共计 57642 个字符，预计需要花费 145 分钟才能阅读完成。
这篇文章主要介绍了 MySQL 高可用架构之 MHA 的原理分析，具有一定借鉴价值，感兴趣的朋友可以参考下，希望大家阅读完这篇文章之后大有收获，下面让丸趣 TV 小编带着大家一起了解一下。
MHA 角色部署
MHA 服务有两种角色，MHA Manager（管理节点）和 MHA Node（数据节点）：
MHA Manager：通常单独部署在一台独立的机器上或者直接部署在其中一台 slave 上（不建议后者），管理多个 master/slave 集群，每个 master/slave 集群称作一个 application；其作用有二：
（1）master 自动切换及故障转移命令运行
（2）其他的帮助脚本运行：手动切换 master；master/slave 状态检测
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。其作用有：
（1）复制主节点的 binlog 数据
（2）对比从节点的中继日志文件
（3）无需停止从节点的 SQL 线程，定时删除中继日志
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，出于机器成本的考虑，淘宝也在该基础上进行了改造，目前淘宝 TMHA 已经支持一主一从。
我们自己使用其实也可以使用 1 主 1 从，但是 master 主机宕机后无法切换，以及无法补全 binlog。master 的 mysqld 进程 crash 后，还是可以切换成功，以及补全 binlog 的。
官方介绍：
https://code.google.com/p/mysql-master-ha/
下图展示了如何通过 MHA Manager 管理多组主从复制。可以将 MHA 工作原理总结为如下：
（1）从宕机崩溃的 master 保存二进制日志事件（binlog events）;
（2）识别含有最新更新的 slave；
（3）应用差异的中继日志（relay log）到其他的 slave；
（4）应用从 master 保存的二进制日志事件（binlog events）；
（5）提升一个 slave 为新的 master；
（6）使其他的 slave 连接新的 master 进行复制；
MHA 组件
(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。
(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)。
(3)、自定义扩展：
-secondary_check_script：通过多条网络路由检测 master 的可用性；
-master_ip_failover_script：更新 application 使用的 masterip； （需要修改）
-shutdown_script：强制关闭 master 节点；
-report_script：发送报告；
-init_conf_load_script：加载初始配置参数；
-master_ip_online_change：更新 master 节点 ip 地址；（需要修改）
MHA 环境准备
OS：CentOS 6.8
MySQL：5.7.18
MHA 软件包：MHA 0.57
角色 ip 地址 主机名 server_id 类型
Master 10.180.2.163 MHA-M1 13306 写入
S1 10.180.2.164 MHA-S1 23306 读 （其实可以一起部署监控，一组 MHA 可以多个监控节点）
S2 10.180.2.165 MHA-S2 33306 读，监控复制组（监控一般不能部署到 master 节点，防止 Master 宕机不能切换）
安装 MHA Node 包
（1）在所有节点安装 MHA node 所需的 perl 模块（DBD:mysql），并下载 MHA 软件包
?
