CentOS7.8-MySQL-MHA高可用部署搭建

目录

Msyql MHA高可用

MHA简介：

MHA 优点：

MHA 故障切换流程

Manager工具：

MHA搭建操作步骤：

一、基础环境搭建

1.1、将主机全部修改主机名&生成密钥

1.2、man-manager修改host同步

1.3、yum安装mysql 5.7版本

1.44、优化MySQL（101、102、103都做）

1.5、将下文导入my.cnf（101、102、103都做）

1.6、mha-slave1

二、开始配置主从复制

2.1、mha-master

2.2、mha-slave1

2.3、mha-slave2

2.4、mha-master确认binlog

2.5、mha-slave1完成从库复制

三、MHA部署

3.1、101-103开始安装mha-node

3.2、配置mha-manager（104）

3.3、测试mha-manager的ssh互信

3.4、mha-manager确认replication状态

3.5、启动mha状态

3.6、配置mha-master-ip信息

---

Msyql MHA高可用

MHA简介：

MHA在mysql故障切换过程中，MHA能做到0-30秒之内自动完成数据库故障切换操作，并且在进行故障切换的过程中，MHA能在做大的程度上保证数据的一致性，以达到真正意义上的高可用，改软件有两部分组成， MHA Manager （管理节点），MHA Node （数据节点），MHA Manager 可以单独部署在一台独立的主机上管理多个master-slave集群，也可以部署在一台slave上， MHA Nodee 运行在每台SQL服务器上，MHA Manager会定时探测集群中的master节点，当master出现故障时，它可以自动将最新的slave提升为新的master节点，然后将所有的slave重新指向新的master，整个故障转移过程对应用程序完全透明。

MHA 优点：

1、主从切换非常迅速，0-30s

2、最大程读上解决数据一致性的问题

3、不需要修改当前已有的mysql架构与配置

4、不需要另外多余的服务器

5、没有性能损耗

6、没有存储引擎的限制

MHA 故障切换流程

1、从dow机的master中保存二进制日志事件

2、识别含有最新更新的slave

3、将更新过数据的slave 中的中继日志 同步到其他的slave

4、应用从master中保存的二进制日志事件

5、提升一个slave 成为新的master

6、使其他的slave 连接新的master 进行复制

Manager工具：

masterha_chexk_ssh :检查MHA的SSH配置

masterha_chexk_repl: 检查Mysql主从复制

masterha_manager :启动MHA

masterha_check_status:检测当前MH的运行状态

masterha_master_monitor:将测master是否宕机

MHA搭建操作步骤：

主机名	主机ip	Server_id	部署应用
mha-master	192.168.200.101	1	mha-node&MySQL-master
mha-slave1	192.168.200.102	2	mha-node&MySQL-slave1
mha-slave2	192.168.200.103	3	mha-node&MySQL-slave2
mha-manager	192.168.200.104	无	mha-node&mha-manager

一、基础环境搭建

1.1、将主机全部修改主机名&生成密钥

 [root@localhost ~]# hostnamectl set-hostname mha-master && bash
 [root@localhost ~]# hostnamectl set-hostname mha-slave1 && bash
 [root@localhost ~]# hostnamectl set-hostname mha-slave2 && bash
 [root@localhost ~]# hostnamectl set-hostname mha-manager && bash
 ssh-keygen   #生成公钥  私钥，
 ssh-copy-id root@192.168.200.% #将公钥发送到目标主机
 #所有主机互相发送ssh公钥，做免密

