disql V8
SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 1);
sp_set_oguid(45330);
DMSQL executed successfully
used time: 9.792(ms). Execute id is 0.
SQL> DMSQL executed successfully
used time: 9.320(ms). Execute id is 1.
SQL> SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 0);
DMSQL executed successfully
used time: 4.399(ms). Execute id is 2.

修改数据库模式

SQL>alter database primary;

配置主库GRP2_MPP_EP02

dm.ini

INSTANCE_NAME = GRP2_MPP_EP02
PORT_NUM = 5237 #数据库实例监听端口
DW_INACTIVE_INTERVAL = 60 #接收守护进程消息超时时间
ALTER_MODE_STATUS = 0 #不允许手工方式修改实例模式/状态/OGUID
ENABLE_OFFLINE_TS = 2 #不允许备库 OFFLINE 表空间
MAL_INI = 1 #打开 MAL 系统
ARCH_INI = 1 #打开归档配置
MPP_INI = 1 #启用 MPP 配置
RLOG_SEND_APPLY_MON = 64 #统计最近 64 次的日志发送信息

dmmal.ini

同GRP1_MPP_EP01.

dmarch.ini

[ARCHIVE_REALTIME1]
ARCH_TYPE = REALTIME #实时归档类型
ARCH_DEST = GRP2_MPP_EP22 #实时归档目标实例名

[ARCHIVE_LOCAL1]
ARCH_TYPE = LOCAL #本地归档类型
ARCH_DEST = /dm8/data/EP02/arch#本地归档文件存放路径
ARCH_FILE_SIZE = 128 #单位 Mb，本地单个归档文件最大值
ARCH_SPACE_LIMIT = 0 #单位 Mb，0 表示无限制，范围 1024~4294967294M

dmmpp.ctl

同GRP1_MPP_EP01.

启动主库

dmserver /dm8/data/EP02/dm.ini mount

设置OGUID

SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 1);
sp_set_oguid(45331);
SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 0);

修改数据库模式

alter database primary;

配置备库GRP1_MPP_EP11

dm.ini

INSTANCE_NAME = GRP1_MPP_EP11
PORT_NUM = 5238 #数据库实例监听端口
DW_INACTIVE_INTERVAL = 60 #接收守护进程消息超时时间
ALTER_MODE_STATUS = 0 #不允许手工方式修改实例模式/状态/OGUID
ENABLE_OFFLINE_TS = 2 #不允许备库 OFFLINE 表空间
MAL_INI = 1 #打开 MAL 系统
ARCH_INI = 1 #打开归档配置
MPP_INI = 1 #打开 MPP 配置
RLOG_SEND_APPLY_MON = 64 #统计最近 64 次的日志重演信息

dmmal.ini

同GRP1_MPP_EP01.

dmarch.ini

[ARCHIVE_REALTIME1]
ARCH_TYPE = REALTIME #实时归档类型
ARCH_DEST = GRP1_MPP_EP01 #实时归档目标实例名

[ARCHIVE_LOCAL1]
ARCH_TYPE = LOCAL #本地归档类型
ARCH_DEST = /dm8/data/EP11/arch#本地归档文件存放路径
ARCH_FILE_SIZE = 128 #单位 Mb，本地单个归档文件最大值
ARCH_SPACE_LIMIT = 0 #单位 Mb，0 表示无限制，范围 1024~4294967294M

dmmpp.ctl

同GRP1_MPP_EP01.

启动备库

dmserver /dm/data/EP11/DAMENG/dm.ini mount

设置OGUID

SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 1);
sp_set_oguid(45330);
SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 0);

修改数据库模式

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

配置备库GRP2_MPP_EP22

dm.ini

INSTANCE_NAME = GRP2_MPP_EP22
PORT_NUM = 5238 #数据库实例监听端口
DW_INACTIVE_INTERVAL = 60 #接收守护进程消息超时时间
ALTER_MODE_STATUS = 0 #不允许手工方式修改实例模式/状态/OGUID
ENABLE_OFFLINE_TS = 2 #不允许备库 OFFLINE 表空间
MAL_INI = 1 #打开 MAL 系统
ARCH_INI = 1 #打开归档配置
MPP_INI = 1 #打开 MPP 配置
RLOG_SEND_APPLY_MON = 64 #统计最近 64 次的日志重演信息

dmmal.ini

同GRP1_MPP_EP01.

