MySQL——MySQL高可用之MHA

MHA简介

MHA（Master High Availability）目前在MySQL高可用方面是一个相对成熟的解决方案，它由日本DeNA公司youshimaton（现就职于Facebook公司）开发，是一套优秀的作为MySQL高可用性环境下故障切换和主从提升的高可用软件。在MySQL故障切换过程中，MHA能做到在0~30秒之内自动完成数据库的故障切换操作，并且在进行故障切换的过程中，MHA能在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里有两个角色一个是MHA Node（数据节点）另一个是MHA Manager（管理节点）。 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 Node运行在每台MySQL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。整个故障转移过程对应用程序完全透明。

在MHA自动故障切换过程中，MHA试图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的不丢失，但这并不总是可行的。例如，如果主服务器硬件故障或无法通过ssh访问，MHA没法保存二进制日志，只进行故障转移而丢失了最新的数据。使用MySQL 5.5的半同步复制，可以大大降低数据丢失的风险。MHA可以与半同步复制结合起来。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。
注：从MySQL5.5开始，MySQL以插件的形式支持半同步复制。

如何理解半同步呢？首先我们来看看异步，全同步的概念：

• 异步复制（Asynchronous replication） MySQL默认的复制即是异步的，主库在执行完客户端提交的事务后会立即将结果返给客户端，并不关心从库是否已经接收并处理，这样就会有一个问题，主如果crash掉了，此时主上已经提交的事务可能并没有传到从上，如果此时，强行将从提升为主，可能导致新主上的数据不完整。
• 全同步复制（Fully synchronous replication） 指当主库执行完一个事务，所有的从库都执行了该事务才返回给客户端。因为需要等待所有从库执行完该事务才能返回，所以全同步复制的性能必然会收到严重的影响。
• 半同步复制（Semisynchronous replication） 介于异步复制和全同步复制之间，主库在执行完客户端提交的事务后不是立刻返回给客户端，而是等待至少一个从库接收到并写到relay log中才返回给客户端。相对于异步复制，半同步复制提高了数据的安全性，同时它也造成了一定程度的延迟，这个延迟最少是一个TCP/IP往返的时间。所以，半同步复制最好在低延时的网络中使用。

半同步复制的原理图：

总结：异步与半同步异同
默认情况下MySQL的复制是异步的，Master上所有的更新操作写入Binlog之后并不确保所有的更新都被复制到Slave之上。异步操作虽然效率高，但是在Master/Slave出现问题的时候，存在很高数据不同步的风险，甚至可能丢失数据。 MySQL5.5引入半同步复制功能的目的是为了保证在master出问题的时候，至少有一台Slave的数据是完整的。在超时的情况下也可以临时转入异步复制，保障业务的正常使用，直到一台salve追赶上之后，继续切换到半同步模式。

工作原理：
相较于其它HA软件，MHA的目的在于维持MySQL Replication中Master库的高可用性，其最大特点是可以修复多个Slave之间的差异日志，最终使所有Slave保持数据一致，然后从中选择一个充当新的Master，并将其它Slave指向它。 -从宕机崩溃的master保存二进制日志事件(binlogevents)。 -识别含有最新更新的slave。-应用差异的中继日志(relay log)到其它slave。 -应用从master保存的二进制日志事件(binlogevents)。 -提升一个slave为新master。 -使其它的slave连接新的master进行复制。

目前MHA主要支持一主多从的架构，要搭建MHA，要求一个复制集群中必须最少有三台数据库服务器，一主二从，即一台充当master，一台充当备用master，另外一台充当从库，因为至少需要三台服务器。

环境：

角色	IP地址	主机名	Server id	类型	os
master1	192.168.1.5	master1	1	主mysql（写入）	centos7x86_64
master2	192.168.1.6	master2	2	从mysql（读）	centos7x86_64
slave	192.168.1.7	slave	3	从mysql（读）	centos7x86_64
Manager	192.168.1.8	Manager		管理节点	centos7x86_64

其中master1 对外提供写服务，备选master2 提供读服务，slave 也提供相关的读服务，一旦master1 宕机，将会把备选master2 提升为新的master，slave指向新的master2，manager作为管理服务器。

一、基础环境准备

1、 在配置好IP地址后检查selinux，iptables设置，关闭 selinux ，iptables 服务以便后期主从同步不出错 注：时间要同步
时间同步

[root@master1 ~]# yum -y install chrony     #安装chrony
[root@master1 ~]# vim /etc/chrony.conf
……
server ntp.aliyun.com iburst
……
allow 192.168.1.0/24
……

[root@master1 ~]# systemctl restart chronyd.service 
[root@master1 ~]# systemctl enable chronyd.service
[root@master1 ~]# date      #验证时间是否同步

2、 在四台机器都配置epel源

# wget -O /etc/yum.repos.d/epel.repo http://mirrors.aliyun.com/repo/epel-7.repo

3、建立ssh无交互登录环境（四台）

# ssh-keygen -t rsa
# for i in 5 6 7 8; do ssh-copy-id -i ~/.ssh/id_rsa.pub root@192.168.1.$i;done

测试ssh无交互登录（四台主机上执行同样的测试操作）

# for i in 5 6 7 8; do ssh root@192.168.1.$i hostname;done
master1
master2
slave
manager

4、配置hosts环境

[root@master1 ~]# vim /etc/hosts
……
192.168.1.5     master1
192.168.1.6     master2
192.168.1.7     slave
192.168.1.8     manager

[root@master1 ~]# for host in master2 slave manager; do scp /etc/hosts $host:/etc/;done

二、配置mysql半同步复制

为了尽可能的减少主库硬件损坏宕机造成的数据丢失，因此在配置MHA的同时建议配置成MySQL的半同步复制。
注：mysql半同步插件是由谷歌提供，具体位置/usr/local/mysql/lib/plugin/下，一个是master用的semisync_master.so，一个是slave用的semisync_slave.so。

如果不清楚Plugin的目录，用如下查找：

mysql> show variables like '%plugin_dir%';
+---------------+------------------------------+
| Variable_name | Value                        |
+---------------+------------------------------+
| plugin_dir    | /usr/local/mysql/lib/plugin/ |
+---------------+------------------------------+

1、分别在主从节点上安装相关的插件（master1, master2,slave）

在MySQL上安装插件需要数据库支持动态载入。检查是否支持，用如下检测：

mysql> show variables like '%have_dynamic%';
+----------------------+-------+
| Variable_name        | Value |
+----------------------+-------+
| have_dynamic_loading | YES   |
+----------------------+-------+

所有mysql数据库服务器，安装半同步插件(semisync_master.so,semisync_slave.so)

mysql> INSTALL PLUGIN rpl_semi_sync_master SONAME 'semisync_master.so';
mysql> INSTALL PLUGIN rpl_semi_sync_slave SONAME 'semisync_slave.so';

检查Plugin是否已正确安装：

mysql> show plugins;
或↑↓
mysql> select * from information_schema.plugins;

查看半同步相关信息

mysql> show variables like '%rpl_semi_sync%';
+-------------------------------------------+------------+
| Variable_name                             | Value      |
+-------------------------------------------+------------+
| rpl_semi_sync_master_enabled              | OFF        |		# 可以看到半同复制插件已经安装，只是还没有启用，所以是off 
| 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 |
| rpl_semi_sync_slave_enabled               | OFF        |
| rpl_semi_sync_slave_trace_level           | 32         |
+-------------------------------------------+------------+

2、修改my.cnf文件，配置主从同步：

注：若主MYSQL服务器已经存在，只是后期才搭建从MYSQL服务器，在置配数据同步前应先将主MYSQL服务器的要同步的数据库拷贝到从MYSQL服务器上（如先在主MYSQL上备份数据库，再用备份在从MYSQL服务器上恢复）

master1主机：

[root@master1 ~]# vim /etc/my.cnf 
……
server_id=1
log_bin=mysql-bin
binlog_format=mixed
log_bin_index=mysql-bin.index
rpl_semi_sync_master_enabled=1
rpl_semi_sync_master_timeout=1000
rpl_semi_sync_slave_enabled=1
relay_log_purge=0
relay_log=relay-bin
relay_log_index=slave-relay-bin.index