1.2、man-manager修改host同步

 [root@mha-manager ~]# cat >> /etc/hosts << EOF
 192.168.200.101 mha-master
 192.168.200.102 mha-slave1
 192.168.200.103 mha-slave2
 192.168.200.104 mha-manager
 EOF
 [root@mha-manager ~]# cat /etc/hosts
 127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
 ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
 192.168.200.101 mha-master
 192.168.200.102 mha-slave1
 192.168.200.103 mha-slave2
 192.168.200.104 mha-manager
 ​
 //ping测试
 [root@mha-manager ~]# for i in {master,slave1,slave2,manager}; do ping -c 1
 mha-$i; done
 PING mha-master (192.168.200.101) 56(84) bytes of data.
 64 bytes from mha-master (192.168.200.101): icmp_seq=1 ttl=64 time=0.330 ms
 --- mha-master ping statistics ---
 1 packets transmitted, 1 received, 0% packet loss, time 0ms
 rtt min/avg/max/mdev = 0.330/0.330/0.330/0.000 ms
 PING mha-slave1 (192.168.200.102) 56(84) bytes of data.
 64 bytes from mha-slave1 (192.168.200.102): icmp_seq=1 ttl=64 time=0.295 ms
 --- mha-slave1 ping statistics ---
 1 packets transmitted, 1 received, 0% packet loss, time 0ms
 rtt min/avg/max/mdev = 0.295/0.295/0.295/0.000 ms
 PING mha-slave2 (192.168.200.103) 56(84) bytes of data.
 64 bytes from mha-slave2 (192.168.200.103): icmp_seq=1 ttl=64 time=0.365 ms
 --- mha-slave2 ping statistics ---
 1 packets transmitted, 1 received, 0% packet loss, time 0ms
 rtt min/avg/max/mdev = 0.365/0.365/0.365/0.000 ms
 PING mha-manager (192.168.200.104) 56(84) bytes of data.
 64 bytes from mha-manager (192.168.200.104): icmp_seq=1 ttl=64 time=0.014 ms
 --- mha-manager ping statistics ---
 1 packets transmitted, 1 received, 0% packet loss, time 0ms
 rtt min/avg/max/mdev = 0.014/0.014/0.014/0.000 ms
 #所有主机同步hosts文件
 [root@mha-manager ~]# for i in {master,slave1,slave2};do echo "#####$i•info#####";ssh mha-$i hostname --all-ip;ssh mha-$i cat /etc/hosts; done
 #####master-info#####
 192.168.200.101
 127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
 ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
 192.168.200.101 mha-master
 192.168.200.102 mha-slave1
 192.168.200.103 mha-slave2
 192.168.200.104 mha-manager
 #####slave1-info#####
 192.168.200.102
 127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
 ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
 192.168.200.101 mha-master
 192.168.200.102 mha-slave1
 192.168.200.103 mha-slave2
 192.168.200.104 mha-manager
 #####slave2-info#####
 192.168.200.103
 127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
 ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
 192.168.200.101 mha-master
 192.168.200.102 mha-slave1
 192.168.200.103 mha-slave2
 192.168.200.104 mha-manager

1.3、yum安装mysql 5.7版本

 #下载yum 源
 yum -y install https://repo.mysql.com/mysql57-community-release-el7-10.noarch.rpm
 #下载mysql
 yum -y install --nogpgcheck mysql-community-server

1.44、优化MySQL（101、102、103都做）

 [root@mha-master ~]# cp /etc/my.cnf{,.back}
 [root@mha-master ~]# mkdir -pv /data/mysql/logs
 mkdir: 已创建目录 "/data"
 mkdir: 已创建目录 "/data/mysql"
 mkdir: 已创建目录 "/data/mysql/logs"
 [root@mha-master ~]# chown -R mysql:mysql /data/mysql/

