MHA 安装与简单使用

MHA 在过去几年一直用的比较火，特别是在在传统复制的那个年代。至从有了GTID好像我们也可以把MHA给忘记了，但是很多企业现在还是在用的比较多。每个公司的MHA玩法也不太一样，但是本质都是差不多了。下面我就带大家简单快速的搭建一个MHA玩一下，看看MHA是怎么玩的。如果要深入的了解MHA我建议去github里面看看一些官方的文档或者别人的一些博客，google一下或者bing一下就能找到了。我也不去重复的写了。

 

MHA 搭建全过程                                                             

wget  http://rpmfind.net/linux/dag/redhat/el6/en/x86_64/dag/RPMS/perl-Log-Dispatch-2.26-1.el6.rf.noarch.rpm

wget ftp:// rpmfind.net/linux/dag/redhat/el6/en/x86_64/dag/RPMS/perl-Parallel-ForkManager-0.7.5-2.2.el6.rf.noarch.rpm

wget ftp:// rpmfind.net/linux/dag/redhat/el6/en/x86_64/dag/RPMS/perl-Mail-Sender-0.8.16-1.el6.rf.noarch.rpm

wget  ftp://rpmfind.net/linux/dag/redhat/el6/en/x86_64/dag/RPMS/perl-Mail-Sendmail-0.79-1.2.el6.rf.noarch.rpm

 

有时候yum localinstall 安装不了原因，可能是上面的几个包没有安装

安装完成以后可以使用如下的管理工具：

#rpm -ql mha4mysql-manager-0.56-0.el6.noarch

/usr/bin/masterha_check_repl

/usr/bin/masterha_check_ssh

/usr/bin/masterha_check_status

/usr/bin/masterha_conf_host

/usr/bin/masterha_manager

/usr/bin/masterha_master_monitor

/usr/bin/masterha_master_switch

/usr/bin/masterha_secondary_check

/usr/bin/masterha_stop

/usr/share/man/man1/masterha_check_repl.1.gz

/usr/share/man/man1/masterha_check_ssh.1.gz

/usr/share/man/man1/masterha_check_status.1.gz

/usr/share/man/man1/masterha_conf_host.1.gz

/usr/share/man/man1/masterha_manager.1.gz

/usr/share/man/man1/masterha_master_monitor.1.gz

 

 环境准备：

首先我们需要准备一下3台机器

 master 192.168.5.100     端口3310

 slave1：192.168.5.101    端口3310

 slave2：192.168.5.102     端口3310

 vip 192.168.5.99

 

 关闭防火墙的和SELINUX

systemctl stop firewalld.service #停止firewall

systemctl disable firewalld.service #禁止firewall开机启动

sed -i 's/SELINUX=.*/SELINUX=disabled/g' /etc/selinux/config

setenforce 0

 

 

1：master 的搭建过程

实例初始化：

/usr/local/mysql/bin/mysqld --defaults-file=/data/mysql/mysql_3310/my_3310.cnf --initialize-insecure

启动实例

/usr/local/mysql/bin/mysqld --defaults-file=/data/mysql/mysql_3310/my_3310.cnf &

 2：创建复制用的账号

#mysql>CREATE USER rpl@'192.168.5.%';

#mysql>GRANT REPLICATION SLAVE ON *.* TO  'rpl'@'192.168.5.%'  identified by '123456';

 

搭建主从环境：

 change master  to master_host='192.168.5.100',master_port=3310, master_user='rpl', master_password='123456',master_auto_position=1;

 

3台机器之间建立机器信任机制建立

 

 使用 ssh-keygen 生成key

 

# ssh-keygen 一路回车就行了。最后会在~/,ssh 下面产生：id_rsa id_rsa.pub 两个文件。

 

生成信任文件：

#cd ~/.ssh/

#cat id_rsa.pub>authorized_keys  ## 每个机器都需要执行下

#chmod 600 *

保留.ssh 下面只有id_rsa,id_rsa.pub 其他文件可以删或是备份移走。

把文件分发到其它三台机器上（需要输出密码）

#cd ~

#scp -r .ssh 192.168.5.102

#scp   -r   .ssh  192.168.5.103

这样就完成了3台机器的ssh信任

 