12yum
install
perl-DBD-MySQL -y (可能需要 epel 源)https://mega.nz/#F!G4oRjARB!SWzFS59bUv9VrKwdAeIGVw（MHA0.57）
（2）在所有的节点安装 mha node（包括 Manager 节点）：
tar xf mha4mysql-node-0.57.tar.gz
cd mha4mysql-node-0.57perl Makefile.PLmake make install安装完成将产生文件如下：
[root@MHA-S1 bin]# ll
total 48-r-xr-xr-x 1 root root 16381 Aug 7 14:06 apply_diff_relay_logs-r-xr-xr-x 1 root root 4807 Aug 7 14:06 filter_mysqlbinlog
lrwxrwxrwx 1 root root 26 Aug 8 17:10 mysql - /usr/local/mysql/bin/mysql
lrwxrwxrwx 1 root root 32 Aug 8 17:09 mysqlbinlog - /usr/local/mysql/bin/mysqlbinlog-r-xr-xr-x 1 root root 8261 Aug 7 14:06 purge_relay_logs-rwxr-xr-x 1 root root 314 Aug 8 16:21 purge_relay.sh-r-xr-xr-x 1 root root 7525 Aug 7 14:06 save_binary_logs
[root@MHA-S1 bin]# pwd/usr/local/bin增加系统环境变量：
echo export PATH=\$PATH:/usr/local/bin /etc/profile
source ~/.bash_profile安装 MHA Manager 包
tar xf mha4mysql-node-0.57.tar.gz
cd mha4mysql-node-0.57perl Makefile.PLmake make install安装完成后会在 /usr/local/bin 目录下面生成以下脚本文件
[root@MHA-S2 bin]# pwd/usr/local/bin
[root@MHA-S2 bin]# ll
total 140-r-xr-xr-x 1 root root 16381 Aug 7 14:07 apply_diff_relay_logs-r-xr-xr-x 1 root root 4807 Aug 7 14:07 filter_mysqlbinlog-rwxr-xr-x 1 root root 166 Aug 9 17:18 manager.sh-r-xr-xr-x 1 root root 1995 Aug 7 17:28 masterha_check_repl-r-xr-xr-x 1 root root 1779 Aug 7 17:28 masterha_check_ssh-r-xr-xr-x 1 root root 1865 Aug 7 17:28 masterha_check_status-r-xr-xr-x 1 root root 3201 Aug 7 17:28 masterha_conf_host-r-xr-xr-x 1 root root 2517 Aug 7 17:28 masterha_manager-r-xr-xr-x 1 root root 2165 Aug 7 17:28 masterha_master_monitor-r-xr-xr-x 1 root root 2373 Aug 7 17:28 masterha_master_switch-r-xr-xr-x 1 root root 5171 Aug 7 17:28 masterha_secondary_check-r-xr-xr-x 1 root root 1739 Aug 7 17:28 masterha_stop-rwxr-xr-x 1 root root 2169 Aug 9 10:49 master_ip_failover-rwxr-xr-x 1 root root 3648 Aug 7 17:30 master_ip_failover.old-rwxr-xr-x 1 root root 10369 Aug 12 21:33 master_ip_online_change-rwxr-xr-x 1 root root 9870 Aug 7 17:30 master_ip_online_change.old
lrwxrwxrwx 1 root root 26 Aug 8 17:10 mysql - /usr/local/mysql/bin/mysql
lrwxrwxrwx 1 root root 32 Aug 8 17:09 mysqlbinlog - /usr/local/mysql/bin/mysqlbinlog-rw------- 1 root root 0 Aug 12 20:04 nohup.out-rwxr-xr-x 1 root root 11867 Aug 7 17:30 power_manager-r-xr-xr-x 1 root root 8261 Aug 7 14:07 purge_relay_logs-rwxr-xr-x 1 root root 314 Aug 8 16:20 purge_relay.sh-r-xr-xr-x 1 root root 7525 Aug 7 14:07 save_binary_logs-rwxr-xr-x 1 root root 1360 Aug 7 17:30 send_report复制相关脚本到 /usr/local/bin 目录 (软件包解压缩后就有了，不是必须，因为这些脚本不完整，需要自己修改，这是软件开发着留给我们自己发挥的, 如果开启下面的任何一个脚本对应的参数，而对应这里的脚本又没有修改，则会抛错，自己被坑的很惨)
[root@MHA-S2 scripts]# ll
total 32
-rwxr-xr-x 1 root root 3443 Jan 8 2012 master_ip_failover # 自动切换时 vip 管理的脚本，不是必须，如果我们使用 keepalived 的，我们可以自己编写脚本完成对 vip 的管理，比如监控 mysql，如果 mysql 异常，我们停止 keepalived 就行，这样 vip 就会自动漂移