dmarch.ini

[ARCHIVE_REALTIME1]
ARCH_TYPE = REALTIME #实时归档类型
ARCH_DEST = GRP2_MPP_EP02 #实时归档目标实例名

[ARCHIVE_LOCAL1]
ARCH_TYPE = LOCAL #本地归档类型
ARCH_DEST = /dm8/data/EP22/arch #本地归档文件存放路径
ARCH_FILE_SIZE = 128 #单位 Mb，本地单个归档文件最大值
ARCH_SPACE_LIMIT = 0 #单位 Mb，0 表示无限制，范围 1024~4294967294M

dmmpp.ctl

同GRP1_MPP_EP01.

启动备库

dmserver /dm8/data/EP22/dm.ini mount

设置OGUID

SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 1);
sp_set_oguid(45331);
SP_SET_PARA_VALUE(1, 'ALTER_MODE_STATUS', 0);

修改数据库模式

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

配置dmwatcher

#A机器
[GRP1]
DW_TYPE = GLOBAL #全局守护类型
DW_MODE = AUTO #自动切换模式
DW_ERROR_TIME = 10 #远程守护进程故障认定时间
INST_RECOVER_TIME = 60 #主库守护进程启动恢复的间隔时间
INST_ERROR_TIME = 10 #本地实例故障认定时间
INST_OGUID = 45330 #守护系统唯一 OGUID 值
INST_INI = /dm8/data/EP01/dm.ini #dm.ini 配置文件路径
INST_AUTO_RESTART = 1 #打开实例的自动启动功能
INST_STARTUP_CMD = /home/dmdba/dmdbms/bin/dmserver #命令行方式启动
RLOG_SEND_THRESHOLD = 0 #指定主库发送日志到备库的时间阈值，默认关闭
RLOG_APPLY_THRESHOLD = 0 #指定备库重演日志的时间阈值，默认关闭
[GRP2]
DW_TYPE = GLOBAL #全局守护类型
DW_MODE = AUTO #自动切换模式
DW_ERROR_TIME = 10 #远程守护进程故障认定时间
INST_RECOVER_TIME = 60 #主库守护进程启动恢复的间隔时间
INST_ERROR_TIME = 10 #本地实例故障认定时间
INST_OGUID = 45331 #守护系统唯一 OGUID 值
INST_INI = /dm8/data/EP22/dm.ini #dm.ini 配置文件路径
INST_AUTO_RESTART = 1 #打开实例的自动启动功能
INST_STARTUP_CMD = /home/dmdba/dmdbms/bin/dmserver #命令行方式启动
RLOG_SEND_THRESHOLD = 0 #指定主库发送日志到备库的时间阈值，默认关闭
RLOG_APPLY_THRESHOLD = 0 #指定备库重演日志的时间阈值，默认关闭

#B机器
[GRP1]
DW_TYPE = GLOBAL #全局守护类型
DW_MODE = AUTO #自动切换模式
DW_ERROR_TIME = 10 #远程守护进程故障认定时间
INST_RECOVER_TIME = 60 #主库守护进程启动恢复的间隔时间
INST_ERROR_TIME = 10 #本地实例故障认定时间
INST_OGUID = 45330 #守护系统唯一 OGUID 值
INST_INI = /dm8/data/EP11/dm.ini #dm.ini 配置文件路径
INST_AUTO_RESTART = 1 #打开实例的自动启动功能
INST_STARTUP_CMD = /home/dmdba/dmdbms/bin/dmserver #命令行方式启动
RLOG_SEND_THRESHOLD = 0 #指定主库发送日志到备库的时间阈值，默认关闭
RLOG_APPLY_THRESHOLD = 0 #指定备库重演日志的时间阈值，默认关闭
[GRP2]
DW_TYPE = GLOBAL #全局守护类型
DW_MODE = AUTO #自动切换模式
DW_ERROR_TIME = 10 #远程守护进程故障认定时间
INST_RECOVER_TIME = 60 #主库守护进程启动恢复的间隔时间
INST_ERROR_TIME = 10 #本地实例故障认定时间
INST_OGUID = 45331 #守护系统唯一 OGUID 值
INST_INI = /dm8/data/EP02/dm.ini #dm.ini 配置文件路径
INST_AUTO_RESTART = 1 #打开实例的自动启动功能
INST_STARTUP_CMD = /home/dmdba/dmdbms/bin/dmserver #命令行方式启动
RLOG_SEND_THRESHOLD = 0 #指定主库发送日志到备库的时间阈值，默认关闭
RLOG_APPLY_THRESHOLD = 0 #指定备库重演日志的时间阈值，默认关闭

