MYSQL集群技术

mysql的部署方法

在企业中对于MySQL安装的方法通常是源码编译的方式

官网：http://www.mysql.com

mysql的源码安装

1、安装依赖性

dnf install cmake gcc-c++ openssl-devel ncurses-devel.x86_64 libtirpc-devel-1.3.3-8.el7_4.x86_64.rpm rpcgen.x86_64

2、将源码包下载并进行解压

 tar zxf mysql-boost-5.7.44.tar.gz

3、源码的编译安装

[root@mysql-node10 mysql-5.7.44]# cmake \
DCMAKE_INSTALL_PREFIX=/usr/local/mysql \       #指定安装路径
DMYSQL_DATADIR=/data/mysql \                   #指定数据目录
DMYSQL_UNIX_ADDR=/data/mysql/mysql.sock \      #指定套接字文件
DWITH_INNOBASE_STORAGE_ENGINE=1 \              #指定启用INNODB存储引擎，默认用myisam
DWITH_EXTRA_CHARSETS=all \                     #扩展字符集
DDEFAULT_CHARSET=utf8mb4 \                     #指定默认字符集
DDEFAULT_COLLATION=utf8mb4_unicode_ci \        #指定默认校验字符集
DWITH_BOOST=/root/mysql-5.7.44/boost/boost_1_59_0/ #指定c++库依赖
[root@mysql-node10 mysql-5.7.44]# make -j2     #-j2 表示有几个核心就跑几个进程
[root@mysql-node10 mysql-5.7.44# make install

mysql的部署

1、启动脚本

dnf install initscripts-10.11.6-1.el9.x86_64 -y
cd /usr/local/mysql/support-files/
cp mysql.server /etc/init.d/mysqld

2、修改环境变量

vim ~/.bash_profile

export PATH=$PATH:/usr/local/mysql/bin     执行的命令

source ~/.bash_profile

3、创建新用户，生成数据目录

useradd -s /sbin/nologin -M mysql
mkdir -p /data/mysql
chown mysql.mysql /data/mysql/

4、修改配置文件

vim /etc/my.cnf

[mysqld]
datadir=/data/mysql                 #指定数据目录
socket=/data/mysql/mysql.sock       #指定套接字
symbolic-links=0                    #数据只能存放到数据目录中，禁止链接到数据目录

5、数据库初始化建立MySQL基本数据库

mysqld --initialize --user=mysql
/etc/init.d/mysqld start
chkconfig mysqld on

6、数据初始化

mysql_secure_installation

Securing the MySQL server deployment.
Enter password for user root:              #输入当前密码
The existing password for the user account root has expired. Please set a new
password.
New password:                              #输入新密码
Re-enter new password:                     #重复密码
VALIDATE PASSWORD PLUGIN can be used to test passwords
and improve security. It checks the strength of password
and allows the users to set only those passwords which are
secure enough. Would you like to setup VALIDATE PASSWORD plugin?
Press y|Y for Yes, any other key for No: no #是否启用密码插件
Using existing password for root.
#是否要重置密码
Change the password for root ? ((Press y|Y for Yes, any other key for No) : no     

... skipping.
By default, a MySQL installation has an anonymous user,
allowing anyone to log into MySQL without having to have
a user account created for them. This is intended only for
testing, and to make the installation go a bit smoother.
You should remove them before moving into a production
environment.
Remove anonymous users? (Press y|Y for Yes, any other key for No) : y
Success.
Normally, root should only be allowed to connect from
'localhost'. This ensures that someone cannot guess at
the root password from the network.
Disallow root login remotely? (Press y|Y for Yes, any other key for No) : y
Success.
By default, MySQL comes with a database named 'test' that
anyone can access. This is also intended only for testing,
and should be removed before moving into a production
environment.
Remove test database and access to it? (Press y|Y for Yes, any other key for No)
: y
- Dropping test database...
Success.
- Removing privileges on test database...
Success.
Reloading the privilege tables will ensure that all changes
made so far will take effect immediately.

Reload privilege tables now? (Press y|Y for Yes, any other key for No) : y
Success.

mysql的组从复制

配置master

1、编辑配置文件vim /etc/my.cnf

[mysqld]

datadir=/data/mysql

socket=/data/mysql/mysql.sock

symbolic-links=0

log-bin=mysql-bin

server-id=1

/etc/init.d/mysqld restart 重启

2、进入数据库配置用户权限

mysql  -uroot    -plee


mysql: [Warning] Using a password on the command line interface can be insecure.
Welcome to the MySQL monitor. Commands end with ; or \g.
Your MySQL connection id is 3
Server version: 5.7.44-log Source distribution
Copyright (c) 2000, 2023, Oracle and/or its affiliates.
Oracle is a registered trademark of Oracle Corporation and/or its
affiliates. Other names may be trademarks of their respective
owners.
Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
mysql> CREATE USER 'repl'@'%' IDENTIFIED BY 'lee'; ##生成专门用来做复制的用
户，此用户是用于slave端做认证用
mysql> GRANT REPLICATION SLAVE ON *.* TO repl@'%'; ##对这个用户进行授权
mysql> SHOW MASTER STATUS; ##查看master的状态
+------------------+----------+--------------+------------------+----------------
---+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
Executed_Gtid_Set |
+------------------+----------+--------------+------------------+----------------
---+
| mysql-bin.000001 | 350 | | |
|
+------------------+----------+--------------+------------------+----------------
---+
1 row in set (0.00 sec)

cd /data/mysql/
mysqlbinlog mysql-bin.000001 -v         #查看二进制日志

配置slave

1、编辑配置文件 vim /etc/my.cnf

[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
symbolic-links=0
server-id=2

