mysql5.6 —>mysql5.7 GTID模式下多源复制之实战案例

千家信息网最后更新 2025年01月21日mysql5.6 —>mysql5.7 GTID模式下多源复制之实战案例背景说明：公司有多个mysql实例，单实例多个数据库，而且版本还是5.6,这给数据查询分析增加了不少繁琐的事情。所以推荐使用mysql5.6的实例多源复制到mysql5.7实例下方便数据的查询、分析以及权限的管理





环境说明：10.89.3.224 mysql5.6.39 master1
10.89.3.225 mysql5.6.39 master2
10.89.3.222 mysql5.7.21 slave
OS:centos 7.4 mysql_port:3309

操作步骤
1、安装mysql5.6 mysql5.7
详细过程不再重复、如有疑问欢迎参考我的博客
http://blog.itpub.net/29987453/viewspace-2149405/


2、配置mysql master01配置
1)配置10.89.3.224 mysql5.6.39 master1
##在一个复制组内的serverid必须不一样
server-id = 123
##gtid 开启gtid mysql5.6支持
gtid_mode=on
##强制事务一致性，保证事务的安全
enforce-gtid-consistency=on
##在从库上记录从主库传过来的日志数据
log-slave-updates=1
##开启binlog 必须的
log_bin = master01-binlog ##强烈建议设置row模式
binlog_format = row
##binlog保存的日期
expire_logs_days = 30
#relay log在从服务器中记录从主服务器传过来的日志数据
skip_slave_start=1
###设置master-info和relay-log 存放在table中
master-info-repository = TABLE
relay-log-info-repository = TABLE

#指定复制的数据库
replicate-do-db=roket1,roket2
##指定忽略的数据库(不需要复制的库)
replicate-ignore- db=mysql,information_schema,performance_schema,test

2)配置完之后重启mysql server
service mysqld restart
3)登录数据库，创建主从同步账号
grant replication slave on *.* to 'repl'@'10.89.3.222' identified by '123456';
flush privileges;
4)重置binlog日志
reset master;
Query OK, 0 rows affected (0.01 sec)

(root@localhost:mysql.sock) [(none)]>show binary logs;
+------------------------+-----------+
| Log_name | File_size |
+------------------------+-----------+
| master01-binlog.000001 | 151 |
+------------------------+-----------+

3、配置mysql master03配置
1)配置10.89.3.225 mysql5.6.39 master2
##在一个复制组内的serverid必须不一样
server-id = 234
##gtid 开启gtid mysql5.6支持
gtid_mode=on
##强制事务一致性，保证事务的安全
enforce-gtid-consistency=on
##在从库上记录从主库传过来的日志数据
log-slave-updates=1
##开启binlog 必须的
log_bin = master01-binlog ##强烈建议设置row模式
binlog_format = row
##binlog保存的日期
expire_logs_days = 30
#relay log在从服务器中记录从主服务器传过来的日志数据
skip_slave_start=1
###设置master-info和relay-log 存放在table中
master-info-repository = TABLE
relay-log-info-repository = TABLE

#指定复制的数据库
replicate-do-db=maya1,maya2
##指定忽略的数据库(不需要复制的库)
replicate-ignore- db=mysql,information_schema,performance_schema,test

2)配置完之后重启mysql server
service mysqld restart
3)登录数据库，创建主从同步账号
grant replication slave on *.* to 'repl'@'10.89.3.222' identified by '123456';
flush privileges;
4)重置binlog日志
reset master;
Query OK, 0 rows affected (0.01 sec)

(root@localhost:mysql.sock) [(none)]>show binary logs;
+------------------------+-----------+
| Log_name | File_size |
+------------------------+-----------+
| master01-binlog.000001 | 151 |
+------------+-----------+

4、配置slave数据库
1）配置配置10.89.3.222 mysql5.7.21 slave
##在一个复制组内的serverid必须不一样
server-id = 234
gtid_mode=on
##强制事务一致性，保证事务的安全
enforce-gtid-consistency=on
##在从库上记录从主库传过来的日志数据
log-slave-updates=1
##开启binlog 必须的
log_bin = master01-binlog ##强烈建议设置row模式
binlog_format = row
##binlog保存的日期
expire_logs_days = 30
#relay log在从服务器中记录从主服务器传过来的日志数据
skip_slave_start=1
###设置master-info和relay-log 存放在table中
master-info-repository = TABLE
relay-log-info-repository = TABLE
#配置只读数据库
read_only=on

4)重置binlog日志
(root@localhost:mysql.sock) [(none)]>reset master;
Query OK, 0 rows affected (0.02 sec)