1.5、将下文导入my.cnf（101、102、103都做）

[mysql]
 port=3306  
 bind-address=0.0.0.0  
 [mysqld]
 character_set_server=utf8 
 init_connect='SET NAMES utf8' 
 user = mysql 
 socket=/var/lib/mysql/mysql.sock
 datadir=/data/mysql/data
 max_connections = 2000
 relay-log=/data/mysql/data/relay-log
 log-bin=/data/mysql/data/mysql-bin
 max_binlog_size = 500M
 binlog_cache_size = 128K 
 binlog-ignore-db = mysql
 expire_logs_days = 7
 binlog_format="ROW"
 server-id=1
 skip-name-resolve
 ​
 #慢查询日志优化
 lower_case_table_names=1 
 slow_query_log=on 
 slow_query_log_file=/data/mysql/logs/slow_query_log.log 
 long_query_time=2 
 symbolic-links=0 
 ​
 log-error=/data/mysql/logs/mysql.err 
 ​
 #mysql优化配置
 max_connect_errors = 100 
 query_cache_type=1 
 query_cache_size=128M 
 innodb_buffer_pool_size=4G 
 innodb_buffer_pool_instances=16
 innodb_open_files=800 #innodb 
 innodb_log_file_size=128M 
 innodb_log_files_in_group=3#redo log
 innodb_log_buffer_size=8M 
 innodb_flush_log_at_trx_commit=2 
 gtid_mode = on 
 ​
 binlog-ignore-db = mysql 
 relay-log-recovery=1
 sync_binlog=1 
 read_only = on 
 enforce_gtid_consistency = 1 
 log_slave_updates = 1 
 ​
 #同步my.cnf
 [root@mha-master ~]# scp /etc/my.cnf mha-slave1:/etc/
 my.cnf
 100% 8916 6.7MB/s 00:00
 [root@mha-master ~]# scp /etc/my.cnf mha-slave2:/etc/
 my.cnf
 100% 8916 9.9MB/s 00:00

1.6、mha-slave1

 [root@mha-slave1 ~]# sed -i 's/server-id=1/server-id=2/g' /etc/my.cnf
 [root@mha-slave1 ~]# mkdir -pv /data/mysql/logs
 mkdir: 已创建目录 "/data"
 mkdir: 已创建目录 "/data/mysql"
 mkdir: 已创建目录 "/data/mysql/logs"
 [root@mha-slave1 ~]# chown -R mysql:mysql /data/mysql/
 [root@mha-slave1 ~]# systemctl start mysqld
 [root@mha-slave1 ~]# systemctl enable mysqld
 [root@mha-slave1 ~]# grep password /data/mysql/logs/mysql.err
 2022-01-18T03:18:30.830531Z 1 [Note] A temporary password is generated for
 root@localhost: BqJDC,lef3R&
 [root@mha-slave1 ~]# mysql -uroot -p'BqJDC,lef3R&'
 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 2
 Server version: 5.7.37-log
 Copyright (c) 2000, 2022, 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> set global validate_password_policy=0;
 Query OK, 0 rows affected (0.00 sec)
 mysql> set global validate_password_length=1;
 Query OK, 0 rows affected (0.00 sec)
 mysql> alter user 'root'@'localhost' identified by '123123';
 Query OK, 0 rows affected (5.00 sec)
 mysql> flush privileges;
 Query OK, 0 rows affected (0.01 sec)
 mysql> \q
 Bye

1.6.1、mha-slave2

[root@mha-slave1 ~]# sed -i 's/server-id=1/server-id=2/g' /etc/my.cnf
 [root@mha-slave1 ~]# mkdir -pv /data/mysql/logs
 mkdir: 已创建目录 "/data"
 mkdir: 已创建目录 "/data/mysql"
 mkdir: 已创建目录 "/data/mysql/logs"
 [root@mha-slave1 ~]# chown -R mysql:mysql /data/mysql/
 [root@mha-slave1 ~]# systemctl start mysqld
 [root@mha-slave1 ~]# systemctl enable mysqld
 [root@mha-slave1 ~]# grep password /data/mysql/logs/mysql.err
 2022-01-18T03:18:30.830531Z 1 [Note] A temporary password is generated for
 root@localhost: BqJDC,lef3R&
 [root@mha-slave1 ~]# mysql -uroot -p'BqJDC,lef3R&'
 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 2
 Server version: 5.7.37-log
 Copyright (c) 2000, 2022, 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> set global validate_password_policy=0;
 Query OK, 0 rows affected (0.00 sec)
 mysql> set global validate_password_length=1;
 Query OK, 0 rows affected (0.00 sec)
 mysql> alter user 'root'@'localhost' identified by '123123';
 Query OK, 0 rows affected (5.00 sec)
 mysql> flush privileges;
 Query OK, 0 rows affected (0.01 sec)
 mysql> \q
 Bye

