主从从，也称为级联主从，数据流向：A(主)->B(从)->C(从从)，主从从级联复制。
应用场景
- 在主从配置的基础上，再增加一个从库，进一步提高数据安全，容灾备份。
- 读写分离，从库只用于查询，提高数据库整体性能。
- 从从库，用于备份，等同在线实时增量备份。
部署环境
注：使用docker部署mysql实例，方便快速搭建演示环境。但本文重点是讲解主从配置，因此简略描述docker环境构建mysql容器实例。
- 数据库：MySQL 5.7.x （相比5.5，5.6而言，5.7同步性能更好，支持多源复制，可实现多主一从，主从库版本应保证一致）
- 操作系统：CentOS 7.x
- 容器：Docker 17.09.0-ce
- 镜像：mysql:5.7
- 主库：IP=192.168.10.212; PORT=4200; server-id=200; database=test; table=user
- 从库：IP=192.168.10.212; PORT=4211; server-id=210; database=test; table=user
- 从从库：IP=192.168.10.212; PORT=4211; server-id=211; database=test; table=userk
配置约束
- 主从库必须保证网络畅通可访问
- 主库必须开启binlog日志
- 主从库的server-id必须不同
事前准备
- 关闭selinux
setenforce 0
vim /etc/sysconfig/selinux
SELINUX=disabled
# 若不关闭，使用docker启动mysql5.7镜像容器时启动不起来，查看日志会有如下错误显示：
ERROR: mysqld failed while attempting to check config
command was: "mysqld --verbose --help"
mysqld: Can't read dir of '/etc/mysql/conf.d/' (Errcode: 13 - Permission denied)
mysqld: [ERROR] Fatal error in defaults handling. Program aborted!
ERROR: mysqld failed while attempting to check config
command was: "mysqld --verbose --help"
mysqld: Can't read dir of '/etc/mysql/conf.d/' (Errcode: 13 - Permission denied)
mysqld: [ERROR] Fatal error in defaults handling. Program aborted!
- 安装并启动docker
yum install epel-release
yum -y install docker
systemctl start docker.service
- 创建目录
mkdir -p /datavol/mysql-200/{mysql,conf}
mkdir -p /datavol/mysql-210/{mysql,conf}
mkdir -p /datavol/mysql-211/{mysql,conf}
【主库】操作及配置
把该文件放到主库所在配置文件路径下：/datavol/mysql-200/conf
配置my.cnf
[client]
port = 3306
default-character-set = utf8mb4
[mysql]
port = 3306
default-character-set = utf8mb4
[mysqld]
##########################
# summary
##########################
#bind-address = 0.0.0.0
#port = 3306
#datadir=/datavol/mysql/data #数据存储目录
##########################
# log bin
##########################
server-id = 200 #必须唯一
log_bin = mysql-bin #开启及设置二进制日志文件名称
binlog_format = MIXED
sync_binlog = 1
expire_logs_days =7 #二进制日志自动删除/过期的天数。默认值为0，表示不自动删除。
#binlog_cache_size = 128m
#max_binlog_cache_size = 512m
#max_binlog_size = 256M
binlog-do-db = test #要同步的数据库
binlog-ignore-db = mysql #不需要同步的数据库
binlog_ignore_db = information_schema
binlog_ignore_db = performation_schema
binlog_ignore_db = sys
##########################
# character set
##########################
character-set-server = utf8mb4
collation-server = utf8mb4_unicode_ci
安装启动主库
[root@localhost ~]# docker run -d -p 4200:3306 --name=mysql-200 -v /datavol/mysql-200/conf:/etc/mysql/conf.d -v /datavol/mysql-200:/datavol/mysql -e MYSQL_ROOT_PASSWORD=123456 mysql:5.7
5691bac538e646db00273e3cad5b350dbe6cce0bd176346b7eefd9a6f9e3a9ad
[root@localhost ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
5691bac538e6 mysql:5.7 "docker-entrypoint..." 44 seconds ago Up 43 seconds 33060/tcp, 0.0.0.0:4200->3306/tcp mysql-200
[root@localhost ~]# docker exec -it mysql-200 /bin/bash
root@5691bac538e6:/# mysql -u root -p
Enter password:
注：若不熟悉docker，可使用传统方式安装mysql，效果相同。
创建授权用户
连接mysql主数据库，键入命令mysql -u root -p，输入密码后登录数据库。创建用户用于从库同步复制，授予复制、同步访问的权限
mysql> CREATE USER 'slave'@'%' IDENTIFIED BY '123456';
Query OK, 0 rows affected (0.01 sec)
mysql> GRANT REPLICATION SLAVE, REPLICATION CLIENT ON *.* TO 'slave'@'%';
