mysql MHA

1，用到的工具
(1)、 Manager工具:
- masterha_check_ssh : 检查MHA的SSH配置。
- masterha_check_repl : 检查MySQL复制。
- masterha_manager : 启动MHA。
- masterha_check_status : 检测当前MHA运行状态。
- masterha_master_monitor : 监测master是否宕机。
- masterha_master_switch : 控制故障转移(自动或手动)。
- masterha_conf_host : 添加或删除配置的server信息。
- masterha_stop : 停止MHA
(2)、 Node工具(这些工具通常由MHAManager的脚本触发,无需人手操作)。
- 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应用于其它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)
2，主机：
ip        MHA    Mysql
192.168.97.118    Node    Master
192.168.97.119    Node    Slave
192.168.97.120    Manager    Slave  软件node也安装

3，ssh
 3.1 各个节点上产生密钥
    cd ~/.ssh
    ssh-keygen -t dsa -P '' -f id_dsa
 3.2 在118上操作，将各个节点上的公钥copy到本机，再操作
   scp 192.168.97.119:/root/.ssh/id_dsa.pub ./id_dsa.pub.119
   scp 192.168.97.120:/root/.ssh/id_dsa.pub ./id_dsa.pub.120
   cat ./id_dsa.pub.119 >> authorized_keys
   cat ./id_dsa.pub.120 >> authorized_keys
 3.3 查看生成的通用密钥，并copy其他节点
   cat authorized_keys
   scp authorized_keys 192.168.97.119:/root/.ssh/
   scp authorized_keys 192.168.97.120:/root/.ssh/
 3.4，验证任意两台节点无密钥登陆
    ssh 192.168.118
    ssh 192.168.119
    ssh 192.168.120
 3.5，主机名登陆
    vim /etc/hosts   编辑追加
    192.168.97.118  m1
    192.168.97.119  s1
    192.168.97.120  s2
    验证主机名无密钥登陆：
    ssh m1
    ssh s1
    ssh s2
4,搭建主从环境、配置账户
  复制环境参考：http://blog.csdn.net/jc_benben/article/details/52836135
   创建用户mha管理的账号，在所有mysql服务器上都需要执行：
   GRANT SUPER,RELOAD,REPLICATION CLIENT,SELECT ON *.* TO  manager@'192.168.97.%'  IDENTIFIED BY 'manager_1234';
   GRANT CREATE,INSERT,UPDATE,DELETE,DROP ON *.* TO manager@'192.168.97.%';

   创建主从账号，在所有mysql服务器上都需要执行，注意如果主从上执行了，这里就不用了，这里也可以执行ip限制：
   GRANT RELOAD, SUPER, REPLICATION SLAVE ON *.* TO 'repl'@'%' IDENTIFIED BY 'repl_1234';

5，安装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masterfailover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。
  为了节省机器，可以从现有复制架构中选一台“闲置”从库作为manager server，比如：某台从库不对外提供读的服务，
  只是作为候选主库，或是专门用于备份
  5.1，mha4mysql-node 所有节点
   tar -zxvpf mha4mysql-node-0.56.tar.gz
   perl Makefile.PL
   make && make install
   需要首先安装一些依赖包，yum install perl-DBI perl-DBD-mysql perl-ExtUtils-CBuilder perl-ExtUtils-MakeMaker cpan   -y
  5.2，mha4mysql-manager (只管理节点)
  yum install -y perl-DBD-MySQL
  yum install -y perl-Config-Tiny
  yum install -y perl-Log-Dispatch
  yum install -y perl-Parallel-ForkManager
  yum install -y perl-Config-IniFiles
 这里有一个地址可以查找rpm包很不错的：https://pkgs.org/
  安装：
     tar -xvf mha4mysql-manager-0.56.tar.gz
    cd mha4mysql-manager-0.56
    perl Makefile.PL
    make && make install
 如果最后还是安装不好，可以尝试一下perl CPAN的方式：
  perl -MCPAN -e shell
  cpan[1]> install Log::Dispatch
6，编辑配置
 mkdir /etc/masterha/
 mkdir /var/log/masterha/app1
 vim /etc/masterha/app1.cnf 添加内容
[server default]
user=manager #ssh免密钥登录的帐号名
password=manager_1234
manager_workdir=/var/log/masterha/app1
manager_log=/var/log/masterha/app1/manager.log
ssh_user=root
repl_user=repl  #mysql复制帐号，用来在主从机之间同步二进制日志等
repl_password=123456
ping_interval=1 #间隔，用来检测master是否正常

