MySQL Replication

MySQL Replication

一般来说，应用系统性能扩展不外乎Scale up如用性能更好的主机，或者Scale out增加主机数量通过负载均衡的方式

MySQL的扩展：Scale up的扩展瓶颈很容易突破，在Scale out的扩展方式下，数据如何同步是其核心的管理模块。如下图所示，slave节点向主节点请求二进制日志并在本地replay。应用访问数据库时必须通过7层的读写分离器R/W spliter，能够把应用使用的SQL语句按照按读写类型，把写的请求只发给Master，把读的请求按照负载均衡的方式发送给多个从节点Slaves。

复制的功用：

1. 数据分布
2. 负载均衡读请求
3. 备份
4. 高可用性和故障切换
5. MySQL升级测试

主Master节点：

1. 二进制日志，主服务器必须开启binary log
2. 开启dump Thread：为每个Slave的I/O thread启动一个dump线程，用于向其发送binary log events
3. 单向复制：主-->从

从Slave节点：

1. IO Thread：从Master请求二进制日志事件，并保存于继日志中
2. SQL Thread：从中继日志中读取日志事件，在本地完成重放

MySQL复制特点：

1. 异步复制
2. 主从数据不一致比较常见

复制架构：

1. 一主多从
2. 从服务器还 可以再有从服务器（第一个从服务器必须开启binary log，同时第一个从服务器使用blackhole存储引擎，不需要存数据，只要记录binary log）
3. 多主环状复制（不常用）

主从复制配置过程：

1. 主节点：
   启动二进制日志；
   为当前节点设置一个全局唯一的ID号；
   创建有复制权限的用户账号(权限是指REPLICATION CLIENT REPLICATION SLAVE )；
2. 从节点
   启动中继日志；
   为当前节点设置一个全局唯一的ID号；
   使用有复制权限的用户账号连接至主服务器，并启动复制线程；

实验1（两个全新的M/S）：

1. 主节点更改/etc/my.cnf配置文件
   log-bin = /data/mysqlbinlog/mysql-bin
   server_id=1
   systemctl restart mysqld

2. 主节点授权
   GRANT REPLICATION SLAVE,REPLICATION CLIENT ON . TO 'repluser'@'172.20.%.%' IDENTIFIED BY 'replpass';
   FLUSH PRIVILEGES;

3. 从节点更改/etc/my.cnf配置文件
   server_id=2 --必须区别于主节目的service_id
   relay_log=relay-log
   relay_log_index=relay-log.index
   read-only
   systemctl restart mysqld

4. 检查从节点，确保配置更改
   MariaDB [(none)]> SHOW VARIABLES LIKE 'server_id';
   +---------------+-------+
   | Variable_name | Value |
   +---------------+-------+
   | server_id | 2 |
   +---------------+-------+
   MariaDB [(none)]> SHOW VARIABLES LIKE 'relay_log';
   +---------------+-----------+
   | Variable_name | Value |
   +---------------+-----------+
   | relay_log | relay-log |
   +---------------+-----------+
   MariaDB [(none)]> SHOW VARIABLES LIKE 'relay_log_index';
   +-----------------+-------------------------------+
   | Variable_name | Value |
   +-----------------+-------------------------------+
   | relay_log_index | /data/mysqldb/relay-log.index |
   +-----------------+-------------------------------+