二、开始配置主从复制

2.1、mha-master

[root@mha-master ~]# mysql -uroot -p123123
 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: 5.7.37-log MySQL Community Server (GPL)
 Copyright (c) 2000, 2022, 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.
 //创建复制用户
 mysql> grant replication slave on *.* to repl@'192.168.200.%' identified by
 'repl_user';
 Query OK, 0 rows affected, 1 warning (0.12 sec)
 //创建监控用户
 mysql> grant all on *.* to 'mha'@'192.168.200.%' identified by 'mha_user';
 Query OK, 0 rows affected, 1 warning (0.00 sec)
 mysql> flush privileges;
 Query OK, 0 rows affected (0.04 sec)
 3.6.2：mha-slave1
 3.6.3：mha-slave2
 mysql> \q
 Bye

2.2、mha-slave1

[root@mha-slave1 ~]# mysql -uroot -p123123
 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: 5.7.37-log MySQL Community Server (GPL)
 Copyright (c) 2000, 2022, 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> grant replication slave on *.* to repl@'192.168.200.%' identified by
 'repl_user';
 Query OK, 0 rows affected, 1 warning (0.03 sec)
 mysql> grant all on *.* to 'mha'@'192.168.200.%' identified by 'mha_user';
 Query OK, 0 rows affected, 1 warning (0.00 sec)
 mysql> flush privileges;
 Query OK, 0 rows affected (0.00 sec)
 mysql> \q
 Bye

2.3、mha-slave2

[root@mha-slave1 ~]# mysql -uroot -p123123
 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: 5.7.37-log MySQL Community Server (GPL)
 Copyright (c) 2000, 2022, 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> grant replication slave on *.* to repl@'192.168.200.%' identified by
 'repl_user';
 Query OK, 0 rows affected, 1 warning (0.03 sec)
 mysql> grant all on *.* to 'mha'@'192.168.200.%' identified by 'mha_user';
 Query OK, 0 rows affected, 1 warning (0.00 sec)
 mysql> flush privileges;
 Query OK, 0 rows affected (0.00 sec)
 mysql> \q
 Bye

完成MySQL主从

2.4、mha-master确认binlog

 mha-master
 [root@mha-master ~]# mysql -uroot -p123123
 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 4
 Server version: 5.7.37-log MySQL Community Server (GPL)
 Copyright (c) 2000, 2022, 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> show master status;
 +------------------+----------+--------------+------------------+---------------
 ---------------------------+
 | File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
 Executed_Gtid_Set |
 +------------------+----------+--------------+------------------+---------------
 ---------------------------+
 | mysql-bin.000002 | 1288 | | mysql,mysql | 5b7281ad-780c-
 11ec-8e49-000c297c8a97:1-5 |
 +------------------+----------+--------------+------------------+---------------
 ---------------------------+
 1 row in set (0.00 sec)
 mysql> \q
 Bye

2.5、mha-slave1完成从库复制

[root@mha-slave1 ~]# mysql -uroot -p123123
 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 5
 Server version: 5.7.37-log MySQL Community Server (GPL)
 Copyright (c) 2000, 2022, 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> change master to
 master_host='192.168.200.101',master_user='repl',master_password='repl_user',mas
 ter_log_file='mysql-bin.000002',master_log_pos=1288;
 Query OK, 0 rows affected, 2 warnings (0.14 sec)
 mysql> commit;
 Query OK, 0 rows affected (0.00 sec)
 mysql> start slave;
 Query OK, 0 rows affected (0.02 sec)
 mysql> show slave status\G;
 *************************** 1. row ***************************
 Slave_IO_State: Waiting for master to send event
 Master_Host: 192.168.200.101
 Master_User: repl
 Master_Port: 3306
 Connect_Retry: 60
 Master_Log_File: mysql-bin.000002
 Read_Master_Log_Pos: 1288
 Relay_Log_File: relay-log.000002
 Relay_Log_Pos: 320
 Relay_Master_Log_File: mysql-bin.000002
 Slave_IO_Running: Yes #关注此项是否为yes
 Slave_SQL_Running: Yes #关注此项是否为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: 1288
 Relay_Log_Space: 521
 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: 5b7281ad-780c-11ec-8e49-000c297c8a97
 3.7.3：mha-slave2完成从库复制
 Master_Info_File: mysql.slave_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: 4f4c120f-780d-11ec-9831-000c292a62e0:1-5
 Auto_Position: 0
 Replicate_Rewrite_DB:
 Channel_Name:
 Master_TLS_Version:
 1 row in set (0.02 sec)
 ERROR:
 No query specified