/etc/init.d/mysqld restart   重启mysql

mysql -plee
mysql> CHANGE MASTER TO
MASTER_HOST='172.25.254.10',MASTER_USER='repl',MASTER_PASSWORD='lee',MASTER_LOG_F
ILE='mysql-bin.000001',MASTER_LOG_POS=350;
Query OK, 0 rows affected, 2 warnings (0.01 sec)
mysql> start slave;
Query OK, 0 rows affected (0.01 sec)
mysql> SHOW SLAVE STATUS\G;          查看slave
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 172.25.254.10
Master_User: repl
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000001
Read_Master_Log_Pos: 350
Relay_Log_File: mysql-node2-relay-bin.000002
Relay_Log_Pos: 320
Relay_Master_Log_File: mysql-bin.000001
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: 350
Relay_Log_Space: 533
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: 1
Master_UUID: 888d2164-4b05-11ef-a049-000c299355ea
Master_Info_File: /data/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

测试：

mysql -plee
mysql> CREATE DATABASE lee;
Query OK, 1 row affected (0.00 sec)
mysql> CREATE TABLE lee.userlist (
-> username varchar(20) not null,
-> password varchar(50) not null
-> );
Query OK, 0 rows affected (0.02 sec)
mysql> INSERT INTO lee.userlist VALUE ('lee','123');
Query OK, 1 row affected (0.02 sec)
mysql> SELECT * FROM lee.userlist;
+----------+----------+
| username | password |
+----------+----------+
| lee | 123 |
+----------+----------+
1 row in set (0.00 sec)

在slave中查看数据是否有同步过来

mysql -plee
mysql> SELECT * FROM lee.userlist;
+----------+----------+
| username | password |
+----------+----------+
| lee | 123 |
+----------+----------+
1 row in set (0.00 sec)

有数据时添加slave2

1、完成基础的配置

vim /etc/my.cnf

[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
symbolic-links=0
server-id=3

重启：

/etc/init.d/mysqld restart

从master节点备份数据：mysqldump -uroot -plee lee > lee.sql

2、利用master节点中备份出来的lee.sql在slave2中拉平数据

[root@mysql-node3 ~]# mysql -uroot -plee -e "create database lee;"
[root@mysql-node3 ~]# mysql -uroot -p lee <lee.sql
[root@mysql-node3 ~]# mysql -uroot -plee -e "select * from lee.userlist;"
mysql: [Warning] Using a password on the command line interface can be insecure.
+----------+----------+
| username | password |
+----------+----------+
| user1 | 123 |
| user2 | 123 |
+----------+----------+

3、配置slave2的slave功能

mysql -uroot -plee -e "SHOW MASTER STATUS;"
mysql: [Warning] Using a password on the command line interface can be insecure.
+------------------+----------+--------------+------------------+----------------
---+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
Executed_Gtid_Set |
+------------------+----------+--------------+------------------+----------------
---+
| mysql-bin.000001 | 1251 | | |
|
+------------------+----------+--------------+------------------+----------------
---+
[root@mysql-node3 ~]# mysql -uroot -plee
mysql> CHANGE MASTER TO MASTER_HOST='172.25.254.10', MASTER_USER='repl',
MASTER_PASSWORD='lee', MASTER_LOG_FILE='mysql-bin.000001', MASTER_LOG_POS=1251;
mysql> start slave;
mysql> SHOW SLAVE STATUS\G;
*************************** 1. row ***************************
Slave_IO_State: Connecting to master
Master_Host: 172.25.254.30
Master_User: repl
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000001
Read_Master_Log_Pos: 1251
Relay_Log_File: mysql-node3-relay-bin.000001
Relay_Log_Pos: 4
Relay_Master_Log_File: mysql-bin.000001
Slave_IO_Running: Yes
Slave_SQL_Running: Yes

测试：

[root@mysql-node1 ~]# mysql -uroot -plee -e "INSERT INTO lee.userlist
VALUES('user3','123');"
[root@mysql-node2 mysql]# mysql -uroot -plee -e 'select * from lee.userlist;'
mysql: [Warning] Using a password on the command line interface can be insecure.
+----------+----------+
| username | password |
+----------+----------+
| user1 | 123 |
| user2 | 123 |
| user3 | 123 |
+----------+----------+
[root@mysql-node3 ~]# mysql -uroot -plee -e 'select * from lee.userlist;'
mysql: [Warning] Using a password on the command line interface can be insecure.
+----------+----------+
| username | password |
+----------+----------+
| user1 | 123 |
| user2 | 123 |
| user3 | 123 |
+----------+----------+

延迟复制

延迟复制时用来控制sql线程的，和i/o线程无关

这个延迟复制不是i/o线程过段时间来复制，i/o是正常工作的

是日志已经保存在slave端了，那个sql要等多久进行回放

当master端误操作，可以在slave端进行数据备份

在master中写入数据后过了延迟时间才能被查询到

mysql> STOP SLAVE SQL_THREAD;
mysql> CHANGE MASTER TO MASTER_DELAY=60;
mysql> START SLAVE SQL_THREAD;
mysql> SHOW SLAVE STATUS\G;
Master_Server_Id: 1
Master_UUID: db2d8c92-4dc2-11ef-b6b0-000c299355ea
Master_Info_File: /data/mysql/master.info
SQL_Delay: 60                 #延迟效果
SQL_Remaining_Delay: NULL
Slave_SQL_Running_State: Slave has read all relay log; waiting for more
updates
Master_Retry_Count: 86400

