Linux--mysql数据库的复制

一.mysql的复制类型
mysql复制类型主要包括异步复制、全同步复制、半同步复制
1.异步复制（Asynchronous replication）
MySQL默认的复制即是异步的，主库在执行完客户端提交的事务后会立即将结果返给给客户端，并不关心从库是否已经接收并处理，这样就会有一个问题，主如果crash掉了，此时主上已经提交的事务可能并没有传到从上，如果此时，强行将从提升为主，可能导致新主上的数据不完整
2.全同步复制（Fully synchronous replication）
当主库执行完一个事务，所有的从库都执行了该事务才返回给客户端。因为需要等待所有从库执行完该事务才能返回，所以全同步复制的性能必然会收到严重的影响
3.半同步复制（Semisynchronous replication）
介于异步复制和全同步复制之间，主库在执行完客户端提交的事务后不是立刻返回给客户端，而是等待至少一个从库接收到并写到relay log中才返回给客户端。相对于异步复制，半同步复制提高了数据的安全性，同时它也造成了一定程度的延迟，这个延迟最少是一个TCP/IP往返的时间。所以，半同步复制最好在低延时的网络中使用。

在2010年MySQL 5.5版本之前，一直采用的是这种异步复制的方式。主库的事务执行不会管备库的同步进度，如果备库落后，主库不幸crash，那么就会导致数据丢失。于是在MySQL在5.5中就顺其自然地引入了半同步复制，主库在应答客户端提交的事务前需要保证至少一个从库接收并写到relay log中
半同步的意义
1）正常的复制为：事务一（t1）写入binlog buffer；dumper线程通知slave有新的事务t1；binlog buffer进行checkpoint；slave的io线程接收到t1并写入到自己的的relay log；slave的sql线程写入到本地数据库。 这时，master和slave都能看到这条新的事务，即使master挂了，slave可以提升为新的master。
2）异常的复制为：事务一（t1）写入binlog buffer；dumper线程通知slave有新的事务t1；binlog buffer进行checkpoint；slave因为网络不稳定，一直没有收到t1；master挂掉，slave提升为新的master，t1丢失。
3）很大的问题是：主机和从机事务更新的不同步，就算是没有网络或者其他系统的异常，当业务并发上来时，slave因为要顺序执行master批量事务，导致很大的延迟。

为了弥补以上几种场景的不足，MySQL从5.5开始推出了半同步复制。相比异步复制，半同步复制提高了数据完整性，因为很明确知道，在一个事务提交成功之后，这个事务就至少会存在于两个地方。即在master的dumper线程通知slave后，增加了一个ack（消息确认），即是否成功收到t1的标志码，也就是dumper线程除了发送t1到slave，还承担了接收slave的ack工作。如果出现异常，没有收到ack，那么将自动降级为普通的复制，直到异常修复后又会自动变为半同步复制。
二.mysql的主从复制
1.主从复制原理（异步）
主master开启bin-log二进制日志功能，此日志会记录所有对数据库进行修改的SQL语句；复制原理即将主库上的日志复制到从库并执行一遍，此时从库与主库的数据就完全相同了。
过程：
要开启3个线程，master IO线程，slave开启 IO线程 SQL线程
（1）Slave 通过IO线程连接master，并且请求某个bin-log，position之后的内容。
（2）MASTER服务器收到slave IO线程发来的日志请求信息
（3）io线程去将bin-log内容，position返回给slave IO线程。
（4）slave服务器收到bin-log日志内容，将bin-log日志内容写入relay-log中继日志，创建一个master.info的文件，该文件记录了master ip 用户名 密码 master bin-log名称，bin-log position。
（5）slave端开启SQL线程，实时监控relay-log日志内容是否有更新，解析文件中的SQL语句，在slave数据库中去执型