[server1]
hostname=192.168.97.118
candidate_master=1 #master机宕掉后,优先启用这台作为新master
master_binlog_dir=/usr/local/mysql/data/

[server2]
hostname=192.168.97.119
#candidate_master=1
master_binlog_dir=/usr/local/mysql/data/

[server3]
hostname=192.168.97.120
candidate_master=1
master_binlog_dir=/usr/local/mysql/data/
#[server4]
#hostname=host4
#no_master=1  #永远不作为master

#以下是vip切换使用内容，可以暂不考虑
#master_ip_failover_script=/etc/masterha/master_ip_failover
#master_ip_online_change_script=/etc/masterha/master_ip_online_change

7，检测（2个就够了）
检查节点连接性：
[root@s2 .ssh]# masterha_check_ssh --conf=/etc/masterha/app1.cnf
检查repl环境：
[root@s2 ~]# masterha_check_repl --conf=/etc/masterha/app1.cnf
Fri Oct 21 11:43:37 2016 - [warning] Global configuration file /etc/masterha_default.cnf not found. Skipping.
Fri Oct 21 11:43:37 2016 - [info] Reading application default configuration from /etc/masterha/app1.cnf..
Fri Oct 21 11:43:37 2016 - [info] Reading server configuration from /etc/masterha/app1.cnf..
Fri Oct 21 11:43:37 2016 - [info] MHA::MasterMonitor version 0.56.
Fri Oct 21 11:43:37 2016 - [info] GTID failover mode = 0
Fri Oct 21 11:43:37 2016 - [info] Dead Servers:
Fri Oct 21 11:43:37 2016 - [info] Alive Servers:
Fri Oct 21 11:43:37 2016 - [info]   192.168.97.118(192.168.97.118:3306)
Fri Oct 21 11:43:37 2016 - [info]   192.168.97.119(192.168.97.119:3306)
Fri Oct 21 11:43:37 2016 - [info]   192.168.97.120(192.168.97.120:3306)
Fri Oct 21 11:43:37 2016 - [info] Alive Slaves:
Fri Oct 21 11:43:37 2016 - [info]   192.168.97.119(192.168.97.119:3306)  Version=5.7.14-log (oldest major version between slaves) log-bin:enabled
Fri Oct 21 11:43:37 2016 - [info]     Replicating from 192.168.97.118(192.168.97.118:3306)
Fri Oct 21 11:43:37 2016 - [info]   192.168.97.120(192.168.97.120:3306)  Version=5.7.14-log (oldest major version between slaves) log-bin:enabled
Fri Oct 21 11:43:37 2016 - [info]     Replicating from 192.168.97.118(192.168.97.118:3306)
Fri Oct 21 11:43:37 2016 - [info]     Primary candidate for the new Master (candidate_master is set)
Fri Oct 21 11:43:37 2016 - [info] Current Alive Master: 192.168.97.118(192.168.97.118:3306)
Fri Oct 21 11:43:37 2016 - [info] Checking slave configurations..
Fri Oct 21 11:43:37 2016 - [info] Checking replication filtering settings..
Fri Oct 21 11:43:37 2016 - [info]  binlog_do_db= t1, binlog_ignore_db= information_schema,mysql,performance_schema,t,test
Fri Oct 21 11:43:37 2016 - [info]  Replication filtering check ok.
Fri Oct 21 11:43:37 2016 - [info] GTID (with auto-pos) is not supported
Fri Oct 21 11:43:37 2016 - [info] Starting SSH connection tests..
Fri Oct 21 11:43:39 2016 - [info] All SSH connection tests passed successfully.
Fri Oct 21 11:43:39 2016 - [info] Checking MHA Node version..
Fri Oct 21 11:43:40 2016 - [info]  Version check ok.
Fri Oct 21 11:43:40 2016 - [info] Checking SSH publickey authentication settings on the current master..
Fri Oct 21 11:43:40 2016 - [info] HealthCheck: SSH to 192.168.97.118 is reachable.
Fri Oct 21 11:43:41 2016 - [info] Master MHA Node version is 0.56.
Fri Oct 21 11:43:41 2016 - [info] Checking recovery script configurations on 192.168.97.118(192.168.97.118:3306)..
Fri Oct 21 11:43:41 2016 - [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=0.56 --start_file=mysql-bin.000006
Fri Oct 21 11:43:41 2016 - [info]   Connecting to root@192.168.97.118(192.168.97.118:22)..
  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.000006
Fri Oct 21 11:43:41 2016 - [info] Binlog setting check done.
Fri Oct 21 11:43:41 2016 - [info] Checking SSH publickey authentication and checking recovery script configurations on all alive slave servers..
Fri Oct 21 11:43:41 2016 - [info]   Executing command : apply_diff_relay_logs --command=test --slave_user='manager' --slave_host=192.168.97.119 --slave_ip=192.168.97.119 --slave_port=3306 --workdir=/var/tmp --target_version=5.7.14-log --manager_version=0.56 --relay_log_info=/usr/local/mysql/data/relay-log.info  --relay_dir=/usr/local/mysql/data/  --slave_pass=xxx
Fri Oct 21 11:43:41 2016 - [info]   Connecting to root@192.168.97.119(192.168.97.119:22)..
  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.000034
    Temporary relay log file is /usr/local/mysql/data/relay-bin.000034
    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.
Fri Oct 21 11:43:41 2016 - [info]   Executing command : apply_diff_relay_logs --command=test --slave_user='manager' --slave_host=192.168.97.120 --slave_ip=192.168.97.120 --slave_port=3306 --workdir=/var/tmp --target_version=5.7.14-log --manager_version=0.56 --relay_log_info=/usr/local/mysql/data/relay-log.info  --relay_dir=/usr/local/mysql/data/  --slave_pass=xxx
Fri Oct 21 11:43:41 2016 - [info]   Connecting to root@192.168.97.120(192.168.97.120:22)..
  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.000038
    Temporary relay log file is /usr/local/mysql/data/relay-bin.000038
    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.
Fri Oct 21 11:43:42 2016 - [info] Slaves settings check done.
Fri Oct 21 11:43:42 2016 - [info]
192.168.97.118(192.168.97.118:3306) (current master)
 +--192.168.97.119(192.168.97.119:3306)
 +--192.168.97.120(192.168.97.120:3306)