慢查询日志

慢查询，顾名思义，执行很慢的查询

当执行SQL超过long_query_time参数设定的时间阈值（默认10s）时，就被认为是慢查询，这个 SQL语句就是需要优化的

慢查询被记录在慢查询日志里

慢查询日志默认是不开启的

如果需要优化SQL语句，就可以开启这个功能，它可以让你很容易地知道哪些语句是需要优化的。

1、开启慢查询日志

mysql> SET GLOBAL slow_query_log=ON;
Query OK, 0 rows affected (0.00 sec)
mysql> SET long_query_time=4;
Query OK, 0 rows affected (0.00 sec)
mysql> SHOW VARIABLES like "long%";
+-----------------+----------+
| Variable_name | Value |
+-----------------+----------+
| long_query_time | 4.000000 |
+-----------------+----------+
1 row in set (0.00 sec)
mysql> SHOW VARIABLES like "slow%";
+---------------------+----------------------------------+
| Variable_name | Value |
+---------------------+----------------------------------+
| slow_launch_time | 2 |
| slow_query_log | ON | ##慢查询日志开启
| slow_query_log_file | /data/mysql/mysql-node1-slow.log |
+---------------------+----------------------------------+
3 rows in set (0.01 sec)
[root@mysql-node1 ~]# cat /data/mysql/mysql-node1-slow.log #慢查询日志
/usr/local/mysql/bin/mysqld, Version: 5.7.44-log (Source distribution). started
with:
Tcp port: 3306 Unix socket: /data/mysql/mysql.sock

2、测试慢查询

mysql> select sleep (10);
[root@mysql-node1 ~]# cat /data/mysql/mysql-node1-slow.log
/usr/local/mysql/bin/mysqld, Version: 5.7.44-log (Source distribution). started
with:
Tcp port: 3306 Unix socket: /data/mysql/mysql.sock
Time Id Command Argument
# Time: 2024-07-29T17:04:07.612704Z
# User@Host: root[root] @ localhost [] Id: 8
# Query_time: 10.000773 Lock_time: 0.000000 Rows_sent: 1 Rows_examined: 0
SET timestamp=1722272647;
select sleep (10);

mysql的并行复制

slave的进程信息：

默认情况下slave中使用的是sql单线程回放 在master中时多用户读写，如果使用sql单线程回放那么会造成组从延迟严重 开启MySQL的多线程回放可以解决上述问题

MySQL 组提交（Group commit）是一个性能优化特性，它允许在一个事务日志同步操作中将多个 事务的日志记录一起写入。这样做可以减少磁盘I/O的次数，从而提高数据库的整体性能。

在slaves中设定

[root@mysql-node2 ~]# vim /etc/my.cnf
[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
server-id=2
gtid_mode=ON
enforce-gtid-consistency=ON
slave-parallel-type=LOGICAL_CLOCK     #基于组提交，
slave-parallel-workers=16             #开启线程数量
master_info_repository=TABLE          #master信息在表中记录，默认记录

在/data/mysql//master.info
relay_log_info_repository=TABLE       #回放日志信息在表中记录，默认记录

在/data/mysql/relay-log.info
relay_log_recovery=ON                 #日志回放恢复功能开启
[root@mysql-node2 ~]# /etc/init.d/mysql start
Starting MySQL. SUCCESS!

半同步模式

半同步模式原理：

1.用户线程写入完成后master中的dump会把日志推送到slave端

2.slave中的io线程接收后保存到relaylog中继日志

3.保存完成后slave向master端返回ack

4.在未接受到slave的ack时master端时不做提交的，一直处于等待当收到ack后提交到存储引擎

5.在5.6版本中用到的时after_commit模式，after_commit模式时先提交在等待ack返回后输出ok

gtid模式：在master端的写入时多用户读写，在slave端的复制时单线程日志回放，所以slave端一定会延迟与 master端 这种延迟在slave端的延迟可能会不一致，当master挂掉后slave接管，一般会挑选一个和master延迟日 志最接近的充当新的master 那么为接管master的主机继续充当slave角色并会指向到新的master上，作为其slave 这时候按照之前的配置我们需要知道新的master上的pos的id，但是我们无法确定新的master和slave之 间差多少

当激活GITD之后：当master出现问题后，slave2和master的数据最接近，会被作为新的master slave1指向新的master，但是他不会去检测新的master的pos id，只需要继续读取自己gtid_next即可

[root@mysql-node1 ~]# mysqlbinlog -vv /data/mysql/mysql-bin.000003

SET @@SESSION.GTID_NEXT= 'AUTOMATIC' /* added by mysqlbinlog */ /*!*/;
DELIMITER ;
# End of log file
/*!50003 SET COMPLETION_TYPE=@OLD_COMPLETION_TYPE*/;
/*!50530 SET @@SESSION.PSEUDO_SLAVE_MODE=0*/;
[root@mysql-node2 ~]# mysql -plee
mysql> select * from mysql.gtid_executed;
+--------------------------------------+----------------+--------------+
| source_uuid | interval_start | interval_end |
+--------------------------------------+----------------+--------------+
| 768c6b91-4c01-11ef-a514-000c299355ea | 1 | 1 |
+--------------------------------------+----------------+--------------+
1 row in set (0.00 sec)

设置gtid：在master端和slave端开启gtid模式

[root@mysql-node1 ~]# vim /etc/my.cnf
[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
server-id=1
log-bin=mysql-bin
gtid_mode=ON
enforce-gtid-consistency=ON
symbolic-links=0
[root@mysql-node1 ~]# /etc/init.d/mysqld restart
[root@mysql-node2 ~]# vim /etc/my.cnf
[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
server-id=2
log-bin=mysql-bin
gtid_mode=ON
enforce-gtid-consistency=ON
symbolic-links=0
[root@mysql-node2 ~]# /etc/init.d/mysqld restart
[root@mysql-node3 ~]# vim /etc/my.cnf
[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
server-id=3
log-bin=mysql-bin
gtid_mode=ON
enforce-gtid-consistency=ON
symbolic-links=0
[root@mysql-node3 ~]# /etc/init.d/mysqld restart
#停止slave端
[root@mysql-node2 ~]# mysql -p
mysql> stop slave;
Query OK, 0 rows affected (0.00 sec)
[root@mysql-node3 ~]# mysql -p
mysql> stop slave;
Query OK, 0 rows affected (0.01 sec)
#开启slave端的gtid
mysql> CHANGE MASTER TO MASTER_HOST='172.25.254.10', MASTER_USER='repl',
MASTER_PASSWORD='lee', MASTER_AUTO_POSITION=1;
mysql> start slave;
mysql> show slave status\G;
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 172.25.254.10
Master_User: repl
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000003
Read_Master_Log_Pos: 154
Relay_Log_File: mysql-node2-relay-bin.000002
Relay_Log_Pos: 367
Relay_Master_Log_File: mysql-bin.000003
Slave_IO_Running: Yes
Slave_SQL_Running: Yes