[root@master1 ~]# systemctl restart mysqld

注：
rpl_semi_sync_master_enabled=1 1表是启用，0表示关闭
rpl_semi_sync_master_timeout=10000：毫秒单位 ，该参数主服务器等待确认消息10秒后，不再等待，变为异步方式。

master2主机：

[root@master2 ~]# vim /etc/my.cnf 
……
server_id=2
log_bin=mysql-bin
binlog_format=mixed
log_bin_index=mysql-bin.index
rpl_semi_sync_master_enabled=1
rpl_semi_sync_master_timeout=10000
rpl_semi_sync_slave_enabled=1
relay_log_purge=0
relay_log=relay-bin
relay_log_index=slave-relay-bin.index

[root@master2 ~]# systemctl restart mysqld

注：
relay_log_purge=0，禁止 SQL 线程在执行完一个 relay log 后自动将其删除，对于MHA场景下，对于某些滞后从库的恢复依赖于其他从库的relay log，因此采取禁用自动删除功能

Slave主机：

[root@slave ~]# vim /etc/my.cnf 
……
server_id=3
log_bin=mysql-bin
relay_log=relay-bin
relay_log_index=slave-relay-bin.index
read_only=1
rpl_semi_sync_slave_enabled=1

[root@slave ~]# systemctl restart mysqld

查看半同步相关信息

mysql> show variables like '%rpl_semi_sync%';
+-------------------------------------------+------------+
| Variable_name                             | Value      |
+-------------------------------------------+------------+
| rpl_semi_sync_master_enabled              | ON         |
| rpl_semi_sync_master_timeout              | 1000       |
| 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 |
| rpl_semi_sync_slave_enabled               | ON         |
| rpl_semi_sync_slave_trace_level           | 32         |
+-------------------------------------------+------------+

查看半同步状态：

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     |
| Rpl_semi_sync_slave_status                 | OFF   |
+--------------------------------------------+-------+

有几个状态参数值得关注的：
rpl_semi_sync_master_status ：显示主服务是异步复制模式还是半同步复制模式
rpl_semi_sync_master_clients ：显示有多少个从服务器配置为半同步复制模式
rpl_semi_sync_master_yes_tx ：显示从服务器确认成功提交的数量
rpl_semi_sync_master_no_tx ：显示从服务器确认不成功提交的数量
rpl_semi_sync_master_tx_avg_wait_time ：事务因开启 semi_sync ，平均需要额外等待的时间
rpl_semi_sync_master_net_avg_wait_time ：事务进入等待队列后，到网络平均等待时间

master1主机：

mysql> grant replication slave on *.* to mharep@'192.168.1.%' identified by '123';

mysql> grant all privileges on *.* to manager@'192.168.1.%' identified by '123';

mysql> show master status;
+------------------+----------+--------------+------------------+-------------------+
| File             | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+------------------+----------+--------------+------------------+-------------------+
| mysql-bin.000001 |      742 |              |                  |                   |
+------------------+----------+--------------+------------------+-------------------+

第一条grant命令是创建一个用于主从复制的帐号，在master和candicate master的主机上创建即可。
第二条grant命令是创建MHA管理账号，所有mysql服务器上都需要执行。MHA会在配置文件里要求能远程登录到数据库，所以要进行必要的赋权。

master2主机：

mysql> grant replication slave on *.* to mharep@'192.168.1.%' identified by '123';

mysql> grant all privileges on *.* to manager@'192.168.1.%' identified by '123';

mysql> change master to master_host='192.168.1.5',master_user='mharep',master_password='123',master_log_file='mysql-bin.000001',master_log_pos=742;
mysql> start slave;
mysql> show slave status\G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.1.5
                  Master_User: mharep
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000001
          Read_Master_Log_Pos: 742
               Relay_Log_File: relay-bin.000003
                Relay_Log_Pos: 320
        Relay_Master_Log_File: mysql-bin.000001
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes

Slave主机：

mysql> grant all privileges on *.* to manager@'192.168.1.%' identified by '123';

mysql> change master to master_host='192.168.1.5',master_user='mharep',master_password='123',master_log_file='mysql-bin.000001',master_log_pos=742;
mysql> start slave;
mysql> show slave status\G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.1.5
                  Master_User: mharep
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000001
          Read_Master_Log_Pos: 742
               Relay_Log_File: relay-bin.000003
                Relay_Log_Pos: 320
        Relay_Master_Log_File: mysql-bin.000001
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes

查看master1服务器的半同步状态：

mysql> show status like '%rpl_semi_sync%';
+--------------------------------------------+-------+
| Variable_name                              | Value |
+--------------------------------------------+-------+
| Rpl_semi_sync_master_clients               | 2     |
| 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              | 1     |
| Rpl_semi_sync_master_no_tx                 | 2     |
| 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     |
| Rpl_semi_sync_slave_status                 | OFF   |
+--------------------------------------------+-------+

三、配置mysql-mha

mha包括manager节点和data节点，data节点包括原有的MySQL复制结构中的主机，至少3台，即1主2从，当master failover后，还能保证主从结构；只需安装node包。manager server：运行监控脚本，负责monitoring 和 auto-failover；需要安装node包和manager包。

1、在所有主机上安装mha所依赖的软件包（需要系统自带的yum源）

# yum -y install perl-DBD-MySQL perl-Config-Tiny perl-Log-Dispatch perl-Parallel-ForkManager perl-Config-IniFiles ncftp perl-Params-Validate perl-CPAN perl-Test-Mock-LWP.noarch perl-LWP-Authen-Negotiate.noarch perl-devel perl-ExtUtils-CBuilder perl-ExtUtils-MakeMaker

2、 以下操作管理节点需要两个都安装，在3台数据库节点只要安装MHA的node节点：

软件下载 ：https://github.com/yoshinorim

在所有数据库节点上安装mha4mysql-node-0.56.tar.gz

# cd /usr/local/src/
# rz mha4mysql-manager-0.58.tar.gz  mha4mysql-node-0.58.tar.gz
# tar zxf mha4mysql-node-0.58.tar.gz 
# cd mha4mysql-node-0.58/
# perl Makefile.PL && make && make install

在管理节点需要两个都安装：mha4mysql-node-0.56.tar.gz和mha4mysql-manager-0.56.tar.gz

安装mha4mysql-node-0.56.tar.gz
[root@manager ~]# cd /usr/local/src/
[root@manager src]# rz mha4mysql-manager-0.58.tar.gz  mha4mysql-node-0.58.tar.gz
[root@manager src]# tar zxf mha4mysql-node-0.58.tar.gz 
[root@manager src]# cd mha4mysql-node-0.58/
[root@manager mha4mysql-node-0.58]# perl Makefile.PL && make && make install


安装mha4mysql-manager-0.56.tar.gz
[root@manager mha4mysql-node-0.58]# cd ..
[root@manager src]# tar zxf mha4mysql-manager-0.58.tar.gz 
[root@manager src]# cd mha4mysql-manager-0.58/
[root@manager mha4mysql-manager-0.58]# perl Makefile.PL && make && make install

error！perl Makefile.PL报错！！

执行以下命令安装依赖
yum install perl-ExtUtils-CBuilder perl-ExtUtils-MakeMaker

三种解决方法↑↓

解决办法:
yum install cpan -y
yum install perl-ExtUtils-CBuilder perl-ExtUtils-MakeMaker -y
cpan ExtUtils::Install

如果不想用cpan安装,那就使用下面这条命令
yum install perl-ExtUtils-Embed -y

或↑↓

执行Perl脚本报错Can’t locate ExtUtils/MakeMaker.pm in @INC 解决如下：
 # wget http://files.directadmin.com/services/9.0/ExtUtils-MakeMaker-6.31.tar.gz
 # tar xvzf ExtUtils-MakeMaker-6.31.tar.gz -C /opt/
 # cd /opt/ExtUtils-MakeMaker-6.31/ 
 # perl Makefile.PL 
 # make 
 # make install 

如果执行报如下错误
# perl Makefile.PL 
Checking if your kit is complete...
Looks good
Could not open '': No such file or directory at lib/ExtUtils/MM_Unix.pm line 2697.