配置dmmonitor

MON_DW_CONFIRM = 1 #确认监视器模式
MON_LOG_PATH = C:\dmdbms #监视器日志文件存放路径
MON_LOG_INTERVAL = 60 #每隔 60s 定时记录系统信息到日志文件
MON_LOG_FILE_SIZE = 32 #每个日志文件最大 32M
MON_LOG_SPACE_LIMIT = 0 #不限定日志文件总占用空间
[GRP1]
MON_INST_OGUID = 45330 #组 GRP1 的唯一 OGUID 值
#以下配置为监视器到组 GRP1 的守护进程的连接信息，以―IP:PORT‖的形式配置
#IP 对应 dmmal.ini 中的 MAL_HOST，PORT 对应 dmmal.ini 中的 MAL_DW_PORT
MON_DW_IP = 8.142.141.128:5253
MON_DW_IP = 39.103.139.80:5254
[GRP2]
MON_INST_OGUID = 45331 #组 GRP2 的唯一 OGUID 值
#以下配置为监视器到组 GRP2 的守护进程的连接信息，以―IP:PORT‖的形式配置
#IP 对应 dmmal.ini 中的 MAL_HOST，PORT 对应 dmmal.ini 中的 MAL_DW_PORT
MON_DW_IP = 8.142.141.128:5254
MON_DW_IP = 39.103.139.80:5253

启动watcher

#A机器，B机器也相同。
dmdba@lxm2:~$ dmwatcher /dm8/data/dmwatcher.ini
DMWATCHER[4.0] V8
DMWATCHER[4.0] IS READY
show
2022-07-10 18:08:44
---------------------------------------------------------------------------
GROUP_NAME       TYPE      MODE      OGUID       MPP_FLAG  AUTO_RESTART DW_STATUS       DW_SUB_STATUS   DW_CTL_STATUS   
GRP1             GLOBAL    AUTO      45330       TRUE      TRUE         STARTUP         SUB_STATE_START VALID           

INST_OK   NAME             SVR_MODE  SYS_STATUS   RTYPE     FSEQ            FLSN            CSEQ            CLSN            DW_STAT_FLAG 
OK        GRP1_MPP_EP01    PRIMARY   MOUNT        REALTIME  5810            39581           5810            39581           0            
---------------------------------------------------------------------------

---------------------------------------------------------------------------
GROUP_NAME       TYPE      MODE      OGUID       MPP_FLAG  AUTO_RESTART DW_STATUS       DW_SUB_STATUS   DW_CTL_STATUS   
GRP2             GLOBAL    AUTO      45331       TRUE      TRUE         STARTUP         SUB_STATE_START VALID           

INST_OK   NAME             SVR_MODE  SYS_STATUS   RTYPE     FSEQ            FLSN            CSEQ            CLSN            DW_STAT_FLAG 
OK        GRP2_MPP_EP22    STANDBY   MOUNT        REALTIME  5084            38632           5084            38632           0            

DATABASE(GRP2_MPP_EP22) APPLY INFO:
REDOS_PARALLEL_NUM (1) 
DSC_SEQNO[0], (RSEQ, SSEQ, KSEQ)[5084, 5084, 5084], (RLSN, SLSN, KLSN)[38632, 38632, 38632], N_TSK[0], TSK_MEM_USE[0]
REDO_LSN_ARR: (38632)

---------------------------------------------------------------------------

启动monitor

C:\dmdbms\bin>dmmonitor.exe C:\dmdbms\data\dmmonitor.ini
[monitor]         2022-07-10 18:11:20: DMMONITOR[4.0] V8
[monitor]         2022-07-10 18:11:20: DMMONITOR[4.0] IS READY.