**启用半同步模式：**在master端配置启用半同步模式

[root@mysql-node1 ~]# vim /etc/my.cnf
[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
server-id=1
log-bin=mysql-bin
gtid_mode=ON
enforce-gtid-consistency=ON
rpl_semi_sync_master_enabled=1 #开启半同步功能
symbolic-links=0
[root@mysql-node1 ~]# mysql -p lee
#安装半同步插件
mysql> INSTALL PLUGIN rpl_semi_sync_master SONAME 'semisync_master.so';
#查看插件情况
mysql> SELECT PLUGIN_NAME, PLUGIN_STATUS
-> FROM INFORMATION_SCHEMA.PLUGINS
-> WHERE PLUGIN_NAME LIKE '%semi%';
+----------------------+---------------+
| PLUGIN_NAME | PLUGIN_STATUS |
+----------------------+---------------+
| rpl_semi_sync_master | ACTIVE |
+----------------------+---------------+
1 row in set (0.01 sec)
#打开半同步功能
mysql> SET GLOBAL rpl_semi_sync_master_enabled = 1;
#查看半同步功能状态
mysql> SHOW VARIABLES LIKE 'rpl_semi_sync%';
+-------------------------------------------+------------+
| Variable_name | Value |
+-------------------------------------------+------------+
| rpl_semi_sync_master_enabled | ON |
| 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 |
+-------------------------------------------+------------+
mysql> SHOW STATUS LIKE 'Rpl_semi_sync%';
+--------------------------------------------+-------+
| 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 |
| 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 |
+--------------------------------------------+-------+
14 rows in set (0.00 sec)
mysql> show plugins

在slave端开启半同步功能：

[root@mysql-node2 ~]# vim /etc/my.cnf
[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
server-id=1
log-bin=mysql-bin
gtid_mode=ON
enforce-gtid-consistency=ON
rpl_semi_sync_master_enabled=1 #开启半同步功能
symbolic-links=0
[root@mysql-node2 ~]# mysql -plee
mysql> INSTALL PLUGIN rpl_semi_sync_slave SONAME 'semisync_slave.so';
Query OK, 0 rows affected (0.01 sec)
mysql> SET GLOBAL rpl_semi_sync_slave_enabled =1;
Query OK, 0 rows affected (0.00 sec)
mysql> STOP SLAVE IO_THREAD; #重启io线程，半同步才能生效
Query OK, 0 rows affected (0.00 sec)
mysql> START SLAVE IO_THREAD; ##重启io线程，半同步才能生效
Query OK, 0 rows affected (0.00 sec)
mysql> SHOW VARIABLES LIKE 'rpl_semi_sync%';
+---------------------------------+-------+
| Variable_name | Value |
+---------------------------------+-------+
| rpl_semi_sync_slave_enabled | ON |
| rpl_semi_sync_slave_trace_level | 32 |
+---------------------------------+-------+
2 rows in set (0.01 sec)
mysql> SHOW STATUS LIKE 'Rpl_semi_sync%';
+----------------------------+-------+
| Variable_name | Value |
+----------------------------+-------+
| Rpl_semi_sync_slave_status | ON |
+----------------------------+-------+
1 row in set (0.00 sec)

mysql高可用的组复制

首先我们将多个节点共同组成一个复制组，在执行读写（RW）事务的时候，需要通过一致性协议层 （Consensus 层）的同意，也就是读写事务想要进行提交，必须要经过组里"大多数人"（对应 Node 节 点）的同意，大多数指的是同意的节点数量需要大于 （N/2+1），这样才可以进行提交，而不是原发起 方一个说了算。而针对只读（RO）事务则不需要经过组内同意，直接 提交 即可