Fri Oct 21 11:43:42 2016 - [info] Checking replication health on 192.168.97.119..
Fri Oct 21 11:43:42 2016 - [info]  ok.
Fri Oct 21 11:43:42 2016 - [info] Checking replication health on 192.168.97.120..
Fri Oct 21 11:43:42 2016 - [info]  ok.
Fri Oct 21 11:43:42 2016 - [warning] master_ip_failover_script is not defined.
Fri Oct 21 11:43:42 2016 - [warning] shutdown_script is not defined.
Fri Oct 21 11:43:42 2016 - [info] Got exit code 0 (Not master dead).

MySQL Replication Health is OK.

8，管理MHA
启动manager
nohup masterha_manager --conf=/etc/masterha/app1.cnf < /dev/null >/logs/mha/app1/manager.log 2>&1 &
查看状态：
masterha_check_status--conf=/etc/masterha/app1.cnf
停止manager：
masterha_stop --conf=/etc/masterha/app1.cnf
9,实现VIP切换
 背景：MHA能够在10~30秒内实现自动故障检测和故障转移，这适用于对高可用性，数据完整性要求较高的场合。要做到无缝切换，
       还需要依赖于VIP漂移。VIP漂移比较常用的方式为使用keepalived或者使用脚本直接实现。脚本方式无须安装及复杂配置，相对简单
       这里采用脚本。
 9.1 在master节点上添加一虚拟ip,有几种方式
 