则需安装依赖环境
# yum install perl-ExtUtils-MakeMaker

管理节点

[root@manager mha4mysql-manager-0.58]# mkdir /etc/masterha
[root@manager mha4mysql-manager-0.58]# mkdir -p /masterha/app1
[root@manager mha4mysql-manager-0.58]# mkdir /scripts
[root@manager mha4mysql-manager-0.58]# cp samples/conf/* /etc/masterha/
[root@manager mha4mysql-manager-0.58]# cp samples/scripts/* /scripts/

3、配置mha

与绝大多数Linux应用程序类似，MHA的正确使用依赖于合理的配置文件。MHA的配置文件与mysql的my.cnf文件配置相似，采取的是param=value的方式来配置，配置文件位于管理节点，通常包括每一个mysql server的主机名，mysql用户名，密码，工作目录等等。 编辑/etc/masterha/app1.conf，内容如下：

[root@manager ~]# vim /etc/masterha/app1.cnf
[server default]
manager_workdir=/masterha/app1
manager_log=/masterha/app1/manager.log
user=manager
password=123
ssh_user=root
repl_user=mharep
repl_password=123
ping_interval=1

[server1]
hostname=192.168.1.5
port=3306
master_binlog_dir=/usr/local/mysql/data
candidate_master=1

[server2]
hostname=192.168.1.6
port=3306
master_binlog_dir=/usr/local/mysql/data
candidate_master=1

[server3]
hostname=192.168.1.7
port=3306
master_binlog_dir=/usr/local/mysql/data
no_master=1

保存退出

[root@manager ~]# >/etc/masterha/masterha_default.cnf

配关配置项的解释：
manager_workdir=/masterha/app1 #设置manager的工作目录
manager_log=/masterha/app1/manager.log #设置manager的日志 user=manager //设置监控用户manager
password=123 #监控用户manager的密码 ssh_user=root //ssh连接用户 repl_user=mharep //主从复制用户
repl_password=123 #主从复制用户密码
ping_interval=1 #设置监控主库，发送ping包的时间间隔，默认是3秒，尝试三次没有回应的时候自动进行
railover master_binlog_dir=/usr/local/mysql/data #设置master 保存 binlog的位置，以便MHA可以找到master的日志，我这里的也就是mysql的数据目录
candidate_master=1 #设置为候选master，如果设置该参数以后，发生主从切换以后将会将此从库提升为主库。

SSH 有效性验证：

[root@manager ~]# masterha_check_ssh --global_conf=/etc/masterha/masterha_default.cnf --conf=/etc/masterha/app1.cnf
Sat Mar 13 06:03:56 2021 - [info] Reading default configuration from /etc/masterha/masterha_default.cnf..
Sat Mar 13 06:03:56 2021 - [info] Reading application default configuration from /etc/masterha/app1.cnf..
Sat Mar 13 06:03:56 2021 - [info] Reading server configuration from /etc/masterha/app1.cnf..
Sat Mar 13 06:03:56 2021 - [info] Starting SSH connection tests..
Sat Mar 13 06:03:57 2021 - [debug] 
Sat Mar 13 06:03:56 2021 - [debug]  Connecting via SSH from root@192.168.1.5(192.168.1.5:22) to root@192.168.1.6(192.168.1.6:22)..
Sat Mar 13 06:03:57 2021 - [debug]   ok.
Sat Mar 13 06:03:57 2021 - [debug]  Connecting via SSH from root@192.168.1.5(192.168.1.5:22) to root@192.168.1.7(192.168.1.7:22)..
Sat Mar 13 06:03:57 2021 - [debug]   ok.
Sat Mar 13 06:03:57 2021 - [debug] 
Sat Mar 13 06:03:57 2021 - [debug]  Connecting via SSH from root@192.168.1.6(192.168.1.6:22) to root@192.168.1.5(192.168.1.5:22)..
Sat Mar 13 06:03:57 2021 - [debug]   ok.
Sat Mar 13 06:03:57 2021 - [debug]  Connecting via SSH from root@192.168.1.6(192.168.1.6:22) to root@192.168.1.7(192.168.1.7:22)..
Sat Mar 13 06:03:57 2021 - [debug]   ok.
Sat Mar 13 06:03:58 2021 - [debug] 
Sat Mar 13 06:03:57 2021 - [debug]  Connecting via SSH from root@192.168.1.7(192.168.1.7:22) to root@192.168.1.5(192.168.1.5:22)..
Sat Mar 13 06:03:58 2021 - [debug]   ok.
Sat Mar 13 06:03:58 2021 - [debug]  Connecting via SSH from root@192.168.1.7(192.168.1.7:22) to root@192.168.1.6(192.168.1.6:22)..
Sat Mar 13 06:03:58 2021 - [debug]   ok.
Sat Mar 13 06:03:58 2021 - [info] All SSH connection tests passed successfully.