2.5.1、mha-slave2完成从库复制

 [root@mha-slave2 ~]# mysql -uroot -p123123
 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: 5.7.37-log MySQL Community Server (GPL)
 Copyright (c) 2000, 2022, 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> change master to
 master_host='192.168.200.101',master_user='repl',master_password='repl_user',mas
 ter_log_file='mysql-bin.000002',master_log_pos=1288;
 Query OK, 0 rows affected, 2 warnings (0.06 sec)
 mysql> commit;
 Query OK, 0 rows affected (0.00 sec)
 mysql> start slave;
 Query OK, 0 rows affected (0.00 sec)
 mysql> show slave status\G;
 *************************** 1. row ***************************
 Slave_IO_State: Waiting for master to send event
 Master_Host: 192.168.200.101
 Master_User: repl
 Master_Port: 3306
 Connect_Retry: 60
 Master_Log_File: mysql-bin.000002
 Read_Master_Log_Pos: 1288
 Relay_Log_File: relay-log.000002
 Relay_Log_Pos: 320
 Relay_Master_Log_File: mysql-bin.000002
 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: 1288
 Relay_Log_Space: 521
 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: 5b7281ad-780c-11ec-8e49-000c297c8a97
 Master_Info_File: mysql.slave_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: f04e0d29-780d-11ec-9506-000c2987e0bd:1-7
 Auto_Position: 0
 Replicate_Rewrite_DB:
 Channel_Name:
 Master_TLS_Version:
 1 row in set (0.00 sec)
 ERROR:
 No query specified
 mysql> quit

三、MHA部署

3.1、101-103开始安装mha-node

 [root@mha-master ~]# mkdir -pv /data/software
 mkdir: 已创建目录 "/data/software"
 [root@mha-master ~]# cd /data/software/
 [root@mha-master software]# yum install -y mha4mysql-node-0.58-
 0.el7.centos.noarch.rpm
 //同步软件包
 [root@mha-master software]# ssh mha-slave1 mkdir -pv /data/software
 mkdir: 已创建目录 "/data/software"
 [root@mha-master software]# scp mha4mysql-node-0.58-0.el7.centos.noarch.rpm mhaslave1:/data/software/
 mha4mysql-node-0.58-0.el7.centos.noarch.rpm
                                                 100% 35KB 1.5MB/s 00:00
 [root@mha-master software]# ssh mha-slave2 mkdir -pv /data/software
 mkdir: 已创建目录 "/data/software"
 [root@mha-master software]# scp mha4mysql-node-0.58-0.el7.centos.noarch.rpm mhaslave2:/data/software/
 mha4mysql-node-0.58-0.el7.centos.noarch.rpm
                                                 100% 35KB 10.9MB/s 00:00
 ​
 mha-slave1安装
 [root@mha-slave1 ~]# cd /data/software/ && yum install -y mha4mysql-node-0.58-
 0.el7.centos.noarch.rpm
 mha-slave2安装
 [root@mha-slave2 ~]# cd /data/software/ && yum install -y mha4mysql-node-0.58-
 0.el7.centos.noarch.rpm