 

配置文件：

[server default]

log_level=debug    ### 开启debug的模式

manager_log=/etc/masterha/app1.log

manager_workdir=/etc/masterha/app1

master_binlog_dir=/data/mysql/mysql_3310/logs

master_ip_failover_script=/etc/masterha/master_ip_failover

master_ip_online_change_script=/etc/masterha/master_ip_failover

password=123456

ping_interval=1

remote_workdir=/etc/masterha/app1

repl_password=123456

repl_user=rpl

ssh_user=root

user=root

 

[server1]

candidate_master=1

hostname=node1

port=3310

 

 

[server2]

candidate_master=1

hostname=node2

port=3310

 

[server3]

hostname=node3

port=3310

  

 

1.检查SSH配置

 

#masterha_check_ssh --conf=/etc/masterha/app1.conf

 

以下是我们能看到的返回信息：

 

#masterha_check_ssh --conf=/etc/masterha/app1.conf

Wed Nov  1 01:56:31 2017 - [info] Reading default configuration from /etc/masterha_default.cnf..

Wed Nov  1 01:56:31 2017 - [info] Reading application default configuration from /etc/masterha/app1.conf..

Wed Nov  1 01:56:31 2017 - [info] Reading server configuration from /etc/masterha/app1.conf..

Wed Nov  1 01:56:31 2017 - [info] Starting SSH connection tests..

Wed Nov  1 01:56:32 2017 - [debug]

Wed Nov  1 01:56:31 2017 - [debug]  Connecting via SSH from root@192.168.5.102(192.168.5.102:22) to root@192.168.5.103(192.168.5.103:22)..

Wed Nov  1 01:56:32 2017 - [debug]   ok.

Wed Nov  1 01:56:32 2017 - [debug]

Wed Nov  1 01:56:32 2017 - [debug]  Connecting via SSH from root@192.168.5.103(192.168.5.103:22) to root@192.168.5.102(192.168.5.102:22)..

Wed Nov  1 01:56:32 2017 - [debug]   ok.

Wed Nov  1 01:56:32 2017 - [info] All SSH connection tests passed successfully.

2. 检查 主从复制是不是正常：masterha_check_repl --conf=/etc/masterha/app1.conf

        以下是我们如果正常，我们会看到一下的内容提示：

# masterha_check_repl --conf=/etc/masterha/app1.conf

Wed Nov  1 01:37:46 2017 - [info] Reading default configuration from /etc/masterha_default.cnf..

Wed Nov  1 01:37:46 2017 - [info] Reading application default configuration from /etc/masterha/app1.conf..

Wed Nov  1 01:37:46 2017 - [info] Reading server configuration from /etc/masterha/app1.conf..

Wed Nov  1 01:37:46 2017 - [info] MHA::MasterMonitor version 0.56.

Wed Nov  1 01:37:47 2017 - [info] GTID failover mode = 1

Wed Nov  1 01:37:47 2017 - [info] Dead Servers:

Wed Nov  1 01:37:47 2017 - [info] Alive Servers:

Wed Nov  1 01:37:47 2017 - [info]   192.168.5.102(192.168.5.102:3310)

Wed Nov  1 01:37:47 2017 - [info]   192.168.5.103(192.168.5.103:3310)

Wed Nov  1 01:37:47 2017 - [info] Alive Slaves:

Wed Nov  1 01:37:47 2017 - [info]   192.168.5.103(192.168.5.103:3310)  Version=5.7.19-17-29.22-log (oldest major version between slaves) log-bin:enabled

Wed Nov  1 01:37:47 2017 - [info]     GTID ON

Wed Nov  1 01:37:47 2017 - [info]     Replicating from 192.168.5.102(192.168.5.102:3310)

Wed Nov  1 01:37:47 2017 - [info]     Primary candidate for the new Master (candidate_master is set)

Wed Nov  1 01:37:47 2017 - [info] Current Alive Master: 192.168.5.102(192.168.5.102:3310)

Wed Nov  1 01:37:47 2017 - [info] Checking slave configurations..