集群复制的有效性验证： mysql必须都启动

[root@manager ~]# masterha_check_repl --global_conf=/etc/masterha/masterha_default.cnf --conf=/etc/masterha/app1.cnf
Sat Mar 13 06:04:41 2021 - [info] Reading default configuration from /etc/masterha/masterha_default.cnf..
Sat Mar 13 06:04:41 2021 - [info] Reading application default configuration from /etc/masterha/app1.cnf..
Sat Mar 13 06:04:41 2021 - [info] Reading server configuration from /etc/masterha/app1.cnf..
Sat Mar 13 06:04:41 2021 - [info] MHA::MasterMonitor version 0.58.
Sat Mar 13 06:04:42 2021 - [info] GTID failover mode = 0
Sat Mar 13 06:04:42 2021 - [info] Dead Servers:
Sat Mar 13 06:04:42 2021 - [info] Alive Servers:
Sat Mar 13 06:04:42 2021 - [info]   192.168.1.5(192.168.1.5:3306)
Sat Mar 13 06:04:42 2021 - [info]   192.168.1.6(192.168.1.6:3306)
Sat Mar 13 06:04:42 2021 - [info]   192.168.1.7(192.168.1.7:3306)
Sat Mar 13 06:04:42 2021 - [info] Alive Slaves:
Sat Mar 13 06:04:42 2021 - [info]   192.168.1.6(192.168.1.6:3306)  Version=5.7.32-log (oldest major version between slaves) log-bin:enabled
Sat Mar 13 06:04:42 2021 - [info]     Replicating from 192.168.1.5(192.168.1.5:3306)
Sat Mar 13 06:04:42 2021 - [info]     Primary candidate for the new Master (candidate_master is set)
Sat Mar 13 06:04:42 2021 - [info]   192.168.1.7(192.168.1.7:3306)  Version=5.7.32-log (oldest major version between slaves) log-bin:enabled
Sat Mar 13 06:04:42 2021 - [info]     Replicating from 192.168.1.5(192.168.1.5:3306)
Sat Mar 13 06:04:42 2021 - [info]     Not candidate for the new Master (no_master is set)
Sat Mar 13 06:04:42 2021 - [info] Current Alive Master: 192.168.1.5(192.168.1.5:3306)
Sat Mar 13 06:04:42 2021 - [info] Checking slave configurations..
Sat Mar 13 06:04:42 2021 - [info]  read_only=1 is not set on slave 192.168.1.6(192.168.1.6:3306).
Sat Mar 13 06:04:42 2021 - [warning]  relay_log_purge=0 is not set on slave 192.168.1.7(192.168.1.7:3306).
Sat Mar 13 06:04:42 2021 - [info] Checking replication filtering settings..
Sat Mar 13 06:04:42 2021 - [info]  binlog_do_db= , binlog_ignore_db= 
Sat Mar 13 06:04:42 2021 - [info]  Replication filtering check ok.
Sat Mar 13 06:04:42 2021 - [info] GTID (with auto-pos) is not supported
Sat Mar 13 06:04:42 2021 - [info] Starting SSH connection tests..
Sat Mar 13 06:04:43 2021 - [info] All SSH connection tests passed successfully.
Sat Mar 13 06:04:43 2021 - [info] Checking MHA Node version..
Sat Mar 13 06:04:44 2021 - [info]  Version check ok.
Sat Mar 13 06:04:44 2021 - [info] Checking SSH publickey authentication settings on the current master..
Sat Mar 13 06:04:44 2021 - [info] HealthCheck: SSH to 192.168.1.5 is reachable.
Sat Mar 13 06:04:44 2021 - [info] Master MHA Node version is 0.58.
Sat Mar 13 06:04:44 2021 - [info] Checking recovery script configurations on 192.168.1.5(192.168.1.5:3306)..
Sat Mar 13 06:04:44 2021 - [info]   Executing command: save_binary_logs --command=test --start_pos=4 --binlog_dir=/usr/local/mysql/data --output_file=/var/tmp/save_binary_logs_test --manager_version=0.58 --start_file=mysql-bin.000001 
Sat Mar 13 06:04:44 2021 - [info]   Connecting to root@192.168.1.5(192.168.1.5:22).. 
  Creating /var/tmp if not exists..    ok.
  Checking output directory is accessible or not..
   ok.
  Binlog found at /usr/local/mysql/data, up to mysql-bin.000001
Sat Mar 13 06:04:44 2021 - [info] Binlog setting check done.
Sat Mar 13 06:04:44 2021 - [info] Checking SSH publickey authentication and checking recovery script configurations on all alive slave servers..
Sat Mar 13 06:04:44 2021 - [info]   Executing command : apply_diff_relay_logs --command=test --slave_user='manager' --slave_host=192.168.1.6 --slave_ip=192.168.1.6 --slave_port=3306 --workdir=/var/tmp --target_version=5.7.32-log --manager_version=0.58 --relay_log_info=/usr/local/mysql/data/relay-log.info  --relay_dir=/usr/local/mysql/data/  --slave_pass=xxx
Sat Mar 13 06:04:44 2021 - [info]   Connecting to root@192.168.1.6(192.168.1.6:22).. 
  Checking slave recovery environment settings..
    Opening /usr/local/mysql/data/relay-log.info ... ok.
    Relay log found at /usr/local/mysql/data, up to relay-bin.000002
    Temporary relay log file is /usr/local/mysql/data/relay-bin.000002
    Checking if super_read_only is defined and turned on.. not present or turned off, ignoring.
    Testing mysql connection and privileges..
mysql: [Warning] Using a password on the command line interface can be insecure.
 done.
    Testing mysqlbinlog output.. done.
    Cleaning up test file(s).. done.
Sat Mar 13 06:04:44 2021 - [info]   Executing command : apply_diff_relay_logs --command=test --slave_user='manager' --slave_host=192.168.1.7 --slave_ip=192.168.1.7 --slave_port=3306 --workdir=/var/tmp --target_version=5.7.32-log --manager_version=0.58 --relay_log_info=/usr/local/mysql/data/relay-log.info  --relay_dir=/usr/local/mysql/data/  --slave_pass=xxx
Sat Mar 13 06:04:44 2021 - [info]   Connecting to root@192.168.1.7(192.168.1.7:22).. 
  Checking slave recovery environment settings..
    Opening /usr/local/mysql/data/relay-log.info ... ok.
    Relay log found at /usr/local/mysql/data, up to relay-bin.000002
    Temporary relay log file is /usr/local/mysql/data/relay-bin.000002
    Checking if super_read_only is defined and turned on.. not present or turned off, ignoring.
    Testing mysql connection and privileges..
mysql: [Warning] Using a password on the command line interface can be insecure.
 done.
    Testing mysqlbinlog output.. done.
    Cleaning up test file(s).. done.
Sat Mar 13 06:04:45 2021 - [info] Slaves settings check done.
Sat Mar 13 06:04:45 2021 - [info] 
192.168.1.5(192.168.1.5:3306) (current master)
 +--192.168.1.6(192.168.1.6:3306)
 +--192.168.1.7(192.168.1.7:3306)

Sat Mar 13 06:04:45 2021 - [info] Checking replication health on 192.168.1.6..
Sat Mar 13 06:04:45 2021 - [info]  ok.
Sat Mar 13 06:04:45 2021 - [info] Checking replication health on 192.168.1.7..
Sat Mar 13 06:04:45 2021 - [info]  ok.
Sat Mar 13 06:04:45 2021 - [warning] master_ip_failover_script is not defined.
Sat Mar 13 06:04:45 2021 - [warning] shutdown_script is not defined.
Sat Mar 13 06:04:45 2021 - [info] Got exit code 0 (Not master dead).

MySQL Replication Health is OK.

验证成功的话会自动识别出所有服务器和主从状况

error! 在验证时，若遇到这个错误：Can’t exec “mysqlbinlog”

解决方法是在所有服务器上执行：

# ln -s /usr/local/mysql/bin/* /usr/local/bin/
# 重新进行集群复制的有效性验证↓
[root@manager ~]# masterha_check_repl --global_conf=/etc/masterha/masterha_default.cnf --conf=/etc/masterha/app1.cnf

启动 manager：

[root@manager ~]# nohup masterha_manager --conf=/etc/masterha/app1.cnf &>/tmp/mha_manager.log &

注：在应用Unix/Linux时，一般想让某个程序在后台运行，常会用 & 在程序结尾来让程序自动运行。比如要运行mysql在后台： /usr/local/mysql/bin/mysqld_safe –user=mysql &。可是有很多程序并不像mysqld一样，这样就需要nohup命令

状态检查：

[root@manager ~]# masterha_check_status --conf=/etc/masterha/app1.cnf
app1 (pid:3610) is running(0:PING_OK), master:192.168.1.5

故障转移验证：

(自动failover) master dead后，MHA当时已经开启，候选Master库（Slave）会自动failover为Master. 验证的方式是先停掉 master1，因为之前的配置文件中，把 master2 作为了候选人，那么就到 slave 上查看 master 的 IP 是否变为了 master2 的 IP

1、停掉 master1 的mysqld

在master（192.168.1.5） 上把 mysql 停掉

[root@master1 ~]# systemctl stop mysqld

2、查看 MHA 日志

上面的配置文件中指定了日志位置为/masterha/app1/manager.log

[root@manager ~]# tailf /masterha/app1/manager.log
----- Failover Report -----

app1: MySQL Master failover 192.168.1.5(192.168.1.5:3306) to 192.168.1.6(192.168.1.6:3306) succeeded

Master 192.168.1.5(192.168.1.5:3306) is down!

Check MHA Manager logs at manager:/masterha/app1/manager.log for details.

Started automated(non-interactive) failover.
The latest slave 192.168.1.6(192.168.1.6:3306) has all relay logs for recovery.
Selected 192.168.1.6(192.168.1.6:3306) as a new master.
192.168.1.6(192.168.1.6:3306): OK: Applying all logs succeeded.
192.168.1.7(192.168.1.7:3306): This host has the latest relay log events.
Generating relay diff files from the latest slave succeeded.
192.168.1.7(192.168.1.7:3306): OK: Applying all logs succeeded. Slave started, replicating from 192.168.1.6(192.168.1.6:3306)
192.168.1.6(192.168.1.6:3306): Resetting slave info succeeded.
Master failover to 192.168.1.6(192.168.1.6:3306) completed successfully.