-rwxr-xr-x 1 root root 9186 Jan 8 2012 master_ip_online_change #在线切换时 vip 的管理，不是必须，同样可以可以自行编写简单的 shell 完成
-rwxr-xr-x 1 root root 11867 Jan 8 2012 power_manager #故障发生后关闭主机的脚本，不是必须
-rwxr-xr-x 1 root root 1360 Jan 8 2012 send_report #因故障切换后发送报警的脚本，不是必须，可自行编写简单的 shell 完成。
[root@MHA-S2 scripts]# cp * /usr/local/bin/
配置 SSH 登录无密码验证
ssh-keygenssh-copy-id root@xxx (XXX 请包括自己，要不然后面 check-ssh 那步要杯具的)搭建主从复制环境
详解之前双主复制环境搭建文档
保证两台 Slave 都搭建成功
Slave_IO_Running: Yes
Slave_SQL_Running: Yes 两台 slave 服务器设置 read_only（从库对外提供读服务，只所以没有写进配置文件，是因为随时 slave 会提升为 master）root@localhost:mysql3306.sock [(none)] set global read_only=1创建监控用户（在 master 上执行）
grant all privileges on *.* to root@ % identified by 123456flush privileges;至此，复制搭建完毕，后面配置 MHA
MHA 环境配置
（1）创建 MHA 工作目录
mkdir -p /etc/mha修改 app1.cnf 配置文件，修改后的文件内容如下：
[root@MHA-S2 ~]# /etc/mha/=/var/log/masterha/app1/=/var/log/masterha/=/data/mysql//=/usr/local/bin/=/usr/local/bin/===/===/usr/local/bin/=/usr/local/bin/masterha_secondary_check -s MHA-S1 -s MHA-===MHA-==// 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，即使这个主库不是集群中事件最新的 slave
check_repl_delay=// 默认情况下如果一个 slave 落后 master 100M 的 relay logs 的话，MHA 将不会选择该 slave 作为一个新的 master，因为对于这个 slave 的恢复需要花费很长时间，通过设置 check_repl_delay=0,MHA 触发切换在选择一个新的 master 的时候将会忽略复制延时，这个参数对于设置了 candidate_master= 1 的主机非常有用，因为这个候选主在切换的过程中一定是新的 master=MHA-S1
port==MHA-=（2）设置 relay log 的清除方式（在每个 slave 节点上）：
?
1 set global relay_log_purge=0
注意：
MHA 在发生切换的过程中，从库的恢复过程中依赖于 relay log 的相关信息，所以这里要将 relay log 的自动清除设置为 OFF，采用手动清除 relay log 的方式。在默认情况下，从服务器上的中继日志会在 SQL 线程执行完毕后被自动删除。但是在 MHA 环境中，这些中继日志在恢复其他从服务器时可能会被用到，因此需要禁用中继日志的自动删除功能。定期清除中继日志需要考虑到复制延时的问题。在 ext3 的文件系统下，删除大的文件需要一定的时间，会导致严重的复制延时。为了避免复制延时，需要暂时为中继日志创建硬链接，因为在 linux 系统中通过硬链接删除大文件速度会很快。（在 mysql 数据库中，删除大表时，通常也采用建立硬链接的方式）
MHA 节点中包含了 pure_relay_logs 命令工具，它可以为中继日志创建硬链接，执行 SET GLOBAL relay_log_purge=1, 等待几秒钟以便 SQL 线程切换到新的中继日志，再执行 SET GLOBAL relay_log_purge=0。
pure_relay_logs 脚本参数如下所示：
--user mysql 用户名 --password mysql 密码 --port 端口号 --workdir 指定创建 relay log 的硬链接的位置，默认是 /var/tmp，由于系统不同分区创建硬链接文件会失败，故需要执行硬链接具体位置，成功执行脚本后，硬链接的中继日志文件被删除 --disable_relay_log_purge 默认情况下，如果 relay_log_purge=1，脚本会什么都不清理，自动退出，通过设定这个参数，当 relay_log_purge= 1 的情况下会将 relay_log_purge 设置为 0。清理 relay log 之后，最后将参数设置为 OFF。（3）设置定期清理 relay 脚本 (例如每天一次，所有服务器）
[root@MHA-S2 bin]# purge_relay.!/bin/====== [ ! -
$log_dir ---user=$user --password=$ --disable_relay_log_purge --port=$port --workdir=$work_dir $log_dir/purge_relay_logs.log 添加到 crontab
[root@MHA-S2 bin]# crontab -l0 4 * * * /bin/bash /root/purge_relay_log.sh可以手工执行以下是否会报错
检查 SSH 配置
[root@MHA-S2 bin]# masterha_check_ssh --conf=/etc/mha/app1.cnf
Mon Aug 14 18:07:02 2017 - [warning] Global configuration file /etc/masterha_default.cnf not found. Skipping.