Wed Nov  1 01:37:47 2017 - [info]  read_only=1 is not set on slave 192.168.5.103(192.168.5.103:3310).

Wed Nov  1 01:37:47 2017 - [info] Checking replication filtering settings..

Wed Nov  1 01:37:47 2017 - [info]  binlog_do_db= , binlog_ignore_db=

Wed Nov  1 01:37:47 2017 - [info]  Replication filtering check ok.

Wed Nov  1 01:37:47 2017 - [info] GTID (with auto-pos) is supported. Skipping all SSH and Node package checking.

Wed Nov  1 01:37:47 2017 - [info] Checking SSH publickey authentication settings on the current master..

Wed Nov  1 01:37:47 2017 - [info] HealthCheck: SSH to 192.168.5.102 is reachable.

Wed Nov  1 01:37:47 2017 - [info]

192.168.5.102(192.168.5.102:3310) (current master)

+--192.168.5.103(192.168.5.103:3310)

Wed Nov  1 01:37:47 2017 - [info] Checking replication health on 192.168.5.103..

Wed Nov  1 01:37:47 2017 - [info]  ok.

Wed Nov  1 01:37:47 2017 - [info] Checking master_ip_failover_script status:

Wed Nov  1 01:37:47 2017 - [info]   /etc/masterha/master_ip_failover --command=status --ssh_user=root --orig_master_host=192.168.5.102 --orig_master_ip=192.168.5.102 --orig_master_port=3310

Wed Nov  1 01:37:47 2017 - [info]  OK.

Wed Nov  1 01:37:47 2017 - [warning] shutdown_script is not defined.

Wed Nov  1 01:37:47 2017 - [info] Got exit code 0 (Not master dead).

MySQL Replication Health is OK.

 3 检查MHA Manager的状态 #masterha_check_status --conf=/etc/masterha/app1.conf

      以下是我们返回的信息：

#masterha_check_status --conf=/etc/masterha/app1.conf

app1 is stopped(2:NOT_RUNNING).

4 启动mha 监控服务：nohup masterha_manager --conf=/etc/masterha/app1.conf --remove_dead_master_conf --ignore_last_failover < /dev/null > /etc/masterha/app1/manager.log 2>&1 &

 

启动后的结果大概是这样的

#nohup masterha_manager --conf=/etc/masterha/app1.conf --remove_dead_master_conf --ignore_last_failover < /dev/null > /etc/masterha/app1/manager.log 2>&1 &

[1] 9440

 

 这个可以查看一下我们的进程：

#ps aux | grep master

root      1108  0.0  0.1  89544  2148 ?        Ss   Oct30   0:00 /usr/libexec/postfix/master -w

root      9440  2.0  1.0 287396 20200 pts/5    S    01:42   0:00 perl /usr/bin/masterha_manager --conf=/etc/masterha/app1.conf --remove_dead_master_conf --ignore_last_failover

root      9463  0.0  0.0 112660   976 pts/5    R+   01:42   0:00 grep --color=auto master

 

 

 

5：关闭MHA Manage监控   masterha_stop --conf=/etc/masterha/app1.conf

 

 

 

手动切换命令：masterha_master_switch --conf=/etc/masterha/app1.conf --master_state=alive --new_master_host=192.168.5.102 --orig_master_is_new_slave --running_updates_limit=10000 --interactive=0

 

#masterha_master_switch --conf=/etc/masterha/app1.conf --master_state=alive --new_master_host=192.168.5.102 --orig_master_is_new_slave --running_updates_limit=10000 --interactive=0

Wed Nov  1 04:40:30 2017 - [info] MHA::MasterRotate version 0.56.

Wed Nov  1 04:40:30 2017 - [info] Starting online master switch..