2.实验环境搭建
两台主机：
172.25.4.111：主数据库
172.25.4.112：从数据库
（1）数据库的安装，两台主机均安装
主库安装

[root@server1 ~]# cd mysql/
[root@server1 mysql]# ls
master_ip_failover       mysql-5.7.24-1.el7.x86_64.rpm-bundle.tar      sysbench
master_ip_online_change  mysql-proxy-0.8.5-linux-el6-x86-64bit.tar.gz
MHA-7                    send_report
[root@server1 mysql]# tar xf mysql-5.7.24-1.el7.x86_64.rpm-bundle.tar  ##解压mysql安装包
[root@server1 mysql]# ls
master_ip_failover
master_ip_online_change
MHA-7
mysql-5.7.24-1.el7.x86_64.rpm-bundle.tar
mysql-community-client-5.7.24-1.el7.x86_64.rpm
mysql-community-common-5.7.24-1.el7.x86_64.rpm
mysql-community-devel-5.7.24-1.el7.x86_64.rpm
mysql-community-embedded-5.7.24-1.el7.x86_64.rpm
mysql-community-embedded-compat-5.7.24-1.el7.x86_64.rpm
mysql-community-embedded-devel-5.7.24-1.el7.x86_64.rpm
mysql-community-libs-5.7.24-1.el7.x86_64.rpm
mysql-community-libs-compat-5.7.24-1.el7.x86_64.rpm
mysql-community-minimal-debuginfo-5.7.24-1.el7.x86_64.rpm
mysql-community-server-5.7.24-1.el7.x86_64.rpm
mysql-community-server-minimal-5.7.24-1.el7.x86_64.rpm
mysql-community-test-5.7.24-1.el7.x86_64.rpm
mysql-proxy-0.8.5-linux-el6-x86-64bit.tar.gz
send_report
sysbench
[root@server1 mysql]# yum install -y mysql-community-client-5.7.24-1.el7.x86_64.rpm mysql-community-common-5.7.24-1.el7.x86_64.rpm mysql-community-libs-5.7.24-1.el7.x86_64.rpm mysql-community-libs-compat-5.7.24-1.el7.x86_64.rpm mysql-community-server-5.7.24-1.el7.x86_64.rpm  ##下载所需软件，解决依赖性

从库安装

[root@server2 ~]# cd mysql/
[root@server2 mysql]# tar xf mysql-5.7.24-1.el7.x86_64.rpm-bundle.tar
[root@server2 mysql]# yum install -y mysql-community-client-5.7.24-1.el7.x86_64.rpm mysql-community-common-5.7.24-1.el7.x86_64.rpm mysql-community-libs-5.7.24-1.el7.x86_64.rpm mysql-community-libs-compat-5.7.24-1.el7.x86_64.rpm mysql-community-server-5.7.24-1.el7.x86_64.rpm

（2）主配置文件的修改
主库

[root@server1 mysql]# vim /etc/my.cnf
log-bin=mysql-bin  ##开启二进制日志功能
server-id=1  ##id号区分
[root@server1 mysql]# systemctl start mysqld  ##开启数据库

从库

[root@server2 mysql]# vim /etc/my.cnf
server-id=2
[root@server2 mysql]# systemctl start mysqld

（3）数据库初始化
主从库均设置

[root@server1 mysql]# cat /var/log/mysqld.log | grep password  ##查看生成的临时密码
2019-07-04T11:20:54.116772Z 1 [Note] A temporary password is generated for root@localhost: ouqp?3T90w2h  ##临时密码为：ouqp?3T90w2h
[root@server1 mysql]# mysql -uroot -p  ##登录数据库
Enter password:   ##输入临时密码
mysql> show databases;
ERROR 1820 (HY000): You must reset your password using ALTER USER statement before executing this statement.  ##查看库出现错误，原因是不够安全
[root@server1 mysql]# mysql_secure_installation  ##数据库进行安全初始化，设置密码为Westos+001
[root@server1 mysql]# mysql -uroot -pWestos+001  ##以设置的新密码登录数据库
mysql> show databases;
+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| performance_schema |
| sys                |
+--------------------+
4 rows in set (0.00 sec)