[monitor]         2022-07-10 18:11:20: Received message from(GRP2_MPP_EP22)
                  WTIME                WSTATUS        INST_OK   INAME            ISTATUS     IMODE     RSTAT    N_OPEN   FLSN            CLSN
                  2022-07-10 18:11:19  OPEN           OK        GRP2_MPP_EP22    OPEN        STANDBY   NULL     3        42257           42257

[monitor]         2022-07-10 18:11:20: Received message from(GRP1_MPP_EP01)
                  WTIME                WSTATUS        INST_OK   INAME            ISTATUS     IMODE     RSTAT    N_OPEN   FLSN            CLSN
                  2022-07-10 18:11:19  OPEN           OK        GRP1_MPP_EP01    OPEN        PRIMARY   VALID    3        41955           41955

[monitor]         2022-07-10 18:11:20: Received message from(GRP2_MPP_EP02)
                  WTIME                WSTATUS        INST_OK   INAME            ISTATUS     IMODE     RSTAT    N_OPEN   FLSN            CLSN
                  2022-07-10 18:11:19  OPEN           OK        GRP2_MPP_EP02    OPEN        PRIMARY   VALID    3        42257           42258

[monitor]         2022-07-10 18:11:20: Received message from(GRP1_MPP_EP11)
                  WTIME                WSTATUS        INST_OK   INAME            ISTATUS     IMODE     RSTAT    N_OPEN   FLSN            CLSN
                  2022-07-10 18:11:19  OPEN           OK        GRP1_MPP_EP11    OPEN        STANDBY   VALID    3        41954           41954

客户端配置

dmsvc.conf

测试验证

(a)测试建表和插入数据是否会根据分布列进行自动分发。
(b)测试故障一个节点后，集群和持续提供服务。

dmdba@lxm2:~$ rlwrap disql SYSDBA/SYSDBA@localhost:5237#"{mpp_type=global}"

Server[localhost:5237]:mode is primary, state is open
login used time : 2.571(ms)
disql V8
SQL> select * from V$INSTANCE;

LINEID     NAME          INSTANCE_NAME INSTANCE_NUMBER HOST_NAME SVR_VERSION                DB_VERSION          START_TIME          STATUS$ MODE$   OGUID       DSC_SEQNO   DSC_ROLE
---------- ------------- ------------- --------------- --------- -------------------------- ------------------- ------------------- ------- ------- ----------- ----------- --------
           BUILD_VERSION                      BUILD_TIME          
           ---------------------------------- --------------------
1          GRP1_MPP_EP01 GRP1_MPP_EP01 1               lxm2      DM Database Server x64 V8  DB Version: 0x7000c 2022-07-10 16:20:56 OPEN    PRIMARY 45330       0           NULL
           1-2-114-22.05.25-161267-10045-ENT  May 25 2022 11:22:12

2          GRP2_MPP_EP02 GRP2_MPP_EP02 2               lxm       DM Database Server x64 V8  DB Version: 0x7000c 2022-07-10 16:40:32 OPEN    PRIMARY 45331       0           NULL
           1-2-114-22.05.25-161267-10045-ENT  May 25 2022 11:22:12


used time: 10.279(ms). Execute id is 954698.
SQL> CREATE TABLE T_RANGE(C1 INT,C2 CHAR(10)) DISTRIBUTED BY RANGE(C1)(VALUES EQU OR LESS THAN (100) ON GRP1_MPP_EP01, VALUES LESS THAN(MAXVALUE) ON GRP2_MPP_EP02);
executed successfully
used time: 31.373(ms). Execute id is 601.
SQL> insert into T_RANGE values(1,'Chris'),(51,'Lily'),(101,'Peter'),(120,'Tim'),(200,'Summy');
affect rows 5

used time: 4.900(ms). Execute id is 954711.
SQL> insert into T_RANGE values(1,'Chris'),(51,'Lily'),(181,'Petr'),(320,'Ti'),(900,'Sumy');
affect rows 5

used time: 3.262(ms). Execute id is 954713.
SQL> COMMIT;
executed successfully
used time: 6.231(ms). Execute id is 954714.
SQL> SP_GET_EP_COUNT('SYSDBA','T_RANGE'); 