Wed Nov  1 04:40:30 2017 - [info]

Wed Nov  1 04:40:30 2017 - [info] * Phase 1: Configuration Check Phase..

Wed Nov  1 04:40:30 2017 - [info]

Wed Nov  1 04:40:30 2017 - [info] Reading default configuration from /etc/masterha_default.cnf..

Wed Nov  1 04:40:30 2017 - [info] Reading application default configuration from /etc/masterha/app1.conf..

Wed Nov  1 04:40:30 2017 - [info] Reading server configuration from /etc/masterha/app1.conf..

Wed Nov  1 04:40:30 2017 - [info] GTID failover mode = 1

Wed Nov  1 04:40:30 2017 - [info] Current Alive Master: 192.168.5.100(192.168.5.100:3310)

Wed Nov  1 04:40:30 2017 - [info] Alive Slaves:

Wed Nov  1 04:40:30 2017 - [info]   192.168.5.102(192.168.5.102:3310)  Version=5.7.19-log (oldest major version between slaves) log-bin:enabled

Wed Nov  1 04:40:30 2017 - [info]     GTID ON

Wed Nov  1 04:40:30 2017 - [info]     Replicating from 192.168.5.100(192.168.5.100:3310)

Wed Nov  1 04:40:30 2017 - [info]     Primary candidate for the new Master (candidate_master is set)

Wed Nov  1 04:40:30 2017 - [info]   192.168.5.103(192.168.5.103:3310)  Version=5.7.19-17-29.22-log (oldest major version between slaves) log-bin:enabled

Wed Nov  1 04:40:30 2017 - [info]     GTID ON

Wed Nov  1 04:40:30 2017 - [info]     Replicating from 192.168.5.100(192.168.5.100:3310)

Wed Nov  1 04:40:30 2017 - [info]     Primary candidate for the new Master (candidate_master is set)

Wed Nov  1 04:40:30 2017 - [info] Executing FLUSH NO_WRITE_TO_BINLOG TABLES. This may take long time..

Wed Nov  1 04:40:30 2017 - [info]  ok.

Wed Nov  1 04:40:30 2017 - [info] Checking MHA is not monitoring or doing failover..

Wed Nov  1 04:40:30 2017 - [error][/usr/share/perl5/vendor_perl/MHA/MasterRotate.pm, ln142] Getting advisory lock failed on the current master. MHA Monitor runs on the current master. Stop MHA Manager/Monitor and try again.

Wed Nov  1 04:40:30 2017 - [error][/usr/share/perl5/vendor_perl/MHA/ManagerUtil.pm, ln177] Got ERROR:  at /usr/bin/masterha_master_switch line 53.

 

       [root@node1 05:57:52 /etc/masterha/app1]

#masterha_master_switch --conf=/etc/masterha/app1.conf --master_state=alive --new_master_host=node2  --running_updates_limit=10000 --interactive=0

Thu Nov  2 05:58:52 2017 - [info] MHA::MasterRotate version 0.56.

Thu Nov  2 05:58:52 2017 - [info] Starting online master switch..

Thu Nov  2 05:58:52 2017 - [info]

Thu Nov  2 05:58:52 2017 - [info] * Phase 1: Configuration Check Phase..

Thu Nov  2 05:58:52 2017 - [info]

Thu Nov  2 05:58:52 2017 - [info] Reading default configuration from /etc/masterha_default.cnf..

Thu Nov  2 05:58:52 2017 - [info] Reading application default configuration from /etc/masterha/app1.conf..

Thu Nov  2 05:58:52 2017 - [info] Reading server configuration from /etc/masterha/app1.conf..

