一、配置组复制的步骤
1、初始化数据目录
2、配置主服务器（primary）
3、配置辅助服务器（secondaries）
4、启动mysql实例
5、安装组复制插件（primary and secondaries）
6、创建复制用户（replication user）的用户
7、在主服务器上启动组复制
8、将辅助（secondaries）服务器连接到主（primary）服务器
9、确认组复制的状态
二、配置过程
1、初始化数据目录
[mysql@node01 ~]$ mysqld --no-defaults --initialize-insecure --user=mysql --basedir=/usr/local/mysql --datadir=/data/mysql3306/data
[mysql@node02 ~]$ mysqld --no-defaults --initialize-insecure --user=mysql --basedir=/usr/local/mysql --datadir=/data/mysql3306/data
[mysql@node03 ~]$ mysqld --no-defaults --initialize-insecure --user=mysql --basedir=/usr/local/mysql --datadir=/data/mysql3306/data
2、配置primary
配置primary节点，我们需要准备好数据目录、端口，GTID以及组复制相关参数选项，还要添加plugin目录确保服务器可以找到组复制相关的插件，并且要打开binlog的日志校验。
因为group_replication_group_seeds参数需要配置组复制初始参与的服务器列表，所以必须提前决定每个服务器将要使用的端口。组复制设置中每个服务器需要配置两个端
口，一个用于客户端连接，一个用于内部组复制的消息通信。
node01:
[mysqld]
datadir=/data/mysql3306/data
basedir=/usr/local/mysql
plugin_dir=/usr/local/mysql/lib/plugin
port=24801
socket=/tmp/mysql.sock
server_id=1
gtid_mode= ON
enforce_gtid_consistency= ON
binlog_checksum=NONE
transaction_write_set_extraction = XXHASH64
loose-group_replication_recovery_use_ssl= ON
loose-group_replication_group_name = '5dbabbe6-8050-49a0-9131-1de449167446'
loose-group_replication_start_on_boot = OFF
loose-group_replication_local_address = '192.168.211.128:24901'
loose-group_replication_group_seeds = '192.168.211.128:24901,192.168.211.129:24902,192.168.211.130:24903'
loose-group_replication_bootstrap_group = OFF
[mysql]
port=24801
socket=/tmp/mysql.sock
这里参数文件没有配置log-bin参数，因为当服务器遇到组复制参数时，它将自动启用binlog，binlog在组复制里面时是必须的。如果你需要命名binlog文件或者将binlog定向
存储到某个目录中，那么需要配置log-bin参数。
3、配置secondaries
secondaries节点的配置与primary节点类似，区别在于对于特定参数（端口信息，数据目录，套接字，服务器ID）等需要设置成特定的值。除此之外，transaction_write_set_extraction参数
只在primary节点上设置，对于secondaries节点，我们添加group_replication_recovery_get_public_key参数并将其设置为ON。这个参数决定，primary节点到secondaries节点的请求是
否为基于RSA密钥对的密码交换所需的公钥。
node02:
[mysqld]
datadir=/data/mysql3306/data
log-bin=/data/mysql3306/binlog
log_slave_updates = ON
basedir=/usr/local/mysql
plugin_dir=/usr/local/mysql/lib/plugin
port=24802
socket=/tmp/mysql.sock
server_id=2
gtid_mode= ON
enforce_gtid_consistency= ON
binlog_checksum=NONE
loose-group_replication_recovery_get_public_key= ON
loose-group_replication_recovery_use_ssl= ON
loose-group_replication_group_name = '5dbabbe6-8050-49a0-9131-1de449167446'
loose-group_replication_start_on_boot = OFF
loose-group_replication_local_address = '192.168.211.129:24902'
loose-group_replication_group_seeds = '192.168.211.128:24901,192.168.211.129:24902,192.168.211.130:24903'
loose-group_replication_bootstrap_group = OFF
[mysql]
port=24802
socket=/tmp/mysql.sock
node03:
[mysqld]
datadir=/data/mysql3306/data
log-bin=/data/mysql3306/binlog
log_slave_updates = ON
basedir=/usr/local/mysql
plugin_dir=/usr/local/mysql/lib/plugin
port=24803
socket=/tmp/mysql.sock
server_id=3
gtid_mode= ON