Mon Aug 14 18:07:02 2017 - [info] Reading application default configuration from /etc/mha/app1.cnf..
Mon Aug 14 18:07:02 2017 - [info] Reading server configuration from /etc/mha/app1.cnf..
Mon Aug 14 18:07:02 2017 - [info] Starting SSH connection tests..
Mon Aug 14 18:07:03 2017 - [debug]
Mon Aug 14 18:07:02 2017 - [debug] Connecting via SSH from root@MHA-M1(10.180.2.163:22) to root@MHA-S1(10.180.2.164:22)..
Mon Aug 14 18:07:02 2017 - [debug] ok.
Mon Aug 14 18:07:02 2017 - [debug] Connecting via SSH from root@MHA-M1(10.180.2.163:22) to root@MHA-S2(10.180.2.165:22)..
Mon Aug 14 18:07:03 2017 - [debug] ok.
Mon Aug 14 18:07:03 2017 - [debug]
Mon Aug 14 18:07:03 2017 - [debug] Connecting via SSH from root@MHA-S1(10.180.2.164:22) to root@MHA-M1(10.180.2.163:22)..
Mon Aug 14 18:07:03 2017 - [debug] ok.
Mon Aug 14 18:07:03 2017 - [debug] Connecting via SSH from root@MHA-S1(10.180.2.164:22) to root@MHA-S2(10.180.2.165:22)..
Mon Aug 14 18:07:03 2017 - [debug] ok.
Mon Aug 14 18:07:04 2017 - [debug]
Mon Aug 14 18:07:03 2017 - [debug] Connecting via SSH from root@MHA-S2(10.180.2.165:22) to root@MHA-M1(10.180.2.163:22)..
Mon Aug 14 18:07:03 2017 - [debug] ok.
Mon Aug 14 18:07:04 2017 - [debug] Connecting via SSH from root@MHA-S2(10.180.2.165:22) to root@MHA-S1(10.180.2.164:22)..
Mon Aug 14 18:07:04 2017 - [debug] ok.
Mon Aug 14 18:07:04 2017 - [info] All SSH connection tests passed successfully.检查整个复制环境状况
发现有报错，
Tue Aug 8 17:46:31 2017 - [info] Checking master_ip_failover_script status:
Tue Aug 8 17:46:31 2017 - [info] /usr/local/bin/master_ip_failover --command=status --ssh_user=root --orig_master_host=MHA-M1 --orig_master_ip=10.180.2.163 --orig_master_port=3306 Bareword FIXME_xxx not allowed while strict subs in use at /usr/local/bin/master_ip_failover line 93.
Execution of /usr/local/bin/master_ip_failover aborted due to compilation errors.
Tue Aug 8 17:46:31 2017 - [error][/usr/local/share/perl5/MHA/MasterMonitor.pm, ln229] Failed to get master_ip_failover_script status with return code 255:0.Tue Aug 8 17:46:31 2017 - [error][/usr/local/share/perl5/MHA/MasterMonitor.pm, ln427] Error happened on checking configurations. at /usr/local/bin/masterha_check_repl line 48Tue Aug 8 17:46:31 2017 - [error][/usr/local/share/perl5/MHA/MasterMonitor.pm, ln525] Error happened on monitoring servers.
Tue Aug 8 17:46:31 2017 - [info] Got exit code 1 (Not master dead).原来 Failover 两种方式：一种是虚拟 IP 地址，一种是全局配置文件。MHA 并没有限定使用哪一种方式，而是让用户自己选择，虚拟 IP 地址的方式会牵扯到其它的软件, 比如 keepalive 软件，而且还要修改脚本 master_ip_failover。这里先把 app1.cnf 里面 master_ip_failover_script= /usr/local/bin/master_ip_failover 这个选项屏蔽才可以通过。
#master_ip_failover_script= /usr/local/bin/master_ip_failover
Tue Aug 8 17:49:40 2017 - [info] Got exit code 0 (Not master dead).