从日志信息中可以看到 master failover 已经成功了，并可以看出故障转移的大体流程

3、检查 slave 的复制

登录 slave（192.168.1.7） 的Mysql，查看 slave 状态

mysql> show slave status\G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.1.6
                  Master_User: mharep
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000001
          Read_Master_Log_Pos: 742
               Relay_Log_File: relay-bin.000002
                Relay_Log_Pos: 320
        Relay_Master_Log_File: mysql-bin.000001
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes

可以看到 master 的 IP 现在为 192.168.1.6，已经切换到和192.168.1.6 同步了，本来是和192.168.1.5 同步的，说明 MHA 已经把 master2 提升为了新的 master，IO线程和SQL线程也正确运行，MHA搭建成功

4、主从切换后续工作

重构： 重构就是你的主挂了，切换到 master2上， master2 变成了主，因此重构的一种方案原主库修复成一个新的slave 主库切换后，把原主库修复成新从库，然后重新检查状态。原主库数据文件完整的情况下，可通过以下方式找出最后执行的CHANGE MASTER命令：

[root@master1 ~]# systemctl start mysqld


[root@manager ~]# grep "CHANGE" /masterha/app1/manager.log 
Sat Mar 13 06:09:58 2021 - [info]  All other slaves should start replication from here. Statement should be: CHANGE MASTER TO MASTER_HOST='192.168.1.6', MASTER_PORT=3306, MASTER_LOG_FILE='mysql-bin.000001', MASTER_LOG_POS=742, MASTER_USER='mharep', MASTER_PASSWORD='xxx';
Sat Mar 13 06:09:59 2021 - [info]  Executed CHANGE MASTER.

[root@master1 ~]# mysql -uroot -p123
mysql> CHANGE MASTER TO MASTER_HOST='192.168.1.6', MASTER_PORT=3306, MASTER_LOG_FILE='mysql-bin.000001', MASTER_LOG_POS=742, MASTER_USER='mharep', MASTER_PASSWORD='123';

mysql> start slave;

mysql> show slave status\G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.1.6
                  Master_User: mharep
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000001
          Read_Master_Log_Pos: 742
               Relay_Log_File: relay-bin.000002
                Relay_Log_Pos: 320
        Relay_Master_Log_File: mysql-bin.000001
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes

corntab -e
0 5 * * * /usr/local/bin/purge_relay_logs --user=root --password=pwd123 --port=3306 --disable_relay_log_purge >> /var/log/purge_relay.log 2>&1

MHA Manager 端日常主要操作步骤

1）检查是否有下列文件，有则删除。

发生主从切换后，MHAmanager服务会自动停掉，且在 manager_workdir（/masterha/app1）目录下面生成文件app1.failover.complete，若要启动MHA，必须先确保无此文件) 如果有这个提示，那么删除此文件
/masterha/app1/app1.failover.complete [error]
[/usr/share/perl5/vendor_perl/MHA/MasterFailover.pm, ln298] Last failover was done at 2015/01/09 10:00:47.Current time is too early to do failover again. If you want to do failover, manually remove /masterha/app1/app1.failover.complete and run this script again.

# ll /masterha/app1/app1.failover.complete
# ll /masterha/app1/app1.failover.error

2）停止MHA：

# masterha_stop --conf=/etc/masterha/app1.cnf

3）启动MHA：

# nohup masterha_manager --conf=/etc/masterha/app1.cnf &>/tmp/mha_manager.log &

当有slave 节点宕掉时，默认是启动不了的，加上 --ignore_fail_on_start 即使有节点宕掉也能启动MHA，如下：

# nohup masterha_manager --conf=/etc/masterha/app1.cnf --ignore_fail_on_start &>/tmp/mha_manager.log &

4） 检查状态：

[root@manager ~]# masterha_check_status --conf=/etc/masterha/app1.cnf
app1 (pid:14923) is running(0:PING_OK), master:192.168.1.6

注意：如果正常，会显示"PING_OK"，否则会显示"NOT_RUNNING"，这代表MHA监控没有开启。定期删除中继日志 在配置主从复制中，slave上设置了参数relay_log_purge=0，所以slave节点需要定期删除中继日志，建议每个slave节点删除中继日志的时间错开。

5） 检查日志：

# tail -f /masterha/app1/manager.log

四、配置VIP

vip配置可以采用两种方式：

• 一种通过keepalived的方式管理虚拟ip的浮动；
• 另外一种通过脚本方式启动虚拟ip的方式（即不需要keepalived或者heartbeat类似的软件）。

为了防止脑裂发生，推荐生产环境采用脚本的方式来管理虚拟ip，而不是使用keepalived来完成

1、keepalived方式管理虚拟ip

keepalived 配置方法如下：下载软件并进行安装（两台master，准确的说一台是master，另外一台是备选master，在没有切换以前是slave）

1）在master01和master02上安装软件包keepalived （也可以通过rpm或yum工具进行安装）

安装keepalived软件包与服务控制在编译安装Keepalived之前，必须先安装内核开发包kernel-devel以及openssl-devel、popt-devel等支持库

[root@master1 ~]# yum install kernel-devel popt-devel openssl-devel gcc -y

[root@master2 ~]# yum install kernel-devel popt-devel openssl-devel gcc -y

编译安装Keepalived
使用指定的linux内核位置对keepalived进行配置，并将安装路径指定为根目录，这样就无需额外创建链接文件了，配置完成后，依次执行make、makeinstall进行安装

[root@master01 ~]# cd /usr/src/
[root@master01 src]# wget https://www.keepalived.org/software/keepalived-2.2.0.tar.gz
[root@master01 src]# tar zxf keepalived-2.2.0.tar.gz 
[root@master01 src]# cd keepalived-2.2.0/
[root@master01 keepalived-2.2.0]# ./configure --prefix=/ && make && make install

[root@master02 ~]# cd /usr/src/
[root@master02 src]# wget https://www.keepalived.org/software/keepalived-2.2.0.tar.gz
[root@master02 src]# tar zxf keepalived-2.2.0.tar.gz 
[root@master02 src]# cd keepalived-2.2.0/
[root@master02 keepalived-2.2.0]# ./configure --prefix=/ && make && make install

使用keepalived服务
执行make install操作之后，会自动生成/etc/init.d/keepalived脚本文件，但还需要手动添加为系统服务，这样就可以使用service、chkconfig工具来对keepalived服务程序进行管理了。

[root@master01 keepalived-2.2.0]# systemctl enable keepalived.service

[root@master01 keepalived-2.2.0]# systemctl enable keepalived.service

注：若开启了防火墙，需要关闭防火墙或创建规则

# firewall-cmd --direct --permanent --add-rule ipv4 filter OUTPUT 0 --in-interface
enp0s3 --destination 224.0.0.18 --protocol vrrp -j ACCEPT
# firewall-cmd --direct --permanent --add-rule ipv4 filter INPUT 0 --in-interface
enp0s3 --destination 224.0.0.18 --protocol vrrp -j ACCEPT
# firewall-cmd --reload

修改Keepalived的配置文件（在master1上配置）

[root@master1 ~]# vim /etc/keepalived/keepalived.conf 

! Configuration File for keepalived

global_defs {
   router_id mysql-ha1
}

vrrp_instance VI_1 {
    state BACKUP
    interface ens33
    virtual_router_id 51
    priority 50
    nopreempt
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        192.168.1.100
    }
}

修改Keepalived的配置文件（在master2上配置）

[root@master2 ~]# vim /etc/keepalived/keepalived.conf 

! Configuration File for keepalived

global_defs {
   router_id mysql-ha2
}

vrrp_instance VI_1 {
    state BACKUP
    interface ens33
    virtual_router_id 51
    priority 100
    nopreempt
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        192.168.1.100
    }
}

启动keepalived服务，在master2 上启动并查看日志

[root@master2 ~]# systemctl start keepalived.service
[root@master2 ~]# ps -ef | grep keep
root      36634      1  0 22:12 ?        00:00:00 //sbin/keepalived -D
root      36635  36634  0 22:12 ?        00:00:00 //sbin/keepalived -D
root      36655   4678  0 22:20 pts/2    00:00:00 grep --color=auto keep
[root@master2 ~]# ip a show dev ens33		#查看eth0网卡绑定情况
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:4c:e2:d0 brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.6/24 brd 192.168.1.255 scope global ens33
       valid_lft forever preferred_lft forever
    inet 192.168.1.100/32 scope global ens33
       valid_lft forever preferred_lft forever
    inet6 fe80::e462:c755:bea9:60ca/64 scope link 
       valid_lft forever preferred_lft forever
    inet6 fe80::2038:9626:f100:f809/64 scope link tentative dadfailed 
       valid_lft forever preferred_lft forever