3.2、配置mha-manager（104）

 [root@mha-manager ~]# mkdir -pv /data/mha/{conf,logs,tools,tmp}
 //配置ip切换脚本
 [root@mha-manager ~]# vim /data/mha/tools/master_ip_failover_script
 #!/usr/bin/env perl
 use strict;
 use warnings FATAL => 'all';
 use Getopt::Long;
 my (
 $command, $ssh_user, $orig_master_host, $orig_master_ip,
 $orig_master_port, $new_master_host, $new_master_ip, $new_master_port
 );
 #定义VIP
 my $vip = '192.168.200.200/24';
 my $key = '1';
 my $ssh_start_vip = "/sbin/ifconfig ens33:$key $vip";
 my $ssh_stop_vip = "/sbin/ifconfig ens33:$key down";
 GetOptions(
 'command=s' => \$command,
 'ssh_user=s' => \$ssh_user,
 'orig_master_host=s' => \$orig_master_host,
 'orig_master_ip=s' => \$orig_master_ip,
 'orig_master_port=i' => \$orig_master_port,
 'new_master_host=s' => \$new_master_host,
 'new_master_ip=s' => \$new_master_ip,
 'new_master_port=i' => \$new_master_port,
 );
 exit &main();
 sub main {
 print "\n\nIN SCRIPT TEST====$ssh_stop_vip==$ssh_start_vip===\n\n";
 if ( $command eq "stop" || $command eq "stopssh" ) {
 my $exit_code = 1;
 eval {
 print "Disabling the VIP on old master: $orig_master_host \n";
 &stop_vip();
 $exit_code = 0;
 };
 if ($@) {
 warn "Got Error: $@\n";
 exit $exit_code;
 }
 exit $exit_code;
 }
 elsif ( $command eq "start" ) {
 my $exit_code = 10;
 eval {
 print "Enabling the VIP - $vip on the new master - $new_master_host
 \n";
 &start_vip();
 $exit_code = 0;
 };
 if ($@) {
 warn $@;
 exit $exit_code;
 }
 exit $exit_code;
 }
 elsif ( $command eq "status" ) {
 print "Checking the Status of the script.. OK \n";
 exit 0;
 }
 else {
 &usage();
 exit 1;
 }
 }
 sub start_vip() {
 `ssh $ssh_user\@$new_master_host \" $ssh_start_vip \"`;
 }
 sub stop_vip() {
 `ssh $ssh_user\@$orig_master_host \" $ssh_stop_vip \"`;
 }
 sub usage {
 print
 "Usage: master_ip_failover --command=start|stop|stopssh|status --
 orig_master_host=host --orig_master_ip=ip --orig_master_port=port --
 new_master_host=host --new_master_ip=ip --new_master_port=port\n";
 }
 ​
 [root@mha-manager ~]# chmod 777 /data/mha/tools/master_ip_failover_script
 ​
 ​
 ###配置mha配置文件
 [root@mha-manager ~]# vim /data/mha/conf/mha192.168.200.101-104.conf
 [server default]
 manager_workdir=/data/mha/tmp
 manager_log=/data/mha/logs/manager.log
 master_binlog_dir=/data/mysql/data
 remote_workdir=/data/mha/tmp
 ssh_user=root
 user=mha
 password=mha_user
 repl_user=repl
 repl_password=repl_user
 ping_interval=1
 master_ip_failover_script=/data/mha/tools/master_ip_failover_script
 [server1]
 hostname=192.168.200.101
 ssh_port=22
 candidate_master=1
 [server2]
 hostname=192.168.200.102
 ssh_port=22
 candidate_master=1
 [server3]
 hostname=192.168.200.103
 ssh_port=22
 candidate_master=0