(root@localhost:mysql.sock) [(none)]>show binary logs;
+---------------------+-----------+
| Log_name | File_size |
+---------------------+-----------+
| slave-binlog.000001 | 154 |
+---------------------+-----------+
1 row in set (0.00 sec)

(root@localhost:mysql.sock) [(none)]>reset slave;
Query OK, 0 rows affected (0.00 sec)
5、开启主从复制
1)关闭master1 和master2的防火墙
2)开启主从复制
(root@localhost:mysql.sock) [(none)]> CHANGE MASTER TO MASTER_HOST = '10.89.3.225', MASTER_PORT = 3309, MASTER_USER = 'repl', MASTER_PASSWORD = '123456', MASTER_AUTO_POSITION = 1 FOR CHANNEL 'master2';
Query OK, 0 rows affected, 2 warnings (0.02 sec)

(root@localhost:mysql.sock) [(none)]> CHANGE MASTER TO MASTER_HOST = '10.89.3.224', MASTER_PORT = 3309, MASTER_USER = 'repl', MASTER_PASSWORD = '123456', MASTER_AUTO_POSITION = 1 FOR CHANNEL 'master1';
Query OK, 0 rows affected, 2 warnings (0.02 sec)
(root@localhost:mysql.sock) [(none)]>start slave;
3)查看主从复制状态
(root@localhost:mysql.sock) [(none)]>show slave status for channel 'master1'\G
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.89.3.224
Master_User: repl
Master_Port: 3309
Connect_Retry: 60
Master_Log_File: master01-binlog.000001
Read_Master_Log_Pos: 151
Relay_Log_File: mysql-relay-bin-master1.000002
Relay_Log_Pos: 376
Relay_Master_Log_File: master01-binlog.000001
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
Replicate_Do_DB:
Replicate_Ignore_DB:
Replicate_Do_Table:
Replicate_Ignore_Table:
Replicate_Wild_Do_Table:
Replicate_Wild_Ignore_Table:
Last_Errno: 0
Last_Error:
Skip_Counter: 0
Exec_Master_Log_Pos: 151
Relay_Log_Space: 591
Until_Condition: None
Until_Log_File:
Until_Log_Pos: 0
Master_SSL_Allowed: No
Master_SSL_CA_File:
Master_SSL_CA_Path:
Master_SSL_Cert:
Master_SSL_Cipher:
Master_SSL_Key:
Seconds_Behind_Master: 0
Master_SSL_Verify_Server_Cert: No
Last_IO_Errno: 0
Last_IO_Error:
Last_SQL_Errno: 0
Last_SQL_Error:
Replicate_Ignore_Server_Ids:
Master_Server_Id: 123
Master_UUID: 3809e1da-25c0-11e8-93b1-080027857522
Master_Info_File: mysql.slave_master_info
SQL_Delay: 0
SQL_Remaining_Delay: NULL
Slave_SQL_Running_State: Slave has read all relay log; waiting for more updates
Master_Retry_Count: 86400
Master_Bind:
Last_IO_Error_Timestamp:
Last_SQL_Error_Timestamp:
Master_SSL_Crl:
Master_SSL_Crlpath:
Retrieved_Gtid_Set:
Executed_Gtid_Set:
Auto_Position: 1
Replicate_Rewrite_DB:
Channel_Name: master1
Master_TLS_Version:
1 row in set (0.00 sec)

(root@localhost:mysql.sock) [(none)]>show slave status for channel 'master2'\G
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.89.3.225
Master_User: repl
Master_Port: 3309
Connect_Retry: 60
Master_Log_File: master02-binlog.000001
Read_Master_Log_Pos: 151
Relay_Log_File: mysql-relay-bin-master2.000002
Relay_Log_Pos: 376
Relay_Master_Log_File: master02-binlog.000001
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
Replicate_Do_DB:
Replicate_Ignore_DB:
Replicate_Do_Table:
Replicate_Ignore_Table:
Replicate_Wild_Do_Table:
Replicate_Wild_Ignore_Table:
Last_Errno: 0
Last_Error:
Skip_Counter: 0
Exec_Master_Log_Pos: 151
Relay_Log_Space: 591
Until_Condition: None
Until_Log_File:
Until_Log_Pos: 0
Master_SSL_Allowed: No
Master_SSL_CA_File:
Master_SSL_CA_Path:
Master_SSL_Cert:
Master_SSL_Cipher:
Master_SSL_Key:
Seconds_Behind_Master: 0
Master_SSL_Verify_Server_Cert: No
Last_IO_Errno: 0
Last_IO_Error:
Last_SQL_Errno: 0
Last_SQL_Error:
Replicate_Ignore_Server_Ids:
Master_Server_Id: 234
Master_UUID: ab8c56e0-25c0-11e8-93b4-0800278c8292
Master_Info_File: mysql.slave_master_info
SQL_Delay: 0
SQL_Remaining_Delay: NULL
Slave_SQL_Running_State: Slave has read all relay log; waiting for more updates
Master_Retry_Count: 86400
Master_Bind:
Last_IO_Error_Timestamp:
Last_SQL_Error_Timestamp:
Master_SSL_Crl:
Master_SSL_Crlpath:
Retrieved_Gtid_Set:
Executed_Gtid_Set:
Auto_Position: 1
Replicate_Rewrite_DB:
Channel_Name: master2
Master_TLS_Version:
1 row in set (0.00 sec)
显示 主从复制正常
Slave_IO_Running: Yes
Slave_SQL_Running: Yes