LINEID     SEQNO       N_ROWS              
---------- ----------- --------------------
1          0           4
2          1           6

used time: 33.026(ms). Execute id is 954715.
SQL>

测试kill掉EP01实例，监视器监控到后，守护进程会把实例自动拉起来。

dmdba@lxm2:~$ ps -ef|grep dms
dmdba    21310 21229  0 16:20 pts/0    00:00:08 dmserver /dm8/data/EP01/dm.ini mount
dmdba    21676 21404  0 17:34 pts/1    00:00:05 dmserver /dm8/data/EP22/dm.ini mount
dmdba    21933 21784  0 18:59 pts/2    00:00:00 grep dms
dmdba@lxm2:~$ kill -9 21310
dmdba@lxm2:~/dmdbms/bin$ ps -ef|grep dms
dmdba    21676 21404  0 17:34 pts/1    00:00:05 dmserver /dm8/data/EP22/dm.ini mount
dmdba    21935     1  1 18:59 ?        00:00:00 /home/dmdba/dmdbms/bin/dmserver /dm8/data/EP01/dm.ini mount
dmdba    22036 21229  0 19:00 pts/0    00:00:00 grep dms

[monitor]         2022-07-10 18:59:37: Instance GRP1_MPP_EP01[PRIMARY, OPEN, ISTAT_SAME:TRUE] error
                  WTIME                WSTATUS        INST_OK   INAME            ISTATUS     IMODE     RSTAT    N_OPEN   FLSN            CLSN
                  2022-07-10 18:59:35  STARTUP        ERROR     GRP1_MPP_EP01    OPEN        PRIMARY   VALID    3        42983           42984

[monitor]         2022-07-10 18:59:37: Dmwatcher process GRP1_MPP_EP01 status switching [OPEN-->STARTUP]
                  WTIME                WSTATUS        INST_OK   INAME            ISTATUS     IMODE     RSTAT    N_OPEN   FLSN            CLSN
                  2022-07-10 18:59:35  STARTUP        ERROR     GRP1_MPP_EP01    OPEN        PRIMARY   VALID    3        42983           42984


--monitor
[monitor]         2022-07-10 18:59:37: Check primary instance error in group(GRP1), start to auto takeover