3.3、测试mha-manager的ssh互信

 [root@mha-manager ~]# masterha_check_ssh --
 conf=/data/mha/conf/mha192.168.200.101-104.conf
 Tue Jan 18 15:33:02 2022 - [warning] Global configuration file
 /etc/masterha_default.cnf not found. Skipping.
 Tue Jan 18 15:33:02 2022 - [info] Reading application default configuration from
 /data/mha/conf/mha192.168.200.101-104.conf..
 Tue Jan 18 15:33:02 2022 - [info] Reading server configuration from
 /data/mha/conf/mha192.168.200.101-104.conf..
 Tue Jan 18 15:33:02 2022 - [info] Starting SSH connection tests..
 Tue Jan 18 15:33:03 2022 - [debug]
 Tue Jan 18 15:33:02 2022 - [debug] Connecting via SSH from
 root@192.168.200.101(192.168.200.101:22) to
 root@192.168.200.102(192.168.200.102:22)..
 Tue Jan 18 15:33:03 2022 - [debug] ok. (确认状态)
 Tue Jan 18 15:33:03 2022 - [debug] Connecting via SSH from
 root@192.168.200.101(192.168.200.101:22) to
 root@192.168.200.103(192.168.200.103:22)..
 Tue Jan 18 15:33:03 2022 - [debug] ok. (确认状态)
 Tue Jan 18 15:33:04 2022 - [debug]
 Tue Jan 18 15:33:03 2022 - [debug] Connecting via SSH from
 root@192.168.200.102(192.168.200.102:22) to
 root@192.168.200.101(192.168.200.101:22)..
 Tue Jan 18 15:33:03 2022 - [debug] ok. (确认状态)
 Tue Jan 18 15:33:03 2022 - [debug] Connecting via SSH from
 root@192.168.200.102(192.168.200.102:22) to
 root@192.168.200.103(192.168.200.103:22)..
 Tue Jan 18 15:33:04 2022 - [debug] ok. (确认状态)
 Tue Jan 18 15:33:05 2022 - [debug]
 Tue Jan 18 15:33:03 2022 - [debug] Connecting via SSH from
 root@192.168.200.103(192.168.200.103:22) to
 root@192.168.200.101(192.168.200.101:22)..
 Tue Jan 18 15:33:04 2022 - [debug] ok. (确认状态)
 Tue Jan 18 15:33:04 2022 - [debug] Connecting via SSH from
 root@192.168.200.103(192.168.200.103:22) to
 root@192.168.200.102(192.168.200.102:22)..
 Tue Jan 18 15:33:04 2022 - [debug] ok. (确认状态)
 Tue Jan 18 15:33:05 2022 - [info] All SSH connection tests passed successfully.

3.4、mha-manager确认replication状态

 [root@mha-manager ~]# masterha_check_repl --
 conf=/data/mha/conf/mha192.168.200.101-104.conf
 Tue Jan 18 15:35:06 2022 - [warning] Global configuration file
 /etc/masterha_default.cnf not found. Skipping.
 Tue Jan 18 15:35:06 2022 - [info] Reading application default configuration from
 /data/mha/conf/mha192.168.200.101-104.conf..
 Tue Jan 18 15:35:06 2022 - [info] Reading server configuration from
 /data/mha/conf/mha192.168.200.101-104.conf..
 Tue Jan 18 15:35:06 2022 - [info] MHA::MasterMonitor version 0.58.
 Tue Jan 18 15:35:08 2022 - [info] GTID failover mode = 1
 Tue Jan 18 15:35:08 2022 - [info] Dead Servers:
 Tue Jan 18 15:35:08 2022 - [info] Alive Servers:
 Tue Jan 18 15:35:08 2022 - [info] 192.168.200.101(192.168.200.101:3306)
 Tue Jan 18 15:35:08 2022 - [info] 192.168.200.102(192.168.200.102:3306)
 Tue Jan 18 15:35:08 2022 - [info] 192.168.200.103(192.168.200.103:3306)
 Tue Jan 18 15:35:08 2022 - [info] Alive Slaves:
 Tue Jan 18 15:35:08 2022 - [info] 192.168.200.102(192.168.200.102:3306)
 Version=5.7.37-log (oldest major version between slaves) log-bin:enabled
 Tue Jan 18 15:35:08 2022 - [info] GTID ON
 Tue Jan 18 15:35:08 2022 - [info] Replicating from
 192.168.200.101(192.168.200.101:3306)
 Tue Jan 18 15:35:08 2022 - [info] Primary candidate for the new Master
 (candidate_master is set)
 Tue Jan 18 15:35:08 2022 - [info] 192.168.200.103(192.168.200.103:3306)
 Version=5.7.37-log (oldest major version between slaves) log-bin:enabled
 Tue Jan 18 15:35:08 2022 - [info] GTID ON
 Tue Jan 18 15:35:08 2022 - [info] Replicating from
 192.168.200.101(192.168.200.101:3306)
 Tue Jan 18 15:35:08 2022 - [info] Current Alive Master:
 192.168.200.101(192.168.200.101:3306)
 Tue Jan 18 15:35:08 2022 - [info] Checking slave configurations..
 Tue Jan 18 15:35:08 2022 - [info] Checking replication filtering settings..
 Tue Jan 18 15:35:08 2022 - [info] binlog_do_db= , binlog_ignore_db= mysql
 Tue Jan 18 15:35:08 2022 - [info] Replication filtering check ok.
 Tue Jan 18 15:35:08 2022 - [info] GTID (with auto-pos) is supported. Skipping
 all SSH and Node package checking.
 Tue Jan 18 15:35:08 2022 - [info] Checking SSH publickey authentication settings
 on the current master..
 Tue Jan 18 15:35:08 2022 - [info] HealthCheck: SSH to 192.168.200.101 is
 reachable.
 Tue Jan 18 15:35:08 2022 - [info]
 192.168.200.101(192.168.200.101:3306) (current master)
 +--192.168.200.102(192.168.200.102:3306)
 +--192.168.200.103(192.168.200.103:3306)
 Tue Jan 18 15:35:08 2022 - [info] Checking replication health on
 192.168.200.102..
 Tue Jan 18 15:35:08 2022 - [info] ok. (确认状态)
 Tue Jan 18 15:35:08 2022 - [info] Checking replication health on
 192.168.200.103..
 Tue Jan 18 15:35:08 2022 - [info] ok. (确认状态)
 Tue Jan 18 15:35:08 2022 - [info] Checking master_ip_failover_script status:
 Tue Jan 18 15:35:08 2022 - [info] /data/mha/tools/master_ip_failover_script --
 command=status --ssh_user=root --orig_master_host=192.168.200.101 --
 orig_master_ip=192.168.200.101 --orig_master_port=3306
 IN SCRIPT TEST====/sbin/ifconfig eth0:1 down==/sbin/ifconfig eth0:1
 192.168.200.200/24===
 Checking the Status of the script.. OK
 Tue Jan 18 15:35:08 2022 - [info] OK.
 Tue Jan 18 15:35:08 2022 - [warning] shutdown_script is not defined.
 Tue Jan 18 15:35:08 2022 - [info] Got exit code 0 (Not master dead).
 MySQL Replication Health is OK. (确认状态)