6、测试核对数据
1)登录master1数据库
create database roket1;
Query OK, 1 row affected (0.00 sec)

(root@localhost:mysql.sock) [(none)]>create database roket2;
Query OK, 1 row affected (0.01 sec)

(root@localhost:mysql.sock) [(none)]>use roket1;
Database changed
(root@localhost:mysql.sock) [roket1]>create table t1(
-> id bigint not null auto_increment,
-> name varchar(20),
-> primary key(id)
-> )engine=innodb;
Query OK, 0 rows affected (0.02 sec)

(root@localhost:mysql.sock) [roket1]>insert into t1 (name) values ('master1'),('master1'),('master1')；
Query OK, 3 rows affected (0.00 sec)
Records: 3 Duplicates: 0 Warnings: 0
(root@localhost:mysql.sock) [roket1]>select * from t1;
+----+---------+
| id | name |
+----+---------+
| 1 | master1 |
| 2 | master1 |
| 3 | master1 |
+----+---------+

2）登录master2数据库
(root@localhost:mysql.sock) [(none)]>create database maya1;
Query OK, 1 row affected (0.00 sec)

(root@localhost:mysql.sock) [(none)]>create database maya2;
Query OK, 1 row affected (0.00 sec)

(root@localhost:mysql.sock) [(none)]>use maya1;
Database changed
(root@localhost:mysql.sock) [maya1]>create table t2(
-> id bigint not null auto_increment,
-> name varchar(20),
-> primary key(id)
-> )engine=innodb;
Query OK, 0 rows affected (0.02 sec)

(root@localhost:mysql.sock) [maya1]>insert into t2 (name) values ('master2'),('master2'),('master2') ;
Query OK, 3 rows affected (0.00 sec)
Records: 3 Duplicates: 0 Warnings: 0
(root@localhost:mysql.sock) [maya1]>select * from t2;
+----+---------+
| id | name |
+----+---------+
| 1 | master2 |
| 2 | master2 |
| 3 | master2 |
+----+---------+
3 rows in set (0.00 sec)

3)登录主库
(root@localhost:mysql.sock) [(none)]>show databases;
+--------------------+
| Database |
+--------------------+
| information_schema |
| maya1 |
| maya2 |
| mysql |
| performance_schema |
| roket1 |
| roket2 |
| sys |

(root@localhost:mysql.sock) [(none)]>
(root@localhost:mysql.sock) [(none)]>use roket1
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A

Database changed
(root@localhost:mysql.sock) [roket1]>show tables;
+------------------+
| Tables_in_roket1 |
+------------------+
| t1 |
+------------------+
1 row in set (0.00 sec)

(root@localhost:mysql.sock) [roket1]>select * from t1;
+----+---------+
| id | name |
+----+---------+
| 1 | master1 |
| 2 | master1 |
| 3 | master1 |
+----+---------+
3 rows in set (0.00 sec)
(root@localhost:mysql.sock) [maya1]>use maya1
Database changed
(root@localhost:mysql.sock) [maya1]>select * from t2;
+----+---------+
| id | name |
+----+---------+
| 1 | master2 |
| 2 | master2 |
| 3 | master2 |
+----+---------+
3 rows in set (0.00 sec)

经核对、master1、master2的实例的数据已经正常同步到slave，多源复制搞定

总结：1、首先为mysql5.7多源复制点歌赞
2、解决了公司业务的需要多实例的情况下，统一分析数据、查询权限控制等众多问题
3、解决了多实例备份麻烦的问题