Thu Nov  2 05:58:52 2017 - [info] GTID failover mode = 1

Thu Nov  2 05:58:52 2017 - [info] Current Alive Master: 192.168.5.100(192.168.5.100:3310)

Thu Nov  2 05:58:52 2017 - [info] Alive Slaves:

Thu Nov  2 05:58:52 2017 - [info]   192.168.5.102(192.168.5.102:3310)  Version=5.7.19-log (oldest major version between slaves) log-bin:enabled

Thu Nov  2 05:58:52 2017 - [info]     GTID ON

Thu Nov  2 05:58:52 2017 - [info]     Replicating from 192.168.5.100(192.168.5.100:3310)

Thu Nov  2 05:58:52 2017 - [info]     Primary candidate for the new Master (candidate_master is set)

Thu Nov  2 05:58:52 2017 - [info] Executing FLUSH NO_WRITE_TO_BINLOG TABLES. This may take long time..

Thu Nov  2 05:58:52 2017 - [info]  ok.

Thu Nov  2 05:58:52 2017 - [info] Checking MHA is not monitoring or doing failover..

Thu Nov  2 05:58:52 2017 - [info] Checking replication health on 192.168.5.102..

Thu Nov  2 05:58:52 2017 - [info]  ok.

Thu Nov  2 05:58:52 2017 - [error][/usr/share/perl5/vendor_perl/MHA/ServerManager.pm, ln1218] node2 is not alive!

Thu Nov  2 05:58:52 2017 - [error][/usr/share/perl5/vendor_perl/MHA/MasterRotate.pm, ln232] Failed to get new master!

Thu Nov  2 05:58:52 2017 - [error][/usr/share/perl5/vendor_perl/MHA/ManagerUtil.pm, ln177] Got ERROR:  at /usr/bin/masterha_master_switch line 53.

[root@node1 05:58:52 /etc/masterha/app1]

#masterha_master_switch --conf=/etc/masterha/app1.conf --master_state=alive --new_master_host=node2  

Thu Nov  2 05:59:00 2017 - [info] MHA::MasterRotate version 0.56.

Thu Nov  2 05:59:00 2017 - [info] Starting online master switch..

Thu Nov  2 05:59:00 2017 - [info]

Thu Nov  2 05:59:00 2017 - [info] * Phase 1: Configuration Check Phase..

Thu Nov  2 05:59:00 2017 - [info]

Thu Nov  2 05:59:00 2017 - [info] Reading default configuration from /etc/masterha_default.cnf..

Thu Nov  2 05:59:00 2017 - [info] Reading application default configuration from /etc/masterha/app1.conf..

Thu Nov  2 05:59:00 2017 - [info] Reading server configuration from /etc/masterha/app1.conf..

Thu Nov  2 05:59:00 2017 - [info] GTID failover mode = 1

Thu Nov  2 05:59:00 2017 - [info] Current Alive Master: 192.168.5.100(192.168.5.100:3310)

Thu Nov  2 05:59:00 2017 - [info] Alive Slaves:

Thu Nov  2 05:59:00 2017 - [info]   192.168.5.102(192.168.5.102:3310)  Version=5.7.19-log (oldest major version between slaves) log-bin:enabled

Thu Nov  2 05:59:00 2017 - [info]     GTID ON

Thu Nov  2 05:59:00 2017 - [info]     Replicating from 192.168.5.100(192.168.5.100:3310)

Thu Nov  2 05:59:00 2017 - [info]     Primary candidate for the new Master (candidate_master is set)

It is better to execute FLUSH NO_WRITE_TO_BINLOG TABLES on the master before switching. Is it ok to execute on 192.168.5.100(192.168.5.100:3310)? (YES/no): yes

Thu Nov  2 05:59:03 2017 - [info] Executing FLUSH NO_WRITE_TO_BINLOG TABLES. This may take long time..

Thu Nov  2 05:59:03 2017 - [info]  ok.

Thu Nov  2 05:59:03 2017 - [info] Checking MHA is not monitoring or doing failover..

Thu Nov  2 05:59:03 2017 - [info] Checking replication health on 192.168.5.102..

Thu Nov  2 05:59:03 2017 - [info]  ok.

Thu Nov  2 05:59:03 2017 - [error][/usr/share/perl5/vendor_perl/MHA/ServerManager.pm, ln1218] node2 is not alive!

Thu Nov  2 05:59:03 2017 - [error][/usr/share/perl5/vendor_perl/MHA/MasterRotate.pm, ln232] Failed to get new master!