3.5、启动mha状态

[root@mha-manager ~]# nohup masterha_manager --
 conf=/data/mha/conf/mha192.168.200.101-104.conf &>
 /data/mha/logs/mha_manager.log &
 [1] 21516
 //检查mha状态
 [root@mha-manager ~]# masterha_check_status --
 conf=/data/mha/conf/mha192.168.200.101-104.conf
 mha192.168.200.101-104 (pid:21516) is running(0:PING_OK), master:192.168.200.101

3.6、配置mha-master-ip信息

 [root@mha-manager ~]# ssh mha-master ifconfig ens33:1 192.168.200.200/24
 [root@mha-manager ~]# ssh mha-master ifconfig
 ens33: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
 inet 192.168.200.101 netmask 255.255.255.0 broadcast 192.168.200.255
 inet6 fe80::20c:29ff:fe7c:8a97 prefixlen 64 scopeid 0x20<link>
 ether 00:0c:29:7c:8a:97 txqueuelen 1000 (Ethernet)
 RX packets 201438 bytes 273581532 (260.9 MiB)
 RX errors 0 dropped 0 overruns 0 frame 0
 TX packets 25843 bytes 2633466 (2.5 MiB)
 TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
 ​
 ens33:1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
 inet 192.168.200.200 netmask 255.255.255.0 broadcast 192.168.200.255
 ether 00:0c:29:7c:8a:97 txqueuelen 1000 (Ethernet)
 lo: flags=73<UP,LOOPBACK,RUNNING> mtu 65536
 inet 127.0.0.1 netmask 255.0.0.0
 inet6 ::1 prefixlen 128 scopeid 0x10<host>
 loop txqueuelen 1000 (Local Loopback)
 RX packets 156 bytes 22326 (21.8 KiB)
 RX errors 0 dropped 0 overruns 0 frame 0
 TX packets 156 bytes 22326 (21.8 KiB)
 TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0