mysql 5.7 MGR

最近看了一下mysql5.7的MGR集群挺不错的，有单主和多主模式，于是乎搭建测试了一下效果还不错，我指的不错是搭建和维护方面都比较简单。网上绝大多数都是单主模式，当然我这里也是，为了加深印象，特意记录一下搭建过程，等以后再去尝试多主模式，相信大家现在数据库的瓶颈基本都是在写，读写分离虽然是一种可行的解决方案，但是如果数据量很大，写一样会有问题，虽然有些解决方案能部署多个主节点，能同时进行读写，但是脑裂又是一个严重的问题，所以这里MGR集群内置了自动化脑裂防护机制又得到了很多人的青睐，这里MGR简称MySQL Group Replication是MySQL官方于2016年12月推出的一个全新的高可用与高扩展的解决方案。注意本文这里不再阐述原理性的东西。
注意：我这里采用编译安装的方式，如果想简单直接yum安装mysql5.7也行，mysql编译安装需要的磁盘空间还是比较大的，一般在7G左右，所以要提前规划好，用三个节点比较接近生产环境，而且更直接清晰。
详细部署信息如下：

主机名	IP地址	安装软件	用途
apache	192.168.2.25	cmake、boost、mysql	节点
nginx	192.168.2.26	cmake、boost、mysql	节点
kibana	192.168.2.30	cmake、boost、mysql	节点

1、三台机器准备工作

1. rpm -qa mysql mariadb

如果有则卸载即可！
写入hosts文件映射关系，集群用得到
192.168.2.25    apache
192.168.2.26    nginx
192.168.2.30    kibana
2、安装依赖包

1. yum install gcc gcc-c++ ncurses-devel -y

3、安装cmake，下载地址：https://cmake.org/download/

1. tar zxvf cmake-3.7.2.tar.gz

2. cd make-3.7.2

3. ./configure

4. gmake && gmake install

4、安装boost，因为mysql5.7需要，注意这里下载版本是1_59_0和mysql版本是对应的，如果你的MySQL版本和我的不一样，不添加-DWITH_BOOST这个参数时它会报错告诉你需要下载boost的哪个版本。

1. tar zxvf boost_1_59_0.tar.gz

2. cp -r boost_1_59_0 /usr/local/boost

5、安装mysql5.7.17及初始化操作

1. groupadd mysql

2. useradd -M -s /sbin/nologin mysql -g mysql

3. tar zxvf mysql-5.7.17.tar.gz

4. cd mysql-5.7.17

5. cmake -DCMAKE_INSTALL_PREFIX=/data/mysql -DSYSCONFDIR=/etc -DDEFAULT_CHARSET=utf8 -DDEFAULT_COLLATION=utf8_general_ci -DWITH_EXTRA_CHARSETS=all -DWITH_BOOST=/usr/local/boost

6. make

7. make install

8. chown -R mysql.mysql /data/mysql

9. mv /etc/my.cnf /etc/my.cnf.default

10. cp /data/mysql/support-files/my-default.cnf /etc/my.cnf

11. /data/mysql/bin/mysqld –initialize –user=mysql –basedir=/data/mysql –datadir=/data/mysql/data                 //注意初始化会生成一个随机的密码，请牢记

12. echo “PATH=$PATH:/data/mysql/bin” >> /etc/profile

13. source /etc/profile

14. cp /data/mysql/support-files/mysql.server /etc/rc.d/init.d/mysqld

15. chmod +x /etc/rc.d/init.d/mysqld

以上步骤在三台机器上都需要执行

6、开始搭建MGR集群环境，修改第一个节点的my.cnf文件，内容如下：

1. # For advice on how to change settings please see

2. # http://dev.mysql.com/doc/refman/5.7/en/server-configuration-defaults.html

3. # *** DO NOT EDIT THIS FILE. It’s a template which will be copied to the

4. # *** default location during install, and will be replaced if you

5. # *** upgrade to a newer version of MySQL.

6. [mysqld]

7. # Remove leading # and set to the amount of RAM for the most important data

8. # cache in MySQL. Start at 70% of total RAM for dedicated server, else 10%.

9. # innodb_buffer_pool_size = 128M

10. # Remove leading # to turn on a very important data integrity option: logging

11. # changes to the binary log between backups.

12. # log_bin

13. # These are commonly set, remove the # and set as required.

14. basedir = /data/mysql

15. datadir = /data/mysql/data

16. port = 3306

17. socket = /data/mysql/data/mysql.sock

18. log-error = /data/mysql/data/mysqld.log

19. pid-file = /data/mysql/data/mysqld.pid

20. # Remove leading # to set options mainly useful for reporting servers.

21. # The server defaults are faster for transactions and fast SELECTs.

22. # Adjust sizes as needed, experiment to find the optimal values.