（4）主从库的设置
主库设置

[root@server1 mysql]# mysql -uroot -pWestos+001
mysql> grant replication slave on *.* to repl@'172.25.4.%' identified by 'Westos+001';  ##在Master的数据库中建立一个备份帐户：每个slave使用标准的MySQL用户名和密码连接master。进行复制操作的用户会授予REPLICATION SLAVE权限。用户名的密码都会存储在文本文件master.info中
mysql> show plugins;  ##查看密码插件状态，状态均为ACTIVE则正常
mysql> show master status;  ##查看主库状态，由此可知日志为mysql-bin.000002；节点为690
+------------------+----------+--------------+------------------+-------------------+
| File             | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+------------------+----------+--------------+------------------+-------------------+
| mysql-bin.000002 |      690 |              |                  |                   |
+------------------+----------+--------------+------------------+-------------------+

从库设置

[root@server2 mysql]# mysql -uroot -pWestos+001
mysql> CHANGE MASTER TO  ##改变主库信息
    -> MASTER_HOST='172.25.4.111',  ##主库IP
    -> MASTER_USER='repl',  ##主库的备份账户
    -> MASTER_PASSWORD='Westos+001',  ##主库备份账户密码
    -> MASTER_LOG_FILE='mysql-bin.000002',  ##主库日志
    -> MASTER_LOG_POS=690;  ##主库日志开始位置

mysql> start slave;  ##保存缓存
Query OK, 0 rows affected (0.15 sec)

mysql> show slave status\G  ##IO线程与SQL线程正常
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes

3.测试主从复制是否完成
主库进行库的创建

mysql> create database westos;  ##创建westos库
Query OK, 1 row affected (0.10 sec)

mysql> show databases;  ##创建成功
+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| performance_schema |
| sys                |
| westos             |
+--------------------+
5 rows in set (0.00 sec)

从库查看

mysql> show databases;  ##同步成功
+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| performance_schema |
| sys                |
| westos             |
+--------------------+
5 rows in set (0.00 sec)

注意：对数据库的操作只可以在主库，因为主库不会去同步从库的信息
三.基于GTID的主从复制
1.基于GTID主从复制的原理
在传统的复制里面，当发生故障，需要主从切换，需要找到binlog和pos点，然后将主节点指向新的主节点，相对来说比较麻烦，也容易出错。在MySQL 5.6里面，不用再找binlog和pos点，我们只需要知道主节点的ip，端口，以及账号密码就行，因为复制是自动的，MySQL会通过内部机制GTID自动找点同步

从服务器连接到主服务器之后，把自己执行过的GTID(Executed_Gtid_Set) 、获取到的GTID(Retrieved_Gtid_Set）发给主服务器，主服务器把从服务器缺少的GTID及对应的transactions发过去补全即可。当主服务器挂掉的时候，找出同步最成功的那台从服务器，直接把它提升为主即可。如果硬要指定某一台不是最新的从服务器提升为主， 先change到同步最成功的那台从服务器， 等把GTID全部补全了，就可以把它提升为主了
2.实验环境的搭建
（1）配置文件更改
主库

[root@server1 ~]# vim /etc/my.cnf
log-bin=mysql-bin
server-id=1
gtid_mode=ON
enforce-gtid-consistency=true
[root@server1 ~]# systemctl start mysqld

从库

[root@server2 ~]# vim /etc/my.cnf
server-id=2
gtid_mode=ON
enforce_gtid_consistency=true
[root@server2 ~]# systemctl start mysqld
[root@server2 ~]# cd /var/lib/mysql
[root@server2 mysql]# cat relay-log.info  ##查看之前的同步信息
7
./server2-relay-bin.000007
367
mysql-bin.000005
154
0
0
1

（2）设置gtid同步
主库

[root@server1 ～]# mysql -uroot -pWestos+001
mysql> grant replication slave on *.* to repl@'172.25.4.%' identified by 'Westos+001'; 

从库

[root@server2 mysql]# mysql -uroot -pWestos+001
mysql> stop slave;
mysql> change master to
    -> master_host='172.25.4.111',
    -> master_user='repl',
    -> master_password='Westos+001',
    -> master_auto_position=1;
mysql> start slave;
mysql> show slave status\G  ##此时GTID的两个参数均为空的
           Retrieved_Gtid_Set: 
            Executed_Gtid_Set: 