  第一：vim /etc/sysconfig/network-scripts/ifcfg-eno16777736 在文件中添加一ip
  IPADDR1=192.168.97.124
  重启网卡
  systemctl restart network或者nmcli conn up "eno16777736"
  第二种，
   [root@m1 ~]# nmcli conn modify "eno16777736" +ipv4.address "192.168.97.129/24 "
   [root@m1 ~]# nmcli conn up "eno16777736"
   Connection successfully activated (D-Bus active path: /org/freedesktop/NetworkManager/ActiveConnection/3)
  centos7之前的临时方式，不推荐：
   /sbin/ifconfig eno16777736:0 192.168.97.126 netmask 255.255.255.0 up
 9.2，在管理节点上重新配置文件app1.cnf
  master_ip_failover_script=/etc/masterha/master_ip_failover
  master_ip_online_change_script=/etc/masterha/master_ip_online_change
  report_script= /etc/masterha/app1/send_report    #master failover时执行，有邮件通知需求是部署
 9.3,部署脚本
 脚本master_ip_failover，内容如下：
#!/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 = '192.168.97.124/24';  # Virtual IP  
my $key = "1";  
my $ssh_start_vip = "/sbin/ifconfig eno16777736:$key $vip";  
my $ssh_stop_vip = "/sbin/ifconfig eno16777736:$key down";  
my $exit_code = 0;  
 
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" ) {  
 
        # $orig_master_host, $orig_master_ip, $orig_master_port are passed.  
        # If you manage master ip address at global catalog database,  
        # invalidate orig_master_ip here.  
        my $exit_code = 1;  
        eval {  
            print "\n\n\n***************************************************************\n";  
            print "Disabling the VIP - $vip on old master: $orig_master_host\n";  
            print "***************************************************************\n\n\n\n";  
&stop_vip();  
            $exit_code = 0;  
        };  
        if ($@) {  
            warn "Got Error: $@\n";  
            exit $exit_code;  
        }  
        exit $exit_code;  
}  
elsif ( $command eq "start" ) {  
 
        # all arguments are passed.  
        # If you manage master ip address at global catalog database,  
        # activate new_master_ip here.  
        # You can also grant write access (create user, set read_only=0, etc) here.  
my $exit_code = 10;  
        eval {  
            print "\n\n\n***************************************************************\n";  
            print "Enabling the VIP - $vip on new master: $new_master_host \n";  
            print "***************************************************************\n\n\n\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 $ssh_user\@$orig_master_host \" $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() {  
`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=po  
rt –new_master_host=host –new_master_ip=ip –new_master_port=port\n";  
}

脚本master_ip_online_change（shell），内容：
#/bin/bash  
source /root/.bash_profile  
 
vip=`echo '192.168.97.124/24'`  # Virtual IP  
key=`echo '1'`  
 
command=`echo "$1" | awk -F = '{print $2}'`  
orig_master_host=`echo "$2" | awk -F = '{print $2}'`  
new_master_host=`echo "$7" | awk -F = '{print $2}'`  
orig_master_ssh_user=`echo "${12}" | awk -F = '{print $2}'`  
new_master_ssh_user=`echo "${13}" | awk -F = '{print $2}'`  
 
stop_vip=`echo "ssh root@$orig_master_host /sbin/ifconfig  eno16777736:$key  down"`  
start_vip=`echo "ssh root@$new_master_host /sbin/ifconfig  eno16777736:$key  $vip"`  
 
if [ $command = 'stop' ]  
   then  
   echo -e "\n\n\n***************************************************************\n"  
   echo -e "Disabling the VIP - $vip on old master: $orig_master_host\n"  
   $stop_vip  
   if [ $? -eq 0 ]  
      then  
      echo "Disabled the VIP successfully"  
   else  
      echo "Disabled the VIP failed"  
   fi  
   echo -e "***************************************************************\n\n\n\n"  
fi  
 
if [ $command = 'start' -o $command = 'status' ]  
   then  
   echo -e "\n\n\n***************************************************************\n"  
   echo -e "Enabling the VIP - $vip on new master: $new_master_host \n"  
   $start_vip  
   if [ $? -eq 0 ]  
      then  
      echo "Enabled the VIP successfully"  
   else  
      echo "Enabled the VIP failed"  
   fi  
   echo -e "***************************************************************\n\n\n\n"  
fi
注意，这个脚本也可以使用perl版本
#!/usr/bin/env perl  
use strict;  
use warnings FATAL =>'all';  
 
use Getopt::Long;  
 
my $vip = '10.1.5.21/24';  # Virtual IP  
my $key = "1";  
my $ssh_start_vip = "/sbin/ifconfig eth0:$key $vip";  
my $ssh_stop_vip = "/sbin/ifconfig eth0:$key down";  
my $exit_code = 0;  
 