**单主模式：**单写模式 group 内只有一台节点可写可读，其他节点只可以读。当主服务器失败时，会自动选择新的主 服务器

**多主模式：**组内的所有机器都是 primary 节点，同时可以进行读写操作，并且数据是最终一致的。

MySQL的组复制

为了避免出错，在所有节点中从新生成数据库数据

在mysql-node10中

[root@mysql-node10 ~]# rm -fr /data/mysql/
[root@mysql-node10 ~]# vim /etc/my.cnf
[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
symbolic-links=0
server-id=1 #配置server唯一标识号
disabled_storage_engines="MyISAM,BLACKHOLE,FEDERATED,ARCHIVE,MEMORY" #禁用指定存储
引擎
gtid_mode=ON #启用全局事件标识
enforce_gtid_consistency=ON #强制gtid一致
master_info_repository=TABLE #复制事件数据到表中而不记录在数据目录中
relay_log_info_repository=TABLE
binlog_checksum=NONE #禁止对二进制日志校验
log_slave_updates=ON #打开数据库中继，
#当slave中sql线程读取日志后也会写入到自己的binlog中
log_bin=binlog #重新指定log名称
binlog_format=ROW #使用行日志格式
plugin_load_add='group_replication.so' #加载组复制插件
transaction_write_set_extraction=XXHASH64 #把每个事件编码为加密散列
group_replication_group_name="aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa" #通知插件正
式加入
#或创建的组
名
#名称为
uuid格式
group_replication_start_on_boot=off #在server启动时不自动启动组复
制
group_replication_local_address="172.25.254.10:33061" #指定插件接受其他成员的信息端
口
group_replication_group_seeds="172.25.254.10:33061,172.25.254.20:33061，
172.25.254.30:33061" #本地地址允许访
问成员列表
group_replication_ip_whitelist="172.25.254.0/24,127.0.0.1/8" #主机白名单
#不随系统自启而启动，只在初始成员主机中手动开启，
#需要在两种情况下做设定：1.初始化建组时 2.关闭并重新启动整个组时
group_replication_bootstrap_group=off
group_replication_single_primary_mode=OFF #使用多主模式
group_replication_enforce_update_everywhere_checks=ON #组同步中有任何改变
检测更新
group_replication_allow_local_disjoint_gtids_join=1 #放弃自己信息以
master事件为主
[root@mysql-node10 ~]# mysqld --user=mysql --initialize
[root@mysql-node10 ~]# /etc/init.d/mysqld start
[root@mysql-node10 ~]# mysql -uroot -p初始化后生成的密码 -e "alter user
root@localhost identified by 'lee';"
#配置sql
[root@mysql-node10 ~]# mysql -plee
mysql> SET SQL_LOG_BIN=0;
Query OK, 0 rows affected (0.00 sec)
mysql> CREATE USER rpl_user@'%' IDENTIFIED BY 'lee';
Query OK, 0 rows affected (0.00 sec)
mysql> GRANT REPLICATION SLAVE ON *.* TO rpl_user@'%';
Query OK, 0 rows affected (0.00 sec)
mysql> FLUSH PRIVILEGES;
Query OK, 0 rows affected (0.00 sec)
mysql> SET SQL_LOG_BIN=1;
Query OK, 0 rows affected (0.00 sec)
mysql> CHANGE MASTER TO MASTER_USER='rpl_user', MASTER_PASSWORD='lee' FOR CHANNEL
'group_replication_recovery';
Query OK, 0 rows affected, 2 warnings (0.00 sec)
mysql> SET GLOBAL group_replication_bootstrap_group=ON; #用以指定初始成员，值在第
一台主机中执行
Query OK, 0 rows affected (0.00 sec)
mysql> START GROUP_REPLICATION;
Query OK, 0 rows affected, 1 warning (2.19 sec)
mysql> SET GLOBAL group_replication_bootstrap_group=OFF;
Query OK, 0 rows affected (0.00 sec)
mysql> SELECT * FROM performance_schema.replication_group_members;

复制文件到20和30上

scp /etc/my.cnf root@172.25.254.20:/etc/my.cnf
scp /etc/my.cnf root@172.25.254.30:/etc/my.cnf

修改20和30的配置

[root@mysql-node20 & 30 ~]# rm -fr /data/mysql/
[root@mysql-node10 ~]# vim /etc/my.cnf
[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
symbolic-links=0
server-id=2                         #在30上写3
disabled_storage_engines="MyISAM,BLACKHOLE,FEDERATED,ARCHIVE,MEMORY"
gtid_mode=ON
enforce_gtid_consistency=ON
master_info_repository=TABLE
relay_log_info_repository=TABLE
测试：
在每个节点都可以完成读写
binlog_checksum=NONE
log_slave_updates=ON
log_bin=binlog
binlog_format=ROW
plugin_load_add='group_replication.so'
transaction_write_set_extraction=XXHASH64
group_replication_group_name="aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"
group_replication_start_on_boot=off
group_replication_local_address="172.25.254.20:33061" #在30上要写30
group_replication_group_seeds="172.25.254.10:33061,172.25.254.20:33061，
172.25.254.30:33061"
group_replication_ip_whitelist="172.25.254.0/24,127.0.0.1/8"
group_replication_bootstrap_group=off
group_replication_single_primary_mode=OFF
group_replication_enforce_update_everywhere_checks=ON
group_replication_allow_local_disjoint_gtids_join=1
[root@mysql-node10 ~]# mysqld --user=mysql --initialize
[root@mysql-node10 ~]# /etc/init.d/mysqld start
[root@mysql-node10 ~]# mysql -uroot -p初始化后生成的密码 -e "alter user
root@localhost identified by 'lee';"

配置mysql：

[root@mysql-node20 & 30 ~]# mysql -plee
mysql> SET SQL_LOG_BIN=0;
Query OK, 0 rows affected (0.00 sec)
mysql> CREATE USER rpl_user@'%' IDENTIFIED BY 'lee';
Query OK, 0 rows affected (0.00 sec)
mysql> GRANT REPLICATION SLAVE ON *.* TO rpl_user@'%';
Query OK, 0 rows affected (0.00 sec)
mysql> FLUSH PRIVILEGES;
Query OK, 0 rows affected (0.00 sec)
mysql> SET SQL_LOG_BIN=1;
Query OK, 0 rows affected (0.00 sec)
mysql> CHANGE MASTER TO MASTER_USER='rpl_user', MASTER_PASSWORD='lee' FOR CHANNEL
'group_replication_recovery';
Query OK, 0 rows affected, 2 warnings (0.00 sec)
mysql> START GROUP_REPLICATION;
Query OK, 0 rows affected, 1 warning (2.19 sec)
mysql> SELECT * FROM performance_schema.replication_group_members;