Query OK, 0 rows affected (0.00 sec)
log_bin是否开启
mysql> show variables like 'log_bin';
+---------------+-------+
| Variable_name | Value |
+---------------+-------+
| log_bin | ON |
+---------------+-------+
1 row in set
查看master状态
mysql> show master status;
+------------------+----------+--------------+--------------------------------------------------+-------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+------------------+----------+--------------+--------------------------------------------------+-------------------+
| mysql-bin.000001 | 154 | test | mysql,information_schema,performation_schema,sys | |
+------------------+----------+--------------+--------------------------------------------------+-------------------+
1 row in set
【从库】配置及操作
配置my.cnf
把该文件放到主库所在配置文件路径下：/datavol/mysql-210/conf/
[client]
port = 3306
default-character-set = utf8mb4
[mysql]
port = 3306
default-character-set = utf8mb4
[mysqld]
##########################
# summary
##########################
#bind-address = 0.0.0.0
#port = 3306
#datadir=/datavol/mysql/data #数据存储目录
##########################
# log bin
##########################
server-id = 210 #必须唯一
log_bin = mysql-bin #开启及设置二进制日志文件名称
binlog_format = MIXED
sync_binlog = 1
expire_logs_days =7 #二进制日志自动删除/过期的天数。默认值为0，表示不自动删除。
log_slave_updates = 1 #【关键点】从主服务器接收到的更新同时要写入二进制日志
#binlog_cache_size = 128m
#max_binlog_cache_size = 512m
#max_binlog_size = 256M
binlog-do-db = test #要同步的数据库
binlog-ignore-db = mysql #不需要同步的数据库
binlog_ignore_db = information_schema
binlog_ignore_db = performation_schema
binlog_ignore_db = sys
##########################
# character set
##########################
character-set-server = utf8mb4
collation-server = utf8mb4_unicode_ci
**log_slave_updates = 1 ** #【关键点】从主服务器接收到的更新同时要写入二进制日志
安装启动从库
[root@localhost ~]# docker run -d -p 4210:3306 --name=mysql-210 -v /datavol/mysql-210/conf:/etc/mysql/conf.d -v /datavol/mysql-210:/datavol/mysql -e MYSQL_ROOT_PASSWORD=123456 mysql:5.7
19e93b6d93ca4e6ca0d540e3f6c831b835cdbb35362733867c3977aee4d33bf7
[root@localhost ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
19e93b6d93ca mysql:5.7 "docker-entrypoint..." 4 seconds ago Up 3 seconds 33060/tcp, 0.0.0.0:4210->3306/tcp mysql-210
[root@localhost ~]# docker exec -it mysql-210 /bin/bash
root@19e93b6d93ca:/# mysql -u root -p
Enter password:
设置主库200信息
登录【从数据库】，进入mysql命令行。
mysql> stop slave;
Query OK, 0 rows affected
mysql> CHANGE MASTER TO
MASTER_HOST='192.168.10.212',
MASTER_PORT=4200,
MASTER_USER='slave',
MASTER_PASSWORD='123456',
MASTER_LOG_FILE='mysql-bin.000001',
MASTER_LOG_POS=154;
Query OK, 0 rows affected
mysql> start slave;
Query OK, 0 rows affected
stop slave; //停止同步
start slave; //开始同步
//必须和【主数据库】的信息匹配。
CHANGE MASTER TO
MASTER_HOST='192.168.10.212', //主库IP
MASTER_PORT=4200, //主库端口
MASTER_USER='slave', //访问主库且有同步复制权限的用户
MASTER_PASSWORD='123456', //登录密码
//【关键处】从主库的该log_bin文件开始读取同步信息，主库show master status返回结果
MASTER_LOG_FILE='mysql-bin.000001',
//【关键处】从文件中指定位置开始读取，主库show master status返回结果
MASTER_LOG_POS=154;
查看同步状态
mysql> show slave status \G
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 192.168.10.212
Master_User: slave
Master_Port: 4200
Connect_Retry: 60