[monitor]         2022-07-10 18:59:37: Notify group(GRP1)'s active dmwatcher to set MID
[monitor]         2022-07-10 18:59:37: Notify group(GRP1)'s active dmwatcher to set MID success
[monitor]         2022-07-10 18:59:37: Notify group(GRP2)'s active dmwatcher to set MID
[monitor]         2022-07-10 18:59:37: Notify group(GRP2)'s active dmwatcher to set MID success
[monitor]         2022-07-10 18:59:37: Start to takeover use instance GRP1_MPP_EP11
[monitor]         2022-07-10 18:59:37: Notify dmwatcher(GRP1_MPP_EP11) switch to TAKEOVER status
[monitor]         2022-07-10 18:59:38: Dmwatcher process GRP1_MPP_EP11 status switching [OPEN-->TAKEOVER]
[monitor]         2022-07-10 18:59:38: Switch dmwatcher GRP1_MPP_EP11 to TAKEOVER status success
[monitor]         2022-07-10 18:59:38: Instance GRP1_MPP_EP11 start to execute sql SP_SET_GLOBAL_DW_STATUS(0, 7)
[monitor]         2022-07-10 18:59:38: Instance GRP1_MPP_EP11 execute sql SP_SET_GLOBAL_DW_STATUS(0, 7) success
[monitor]         2022-07-10 18:59:38: Instance GRP1_MPP_EP11 start to execute sql SP_APPLY_KEEP_PKG()
[monitor]         2022-07-10 18:59:38: Instance GRP1_MPP_EP11 execute sql SP_APPLY_KEEP_PKG() success
[monitor]         2022-07-10 18:59:38: Instance GRP1_MPP_EP11 start to execute sql ALTER DATABASE MOUNT
[monitor]         2022-07-10 18:59:39: Instance GRP1_MPP_EP11 execute sql ALTER DATABASE MOUNT success
[monitor]         2022-07-10 18:59:39: Instance GRP1_MPP_EP11 start to execute sql ALTER DATABASE PRIMARY
[monitor]         2022-07-10 18:59:39: Instance GRP1_MPP_EP11 execute sql ALTER DATABASE PRIMARY success
[monitor]         2022-07-10 18:59:39: Notify instance GRP1_MPP_EP11 to change all arch status to be invalid
[monitor]         2022-07-10 18:59:39: Succeed to change all instances arch status to be invalid
[monitor]         2022-07-10 18:59:39: Build new mppctl file success
[monitor]         2022-07-10 18:59:39: Notify to update instance GRP1_MPP_EP11[PRIMARY, MOUNT, ISTAT_SAME:TRUE] mppctl file
[monitor]         2022-07-10 18:59:39: Notify instance GRP1_MPP_EP11[PRIMARY, MOUNT, ISTAT_SAME:TRUE] update mppctl file success
[monitor]         2022-07-10 18:59:39: Notify to update instance GRP2_MPP_EP02[PRIMARY, OPEN, ISTAT_SAME:TRUE] mppctl file
[monitor]         2022-07-10 18:59:39: Notify dmwatcher(GRP2_MPP_EP02) switch to MPPCTL UPDATE status
[monitor]         2022-07-10 18:59:40: Dmwatcher process GRP2_MPP_EP02 status switching [OPEN-->MPPCTL UPDATE]
[monitor]         2022-07-10 18:59:40: Switch dmwatcher GRP2_MPP_EP02 to MPPCTL UPDATE status success
[monitor]         2022-07-10 18:59:40: Instance GRP2_MPP_EP02 start to execute sql SP_SET_GLOBAL_DW_STATUS(0, 12)
[monitor]         2022-07-10 18:59:40: Instance GRP2_MPP_EP02 execute sql SP_SET_GLOBAL_DW_STATUS(0, 12) success
[monitor]         2022-07-10 18:59:40: Instance GRP2_MPP_EP02 start to execute sql SP_SET_GLOBAL_DW_STATUS(12, 0)
[monitor]         2022-07-10 18:59:40: Instance GRP2_MPP_EP02 execute sql SP_SET_GLOBAL_DW_STATUS(12, 0) success
[monitor]         2022-07-10 18:59:40: Notify dmwatcher(GRP2_MPP_EP02) switch to OPEN status
[monitor]         2022-07-10 18:59:41: Dmwatcher process GRP2_MPP_EP02 status switching [MPPCTL UPDATE-->OPEN]
[monitor]         2022-07-10 18:59:42: Switch dmwatcher GRP2_MPP_EP02 to OPEN status success
[monitor]         2022-07-10 18:59:42: Notify instance GRP2_MPP_EP02[PRIMARY, OPEN, ISTAT_SAME:TRUE] update mppctl file success
[monitor]         2022-07-10 18:59:42: Instance GRP1_MPP_EP11 start to execute sql ALTER DATABASE OPEN FORCE
[monitor]         2022-07-10 18:59:42: Instance GRP1_MPP_EP11 execute sql ALTER DATABASE OPEN FORCE success
[monitor]         2022-07-10 18:59:42: Instance GRP1_MPP_EP11 start to execute sql SP_SET_GLOBAL_DW_STATUS(7, 0)
[monitor]         2022-07-10 18:59:42: Instance GRP1_MPP_EP11 execute sql SP_SET_GLOBAL_DW_STATUS(7, 0) success
[monitor]         2022-07-10 18:59:42: Notify dmwatcher(GRP1_MPP_EP11) switch to OPEN status
[monitor]         2022-07-10 18:59:42: Dmwatcher process GRP1_MPP_EP11 status switching [TAKEOVER-->OPEN]
[monitor]         2022-07-10 18:59:43: Switch dmwatcher GRP1_MPP_EP11 to OPEN status success
[monitor]         2022-07-10 18:59:43: Notify group(GRP1)'s dmwatcher to do clear
[monitor]         2022-07-10 18:59:43: Clean request of dmwatcher processer GRP1_MPP_EP01 success
[monitor]         2022-07-10 18:59:43: Clean request of dmwatcher processer GRP1_MPP_EP11 success
[monitor]         2022-07-10 18:59:43: Notify group(GRP2)'s dmwatcher to do clear
[monitor]         2022-07-10 18:59:43: Clean request of dmwatcher processer GRP2_MPP_EP22 success
[monitor]         2022-07-10 18:59:43: Clean request of dmwatcher processer GRP2_MPP_EP02 success
[monitor]         2022-07-10 18:59:43: Success to takeover use instance GRP1_MPP_EP11