测试：

node1中

[root@mysql-node10 ~]# mysql -p1
mysql> CREATE DATABASE lee;
Query OK, 1 row affected (0.00 sec)
mysql> CREATE TABLE lee.userlist(
-> username VARCHAR(10) PRIMARY KEY NOT NULL,
-> password VARCHAR(50) NOT NULL
-> );
Query OK, 0 rows affected (0.01 sec)
mysql> INSERT INTO lee.userlist VALUES ('user1','111');
Query OK, 1 row affected (0.00 sec)
mysql> SELECT * FROM lee.userlist;
+----------+----------+
| username | password |
+----------+----------+
| user1 | 111 |
+----------+----------+
1 row in set (0.00 sec)

node2

[root@mysql-node20 ~]# mysql -p1
mysql> INSERT INTO lee.userlist values ('user2','222');
Query OK, 1 row affected (0.00 sec)
mysql> select * from lee.userlist;
+----------+----------+
| username | password |
+----------+----------+
| user1 | 111 |
| user2 | 222 |
+----------+----------+
2 rows in set (0.00 sec)

node3

[root@mysql-node30 ~]# mysql -p1
mysql> INSERT INTO lee.userlist values ('user3','333');
Query OK, 1 row affected (0.00 sec)
mysql> select * from lee.userlist;
+----------+----------+
| username | password |
+----------+----------+
| user1 | 111 |
| user2 | 222 |
| user3 | 333 |
+----------+----------+
3 rows in set (0.00 sec)

mysql路由

mysql-router是一个对应用程序透明的InnoDB Cluster连接路由服务，提供负载均衡、应用连接故障转移和客户端路 由。 利用路由器的连接路由特性，用户可以编写应用程序来连接到路由器，并令路由器使用相应的路由策略 来处理连接，使其连接到正确的MySQL数据库服务器

重新打开一台主机：安装mysql-router

[root@mysql-router ~]# rpm -ivh mysql-router-community-8.4.0-1.el7.x86_64.rpm
#配置mysql-router
[root@mysql-router ~]# vim /etc/mysqlrouter/mysqlrouter.conf
[routing:ro]
bind_address = 0.0.0.0
bind_port = 7001
destinations = 172.25.254.10:3306,172.25.254.20:3306,172.25.254.30:3306
routing_strategy = round-robin
[routing:rw]
bind_address = 0.0.0.0
bind_port = 7002
destinations = 172.25.254.30:3306,172.25.254.20:3306,172.25.254.10:3306
routing_strategy = first-available
[root@mysql-router ~]# systemctl start mysqlrouter.service

测试：

#建立测试用户
mysql> CREATE USER lee@'%' IDENTIFIED BY 'lee';
mysql> GRANT ALL ON lee.* TO lee@'%';
#查看调度效果
[root@mysql-node10 & 20 & 30 ~]# watch -1 lsof -i :3306
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
mysqld 9879 mysql 22u IPv6 56697 0t0 TCP *:mysql (LISTEN)
[root@mysql-router ~]# mysql -ulee -plee -h 172.25.254.40 -P 700

mysql高可用MHA

MHA时解决Master的单点故障问题

MHA（Master High Availability）是一套优秀的MySQL高可用环境下故障切换和主从复制的软件。 MHA 的出现就是解决MySQL 单点的问题。 MySQL故障切换过程中，MHA能做到0-30秒内自动完成故障切换操作。 MHA能在故障切换的过程中最大程度上保证数据的一致性，以达到真正意义上的高可用

MHA由两部分组成:

MHAManager (管理节点) MHA Node (数据库节点), MHA Manager 可以单独部署在一台独立的机器上管理多个master-slave集群，也可以部署在一台 slave 节点上。 MHA Manager 会定时探测集群中的 master 节点。 当 master 出现故障时，它可以自动将最新数据的 slave 提升为新的 master， 然后将所有其他的 slave 重新指向新的 master。

特点：

自动故障切换过程中，MHA从宕机的主服务器上保存二进制日志，最大程度的保证数据不丢失 使用半同步复制，可以大大降低数据丢失的风险，如果只有一个slave已经收到了最新的二进制日 志，MHA可以将最新的二进制日志应用于其他所有的slave服务器上，因此可以保证所有节点的数 据一致性 目前MHA支持一主多从架构，最少三台服务，即一主两从

原理：

目前MHA主要支持一主多从的架构，要搭建MHA,要求一个复制集群必须最少有3台数据库服务器， 一主二从，即一台充当Master，台充当备用Master，另一台充当从库。 MHA Node 运行在每台 MySQL 服务器上 MHAManager 会定时探测集群中的master 节点 当master 出现故障时，它可以自动将最新数据的slave 提升为新的master 然后将所有其他的slave 重新指向新的master，VIP自动漂移到新的master。 整个故障转移过程对应用程序完全透明。

MHA的部署

1、主从架构：

在master节点中

[root@mysql-node10 ~]# /etc/init.d/mysqld stop
[root@mysql-node10 ~]# rm -fr /data/mysql/*
[root@mysql-node10 ~]# vim /etc/my.cnf
[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
server-id=1
log-bin=mysql-bin
gtid_mode=ON
log_slave_updates=ON
enforce-gtid-consistency=ON
symbolic-links=0
[root@mysql-node10 ~]# mysqld --user mysql --initialize
[root@mysql-node10 ~]# /etc/init.d/mysqld start
[root@mysql-node10 ~]# mysql_secure_installation
[root@mysql-node10 ~]# mysql -p
mysql> CREATE USER 'repl'@'%' IDENTIFIED BY 'lee';
Query OK, 0 rows affected (0.00 sec)
mysql> GRANT REPLICATION SLAVE ON *.* TO repl@'%';
Query OK, 0 rows affected (0.00 sec)
mysql> INSTALL PLUGIN rpl_semi_sync_master SONAME 'semisync_master.so';
Query OK, 0 rows affected (0.02 sec)
mysql> SET GLOBAL rpl_semi_sync_master_enabled = 1;
Query OK, 0 rows affected (0.00 sec)