Master_Log_File: mysql-bin.000003
Read_Master_Log_Pos: 617
Relay_Log_File: 5f0d9861bcb1-relay-bin.000002
Relay_Log_Pos: 320
Relay_Master_Log_File: mysql-bin.000003
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: 617
Relay_Log_Space: 534
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: 200
Master_UUID: b6df98cf-e71f-11e8-9987-0242ac110007
Master_Info_File: /var/lib/mysql/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: 0
Replicate_Rewrite_DB:
Channel_Name:
Master_TLS_Version:
1 row in set (0.00 sec)
只有【Slave_IO_Running】和【Slave_SQL_Running】都是Yes，则同步是正常的。
如果是No或者Connecting都不行，可查看mysql-error.log，以排查问题。
mysql> show variables like 'log_error%';
+---------------------+--------+
| Variable_name | Value |
+---------------------+--------+
| log_error | stderr |
| log_error_verbosity | 3 |
+---------------------+--------+
2 rows in set
配置完成，则主从数据库开始自动同步。此外，【从库】相当于【从从库】的主库，也需要设置账号及同步权限。
创建授权用户
创建用户用于从库同步复制，授予复制、同步访问的权限
mysql> CREATE USER 'slave'@'%' IDENTIFIED BY '123456';
Query OK, 0 rows affected (0.01 sec)
mysql> GRANT REPLICATION SLAVE, REPLICATION CLIENT ON *.* TO 'slave'@'%';
Query OK, 0 rows affected (0.00 sec)
log_bin是否开启
mysql> show variables like 'log_bin';
+---------------+-------+
| Variable_name | Value |
+---------------+-------+
| log_bin | ON |
+---------------+-------+
1 row in set
查看master状态
mysql> show master status;
+------------------+----------+--------------+--------------------------------------------------+-------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+------------------+----------+--------------+--------------------------------------------------+-------------------+
| mysql-bin.000003 | 617 | test | mysql,information_schema,performation_schema,sys | |
+------------------+----------+--------------+--------------------------------------------------+-------------------+
1 row in set
【从从库】配置及操作
配置my.cnf
把该文件放到主库所在配置文件路径下：/datavol/mysql-211/conf/
[client]
port = 3306
default-character-set = utf8mb4
[mysql]
port = 3306
default-character-set = utf8mb4
[mysqld]
##########################
# summary
##########################
#bind-address = 0.0.0.0
#port = 3306
#datadir=/datavol/mysql/data #数据存储目录
##########################
# log bin
##########################
server-id = 211
##########################
# character set
##########################
character-set-server = utf8mb4
collation-server = utf8mb4_unicode_ci
安装启动从从库
[root@localhost ~]# docker run -d -p 4211:3306 --name=mysql-211 -v /datavol/mysql-211/conf:/etc/mysql/conf.d -v /datavol/mysql-211:/datavol/mysql -e MYSQL_ROOT_PASSWORD=123456 mysql:5.7
19e93b6d93ca4e6ca0d540e3f6c831b835cdbb35362733867c3977aee4d33bf7
[root@localhost ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
19e93b6d93ca mysql:5.7 "docker-entrypoint..." 4 seconds ago Up 3 seconds 33060/tcp, 0.0.0.0:4211->3306/tcp mysql-211
[root@localhost ~]# docker exec -it mysql-211 /bin/bash
root@19e93b6d93ca:/# mysql -u root -p
Enter password:
设置从库210信息
登录【从数据库】，进入mysql命令行。
mysql> stop slave;
Query OK, 0 rows affected
mysql> CHANGE MASTER TO MASTER_HOST='192.168.10.212',
MASTER_PORT=4210,
MASTER_USER='slave',
MASTER_PASSWORD='123456',
MASTER_LOG_FILE='mysql-bin.000003',
