MGR架构~单写模式架构的搭建

一 简介 :MGR一直没有时间测试,今天咱们来初步了解搭建一下呗

二 环境: mysql5.7.20  单台机器 启动三实例

三  mysql 搭建:

1 建立相关目录+

mkdir -p /data/mysql/data_3306

mkdir -p /data/mysql/data_3307

mkdir -p /data/mysql/data_3308

2 脚本初始化

/usr/local/mysql/bin/mysqld --initialize-insecure --user=mysql --basedir=/usr/local/mysql --datadir=/data/mysql/data_3306

/usr/local/mysql/bin/mysqld --initialize-insecure --user=mysql --basedir=/usr/local/mysql --datadir=/data/mysql/data_3307
    /usr/local/mysql/bin/mysqld --initialize-insecure --user=mysql --basedir=/usr/local/mysql --datadir=/data/mysql/data_3308

3 修改密码

mysql -uroot -S /tmp/mysql_3306.sock -e "set password for 'root'@'localhost' = password('test');"

mysql -uroot -ptest -S /tmp/mysql_3306.sock -e "flush privileges"

mysql -uroot -S /tmp/mysql_3307.sock -e "set password for 'root'@'localhost' = password('test');"
   mysql -uroot -ptest -S /tmp/mysql_3307.sock -e "flush privileges"

mysql -uroot -S /tmp/mysql_3308.sock -e "set password for 'root'@'localhost' = password('test');"
   mysql -uroot -ptest -S /tmp/mysql_3308.sock -e "flush privileges"

4  启动命令

nohup /usr/local/mysql/bin/mysqld_safe --defaults-file=/etc/my_3306.cnf &

nohup /usr/local/mysql/bin/mysqld_safe --defaults-file=/etc/my_3307.cnf &
    nohup /usr/local/mysql/bin/mysqld_safe --defaults-file=/etc/my_3308.cnf &

5 三节点安装插件

change master to master_user='repl',master_password='repl' for channel 'group_replication_recovery';
          install plugin group_replication soname 'group_replication.so';

6  重启服务,分别创建复制账户

set sql_log_bin=0;

GRANT REPLICATION SLAVE,REPLICATION CLIENT ON *.* TO 'repl'@'%' IDENTIFIED BY 'repl';
       flush privileges;
       set sql_log_bin=1;

四 MGR配置

1 配置文件添加

report_host=IP //VIEW显示具体IP,而非域名

gtid_mode = ON
enforce_gtid_consistency = ON
master_info_repository = TABLE
relay_log_info_repository = TABLE
binlog_checksum = NONE
log_slave_updates = ON
slave_parallel_type=LOGICAL_CLOCK //并行复制
slave_parallel_workers=8
slave-preserve-commit-order=on

group_replication_unreachable_majority_timeout=5 //当节点存在不可达状态时等待5S就不再等待,防止影响集群服务,如果仍保持UNREACHABLE，则将节点置为ERROR状态.默认无限等待
group_replication_compression_threshold=131072 //开启压缩针对大事务->这里设置是2M
group_replication_transaction_size_limit=20971520//对于最大事务的限制(2g),超过此大小会导致同步失败->默认是0,推荐设置成20M
log_timestamps=SYSTEM//时区
transaction_write_set_extraction = XXHASH64
loose-group_replication_group_name="01e5fb97-be64-41f7-bafd-3afc7a6ab555"
loose-group_replication_start_on_boot=off
loose-group_replication_local_address="127.0.0.1:13306"//除了此处不同 其他都相同
loose-group_replication_group_seeds="127.0.0.1:13306,127.0.0.1:13307,127.0.0.1:13308"
loose-group_replication_bootstrap_group = off
loose-group_replication_ip_whitelist='IP1/24,IP2/24'

2 主节点执行

set global group_replication_bootstrap_group=ON;
      START group_replication;
      set global group_replication_bootstrap_group=OFF;

3 从节点执行

      START group_replication;

4 查看状态

select * from performance_schema.replication_group_members;

三个成员均为online 则代表搭建成功

6 查看主成员

1 select variable_value from performance_schema.global_status where variable_name ='group_replication_primary_member';  获取主UUID

2  select * from performance_schema.replication_group_members; 根据UUID确定primary

7 测试

1 建立一个拥有主键的表并插入数据(MGR架构下表必须拥有主键)

2 查看其他节点是否拥有数据

五 说明

这篇文章是单主的MGR的第一次尝试,以后会继续补充