MySQL Replication Health is OK.检查 MHA Manager 的状态
[root@MHA-S2 mha]# masterha_check_status --conf=/etc/mha/app1.cnf
app1 is stopped(2:NOT_RUNNING).手动启动
[root@MHA-S2 mha]# nohup masterha_manager --conf=/etc/mha/app1.cnf --remove_dead_master_conf --ignore_last_failover /dev/null /var/log/masterha/app1/manager.log 2 1 [1] 16774[root@MHA-S2 mha]# ps -ef|grep masterha
root 16774 15297 4 17:52 pts/3 00:00:00 perl /usr/local/bin/masterha_manager --conf=/etc/mha/app1.cnf --remove_dead_master_conf --ignore_last_failover
[root@MHA-S2 mha]# masterha_check_status --conf=/etc/mha/app1.cnf
app1 (pid:16774) is running(0:PING_OK), master:MHA-M1–remove_dead_master_conf 该参数代表当发生主从切换后，老的主库的 ip 将会从配置文件中移除。（如果发生异常切换之后修复了旧的 master，要加进去新的 MHA 的话，必须记得 app1.cnf 回补 server1 的信息）
–manger_log 日志存放位置
–ignore_last_failover 在缺省情况下，如果 MHA 检测到连续发生宕机，且两次宕机间隔不足 8 小时的话，则不会进行 Failover，之所以这样限制是为了避免 ping-pong 效应。该参数代表忽略上次 MHA 触发切换产生的文件，默认情况下，MHA 发生切换后会在日志目录，也就是上面我设置的 /data 产生 app1.failover.complete 文件，下次再次切换的时候如果发现该目录下存在该文件将不允许触发切换，除非在第一次切换后收到删除该文件，为了方便，这里设置为 –ignore_last_failover。
检查启动日志
[root@MHA-S2 app1]# vi manager.log
Tue Aug 8 17:52:37 2017 - [warning] Global configuration file /etc/masterha_default.cnf not found. Skipping.
Tue Aug 8 17:52:37 2017 - [info] Reading application default configuration from /etc/mha/app1.cnf..
Tue Aug 8 17:52:37 2017 - [info] Reading server configuration from /etc/mha/app1.cnf..
Tue Aug 8 17:52:37 2017 - [info] MHA::MasterMonitor version 0.57.
Tue Aug 8 17:52:38 2017 - [info] GTID failover mode = 1Tue Aug 8 17:52:38 2017 - [info] Dead Servers:
Tue Aug 8 17:52:38 2017 - [info] Alive Servers:
Tue Aug 8 17:52:38 2017 - [info] MHA-M1(10.180.2.163:3306)
Tue Aug 8 17:52:38 2017 - [info] MHA-S1(10.180.2.164:3306)
Tue Aug 8 17:52:38 2017 - [info] MHA-S2(10.180.2.165:3306)
Tue Aug 8 17:52:38 2017 - [info] Alive Slaves:
Tue Aug 8 17:52:38 2017 - [info] MHA-S1(10.180.2.164:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Tue Aug 8 17:52:38 2017 - [info] GTID ON
Tue Aug 8 17:52:38 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Tue Aug 8 17:52:38 2017 - [info] Primary candidate for the new Master (candidate_master is set)
Tue Aug 8 17:52:38 2017 - [info] MHA-S2(10.180.2.165:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Tue Aug 8 17:52:38 2017 - [info] GTID ON
Tue Aug 8 17:52:38 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Tue Aug 8 17:52:38 2017 - [info] Current Alive Master: MHA-M1(10.180.2.163:3306)
Tue Aug 8 17:52:38 2017 - [info] Checking slave configurations..
Tue Aug 8 17:52:38 2017 - [info] Checking replication filtering settings..
Tue Aug 8 17:52:38 2017 - [info] binlog_do_db= , binlog_ignore_db=Tue Aug 8 17:52:38 2017 - [info] Replication filtering check ok.
Tue Aug 8 17:52:38 2017 - [info] GTID (with auto-pos) is supported. Skipping all SSH and Node package checking.
Tue Aug 8 17:52:38 2017 - [info] Checking SSH publickey authentication settings on the current master..
Tue Aug 8 17:52:38 2017 - [info] HealthCheck: SSH to MHA-M1 is reachable.