（3）测试GTID同步
主库创建数据

mysql> show databases;  ##查看数据库
+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| performance_schema |
| sys                |
| westos             |
+--------------------+
5 rows in set (0.00 sec)

mysql> use westos;  ##使用数据库
Database changed
mysql> create table usertb(  ##创建表格
    -> username varchar(10) not null,
    -> password varchar(15) not null);
Query OK, 0 rows affected (0.82 sec)

mysql> desc usertb;  ##查看表格
+----------+-------------+------+-----+---------+-------+
| Field    | Type        | Null | Key | Default | Extra |
+----------+-------------+------+-----+---------+-------+
| username | varchar(10) | NO   |     | NULL    |       |
| password | varchar(15) | NO   |     | NULL    |       |
+----------+-------------+------+-----+---------+-------+
2 rows in set (0.02 sec)

mysql> insert into usertb values ('user1','11111');  ##表格内插入数据
Query OK, 1 row affected (0.18 sec)

mysql> select from usertb;  ##查看
ERROR 1064 (42000): You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'from usertb' at line 1
mysql> select * from usertb;
+----------+----------+
| username | password |
+----------+----------+
| user1    | 11111    |
+----------+----------+
1 row in set (0.00 sec)

从库查看

mysql> use westos;
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
mysql> select * from usertb;  ##同步成功
+----------+----------+
| username | password |
+----------+----------+
| user1    | 11111    |
+----------+----------+
1 row in set (0.00 sec)
mysql> show slave status\G
           Retrieved_Gtid_Set: c96e01c1-9e4d-11e9-a43a-5254002a986a:1-2
            Executed_Gtid_Set: c96e01c1-9e4d-11e9-a43a-5254002a986a:1-2
mysql> use mysql;

Database changed
mysql> select * from gtid_executed;  ##查看gtid模式复制的起始和结束位置
+--------------------------------------+----------------+--------------+
| source_uuid                          | interval_start | interval_end |
+--------------------------------------+----------------+--------------+
| c96e01c1-9e4d-11e9-a43a-5254002a986a |              1 |            1 |
| c96e01c1-9e4d-11e9-a43a-5254002a986a |              2 |            2 |
+--------------------------------------+----------------+--------------+

三.mysql的半同步
1.半同步的设置
主库

[root@server1 ~]# mysql -uroot -pWestos+001
mysql> INSTALL PLUGIN rpl_semi_sync_master SONAME 'semisync_master.so';  ##安扎插件
mysql> SELECT PLUGIN_NAME, PLUGIN_STATUS FROM INFORMATION_SCHEMA.PLUGINS WHERE PLUGIN_NAME LIKE '%semi%';  ##查看插件是否安装
+----------------------+---------------+
| PLUGIN_NAME          | PLUGIN_STATUS |
+----------------------+---------------+
| rpl_semi_sync_master | ACTIVE        |
+----------------------+---------------+
mysql> SET GLOBAL rpl_semi_sync_master_enabled =1;  ##激活插件
mysql> show status like '%rpl%';
+--------------------------------------------+-------+
| Variable_name                              | Value |
+--------------------------------------------+-------+
| Rpl_semi_sync_master_clients               | 0     |
| Rpl_semi_sync_master_net_avg_wait_time     | 0     |
| Rpl_semi_sync_master_net_wait_time         | 0     |
| Rpl_semi_sync_master_net_waits             | 0     |
| Rpl_semi_sync_master_no_times              | 0     |
| Rpl_semi_sync_master_no_tx                 | 0     |
| Rpl_semi_sync_master_status                | ON    |  ##开启
| Rpl_semi_sync_master_timefunc_failures     | 0     |
| Rpl_semi_sync_master_tx_avg_wait_time      | 0     |
| Rpl_semi_sync_master_tx_wait_time          | 0     |
| Rpl_semi_sync_master_tx_waits              | 0     |
| Rpl_semi_sync_master_wait_pos_backtraverse | 0     |
| Rpl_semi_sync_master_wait_sessions         | 0     |
| Rpl_semi_sync_master_yes_tx                | 0     |
+--------------------------------------------+-------+
mysql> show variables like '%rpl%';
+-------------------------------------------+------------+
| Variable_name                             | Value      |
+-------------------------------------------+------------+
| rpl_semi_sync_master_enabled              | ON         |  ##是否开启半同步
| rpl_semi_sync_master_timeout              | 10000      |  ##切换复制的延迟时间为10s
| rpl_semi_sync_master_trace_level          | 32         |  ##开启半同步复制模式的调试级别默认32
| rpl_semi_sync_master_wait_for_slave_count | 1          |  ##至少有1个slave接收到日志
| rpl_semi_sync_master_wait_no_slave        | ON         |  ##是否允许master的每个事物提交后都等待slave的回复信号
| rpl_semi_sync_master_wait_point           | AFTER_SYNC |  ##等待的point
| rpl_stop_slave_timeout                    | 31536000   |  ##控制stop slave 的执行时间，在重放一个大的事务的时候,突然执行stop slave,命令 stop slave会执行很久,这个时候可能产生死锁或阻塞,严重影响性能
+-------------------------------------------+------------+