enforce_gtid_consistency= ON
binlog_checksum=NONE
loose-group_replication_recovery_get_public_key= ON
loose-group_replication_recovery_use_ssl= ON
loose-group_replication_group_name = '5dbabbe6-8050-49a0-9131-1de449167446'
loose-group_replication_start_on_boot = OFF
loose-group_replication_local_address = '192.168.211.130:24903'
loose-group_replication_group_seeds = '192.168.211.128:24901,192.168.211.129:24902,192.168.211.130:24903'
loose-group_replication_bootstrap_group = OFF
[mysql]
port=24803
socket=/tmp/mysql.sock
4、启动实例
[mysql@node01 ~]$ mysqld --defaults-file=/etc/primary.cnf &
[mysql@node02 ~]$ mysqld --defaults-file=/etc/slave1.cnf &
[mysql@node03 ~]$ mysqld --defaults-file=/etc/slave2.cnf &
启动实例并且输出日志：
$ mysqld --defaults-file=/etc/primary.cnf > primary_output.log 2>&1 &
$ mysqld --defaults-file=/etc/slave1.cnf > slave1_output.log 2>&1 &
$ mysqld --defaults-file=/etc/slave2.cnf > slave2_output.log 2>&1 &
5、安装组复制插件（所有节点）
mysql实例启动后，必须安装组复制插件。
mysql> INSTALL PLUGIN group_replication SONAME 'group_replication.so';
mysql> INSTALL PLUGIN group_replication SONAME 'group_replication.so';
mysql> INSTALL PLUGIN group_replication SONAME 'group_replication.so';
查看插件状态show plugins，或者从information_schema.plugins系统表中查询插件信息。
select * from information_schema.plugins where plugin_name = 'group_replication'\G;
6、创建复制组用户（所有节点）
在启动mysql实例之后，必须创建一个用户，供服务器用于彼此连接。在组复制中，服务器都是互相“交流”的。这些命令与我们在mysql复制中使用的命令相同，我们需要在所有服务器
实例上创建这个用户。
set sql_log_bin=0;
create user rpl@'%' identified by 'mysql';
grant replication slave on *.* to rpl@'%';
flush privileges;
set sql_log_bin=1;
7、在primary上启动复制（primary上执行）
group_replication_bootstrap_group参数通常在第一次启动的时候设置成OFF，因为复制组还没有启动，我们在primary上启动复制组。参数group_replication_boostrap_group是可以动态设置的，我们可以动态的打开和关闭这个参数。
set global group_replication_bootstrap_group=ON;
start group_replication;
set global group_replication_bootstrap_group=OFF;
8、连接secondaries到primary节点（secondaries上执行）
mysql> change master to master_user="rpl",master_password='mysql' for channel 'group_replication_recovery';
Query OK, 0 rows affected, 2 warnings (0.01 sec)
现在我们已经将辅助设备配置为连接到主服务器，但是连接还没建立，我们必须通过启动组复制来完成此过程。
9、在secondaries上启动组复制
mysql> start group_replication;
2020-09-29T00:17:15.035829Z 10 [System] [MY-010597] [Repl] 'CHANGE MASTER TO FOR CHANNEL 'group_repl ication_applier' executed'. Previous state master
_host='', master_port= 3306, master_log_file='', ma ster_log_pos= 4, master_bind=''. New state master_host='<NULL>', master_port= 0, master_log_file=
'', master_log_pos= 4, master_bind=''.
Query OK, 0 rows affected (4.30 sec)
mysql> 2020-09-29T00:17:19.332065Z 17 [System] [MY-010597] [Repl] 'CHANGE MASTER TO FOR CHANNEL 'gro up_replication_recovery' executed'. Previous state
master_host='', master_port= 3306, master_log_fil e='', master_log_pos= 4, master_bind=''. New state master_host='node01', master_port= 24801, master_
log_file='', master_log_pos= 4, master_bind=''.