my (  
  $command,              $orig_master_is_new_slave, $orig_master_host,  
  $orig_master_ip,       $orig_master_port,         $orig_master_user,  
  $orig_master_password, $orig_master_ssh_user,     $new_master_host,  
  $new_master_ip,        $new_master_port,          $new_master_user,  
  $new_master_password,  $new_master_ssh_user,  
);  
GetOptions(  
  'command=s'                => \$command,  
  'orig_master_is_new_slave' => \$orig_master_is_new_slave,  
  'orig_master_host=s'       => \$orig_master_host,  
  'orig_master_ip=s'         => \$orig_master_ip,  
  'orig_master_port=i'       => \$orig_master_port,  
  'orig_master_user=s'       => \$orig_master_user,  
  'orig_master_password=s'   => \$orig_master_password,  
  'orig_master_ssh_user=s'   => \$orig_master_ssh_user,  
  'new_master_host=s'        => \$new_master_host,  
  'new_master_ip=s'          => \$new_master_ip,  
  'new_master_port=i'        => \$new_master_port,  
  'new_master_user=s'        => \$new_master_user,  
  'new_master_password=s'    => \$new_master_password,  
  'new_master_ssh_user=s'    => \$new_master_ssh_user,  
);  
 
 
exit &main();  
 
sub main {  
 
#print "\n\nIN SCRIPT TEST====$ssh_stop_vip==$ssh_start_vip===\n\n";  
 
if ( $command eq "stop" || $command eq "stopssh" ) {  
 
        # $orig_master_host, $orig_master_ip, $orig_master_port are passed.  
        # If you manage master ip address at global catalog database,  
        # invalidate orig_master_ip here.  
        my $exit_code = 1;  
        eval {  
            print "\n\n\n***************************************************************\n";  
            print "Disabling the VIP - $vip on old master: $orig_master_host\n";  
            print "***************************************************************\n\n\n\n";  
&stop_vip();  
            $exit_code = 0;  
        };  
        if ($@) {  
            warn "Got Error: $@\n";  
            exit $exit_code;  
        }  
        exit $exit_code;  
}  
elsif ( $command eq "start" ) {  
 
        # all arguments are passed.  
        # If you manage master ip address at global catalog database,  
        # activate new_master_ip here.  
        # You can also grant write access (create user, set read_only=0, etc) here.  
my $exit_code = 10;  
        eval {  
            print "\n\n\n***************************************************************\n";  
            print "Enabling the VIP - $vip on new master: $new_master_host \n";  
            print "***************************************************************\n\n\n\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 $orig_master_ssh_user\@$orig_master_host \" $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 $new_master_ssh_user\@$new_master_host \" $ssh_start_vip \"`;  
}  
# A simple system call that disable the VIP on the old_master  
sub stop_vip() {  
`ssh $orig_master_ssh_user\@$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=po  
rt –new_master_host=host –new_master_ip=ip –new_master_port=port\n";  
}

脚本send_report（shell），内容：
    #/bin/bash  
    source /root/.bash_profile  
      
    orig_master_host=`echo "$1" | awk -F = '{print $2}'`  
    new_master_host=`echo "$2" | awk -F = '{print $2}'`  
    new_slave_hosts=`echo "$3" | awk -F = '{print $2}'`  
    subject=`echo "$4" | awk -F = '{print $2}'`  
    body=`echo "$5" | awk -F = '{print $2}'`  
      
    #判断日志结尾是否有successfully，有则表示切换成功，成功与否都发邮件。  
    tac /etc/masterha/app1/manager.log | sed -n 2p | grep 'successfully' > /dev/null  
    if [ $? -eq 0 ]  
        then  
        echo -e "MHA $subject 主从切换成功\n master:$orig_master_host --> $new_master_host \n $body \n 当前从库:$new_slave_hosts" | mutt  
     -s "MySQL实例宕掉,MHA $subject 切换成功" -- [e-mail]  
    else  
        echo -e "MHA $subject 主从切换失败\n master:$orig_master_host --> $new_master_host \n $body" | mutt -s "MySQL实例宕掉,MHA $subje  
    ct 切换失败" -- [e-mail]  
    fi   