从库

mysql> INSTALL PLUGIN rpl_semi_sync_master SONAME 'semisync_master.so';
mysql> SELECT PLUGIN_NAME, PLUGIN_STATUS FROM INFORMATION_SCHEMA.PLUGINS WHERE PLUGIN_NAME LIKE '%semi%';
+----------------------+---------------+
| PLUGIN_NAME          | PLUGIN_STATUS |
+----------------------+---------------+
| rpl_semi_sync_master | ACTIVE        |
+----------------------+---------------+

mysql> SET GLOBAL rpl_semi_sync_master_enabled =1;
mysql> STOP SLAVE IO_THREAD; 
mysql> START SLAVE IO_THREAD;  ##开启IO线程使半同步生效
mysql> show variables like '%rpl%';
+-------------------------------------------+------------+
| Variable_name                             | Value      |
+-------------------------------------------+------------+
| rpl_semi_sync_master_enabled              | ON         |  ##是否开启同步
| rpl_semi_sync_master_timeout              | 10000      |
| rpl_semi_sync_master_trace_level          | 32         |
| rpl_semi_sync_master_wait_for_slave_count | 1          |
| rpl_semi_sync_master_wait_no_slave        | ON         |
| rpl_semi_sync_master_wait_point           | AFTER_SYNC |
| rpl_stop_slave_timeout                    | 31536000   |
+-------------------------------------------+------------+
mysql> STOP SLAVE IO_THREAD;

2.测试半同步是否成功
主库

mysql> use westos;
Database changed
mysql> insert into usertb values ('user2','22222');  ##插入数据
Query OK, 1 row affected (10.14 sec)  ##等待10s才成功，因为上面超时时间是10s，10s后如果没有收到slave节点的返回，就会切换到异步复制
mysql> show status like '%rpl%';
+--------------------------------------------+-------+
| Variable_name                              | Value |
+--------------------------------------------+-------+
| Rpl_semi_sync_master_clients               | 0     |
| Rpl_semi_sync_master_net_avg_wait_time     | 0     |
| Rpl_semi_sync_master_net_wait_time         | 0     |
| Rpl_semi_sync_master_net_waits             | 0     |
| Rpl_semi_sync_master_no_times              | 1     |
| Rpl_semi_sync_master_no_tx                 | 1     |  ##待同步的事物为1
| Rpl_semi_sync_master_status                | OFF   |  ##此时半同步状态为OFF
| Rpl_semi_sync_master_timefunc_failures     | 0     |
| Rpl_semi_sync_master_tx_avg_wait_time      | 0     |
| Rpl_semi_sync_master_tx_wait_time          | 0     |
| Rpl_semi_sync_master_tx_waits              | 0     |
| Rpl_semi_sync_master_wait_pos_backtraverse | 0     |
| Rpl_semi_sync_master_wait_sessions         | 0     |
| Rpl_semi_sync_master_yes_tx                | 0     |
+--------------------------------------------+-------+