2020-09-29T00:17:19.375174Z 18 [Warning] [MY-010897] [Repl] Storing MySQL user name or password info rmation in the master info repository is not secure
and is therefore not recommended. Please conside r using the USER and PASSWORD connection options for START SLAVE; see the 'START SLAVE Syntax' in the
MySQL Manual for more information.
2020-09-29T00:17:19.395921Z 18 [System] [MY-010562] [Repl] Slave I/O thread for channel 'group_repli cation_recovery': connected to master 'rpl@node01:24
801',replication started in log 'FIRST' at posit ion 4
2020-09-29T00:17:19.511547Z 17 [System] [MY-010597] [Repl] 'CHANGE MASTER TO FOR CHANNEL 'group_repl ication_recovery' executed'. Previous state master_ho
st='node01', master_port= 24801, master_log_fil e='', master_log_pos= 4, master_bind=''. New state master_host='<NULL>', master_port= 0, master_log_ file=
'', master_log_pos= 4, master_bind=''.
start group_replication命令通常不会报告任何错误，并且可能需要更长的时间才能返回，这是因为当secondaries连接到主节点并开始与主节点协调时，很多工作都是在后台进行的。
10、确认集群状态
select * from performance_schema.replication_group_members\G;
mysql> select * from performance_schema.replication_group_members\G;
*************************** 1. row ***************************
CHANNEL_NAME: group_replication_applier
MEMBER_ID: 577274f9-00e7-11eb-89e9-00505625f5f5
MEMBER_HOST: node01
MEMBER_PORT: 24801
MEMBER_STATE: ONLINE
MEMBER_ROLE: PRIMARY
MEMBER_VERSION: 8.0.18
*************************** 2. row ***************************
CHANNEL_NAME: group_replication_applier
MEMBER_ID: 75ad0e91-00e7-11eb-a505-005056320441
MEMBER_HOST: node02
MEMBER_PORT: 24802
MEMBER_STATE: ONLINE
MEMBER_ROLE: SECONDARY
MEMBER_VERSION: 8.0.18
*************************** 3. row ***************************
CHANNEL_NAME: group_replication_applier
MEMBER_ID: 7bf9c4cb-00e7-11eb-8cc3-00505625d41f
MEMBER_HOST: node03
MEMBER_PORT: 24803
MEMBER_STATE: ONLINE
MEMBER_ROLE: SECONDARY
MEMBER_VERSION: 8.0.18
3 rows in set (0.01 sec)
ERROR:
No query specified
10、关闭MGR集群（secondaries先，primary后）
node02:
mysql> stop group_replication;
Query OK, 0 rows affected (4.68 sec)
node03:
mysql> stop group_replication;
Query OK, 0 rows affected (4.68 sec)
node01:
mysql> stop group_replication;
Query OK, 0 rows affected (5.17 sec)
开启MGR集群（primary先，secondaries后）
node01：
mysql> SET GLOBAL group_replication_bootstrap_group=ON;
Query OK, 0 rows affected (0.00 sec)
node02：
mysql> start group_replication;
node03：
mysql> start group_replication;
node01：
mysql> SET GLOBAL group_replication_bootstrap_group=OFF;
Query OK, 0 rows affected (0.00 sec)
node01查看集群状态：
mysql> select * from performance_schema.replication_group_members;
+---------------------------+--------------------------------------+-------------+-------------+--------------+-------------+----------------+
| CHANNEL_NAME | MEMBER_ID | MEMBER_HOST | MEMBER_PORT | MEMBER_STATE | MEMBER_ROLE | MEMBER_VERSION |
+---------------------------+--------------------------------------+-------------+-------------+--------------+-------------+----------------+
| group_replication_applier | 577274f9-00e7-11eb-89e9-00505625f5f5 | node01 | 24801 | ONLINE | PRIMARY | 8.0.18 |
| group_replication_applier | 75ad0e91-00e7-11eb-a505-005056320441 | node02 | 24802 | ONLINE | SECONDARY | 8.0.18 |
| group_replication_applier | 7bf9c4cb-00e7-11eb-8cc3-00505625d41f | node03 | 24803 | ONLINE | SECONDARY | 8.0.18 |
+---------------------------+--------------------------------------+-------------+-------------+--------------+-------------+----------------+
3 rows in set (0.00 sec)