9，测试
  9.1在原主节点97.118上执行停止操作，删除虚拟ip
    systemctl stop mysqld_safe
  9.2，在管理节点上查看状态
  [root@s2 masterha]# tail -10f /var/log/masterha/app1/manager.log
Started automated(non-interactive) failover.
The latest slave 192.168.97.119(192.168.97.119:3306) has all relay logs for recovery.
Selected 192.168.97.120(192.168.97.120:3306) as a new master.
192.168.97.120(192.168.97.120:3306): OK: Applying all logs succeeded.
192.168.97.120(192.168.97.120:3306): OK: Activated master IP address.
192.168.97.119(192.168.97.119:3306): This host has the latest relay log events.
Generating relay diff files from the latest slave succeeded.
192.168.97.119(192.168.97.119:3306): OK: Applying all logs succeeded. Slave started, replicating from 192.168.97.120(192.168.97.120:3306)
192.168.97.120(192.168.97.120:3306): Resetting slave info succeeded.
Master failover to 192.168.97.120(192.168.97.120:3306) completed successfully.
可以看到，master正常切换到了97.120
  9.3，使用虚拟地址连接mysql验证切换正常
  mysql -uroot -p123456 -h 192.168.97.124
  OK 正常
  9.4，连接s1节点查看
mysql> show slave status \G;
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.97.120
                  Master_User: repl
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000007
          Read_Master_Log_Pos: 154
               Relay_Log_File: relay-bin.000002
                Relay_Log_Pos: 320
        Relay_Master_Log_File: mysql-bin.000007
             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: 154
              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: 137
                  Master_UUID: 49bb8923-8eb2-11e6-b687-000c29b86833
             Master_Info_File: /usr/local/mysql/data/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)
 看到心得master节点为192.168.97.120

10，一些错误
  10.1，执行：masterha_check_repl --conf=/etc/masterha/app1.cnf出现
Can't exec "mysqlbinlog": No suchfile or directory at /usr/local/share/perl5/MHA/BinlogManager.pm line 106.
mysqlbinlog version command failed with rc1:0, please verify PATH, LD_LIBRARY_PATH, and client options
解决：
type mysqlbinlog
然后：ln -s /usr/local/mysql/bin/mysqlbinlog /usr/bin/mysqlbinlog  
注意/usr/local/mysql/bin/mysqlbinlog是type mysqlbinlog的结果

  10.2，执行：masterha_check_repl --conf=/etc/masterha/app1.cnf出现
Testing mysql connection and privileges..sh: mysql: command not found
mysql command failed with rc 127:0!
解决：ln -s /usr/local/mysql/bin/mysql /usr/bin/mysql

11，
 1，简介
1.1mha简介
MHA，即MasterHigh Availability Manager and Tools for MySQL，是日本的一位MySQL专家采用Perl语言编写的一个脚本管理工具，该工具仅适用于MySQLReplication（二层）环境，目的在于维持Master主库的高可用性。
MHA(Master High Availability)是自动的master故障转移和Slave提升的软件包.它是基于标准的MySQL复制(异步/半同步).
     MHA有两部分组成:MHA Manager(管理节点)和MHA Node(数据节点).
     MHA Manager可以单独部署在一台独立机器上管理多个master-slave集群,也可以部署在一台slave上.MHA Manager探测集群的node节点,当发现master出现故障的时候,它可以自动将具有最新数据的slave提升为新的master,然后将所有其它的slave导向新的master上.整个故障转移过程对应用程序是透明的。
     MHA node运行在每台MySQL服务器上(master/slave/manager),它通过监控具备解析和清理logs功能的脚本来加快故障转移的。
 
1.2，mha特点
（1）. 10-30s实现master failover（9-12s可以检测到主机故障，7-10s可以关闭主机避免SB，在用很短的时间应用差异日志）
（2） 部署简单，无需对现有M-S结构做任何改动（至少3台，保证切换后仍保持M-S结构）
（3）. 支持手动在线切换（主机硬件维护），downtime几乎很短0.5-2s
（4）. 保证故障切换后多从库数据的一致性
（5）. 完全自动化的failover及快速复制架构恢复方案（一主多从）
（6）. 恢复过程包括：选择新主库、确认从库间relaylog差异、新主库应用必要语句、其他从库同步差异语句、重新建立复制连接
1.3，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。
说明：本文参考互联网