Thu Nov  2 05:59:03 2017 - [error][/usr/share/perl5/vendor_perl/MHA/ManagerUtil.pm, ln177] Got ERROR:  at /usr/bin/masterha_master_switch line 53.

[root@node1 05:59:03 /etc/masterha/app1]

#masterha_master_switch --conf=/etc/masterha/app1.conf --master_state=alive

Thu Nov  2 05:59:26 2017 - [info] MHA::MasterRotate version 0.56.

Thu Nov  2 05:59:26 2017 - [info] Starting online master switch..

Thu Nov  2 05:59:26 2017 - [info]

Thu Nov  2 05:59:26 2017 - [info] * Phase 1: Configuration Check Phase..

Thu Nov  2 05:59:26 2017 - [info]

Thu Nov  2 05:59:26 2017 - [info] Reading default configuration from /etc/masterha_default.cnf..

Thu Nov  2 05:59:26 2017 - [info] Reading application default configuration from /etc/masterha/app1.conf..

Thu Nov  2 05:59:26 2017 - [info] Reading server configuration from /etc/masterha/app1.conf..

Thu Nov  2 05:59:26 2017 - [info] GTID failover mode = 1

Thu Nov  2 05:59:26 2017 - [info] Current Alive Master: 192.168.5.100(192.168.5.100:3310)

Thu Nov  2 05:59:26 2017 - [info] Alive Slaves:

Thu Nov  2 05:59:26 2017 - [info]   192.168.5.102(192.168.5.102:3310)  Version=5.7.19-log (oldest major version between slaves) log-bin:enabled

Thu Nov  2 05:59:26 2017 - [info]     GTID ON

Thu Nov  2 05:59:26 2017 - [info]     Replicating from 192.168.5.100(192.168.5.100:3310)

Thu Nov  2 05:59:26 2017 - [info]     Primary candidate for the new Master (candidate_master is set)

It is better to execute FLUSH NO_WRITE_TO_BINLOG TABLES on the master before switching. Is it ok to execute on 192.168.5.100(192.168.5.100:3310)? (YES/no): yes

Thu Nov  2 05:59:28 2017 - [info] Executing FLUSH NO_WRITE_TO_BINLOG TABLES. This may take long time..

Thu Nov  2 05:59:28 2017 - [info]  ok.

Thu Nov  2 05:59:28 2017 - [info] Checking MHA is not monitoring or doing failover..

Thu Nov  2 05:59:28 2017 - [info] Checking replication health on 192.168.5.102..

Thu Nov  2 05:59:28 2017 - [info]  ok.

Thu Nov  2 05:59:28 2017 - [info] Searching new master from slaves..

Thu Nov  2 05:59:28 2017 - [info]  Candidate masters from the configuration file:

Thu Nov  2 05:59:28 2017 - [info]   192.168.5.100(192.168.5.100:3310)  Version=5.7.19-log log-bin:enabled

Thu Nov  2 05:59:28 2017 - [info]     GTID ON

Thu Nov  2 05:59:28 2017 - [info]   192.168.5.102(192.168.5.102:3310)  Version=5.7.19-log (oldest major version between slaves) log-bin:enabled

Thu Nov  2 05:59:28 2017 - [info]     GTID ON

Thu Nov  2 05:59:28 2017 - [info]     Replicating from 192.168.5.100(192.168.5.100:3310)

Thu Nov  2 05:59:28 2017 - [info]     Primary candidate for the new Master (candidate_master is set)

Thu Nov  2 05:59:28 2017 - [info]  Non-candidate masters:

Thu Nov  2 05:59:28 2017 - [info]  Searching from candidate_master slaves which have received the latest relay log events..

Thu Nov  2 05:59:28 2017 - [info]

From:

192.168.5.100(192.168.5.100:3310) (current master)

+--192.168.5.102(192.168.5.102:3310)

To:

192.168.5.102(192.168.5.102:3310) (new master)

Starting master switch from 192.168.5.100(192.168.5.100:3310) to 192.168.5.102(192.168.5.102:3310)? (yes/NO): yes

Thu Nov  2 05:59:30 2017 - [info] Checking whether 192.168.5.102(192.168.5.102:3310) is ok for the new master..