在slave1和slave2中

 /etc/init.d/mysqld stop
 rm -fr /data/mysql/*
 vim /etc/my.cnf

[mysqld]
datadir=/data/mysql
socket=/data/mysql/mysql.sock
server-id=1
log-bin=mysql-bin
gtid_mode=ON
log_slave_updates=ON
enforce-gtid-consistency=ON
symbolic-links=0
[root@mysql-node20  ~]# mysqld --user mysql --initialize
[root@mysql-node20  ~]# /etc/init.d/mysqld start
[root@mysql-node20  ~]# mysql_secure_installation
[root@mysql-node20  ~]# mysql -p
mysql> CHANGE MASTER TO MASTER_HOST='172.25.254.10', MASTER_USER='repl',
MASTER_PASSWORD='lee', MASTER_AUTO_POSITION=1;
Query OK, 0 rows affected, 2 warnings (0.00 sec)
mysql> start slave;
Query OK, 0 rows affected (0.00 sec)
mysql> INSTALL PLUGIN rpl_semi_sync_slave SONAME 'semisync_slave.so';
Query OK, 0 rows affected (0.01 sec)
mysql> SET GLOBAL rpl_semi_sync_slave_enabled =1;
Query OK, 0 rows affected (0.00 sec)
mysql> STOP SLAVE IO_THREAD;
Query OK, 0 rows affected (0.00 sec)
mysql> START SLAVE IO_THREAD;
Query OK, 0 rows affected (0.00 sec)
mysql> SHOW STATUS LIKE 'Rpl_semi_sync%';
+----------------------------+-------+
| Variable_name | Value |
+----------------------------+-------+
| Rpl_semi_sync_slave_status | ON |
+----------------------------+-------+
1 row in set (0.01 sec)

在MHA中安装MHA所需要的软件：

[root@mysql-mha ~]# unzip MHA-7.zip
[root@mysql-mha MHA-7]# ls
mha4mysql-manager-0.58-0.el7.centos.noarch.rpm perl-Mail-Sender-0.8.23-
1.el7.noarch.rpm
mha4mysql-manager-0.58.tar.gz perl-Mail-Sendmail-0.79-
21.el7.noarch.rpm
mha4mysql-node-0.58-0.el7.centos.noarch.rpm perl-MIME-Lite-3.030-
1.el7.noarch.rpm
perl-Config-Tiny-2.14-7.el7.noarch.rpm perl-MIME-Types-1.38-
2.el7.noarch.rpm
perl-Email-Date-Format-1.002-15.el7.noarch.rpm perl-Net-Telnet-3.03-
19.el7.noarch.rpm
perl-Log-Dispatch-2.41-1.el7.1.noarch.rpm perl-Parallel-ForkManager-1.18-
2.el7.noarch.rpm
[root@mysql-mha MHA-7]# yum install *.rpm -y
[root@mysql-mha MHA-7]# scp mha4mysql-node-0.58-0.el7.centos.noarch.rpm
root@172.25.254.10:/mnt
[root@mysql-mha MHA-7]# scp mha4mysql-node-0.58-0.el7.centos.noarch.rpm
root@172.25.254.20:/mnt
[root@mysql-mha MHA-7]# scp mha4mysql-node-0.58-0.el7.centos.noarch.rpm
root@172.25.254.30:/mnt
#在sql-node中
[root@mysql-node10 ~]# yum install /mnt/mha4mysql-node-0.58-
0.el7.centos.noarch.rpm -y
[root@mysql-node20 ~]# yum install /mnt/mha4mysql-node-0.58-
0.el7.centos.noarch.rpm -y
[root@mysql-node30 ~]# yum install /mnt/mha4mysql-node-0.58-
0.el7.centos.noarch.rpm -y

配置MHA 的管理环境

因为我们当前只有一套主从，所以我们只需要写一个配置文件即可 rpm包中没有为我们准备配置文件的模板 可以解压源码包后在samples中找到配置文件的模板文件

生成配置文件：

[root@mysql-mha ~]# mkdir /etc/masterha
[root@mysql-mha MHA-7]# tar zxf mha4mysql-manager-0.58.tar.gz
[root@mysql-mha MHA-7]# cd mha4mysql-manager-0.58/samples/conf/
[root@mysql-mha conf]# cat masterha_default.cnf app1.cnf >
/etc/masterha/app1.cnf
#编辑配置文件
[root@mysql-mha ~]# vim /etc/masterha/app1.cnf
[server default]
user=root #mysql管理员用户，因为需要做自动化配置
password=lee #mysql密码
ssh_user=root #ssh远程登陆用户
repl_user=repl #mysql主从复制中负责认证的用户
repl_password=lee #mysql主从复制中负责认证的用户密码
master_binlog_dir= /data/mysql #二进制日志目录
remote_workdir=/tmp #远程工作目录
#此参数使为了提供冗余检测，方式是mha主机网络自身的问题无法连接数据库节点，应为集群之外的主机
secondary_check_script= masterha_secondary_check -s 172.25.254.10 -s
172.25.254.11
ping_interval=3 #每隔3秒检测一次
#发生故障后调用的脚本，用来迁移vip
# master_ip_failover_script= /script/masterha/master_ip_failover
#电源管理脚本
# shutdown_script= /script/masterha/power_manager
#当发生故障后用此脚本发邮件或者告警通知
# report_script= /script/masterha/send_report
#在线切换时调用的vip迁移脚本，手动
# master_ip_online_change_script= /script/masterha/master_ip_online_change
manager_workdir=/etc/masterha #mha工作目录
manager_log=/var/etc/masterha/manager.log #mha日志
[server1]
hostname=172.25.254.10
candidate_master=1 #可能作为master的主机
check_repl_delay=0 ##默认情况下如果一个slave落后master 100M的relay logs的话
#MHA将不会选择该slave作为一个新的master
#因为对于这个slave的恢复需要花费很长时间
#通过设置check_repl_delay=0
#MHA触发切换在选择一个新的master的时候将会忽略复制延时
#这个参数对于设置了candidate_master=1的主机非常有用
#因为这个候选主在切换的过程中一定是新的master
[server2]
hostname=172.25.254.20
candidate_master=1 #可能作为master的主机
check_repl_delay=0
[server3]
hostname=172.25.254.30
no_master=1 #不会作为master的主机