[root@master2 ~]# tail -f /var/log/messages
Mar 14 22:19:55 master1 Keepalived_vrrp[36635]: Sending gratuitous ARP on ens33 for 192.168.1.100
Mar 14 22:19:55 master1 Keepalived_vrrp[36635]: Sending gratuitous ARP on ens33 for 192.168.1.100
Mar 14 22:19:55 master1 Keepalived_vrrp[36635]: Sending gratuitous ARP on ens33 for 192.168.1.100
Mar 14 22:19:55 master1 Keepalived_vrrp[36635]: Sending gratuitous ARP on ens33 for 192.168.1.100
Mar 14 22:20:00 master1 Keepalived_vrrp[36635]: (VI_1) Sending/queueing gratuitous ARPs on ens33 for 192.168.1.100
Mar 14 22:20:00 master1 Keepalived_vrrp[36635]: Sending gratuitous ARP on ens33 for 192.168.1.100
Mar 14 22:20:00 master1 Keepalived_vrrp[36635]: Sending gratuitous ARP on ens33 for 192.168.1.100

发现已经将虚拟ip 192.168.1.100绑定了网卡ens33上

在另外一台服务器，候选master1 上启动keepalived服务，并观察

[root@master1 ~]# systemctl start keepalived
[root@master1 ~]# ps -ef | grep keep
root      36743      1  0 22:20 ?        00:00:00 //sbin/keepalived -D
root      36744  36743  0 22:20 ?        00:00:00 //sbin/keepalived -D
root      36746   4735  0 22:22 pts/2    00:00:00 grep --color=auto keep
[root@master1 ~]# ip a show dev ens33		#查看eth0网卡绑定情况
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:da:fe:ec brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.5/24 brd 192.168.1.255 scope global ens33
       valid_lft forever preferred_lft forever
    inet6 fe80::2038:9626:f100:f809/64 scope link 
       valid_lft forever preferred_lft forever

[root@master1 ~]# tail -f /var/log/messages
Mar 14 22:20:25 master1 Keepalived[36743]: Starting VRRP child process, pid=36744
Mar 14 22:20:25 master1 systemd: Started LVS and VRRP High Availability Monitor.
Mar 14 22:20:25 master1 Keepalived_vrrp[36744]: Registering Kernel netlink reflector
Mar 14 22:20:25 master1 Keepalived_vrrp[36744]: Registering Kernel netlink command channel
Mar 14 22:20:25 master1 Keepalived_vrrp[36744]: Assigned address 192.168.1.5 for interface ens33
Mar 14 22:20:25 master1 Keepalived_vrrp[36744]: Assigned address fe80::2038:9626:f100:f809 for interface ens33
Mar 14 22:20:25 master1 Keepalived_vrrp[36744]: Registering gratuitous ARP shared channel
Mar 14 22:20:25 master1 Keepalived_vrrp[36744]: (VI_1) removing VIPs.
Mar 14 22:20:25 master1 Keepalived_vrrp[36744]: (VI_1) Entering BACKUP STATE (init)
Mar 14 22:20:25 master1 Keepalived_vrrp[36744]: VRRP sockpool: [ifindex(  2), family(IPv4), proto(112), fd(12,13)]

从上面的信息可以看到keepalived已经配置成功

注意： 上面两台服务器的keepalived都设置为了BACKUP模式，在keepalived中2种模式，分别是master->backup模式和backup->backup模式。这两种模式有很大区别。
在master->backup模式下，一旦主库宕机，虚拟ip会自动漂移到从库，当主库修复后，keepalived启动后，还会把虚拟ip抢占过来，即使设置了非抢占模式（nopreempt）抢占ip的动作也会发生。
在backup->backup模式下，当主库宕机后虚拟ip会自动漂移到从库上，当原主库恢复和keepalived服务启动后，并不会抢占新主的虚拟ip，即使是优先级高于从库的优先级别，也不会发生抢占。为了减少ip漂移次数，通常是把修复好的主库当做新的备库。

2）MHA引入keepalived（MySQL服务进程挂掉时通过MHA 停止keepalived）

要想把keepalived服务引入MHA，只需要修改切换时触发的脚本文件master_ip_failover即可，在该脚本中添加在master发生宕机时对keepalived的处理。

编辑脚本/scripts/master_ip_failover，修改后如下

[root@manager ~]# vim /scripts/master_ip_failover
#!/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
);
my $vip = '192.168.1.100';
my $ssh_start_vip = "systemctl start keepalived.service";
my $ssh_stop_vip = "systemctl stop keepalived.service";
GetOptions(
'command=s' => \$command,
'ssh_user=s' => \$ssh_user,
'orig_master_host=s' => \$orig_master_host,
'orig_master_ip=s' => \$orig_master_ip,
'orig_master_port=i' => \$orig_master_port,
'new_master_host=s' => \$new_master_host,
'new_master_ip=s' => \$new_master_ip,
'new_master_port=i' => \$new_master_port,
);
exit &main();
sub main {
print "\n\nIN SCRIPT TEST====$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";
#`ssh $ssh_user\@cluster1 \" $ssh_start_vip \"`;
exit 0;
}
else {
&usage();
exit 1;
}
}
# A simple system call that enable the VIP on the new master
sub start_vip() {
`ssh $ssh_user\@$new_master_host \" $ssh_start_vip \"`;
}
# A simple system call that disable the VIP on the old_master
sub stop_vip() {
return 0 unless ($ssh_user);
`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";
}

修改完脚本，接下来在/etc/masterha/app1.cnf 中调用故障切换脚本

停止MHA：

[root@manager ~]# masterha_stop --conf=/etc/masterha/app1.cnf

在配置文件/etc/masterha/app1.cnf 中启用下面的参数

[root@manager ~]# vim /etc/masterha/app1.cnf
[server default]
……
master_ip_failover_script=/scripts/master_ip_failover		#(在[server default下面添加])
……

启动MHA：

[root@manager ~]# chmod +x /scripts/master_ip_failover

[root@manager ~]# nohup masterha_manager --conf=/etc/masterha/app1.cnf &>/tmp/mha_manager.log &

检查状态：

[root@manager ~]# masterha_check_status --conf=/etc/masterha/app1.cnf
app1 (pid:16786) is running(0:PING_OK), master:192.168.1.6

再检查集群状态，看是否会报错

[root@manager ~]# masterha_check_repl --conf=/etc/masterha/app1.cnf
……
Sun Mar 14 23:09:28 2021 - [info] Slaves settings check done.
Sun Mar 14 23:09:28 2021 - [info] 
192.168.1.6(192.168.1.6:3306) (current master)
 +--192.168.1.5(192.168.1.5:3306)
 +--192.168.1.7(192.168.1.7:3306)

Sun Mar 14 23:09:28 2021 - [info] Checking replication health on 192.168.1.5..
Sun Mar 14 23:09:28 2021 - [info]  ok.
Sun Mar 14 23:09:28 2021 - [info] Checking replication health on 192.168.1.7..
Sun Mar 14 23:09:28 2021 - [info]  ok.
Sun Mar 14 23:09:28 2021 - [info] Checking master_ip_failover_script status:
Sun Mar 14 23:09:28 2021 - [info]   /scripts/master_ip_failover --command=status --ssh_user=root --orig_master_host=192.168.1.6 --orig_master_ip=192.168.1.6 --orig_master_port=3306 


IN SCRIPT TEST====systemctl stop keepalived.service==systemctl start keepalived.service===

Checking the Status of the script.. OK 
Sun Mar 14 23:09:28 2021 - [info]  OK.
Sun Mar 14 23:09:28 2021 - [warning] shutdown_script is not defined.
Sun Mar 14 23:09:28 2021 - [info] Got exit code 0 (Not master dead).

MySQL Replication Health is OK.