[monitor]         2022-07-10 18:59:43: Group(GRP1) use instance GRP1_MPP_EP11 auto takeover success

[monitor]         2022-07-10 18:59:50: Instance GRP1_MPP_EP01[PRIMARY, AFTER REDO, ISTAT_SAME:TRUE] recover to OK
                  WTIME                WSTATUS        INST_OK   INAME            ISTATUS     IMODE     RSTAT    N_OPEN   FLSN            CLSN
                  2022-07-10 18:59:48  STARTUP        OK        GRP1_MPP_EP01    AFTER REDO  PRIMARY   VALID    3        42983           42983

[monitor]         2022-07-10 18:59:50: Dmwatcher process GRP1_MPP_EP01 status switching [STARTUP-->OPEN]
                  WTIME                WSTATUS        INST_OK   INAME            ISTATUS     IMODE     RSTAT    N_OPEN   FLSN            CLSN
                  2022-07-10 18:59:48  OPEN           OK        GRP1_MPP_EP01    OPEN        STANDBY   INVALID  3        42983           42983

[monitor]         2022-07-10 18:59:50: Dmwatcher process GRP1_MPP_EP11 status switching [OPEN-->RECOVERY]
                  WTIME                WSTATUS        INST_OK   INAME            ISTATUS     IMODE     RSTAT    N_OPEN   FLSN            CLSN
                  2022-07-10 18:59:48  RECOVERY       OK        GRP1_MPP_EP11    OPEN        PRIMARY   VALID    4        45431           45432

[monitor]         2022-07-10 18:59:51: Dmwatcher process GRP1_MPP_EP11 status switching [RECOVERY-->OPEN]
                  WTIME                WSTATUS        INST_OK   INAME            ISTATUS     IMODE     RSTAT    N_OPEN   FLSN            CLSN
                  2022-07-10 18:59:50  OPEN           OK        GRP1_MPP_EP11    OPEN        PRIMARY   VALID    4        45432           45432

更多内容参见达梦技术社区：https://eco.dameng.com

posted on 2022-07-10 20:24 真喜欢落叶与飞花起舞阅读(117) 评论(0) 编辑收藏举报

刷新页面返回顶部

登录后才能查看或发表评论，立即登录或者逛逛博客园首页

相关博文：

· 读写分离三节点集群环境搭建

· 数据守护主备集群搭建

· MPP集群的搭建

· 大规模并行处理MPP的手动搭建

· 达梦数据库dm8版本Mpp集群原理系统架构和搭建测试

阅读排行：
· 全程不用写代码，我用AI程序员写了一个飞机大战
· DeepSeek 开源周回顾「GitHub 热点速览」
· MongoDB 8.0这个新功能碉堡了，比商业数据库还牛
· 记一次.NET内存居高不下排查解决与启示
· 白话解读 Dapr 1.15：你的「微服务管家」又秀新绝活了

导航

统计

搜索

常用链接

我的标签

合集

随笔分类

随笔档案

阅读排行榜

MPP主备集群搭建

部署环境

部署规划

主库

备库

守护进程规划

数据准备

初始化实例

主库

备库

注册服务

主库

备库

启动主库

配置归档

备份还原

联机备份EP01并脱机还原到EP11

联机备份EP02并脱机还原到EP22

配置文件

配置主库 GRP1_MPP_EP01

dm.ini

dmmal.ini

dmarch.ini

dmmpp.ctl

启动主库

设置OGUID

修改数据库模式

配置主库GRP2_MPP_EP02

dm.ini

dmmal.ini

dmarch.ini

dmmpp.ctl

启动主库

设置OGUID

修改数据库模式

配置备库GRP1_MPP_EP11

dm.ini

dmmal.ini

dmarch.ini

dmmpp.ctl

启动备库

设置OGUID

修改数据库模式

配置备库GRP2_MPP_EP22

dm.ini

dmmal.ini

dmarch.ini

dmmpp.ctl

启动备库

设置OGUID

修改数据库模式

配置dmwatcher

配置dmmonitor

启动watcher

启动monitor

客户端配置

测试验证