从库

mysql> use westos;
mysql> select * from usertb;  ##查看表信息，无半同步插入的数据
+----------+----------+
| username | password |
+----------+----------+
| user1    | 11111    |
+----------+----------+
mysql> START SLAVE IO_THREAD;  ##开启IO线程使半同步生效
mysql> select * from usertb;  ##再次查看，半同步的数据同步成功
+----------+----------+
| username | password |
+----------+----------+
| user1    | 11111    |
| user2    | 22222    |
+----------+----------+

3.进程的查看

mysql> show processlist;
+----+-------------+-----------+--------+---------+------+--------------------------------------------------------+------------------+
| Id | User        | Host      | db     | Command | Time | State                                                  | Info             |
+----+-------------+-----------+--------+---------+------+--------------------------------------------------------+------------------+
|  6 | system user |           | NULL   | Connect | 1332 | Slave has read all relay log; waiting for more updates | NULL             |
|  7 | root        | localhost | westos | Query   |    0 | starting                                               | show processlist |
|  9 | system user |           | NULL   | Connect | 1216 | Waiting for master to send event                       | NULL             |
+----+-------------+-----------+--------+---------+------+--------------------------------------------------------+------------------+

参数的含义
id列:一个标识
user列: 显示当前用户，如果不是root，这个命令就只显示你权限范围内的sql语句
host列:显示这个语句是从哪个ip 的哪个端口上发出的。可用来追踪出问题语句的用户
db列:显示这个进程目前连接的是哪个数据库
command列:显示当前连接的执行的命令，一般就是休眠（sleep），查询（query），连接（connect）
time列:此这个状态持续的时间，单位是秒
state列:显示使用当前连接的sql语句的状态
四.mysql的组复制（全同步复制）
三台主机
172.25.4.111 server1
172.25.4.112 server2
172.25.4.113 server3
1.对3台主机的数据库进行初始化
server1

[root@server1 ~]# cd /var/lib/mysql
[root@server1 mysql]# cat auto.cnf 
[auto]
server-uuid=c96e01c1-9e4d-11e9-a43a-5254002a986a  ##复制uuid
[root@server1 mysql]# rm -fr /var/lib/mysql/*  ##删除数据

server2

[root@server2 ~]# rm -fr /var/lib/mysql/*

2.主配置文件的更改
server1

server-id=1
gtid_mode=ON
enforce_gtid_consistency=ON
master_info_repository=TABLE
relay_log_info_repository=TABLE
binlog_checksum=NONE  ##关闭binlog校验
log_slave_updates=ON
log_bin=binlog
binlog_format=ROW  ##组复制依赖于行的复制格式

transaction_write_set_extraction=XXHASH64
loose-group_replication_group_name="c96e01c1-9e4d-11e9-a43a-5254002a986a"  ##uuid
loose-group_replication_start_on_boot=off
loose-group_replication_local_address= "172.25.4.111:24901"
loose-group_replication_group_seeds= "172.25.4.111:24901,172.25.4.112:24901,172.25.4.113:24901"
loose-group_replication_bootstrap_group=off  ##插件是否自动引导，这个选项一般都要off掉，只需要由发起组复制的节点开启，并只启动一次，如果是on，下次再启动时，会生成一个同名的组，可能会发生脑裂
loose-group_replication_ip_whitelist="127.0.0.1,172.25.4.0/24"
loose-group_replication_enforce_update_everywhere_checks=ON
loose-group_replication_single_primary_mode=OFF  ##后两行是开启多主模式的参数
[root@server1 mysql]# systemctl start mysqld

server2

server-id=2
gtid_mode=ON
enforce_gtid_consistency=ON
master_info_repository=TABLE
relay_log_info_repository=TABLE
binlog_checksum=NONE
log_slave_updates=ON
log_bin=binlog
binlog_format=ROW