23. # join_buffer_size = 128M

24. # sort_buffer_size = 2M

25. # read_rnd_buffer_size = 2M

26. sql_mode=NO_ENGINE_SUBSTITUTION,STRICT_TRANS_TABLES

27. # Group Replication

28. server_id = 1

29. gtid_mode = ON

30. enforce_gtid_consistency = ON

31. master_info_repository = TABLE

32. relay_log_info_repository = TABLE

33. binlog_checksum = NONE

34. log_slave_updates = ON

35. log_bin = binlog

36. binlog_format= ROW

37. transaction_write_set_extraction = XXHASH64

38. loose-group_replication_group_name = ‘ce9be252-2b71-11e6-b8f4-00212844f856’

39. loose-group_replication_start_on_boot = off

40. loose-group_replication_local_address = ‘192.168.2.25:33061’

41. loose-group_replication_group_seeds =’192.168.2.25:33061,192.168.2.26:33061,192.168.2.30:33061′

42. loose-group_replication_bootstrap_group = off

43. [client]

44. socket = /data/mysql/data/mysql.sock

启动mysql服务
/etc/init.d/mysqld start

1. set sql_log_bin=0;

2. create user rpl_user@’%’;

3. grant replication slave on *.* to rpl_user@’%’ identified by ‘rpl_pass’;

4. flush privileges;

5. set sql_log_bin=1;

6. change master to master_user=’rpl_user’,master_password=’rpl_pass’ for channel ‘group_replication_recovery’;

7. install PLUGIN group_replication SONAME ‘group_replication.so’;

8. set global group_replication_bootstrap_group=ON;

9. start group_replication;

10. set global group_replication_bootstrap_group=OFF;

11. select * from performance_schema.replication_group_members;

显示结果如下：

如果出现ONLINE，说明正常，这就是主节点，再搭建两个从节点。

7、第二个节点加入集群，复制刚刚的第一个节点的主配置文件my.cnf，只需要修改两个地方就行，已经用红色标注

1. # For advice on how to change settings please see

2. # http://dev.mysql.com/doc/refman/5.7/en/server-configuration-defaults.html

3. # *** DO NOT EDIT THIS FILE. It’s a template which will be copied to the

4. # *** default location during install, and will be replaced if you

5. # *** upgrade to a newer version of MySQL.

6. [mysqld]

7. # Remove leading # and set to the amount of RAM for the most important data

8. # cache in MySQL. Start at 70% of total RAM for dedicated server, else 10%.

9. # innodb_buffer_pool_size = 128M

10. # Remove leading # to turn on a very important data integrity option: logging

11. # changes to the binary log between backups.

12. # log_bin

13. # These are commonly set, remove the # and set as required.

14. basedir = /data/mysql

15. datadir = /data/mysql/data

16. port = 3306

17. socket = /data/mysql/data/mysql.sock

18. log-error = /data/mysql/data/mysqld.log

19. pid-file = /data/mysql/data/mysqld.pid

20. # Remove leading # to set options mainly useful for reporting servers.

21. # The server defaults are faster for transactions and fast SELECTs.

22. # Adjust sizes as needed, experiment to find the optimal values.

23. # join_buffer_size = 128M

24. # sort_buffer_size = 2M

25. # read_rnd_buffer_size = 2M

26. sql_mode=NO_ENGINE_SUBSTITUTION,STRICT_TRANS_TABLES

27. # Group Replication

28. server_id = 2

29. gtid_mode = ON

30. enforce_gtid_consistency = ON

31. master_info_repository = TABLE

32. relay_log_info_repository = TABLE

33. binlog_checksum = NONE

34. log_slave_updates = ON

35. log_bin = binlog

36. binlog_format= ROW

37. transaction_write_set_extraction = XXHASH64

38. loose-group_replication_group_name = ‘ce9be252-2b71-11e6-b8f4-00212844f856’

39. loose-group_replication_start_on_boot = off

40. loose-group_replication_local_address = ‘192.168.2.26:33061’

41. loose-group_replication_group_seeds =’192.168.2.25:33061,192.168.2.26:33061,192.168.2.30:33061′

42. loose-group_replication_bootstrap_group = off

43. [client]

44. socket = /data/mysql/data/mysql.sock

第二个节点执行如下命令：

1. set sql_log_bin=0;

2. create user rpl_user@’%’;

3. grant replication slave on *.* to rpl_user@’%’ identified by ‘rpl_pass’;

4. set sql_log_bin=1;

5. change master to master_user=’rpl_user’,master_password=’rpl_pass’ for channel ‘group_replication_recovery’;

6. install plugin group_replication SONAME ‘group_replication.so’;

7. set global group_replication_allow_local_disjoint_gtids_join=ON;

8. start group_replication;

显示结果如下：

同理第三个节点加入操作方法也和第二个节点一样。
截图如下：

查询哪个是主节点：

从上图来看很明显apache主机是主节点。

测试步骤：
1、在主库上创建一个库，然后创建表，在两个从库上查询数据是否同步？
2、两个从库只能执行查询操作？
2、手动关闭主库，确认两个从库其中一个是否会变成主库？而且是MEMBER_ID第一个字母按优先级排列的接管主库？

日常维护步骤：
1、如果从库某一节点关闭

1. start group_replication;

2、如果所有的库都关闭后，第一个库作为主库首先执行

1. set global group_replication_bootstrap_group=ON;

2. start group_replication;

剩下的库直接执行即可！

1. start group_replication;

3、如果主库故障，会自动从两个从库选出一个主库，主库启动后再次执行如下命令后会变成从库

1. start group_replication;

转载于:https://blog.51cto.com/12107560/1934073