可以看见已经没有报错了

/scripts/master_ip_failover添加或者修改的内容意思是当主库数据库发生故障时，会触发MHA切换，MHA Manager会停掉主库上的keepalived服务，触发虚拟ip漂移到备选从库，从而完成切换。 当然可以在keepalived里面引入脚本，这个脚本监控mysql是否正常运行，如果不正常，则调用该脚本杀掉keepalived进程（参考MySQL 高可用性keepalived+mysql双主）。

3）测试：

在master2上停止mysqld服务

[root@master2 ~]# systemctl stop mysqld

到slave(192.168.1.7)查看slave的状态：

mysql> show slave status\G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.1.5
                  Master_User: mharep
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000002
          Read_Master_Log_Pos: 154
               Relay_Log_File: relay-bin.000002
                Relay_Log_Pos: 320
        Relay_Master_Log_File: mysql-bin.000002
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes

从上图可以看出slave指向了新的master服务器192.168.1.5(在故障切换前指向的是192.168.1.6)

查看VIP绑定：

在192.168.1.5上查看vip绑定
[root@master1 ~]# ip a show dev ens33
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:da:fe:ec brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.5/24 brd 192.168.1.255 scope global ens33
       valid_lft forever preferred_lft forever
    inet 192.168.1.100/32 scope global ens33
       valid_lft forever preferred_lft forever
    inet6 fe80::2038:9626:f100:f809/64 scope link 
       valid_lft forever preferred_lft forever



在192.168.1.6上查看vip绑定
[root@master2 ~]# ip a show dev ens33
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:4c:e2:d0 brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.6/24 brd 192.168.1.255 scope global ens33
       valid_lft forever preferred_lft forever
    inet6 fe80::e462:c755:bea9:60ca/64 scope link 
       valid_lft forever preferred_lft forever
    inet6 fe80::2038:9626:f100:f809/64 scope link tentative dadfailed 
       valid_lft forever preferred_lft forever

从上面的显示结果可以看出vip地址漂移到了192.168.1.5

4）主从切换后续工作（把原主库修复成新从库）

重构： 重构就是你的主挂了，切换到 master1 上，master1 变成了主，因此重构的一种方案原主库修复成一个新的slave 主库切换后，把原主库修复成新从库，原主库数据文件完整的情况下，可通过以下方式找出最后执行的CHANGE MASTER命令：

[root@manager ~]# grep "CHANGE MASTER TO MASTER_" /masterha/app1/manager.log | tail -1
Sun Mar 14 23:13:36 2021 - [info]  All other slaves should start replication from here. Statement should be: CHANGE MASTER TO MASTER_HOST='192.168.1.5', MASTER_PORT=3306, MASTER_LOG_FILE='mysql-bin.000002', MASTER_LOG_POS=154, MASTER_USER='mharep', MASTER_PASSWORD='xxx';

将192.168.1.6（原主库）修复成从库

[root@master2 ~]# systemctl start mysqld
[root@master2 ~]# mysql -uroot -p123
mysql> CHANGE MASTER TO MASTER_HOST='192.168.1.5', MASTER_PORT=3306, MASTER_LOG_FILE='mysql-bin.000002', MASTER_LOG_POS=154, MASTER_USER='mharep', MASTER_PASSWORD='123';
mysql> start slave;
mysql> show slave status\G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.1.5
                  Master_User: mharep
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000002
          Read_Master_Log_Pos: 154
               Relay_Log_File: relay-bin.000002
                Relay_Log_Pos: 320
        Relay_Master_Log_File: mysql-bin.000002
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes
            
mysql> Bye

[root@master2 ~]# systemctl start keepalived
[root@master2 ~]# systemctl status keepalived
● keepalived.service - LVS and VRRP High Availability Monitor
   Loaded: loaded (/usr/lib/systemd/system/keepalived.service; enabled; vendor preset: disabled)
   Active: active (running) since Sun 2021-03-14 23:20:28 CST; 8s ago
……

启动 mha manager

[root@manager ~]# rm -rf /masterha/app1/app1.failover.complete 
[root@manager ~]# nohup masterha_manager --conf=/etc/masterha/app1.cnf --ignore_fail_on_start &>/tmp/mha_manager.log &
[1] 17294
[root@manager ~]# masterha_check_status --conf=/etc/masterha/app1.cnf
app1 (pid:17294) is running(0:PING_OK), master:192.168.1.5

[root@manager ~]# masterha_check_repl --conf=/etc/masterha/app1.cnf
Sun Mar 14 23:22:54 2021 - [info] Slaves settings check done.
Sun Mar 14 23:22:54 2021 - [info] 
192.168.1.5(192.168.1.5:3306) (current master)
 +--192.168.1.6(192.168.1.6:3306)
 +--192.168.1.7(192.168.1.7:3306)

Sun Mar 14 23:22:54 2021 - [info] Checking replication health on 192.168.1.6..
Sun Mar 14 23:22:54 2021 - [info]  ok.
Sun Mar 14 23:22:54 2021 - [info] Checking replication health on 192.168.1.7..
Sun Mar 14 23:22:54 2021 - [info]  ok.
Sun Mar 14 23:22:54 2021 - [info] Checking master_ip_failover_script status:
Sun Mar 14 23:22:54 2021 - [info]   /scripts/master_ip_failover --command=status --ssh_user=root --orig_master_host=192.168.1.5 --orig_master_ip=192.168.1.5 --orig_master_port=3306 


IN SCRIPT TEST====systemctl stop keepalived.service==systemctl start keepalived.service===

Checking the Status of the script.. OK 
Sun Mar 14 23:22:54 2021 - [info]  OK.
Sun Mar 14 23:22:54 2021 - [warning] shutdown_script is not defined.
Sun Mar 14 23:22:54 2021 - [info] Got exit code 0 (Not master dead).

MySQL Replication Health is OK.

1.5、停止keepalived为下边做环境

[root@master1 ~]# systemctl stop keepalived

[root@master2 ~]# systemctl stop keepalived

2、通过脚本实现VIP切换

通过脚本的方式管理VIP。这里是修改/scripts/master_ip_failover，也可以使用其他的语言完成，比如php语言。使用php脚本编写的failover这里就不介绍了。

1）使用脚本管理vip的话，需要手动在master服务器上绑定一个vip

[root@master2 ~]# ifconfig ens33:0 192.168.1.200/24

[root@master2 ~]# ip a show dev ens33
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:4c:e2:d0 brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.6/24 brd 192.168.1.255 scope global ens33
       valid_lft forever preferred_lft forever
    inet 192.168.1.200/24 brd 192.168.1.255 scope global secondary ens33:0
       valid_lft forever preferred_lft forever
    inet6 fe80::e462:c755:bea9:60ca/64 scope link 
       valid_lft forever preferred_lft forever
    inet6 fe80::2038:9626:f100:f809/64 scope link tentative dadfailed 
       valid_lft forever preferred_lft forever

2）在mha-manager上修改/scripts/ master_ip_failover，内容如下

[root@manager ~]# vim /scripts/master_ip_failover
#!/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
);
my $vip = '192.168.1.200';
my $key = '0';
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";
#`ssh $ssh_user\@cluster1 \" $ssh_start_vip \"`;
exit 0;
}
else {
&usage();
exit 1;
}
}
# A simple system call that enable the VIP on the new master
sub start_vip() {
`ssh $ssh_user\@$new_master_host \" $ssh_start_vip \"`;
}
# A simple system call that disable the VIP on the old_master
sub stop_vip() {
return 0 unless ($ssh_user);
`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";
}

停止MHA：

[root@manager ~]# masterha_stop --conf=/etc/masterha/app1.cnf

启动MHA：

[root@manager ~]# nohup masterha_manager --conf=/etc/masterha/app1.cnf &>/tmp/mha_manager.log &

检查状态：

[root@manager ~]# masterha_check_status --conf=/etc/masterha/app1.cnf
app1 (pid:17962) is running(0:PING_OK), master:192.168.1.5