transaction_write_set_extraction=XXHASH64
loose-group_replication_group_name="c96e01c1-9e4d-11e9-a43a-5254002a986a"  ##与server1的uuid相同
loose-group_replication_start_on_boot=off
loose-group_replication_local_address= "172.25.4.112:24901"
loose-group_replication_group_seeds= "172.25.4.111:24901,172.25.4.112:24901,172.25.4.113:24901"
loose-group_replication_bootstrap_group=off
loose-group_replication_ip_whitelist="127.0.0.1,172.25.4.0/24"
loose-group_replication_enforce_update_everywhere_checks=ON
loose-group_replication_single_primary_mode=OFF
[root@server2 ~]# systemctl start mysqld

server3

server-id=3
gtid_mode=ON
enforce_gtid_consistency=ON
master_info_repository=TABLE
relay_log_info_repository=TABLE
binlog_checksum=NONE
log_slave_updates=ON
log_bin=binlog
binlog_format=ROW

transaction_write_set_extraction=XXHASH64
loose-group_replication_group_name="c96e01c1-9e4d-11e9-a43a-5254002a986a"
loose-group_replication_start_on_boot=off
loose-group_replication_local_address= "172.25.4.113:24901"
loose-group_replication_group_seeds= "172.25.4.111:24901,172.25.4.112:24901,172.25.4.113:24901"
loose-group_replication_bootstrap_group=off
loose-group_replication_ip_whitelist="127.0.0.1,172.25.4.0/24"
loose-group_replication_enforce_update_everywhere_checks=ON
loose-group_replication_single_primary_mode=OFF
[root@server3 mysql]# systemctl start mysqld

3.密码初始化
server1

[root@server1 mysql]# cat /var/log/mysqld.log | grep password
2019-07-04T15:21:25.584994Z 1 [Note] A temporary password is generated for root@localhost: wO;79+Tqx*lJ
[root@server1 mysql]# mysql_secure_installation

server2

[root@server2 ~]# cat /var/log/mysqld.log |grep password
2019-07-04T15:31:14.238548Z 1 [Note] A temporary password is generated for root@localhost: _+8xha,qlP:V
[root@server2 ~]# mysql_secure_installation

server3

[root@server3 mysql]# cat /var/log/mysqld.log | grep password
2019-07-04T15:41:44.247994Z 1 [Note] A temporary password is generated for root@localhost: aahJwqgSf1!f
[root@server3 mysql]# mysql_secure_installation[root@server3 mysql]# cat /var/log/mysqld.log | grep password
2019-07-04T15:41:44.247994Z 1 [Note] A temporary password is generated for root@localhost: aahJwqgSf1!f
[root@server3 mysql]# mysql_secure_installation

4.设置组同步
server1

[root@server1 mysql]# mysql -uroot -pWestos+001
mysql> SET SQL_LOG_BIN=0;  ##关闭二进制日志，防止传到其他server上
mysql> CREATE USER rpl_user@'%' IDENTIFIED BY 'Westos+001';
mysql> GRANT REPLICATION SLAVE ON *.* TO rpl_user@'%';
mysql> FLUSH PRIVILEGES;
mysql> SET SQL_LOG_BIN=1;  ##开启日志
mysql> CHANGE MASTER TO MASTER_USER='rpl_user', MASTER_PASSWORD='Westos+001' FOR CHANNEL 'group_replication_recovery';
mysql> INSTALL PLUGIN group_replication SONAME 'group_replication.so';  ##插入插件模块
mysql> SHOW PLUGINS;  ##查看插件
| group_replication          | ACTIVE   | GROUP REPLICATION  | group_replication.so | GPL     |
+----------------------------+----------+--------------------+----------------------+---------+
mysql> SET GLOBAL group_replication_bootstrap_group=ON;  ##组复制发起节点则开启此参
mysql> START GROUP_REPLICATION;
mysql> SET GLOBAL group_replication_bootstrap_group=OFF;