5. 从节点，开始同步
   ```
   MariaDB [(none)]> HELP CHANGE MASTER TO
   MariaDB [(none)]> CHANGE MASTER TO MASTER_HOST='172.20.42.204',MASTER_USER='repluser',MASTER_PASSWORD='replpass',MASTER_LOG_FILE='mysql-bin.000015',MASTER_LOG_POS=685;
   MariaDB [(none)]> SHOW SLAVE STATUSG;
   *************************** 1. row ***************************
   Slave_IO_State:
   Master_Host: 172.20.42.204
   Master_User: repluser
   Master_Port: 3306
   Connect_Retry: 60
   Master_Log_File: mysql-bin.000015
   Read_Master_Log_Pos: 685
   Relay_Log_File: relay-log.000001
   Relay_Log_Pos: 4
   Relay_Master_Log_File: mysql-bin.000015
   Slave_IO_Running: No
   Slave_SQL_Running: No
   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: 685
   Relay_Log_Space: 256
   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: NULL
   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: 0
   Master_SSL_Crl:
   Master_SSL_Crlpath:
   Using_Gtid: No
   Gtid_IO_Pos:
   Replicate_Do_Domain_Ids:
   Replicate_Ignore_Domain_Ids:
   Parallel_Mode: conservative
   SQL_Delay: 0
   SQL_Remaining_Delay: NULL
   Slave_SQL_Running_State:

   MariaDB [(none)]> START SLAVE;
   MariaDB [(none)]> SHOW SLAVE STATUSG;
   *************************** 1. row ***************************
                  Slave_IO_State: Waiting for master to send event
                     Master_Host: 172.20.42.204
                     Master_User: repluser
                     Master_Port: 3306
                   Connect_Retry: 60
                 Master_Log_File: mysql-bin.000015
             Read_Master_Log_Pos: 685
                  Relay_Log_File: relay-log.000002
                   Relay_Log_Pos: 555
           Relay_Master_Log_File: mysql-bin.000015
                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: 685
                 Relay_Log_Space: 858
                 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: 1
                  Master_SSL_Crl:
              Master_SSL_Crlpath:
                      Using_Gtid: No
                     Gtid_IO_Pos:
         Replicate_Do_Domain_Ids:
     Replicate_Ignore_Domain_Ids:
                   Parallel_Mode: conservative
                       SQL_Delay: 0
             SQL_Remaining_Delay: NULL
         Slave_SQL_Running_State: Slave has read all relay log; waiting for the slave I/O thread to update it
   1 row in set (0.00 sec)
   

6. 主节点模拟数据变化
   MariaDB [mydb]>create table testtb(name varchar(10),age int);
   调用存储过程生成10万行数据：
   MariaDB [mydb]> DELIMITER // #修改语句结束符为“//”
   MariaDB [mydb]> CREATE PROCEDURE pro_testtb() #写一个存储过程，目的是生成十万条记录测试用
   -> BEGIN
   -> declare i int;
   -> set i = 1;
   -> while i < 100000
   -> do INSERT INTO testtb(name,age) VALUES (CONCAT('testuser',i),i);
   -> SET i = i + 1;
   -> END while;
   -> END//
   MariaDB [mydb]> DELIMITER ; #记得将语句结束符再改回来
   MariaDB [mydb]> CALL pro_testtb; #调用存储过程来
   MariaDB [mydb]> SELECT COUNT() FROM testtb; #查看一下表中有十万条记录
   MariaDB [mydb]> select count() from testtb;
   +----------+
   | count(*) |
   +----------+
   | 100000 |
   +----------+

7. 从节点上查看同步状态
   MariaDB [mydb]> SHOW GLOBAL VARIABLES LIKE 'server_id'; +---------------+-------+ | Variable_name | Value | +---------------+-------+ | server_id | 2 | +---------------+-------+ MariaDB [mydb]> SHOW SLAVE STATUSG; MariaDB [mydb]> SELECT COUNT(*) FROM testtb; +----------+ | COUNT(*) | +----------+ | 100000| +----------+

实验二（模拟生产环境，主库已经有很多的数据，需要扩容只读库满足性能需求）

1. 更改从节点上配置/etc/my.cnf
   server_id=3
   relay_log=relay-log
   relay_log_index=relay-log.index
   systemctl restart mysqld
2. 检查从节点，确保配置更改
3. 主节点上做全备
   mysqldump -A -F --single-transaction --master-data=2 --flush-privileges > fullbackup.sql
   把fullbackup.sql复制到从节点
4. 从节点
   更改fullbackup.sql
   CHANGE MASTER TO MASTER_HOST='172.20.42.204',MASTER_USER='repluser',MASTER_PASSWORD='replpass',MASTER_LOG_FILE='mysql-bin.000019', MASTER_LOG_POS=385;
   mysql < fullbackup.sql
   START SLAVE;
   检查从同步状态：
   MariaDB [mydb]> SHOW SLAVE STATUSG;
   Slave_IO_Running: Yes
   Slave_SQL_Running: Yes