检测网络和ssh免密：

[root@mysql-mha ~]# masterha_check_ssh --conf=/etc/masterha/app1.cnf
#在数据节点master端
mysql> GRANT ALL ON *.* TO root@'%' identified by 'lee'; #允许root远程登陆
#执行检测
[root@mysql-mha ~]# masterha_check_repl --conf=/etc/masterha/app1.cnf

MHA的故障切换

MHA的故障切换过程 共包括以下的步骤：

1.配置文件检查阶段，这个阶段会检查整个集群配置文件配置

2.宕机的master处理，这个阶段包括虚拟ip摘除操作，主机关机操作

3.复制dead master和最新slave相差的relay log，并保存到MHA Manger具体的目录下

4.识别含有最新更新的slave

5.应用从master保存的二进制日志事件（binlog events）

6.提升一个slave为新的master进行复制

7.使其他的slave连接新的master进行复制

切换方式： master未出现故障手动切换
在master数据节点还在正常工作情况下

[root@mysql-mha ~]# masterha_master_switch \
--conf=/etc/masterha/app1.cnf \ #指定配置文件
--master_state=alive \ #指定master节点状态
--new_master_host=172.25.254.20 \ #指定新master节点
--new_master_port=3306 \ #执行新master节点端口
--orig_master_is_new_slave \ #原始master会变成新的slave
--running_updates_limit=10000 #切换的超时时间

#切换过程如下：
[root@mysql-mha masterha]# masterha_master_switch --conf=/etc/masterha/app1.cnf -
-master_state=alive --new_master_host=172.25.254.20 --new_master_port=3306 --
orig_master_is_new_slave --running_updates_limit=10000



检测：
[root@mysql-mha masterha]# masterha_check_repl --conf=/etc/masterha/app1.cnf
Fri Aug 2 18:33:46 2024 - [warning] Global
configuration file /etc/masterha_default.cnf not found. Skipping.
Fri Aug 2 18:33:46 2024 - [info] Reading application default configuration from
/etc/masterha/app1.cnf..
Fri Aug 2 18:33:46 2024 - [info] Reading server configuration from
/etc/masterha/app1.cnf..
Fri Aug 2 18:33:46 2024 - [info] MHA::MasterMonitor version 0.58.
Fri Aug 2 18:33:47 2024 - [info] GTID failover mode = 1
Fri Aug 2 18:33:47 2024 - [info] Dead Servers:
Fri Aug 2 18:33:47 2024 - [info] Alive Servers:
Fri Aug 2 18:33:47 2024 - [info] 172.25.254.10(172.25.254.10:3306)
Fri Aug 2 18:33:47 2024 - [info] 172.25.254.20(172.25.254.20:3306)
Fri Aug 2 18:33:47 2024 - [info] 172.25.254.30(172.25.254.30:3306)
Fri Aug 2 18:33:47 2024 - [info] Alive Slaves:
Fri Aug 2 18:33:47 2024 - [info] 172.25.254.10(172.25.254.10:3306)
Version=5.7.44-log (oldest major version between slaves) log-bin:enabled
Fri Aug 2 18:33:47 2024 - [info] GTID O

master故障手动切换

模拟master故障

/etc/init.d/mysqld stop

在MHA-master中做故障切换

--ignore_last_failover 表示忽略在/etc/masterha/目录中在切换过程中生成的锁文件

 [root@mysql-mha masterha]# masterha_master_switch --master_state=dead -- conf=/etc/masterha/app1.cnf --dead_master_host=192.168.56.12 -- dead_master_port=3306 --new_master_host=192.168.56.11 --new_master_port=3306 -- ignore_last_failover --ignore_last_failover 

恢复故障mysql节点

[root@mysql-node20 tmp]# /etc/init.d/mysqld start
Starting MySQL. SUCCESS!
[root@mysql-node20 tmp]# mysql -p
mysql> CHANGE MASTER TO MASTER_HOST='172.25.254.10', MASTER_USER='repl',
MASTER_PASSWORD='lee', MASTER_AUTO_POSITION=1;
mysql> show slave\G
#测试一主两从是否正常
[root@mysql-mha masterha]# masterha_check_repl --conf=/etc/masterha/app1.cnf
Fri Aug 2 20:15:29 2024 - [info] Checking replication health on 172.25.254.20..
Fri Aug 2 20:15:29 2024 - [info] ok.
Fri Aug 2 20:15:29 2024 - [info] Checking replication health on 172.25.254.30..
Fri Aug 2 20:15:29 2024 - [info] ok.
Fri Aug 2 20:15:29 2024 - [warning] master_ip_failover_script is not defined.
Fri Aug 2 20:15:29 2024 - [warning] shutdown_script is not defined.
Fri Aug 2 20:15:29 2024 - [info] Got exit code 0 (Not master dead).
MySQL Replication Health is OK.

自动切换

[root@mysql-mha masterha]# rm -fr app1.failover.complete

#删掉切换锁文件 #监控程序通过指定配置文件监控master状态，当master出问题后自动切换并退出避免重复做故障切换

[root@mysql-mha masterha]# masterha_manager --conf=/etc/masterha/app1.cnf 
[root@mysql-mha masterha]# cat /etc/masterha/manager.log

恢复故障节点
[root@mysql-node20 mysql]# /etc/init.d/mysqld start
mysql> CHANGE MASTER TO MASTER_HOST='172.25.254.10', MASTER_USER='repl',
MASTER_PASSWORD='lee', MASTER_AUTO_POSITION=1