mysql> SELECT * FROM  performance_schema.replication_group_members;  ##查看server1是否上线
+---------------------------+--------------------------------------+-------------+-------------+--------------+
| CHANNEL_NAME              | MEMBER_ID                            | MEMBER_HOST | MEMBER_PORT | MEMBER_STATE |
+---------------------------+--------------------------------------+-------------+-------------+--------------+
| group_replication_applier | 6342afd9-9e6f-11e9-9c01-5254002a986a | server1     |        3306 | ONLINE       |
+---------------------------+--------------------------------------+-------------+-------------+--------------+

server2

[root@server2 ~]# mysql -uroot -pWestos+001
mysql> SET SQL_LOG_BIN=0;
CREATE USER rpl_user@'%' IDENTIFIED BY 'Westos+001';
GRANT REPLICATION SLAVE ON *.* TO rpl_user@'%';
mysql> FLUSH PRIVILEGES;
mysql> SET SQL_LOG_BIN=1;
mysql> CHANGE MASTER TO MASTER_USER='rpl_user', MASTER_PASSWORD='Westos+001' FOR CHANNEL 'group_replication_recovery';
INSTALL PLUGIN group_replication SONAME 'group_replication.so';
INSTALL PLUGIN group_replication SONAME 'group_replication.so';
START GROUP_REPLICATION;
ERROR 3092 (HY000): The server is not configured properly to be an active member of the group. Please see more details on error log.
mysql> STOP GROUP_REPLICATION;  ##此时开启组复制报错

查看日志

[root@server2 ~]# cat /var/log/mysqld.log 
2019-07-04T15:36:13.066851Z 0 [Note] Plugin group_replication reported: 'To force this member into the group you can use the group_replication_allow_local_disjoint_gtids_join option'

解决此问题

mysql> STOP GROUP_REPLICATION;
mysql> set global group_replication_allow_local_disjoint_gtids_join=on;
mysql> START GROUP_REPLICATION;

server3与server2配置完全相同
查看3台主机是否都上线

mysql> SELECT * FROM performance_schema.replication_group_members;
+---------------------------+--------------------------------------+-------------+-------------+--------------+
| CHANNEL_NAME              | MEMBER_ID                            | MEMBER_HOST | MEMBER_PORT | MEMBER_STATE |
+---------------------------+--------------------------------------+-------------+-------------+--------------+
| group_replication_applier | 399f228a-9e72-11e9-8b6f-5254009c245f | server3     |        3306 | ONLINE       |
| group_replication_applier | 6342afd9-9e6f-11e9-9c01-5254002a986a | server1     |        3306 | ONLINE       |
| group_replication_applier | c2205e49-9e70-11e9-9f73-525400762b1a | server2     |        3306 | ONLINE       |
+---------------------------+--------------------------------------+-------------+-------------+--------------+

5.测试组同步是否成功
（1）在server1进行数据库编辑

mysql> creatase database test;

mysql> use test;
Database changed
mysql> create table t1 (c1 int primary key,c2 text not null);
Query OK, 0 rows affected (0.61 sec)

mysql> insert into t1 values (1,'luis');
Query OK, 1 row affected (0.61 sec)

mysql> select * from t1;
+----+------+
| c1 | c2   |
+----+------+
|  1 | luis |
+----+------+
1 row in set (0.00 sec)

在其他主机查看
server2和server3均可以看到

mysql> use test;
mysql> select * from t1;
+----+------+
| c1 | c2   |
+----+------+
|  1 | luis |
+----+------+

（2）在sever3上进行数据库的编辑

mysql> insert into t1 values (2,'wsp');
Query OK, 1 row affected (0.61 sec)

mysql> select * from t1;
+----+------+
| c1 | c2   |
+----+------+
|  1 | luis |
|  2 | wsp  |
+----+------+

其他主机进行查看，均可以查看

mysql> select * from t1;
+----+------+
| c1 | c2   |
+----+------+
|  1 | luis |
|  2 | wsp  |
+----+------+

由此可知：数据库组复制，可以在组中的任何一台主机中进行读写