再检查集群状态，看是否会报错

[root@manager ~]# masterha_check_repl --conf=/etc/masterha/app1.cnf
Sun Mar 14 23:32:08 2021 - [info] Slaves settings check done.
Sun Mar 14 23:32:08 2021 - [info] 
192.168.1.5(192.168.1.5:3306) (current master)
 +--192.168.1.6(192.168.1.6:3306)
 +--192.168.1.7(192.168.1.7:3306)

Sun Mar 14 23:32:08 2021 - [info] Checking replication health on 192.168.1.6..
Sun Mar 14 23:32:08 2021 - [info]  ok.
Sun Mar 14 23:32:08 2021 - [info] Checking replication health on 192.168.1.7..
Sun Mar 14 23:32:08 2021 - [info]  ok.
Sun Mar 14 23:32:08 2021 - [info] Checking master_ip_failover_script status:
Sun Mar 14 23:32:08 2021 - [info]   /scripts/master_ip_failover --command=status --ssh_user=root --orig_master_host=192.168.1.5 --orig_master_ip=192.168.1.5 --orig_master_port=3306 


IN SCRIPT TEST====/sbin/ifconfig ens33:0 down==/sbin/ifconfig ens33:0 192.168.1.100===

Checking the Status of the script.. OK 
Sun Mar 14 23:32:08 2021 - [info]  OK.
Sun Mar 14 23:32:08 2021 - [warning] shutdown_script is not defined.
Sun Mar 14 23:32:08 2021 - [info] Got exit code 0 (Not master dead).

MySQL Replication Health is OK.

3）测试：

在master1 上停掉mysql服务

[root@master1 ~]# systemctl stop mysqld

到slave(192.168.1.7)查看slave的状态：

mysql> show slave status\G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.1.6
                  Master_User: mharep
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000002
          Read_Master_Log_Pos: 154
               Relay_Log_File: relay-bin.000002
                Relay_Log_Pos: 320
        Relay_Master_Log_File: mysql-bin.000002
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes

从上图可以看出slave指向了新的master服务器（192.168.1.6） 查看VIP

[root@master2 ~]# ip a show dev ens33
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:4c:e2:d0 brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.6/24 brd 192.168.1.255 scope global ens33
       valid_lft forever preferred_lft forever
    inet 192.168.1.200/24 brd 192.168.1.255 scope global secondary ens33:0
       valid_lft forever preferred_lft forever
    inet6 fe80::e462:c755:bea9:60ca/64 scope link 
       valid_lft forever preferred_lft forever
    inet6 fe80::2038:9626:f100:f809/64 scope link tentative dadfailed 
       valid_lft forever preferred_lft forever

从上边可以看到master1(原来的master)释放了VIP，master2(新的master)接管了VIP地址

4）主从切换后续工作（把原主库修复成新从库）

重构： 重构就是你的主挂了，切换到 master1 上，master1 变成了主，因此重构的一种方案原主库修复成一个新的slave 主库切换后，把原主库修复成新从库，原主库数据文件完整的情况下，可通过以下方式找出最后执行的CHANGE MASTER命令：

[root@manager ~]# grep "CHANGE MASTER TO MASTER_" /masterha/app1/manager.log | tail -1
Mon Mar 15 00:05:29 2021 - [info]  All other slaves should start replication from here. Statement should be: CHANGE MASTER TO MASTER_HOST='192.168.1.6', MASTER_PORT=3306, MASTER_LOG_FILE='mysql-bin.000002', MASTER_LOG_POS=154, MASTER_USER='mharep', MASTER_PASSWORD='xxx';
[1]+  Done                    nohup masterha_manager --conf=/etc/masterha/app1.cnf &>/tmp/mha_manager.log

将192.168.1.5（原主库）修复成从库

[root@master1 ~]# systemctl start mysqld
[root@master1 ~]# mysql -uroot -p123
mysql> CHANGE MASTER TO MASTER_HOST='192.168.1.6', MASTER_PORT=3306, MASTER_LOG_FILE='mysql-bin.000002', MASTER_LOG_POS=154, MASTER_USER='mharep', MASTER_PASSWORD='123';
mysql> start slave;
mysql> show slave status\G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.1.6
                  Master_User: mharep
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000002
          Read_Master_Log_Pos: 154
               Relay_Log_File: relay-bin.000002
                Relay_Log_Pos: 320
        Relay_Master_Log_File: mysql-bin.000002
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes

启动 mha manager

[root@manager ~]# rm -rf /masterha/app1/app1.failover.complete 
[root@manager ~]# nohup masterha_manager --conf=/etc/masterha/app1.cnf --ignore_fail_on_start &>/tmp/mha_manager.log &
[root@manager ~]# masterha_check_status --conf=/etc/masterha/app1.cnf
app1 (pid:5415) is running(0:PING_OK), master:192.168.1.6

[root@manager ~]# masterha_check_repl --conf=/etc/masterha/app1.cnf
Mon Mar 15 00:10:21 2021 - [info] Slaves settings check done.
Mon Mar 15 00:10:21 2021 - [info] 
192.168.1.6(192.168.1.6:3306) (current master)
 +--192.168.1.5(192.168.1.5:3306)
 +--192.168.1.7(192.168.1.7:3306)

Mon Mar 15 00:10:21 2021 - [info] Checking replication health on 192.168.1.5..
Mon Mar 15 00:10:21 2021 - [info]  ok.
Mon Mar 15 00:10:21 2021 - [info] Checking replication health on 192.168.1.7..
Mon Mar 15 00:10:21 2021 - [info]  ok.
Mon Mar 15 00:10:21 2021 - [info] Checking master_ip_failover_script status:
Mon Mar 15 00:10:21 2021 - [info]   /scripts/master_ip_failover --command=status --ssh_user=root --orig_master_host=192.168.1.6 --orig_master_ip=192.168.1.6 --orig_master_port=3306 


IN SCRIPT TEST====/sbin/ifconfig ens33:0 down==/sbin/ifconfig ens33:0 192.168.1.200===

Checking the Status of the script.. OK 
Mon Mar 15 00:10:21 2021 - [info]  OK.
Mon Mar 15 00:10:21 2021 - [warning] shutdown_script is not defined.
Mon Mar 15 00:10:21 2021 - [info] Got exit code 0 (Not master dead).

MySQL Replication Health is OK.

总结：MHA软件由两部分组成，Manager工具包和Node工具包，具体的说明如下

Manager工具包主要包括以下几个工具：
masterha_check_ssh 检查MHA的SSH配置状况
masterha_check_repl 检查MySQL复制状况
masterha_manger 启动MHA
masterha_check_status 检测当前MHA运行状态
masterha_master_monitor 检测 master是否宕机
masterha_master_switch 控制故障转移（自动或者手动）
masterha_conf_host 添加或删除配置的server信息
Node工具包（这些工具通常由MHA Manager的脚本触发，无需人为操作）主要包括以下几个工具：
save_binary_logs 保存和复制master的二进制日志
apply_diff_relay_logs 识别差异的中继日志事件并将其差异的事件应用于其他的slave
filter_mysqlbinlog 去除不必要的ROLLBACK事件（MHA已不再使用这个工具）
purge_relay_logs 清除中继日志（不会阻塞SQL线程）

mysql必备技能掌握：

1、MySQL架构：对mysql的架构，整体有个印象，才能不断的加深对mysql的理解和后继的学习
2、用各种姿势备份MySQL数据库 数据备份是DBA或运维工程师日常工作之一，如果让你来备份，你
会用什么方式备份，在时间时间备份，使用什么策略备份
3、mysql主从复制及读写分离 mysql的主从复制及读写分离是DBA必备技能之一
4、MySQL/MariaDB数据库基于SSL实现主从复制 加强主从复制的安全性
5、MySQL高可用 数据的高可用如何保证
6、数据库Sharding的基本思想和切分策略 随着数据量的不断攀升，从性能和可维护的角度，需要进行一些Sharding，也就是数据库的切分，有垂直切分和水平切分
7、MySQL/MariaDB 性能调整和优化技巧 掌握优化思路和技巧，对数据库的不断优化是一项长期工程