Thu Nov  2 05:59:30 2017 - [info]  ok.

Thu Nov  2 05:59:30 2017 - [info] ** Phase 1: Configuration Check Phase completed.

Thu Nov  2 05:59:30 2017 - [info]

Thu Nov  2 05:59:30 2017 - [info] * Phase 2: Rejecting updates Phase..

Thu Nov  2 05:59:30 2017 - [info]

master_ip_online_change_script is not defined. If you do not disable writes on the current master manually, applications keep writing on the current master. Is it ok to proceed? (yes/NO): yes

Thu Nov  2 05:59:33 2017 - [info] Locking all tables on the orig master to reject updates from everybody (including root):

Thu Nov  2 05:59:33 2017 - [info] Executing FLUSH TABLES WITH READ LOCK..

Thu Nov  2 05:59:33 2017 - [info]  ok.

Thu Nov  2 05:59:33 2017 - [info] Orig master binlog:pos is mysql-bin.000013:190.

Thu Nov  2 05:59:33 2017 - [info]  Waiting to execute all relay logs on 192.168.5.102(192.168.5.102:3310)..

Thu Nov  2 05:59:33 2017 - [info]  master_pos_wait(mysql-bin.000013:190) completed on 192.168.5.102(192.168.5.102:3310). Executed 0 events.

Thu Nov  2 05:59:33 2017 - [info]   done.

Thu Nov  2 05:59:33 2017 - [info] Getting new master's binlog name and position..

Thu Nov  2 05:59:33 2017 - [info]  mysql-bin.000011:1167

Thu Nov  2 05:59:33 2017 - [info]  All other slaves should start replication from here. Statement should be: CHANGE MASTER TO MASTER_HOST='192.168.5.102', MASTER_PORT=3310, MASTER_AUTO_POSITION=1, MASTER_USER='rpl', MASTER_PASSWORD='xxx';

Thu Nov  2 05:59:33 2017 - [info]

Thu Nov  2 05:59:33 2017 - [info] * Switching slaves in parallel..

Thu Nov  2 05:59:33 2017 - [info]

Thu Nov  2 05:59:33 2017 - [info] Unlocking all tables on the orig master:

Thu Nov  2 05:59:33 2017 - [info] Executing UNLOCK TABLES..

Thu Nov  2 05:59:33 2017 - [info]  ok.

Thu Nov  2 05:59:33 2017 - [info] All new slave servers switched successfully.

Thu Nov  2 05:59:33 2017 - [info]

Thu Nov  2 05:59:33 2017 - [info] * Phase 5: New master cleanup phase..

Thu Nov  2 05:59:33 2017 - [info]

Thu Nov  2 05:59:33 2017 - [info]  192.168.5.102: Resetting slave info succeeded.

Thu Nov  2 05:59:33 2017 - [info] Switching master to 192.168.5.102(192.168.5.102:3310) completed successfully.

[root@node1 05:59:33 /etc/masterha/app1]

 

几个脚本的：注意脚本的里面有些地方是自己需要去修改的，比如vip

1.绑定 vip的 shell 脚本（当然，也可以自己手动的去绑定或者删除）

vip="192.168.5.99/24"   #####注意我们的子网掩码还是需要加上的

/sbin/ip addr add $vip dev enp0s8

 

2.删除vip 的shell 脚本

vip="192.168.1.99"

/sbin/ip addr del $vip dev enp0s8

 

手动切换的时候：

手动failover，这种场景意味着在业务上没有启用MHA自动切换功能，当主服务器故障时，人工手动调用MHA来进行故障切换操作，具体命令如下：

注意：如果，MHA manager检测到没有dead的server，将报错，并结束failover： 

转载于:https://www.cnblogs.com/gostars/p/8745583.html