MASTER_LOG_POS=617;
Query OK, 0 rows affected
mysql> start slave;
Query OK, 0 rows affected
stop slave; //停止同步
start slave; //开始同步
//必须和【主数据库】的信息匹配。
CHANGE MASTER TO
MASTER_HOST='192.168.10.212', //主库IP
MASTER_PORT=4210, //主库端口(在这里是指从库)
MASTER_USER='slave', //访问主库且有同步复制权限的用户
MASTER_PASSWORD='123456', //登录密码
//【关键处】从主库的该log_bin文件开始读取同步信息，主库show master status返回结果
MASTER_LOG_FILE='mysql-bin.000003',
//【关键处】从文件中指定位置开始读取，主库show master status返回结果
MASTER_LOG_POS=617;
查看同步状态
mysql> show slave status \G
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 192.168.10.212
Master_User: slave
Master_Port: 4210
Connect_Retry: 60
Master_Log_File: mysql-bin.000003
Read_Master_Log_Pos: 617
Relay_Log_File: a3d52e314640-relay-bin.000002
Relay_Log_Pos: 783
Relay_Master_Log_File: mysql-bin.000003
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: 617
Relay_Log_Space: 997
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: 210
Master_UUID: 088c2c93-e720-11e8-8e80-0242ac110009
Master_Info_File: /var/lib/mysql/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: 0
Replicate_Rewrite_DB:
Channel_Name:
Master_TLS_Version:
1 row in set (0.00 sec)
只有【Slave_IO_Running】和【Slave_SQL_Running】都是Yes，则同步是正常的。
如果是No或者Connecting都不行，可查看mysql-error.log，以排查问题。
mysql> show variables like 'log_error%';
+---------------------+--------+
| Variable_name | Value |
+---------------------+--------+
| log_error | stderr |
| log_error_verbosity | 3 |
+---------------------+--------+
2 rows in set
配置完成，则从从数据库开始自动同步。
验证数据同步
建库
使用root账号登录【主服务器】，创建test数据库
mysql> CREATE DATABASE test;
Query OK, 1 row affected (0.00 sec)
mysql> USE test;
Database changed
建表
在【主库】中创建user表
CREATE TABLE `user` (
`id` bigint(20) NOT NULL AUTO_INCREMENT,
`name` varchar(20) COLLATE utf8mb4_unicode_ci NOT NULL,
`age` tinyint(3) unsigned NOT NULL,
PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;
新增
在【主库】中向user表插入一条数据：
mysql> use test;
Database changed
mysql> INSERT INTO user (id, name, age) VALUES (1, '张三', 18);
Database changed
mysql> SELECT * FROM user;
+----+------+-----+
| id | name | age |
+----+------+-----+
| 1 | 张三 | 18 |
+----+------+-----+
1 row in set
在【从库】及【从从库】中查询user表数据：
mysql> use test;
Database changed
mysql> SELECT * FROM user;
+----+------+-----+
| id | name | age |
+----+------+-----+
| 1 | 张三 | 18 |
+----+------+-----+
1 row in set
新增记录同步成功。
更新
在【主库】中修改刚才插入的数据：
mysql> UPDATE user SET name='李四' where id=1;
Database changed
Rows matched: 1 Changed: 1 Warnings: 0
mysql> SELECT * FROM user;
+----+------+-----+
| id | name | age |
+----+------+-----+
| 1 | 李四 | 18 |
+----+------+-----+
1 row in set
在【从库】及【从从库】中查询user表数据：
mysql> SELECT * FROM user;
+----+------+-----+
| id | name | age |
+----+------+-----+
| 1 | 李四 | 18 |
+----+------+-----+
1 row in set
更新记录同步成功。
删除
在【主库】中删除刚才更新的数据：
mysql> DELETE FROM user WHERE id=1;
Database changed
mysql> SELECT * FROM user;
Empty set
在【从库】及【从从库】中查询user表数据：
mysql> SELECT * FROM user;
Empty set
删除记录同步成功。
补充：
- 如果【主服务器】重启mysql服务，【从服务器】会等待与【主服务器】重连。当主服务器恢复正常后，从服务器会自动重新连接上主服务器，并正常同步数据。
- 如果某段时间内，【从数据库】服务器异常导致同步中断（可能是同步点位置不匹配），可以尝试以下恢复方法：进入【主数据库】服务器（正常），在bin-log中找到【从数据库】出错前的position，然后在【从数据库】上执行change master，将master_log_file和master_log_pos重新指定后，开始同步。