Tue Aug 8 17:52:38 2017 - [info]
MHA-M1(10.180.2.163:3306) (current master) +--MHA-S1(10.180.2.164:3306) +--MHA-S2(10.180.2.165:3306)
Tue Aug 8 17:52:38 2017 - [warning] master_ip_failover_script is not defined.
Tue Aug 8 17:52:38 2017 - [warning] shutdown_script is not defined.
Tue Aug 8 17:52:38 2017 - [info] Set master ping interval 1 seconds.
Tue Aug 8 17:52:38 2017 - [info] Set secondary check script: /usr/local/bin/masterha_secondary_check -s MHA-S1 -s MHA-S2
Tue Aug 8 17:52:38 2017 - [info] Starting ping health check on MHA-M1(10.180.2.163:3306)..Tue Aug 8 17:52:38 2017 - [info] Ping(SELECT) succeeded, waiting until MySQL doesn t respond..
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）。
这里仅演示使用脚本管理 VIP 的方式， 修改 master_ip_failover 脚本，使用脚本管理 VIP
[root@MHA-M1 ~]# /sbin/ifconfig eth2:1 10.180.2.168/19脚本：
[root@MHA-S2 bin]# cat master_ip_failover
#!/usr/bin/env perluse 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 = 10.180.2.168/19
my $key = 1
my $ssh_start_vip = /sbin/ifconfig eth2:$key $vip
my $ssh_stop_vip = /sbin/ifconfig eth2:$key down
GetOptions( command=s = \$command, ssh_user=s = \$ssh_user, orig_master_host=s = \$orig_master_host, orig_master_ip=s = \$orig_master_ip, orig_master_port=i = \$orig_master_port, new_master_host=s = \$new_master_host, new_master_ip=s = \$new_master_ip, new_master_port=i = \$new_master_port,
exit main();
sub main { print \n\nIN SCRIPT TEST====$ssh_stop_vip==$ssh_start_vip===\n\n if ( $command eq stop || $command eq stopssh ) {
my $exit_code = 1;
eval { print Disabling the VIP on old master: $orig_master_host \n stop_vip();
$exit_code = 0;
}; if ($@) {
warn Got Error: $@\n
exit $exit_code;
}
exit $exit_code;
}
elsif ( $command eq start ) {
my $exit_code = 10;
eval { print Enabling the VIP - $vip on the new master - $new_master_host \n start_vip();
$exit_code = 0;
}; if ($@) {
warn $@;
exit $exit_code;
}
exit $exit_code;
}
elsif ( $command eq status ) {
print Checking the Status of the script.. OK \n
exit 0;
} else { usage();
exit 1;
}
sub start_vip() { `ssh $ssh_user\@$new_master_host \ $ssh_start_vip \}
sub stop_vip() { return 0 unless ($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}
在 app1.cnf 文件中取消刚刚对 master_ip_online_failover 的注释并测试：
再次检查 MHA check
[root@MHA-S2 bin]# masterha_check_repl --conf=/etc/mha/app1.cnf
Wed Aug 9 10:49:42 2017 - [warning] Global configuration file /etc/masterha_default.cnf not found. Skipping.
Wed Aug 9 10:49:42 2017 - [info] Reading application default configuration from /etc/mha/app1.cnf..
Wed Aug 9 10:49:42 2017 - [info] Reading server configuration from /etc/mha/app1.cnf..
Wed Aug 9 10:49:42 2017 - [info] MHA::MasterMonitor version 0.57.
Wed Aug 9 10:49:43 2017 - [info] GTID failover mode = 1Wed Aug 9 10:49:43 2017 - [info] Dead Servers:
Wed Aug 9 10:49:43 2017 - [info] Alive Servers:
Wed Aug 9 10:49:43 2017 - [info] MHA-M1(10.180.2.163:3306)
Wed Aug 9 10:49:43 2017 - [info] MHA-S1(10.180.2.164:3306)
Wed Aug 9 10:49:43 2017 - [info] MHA-S2(10.180.2.165:3306)
Wed Aug 9 10:49:43 2017 - [info] Alive Slaves:
Wed Aug 9 10:49:43 2017 - [info] MHA-S1(10.180.2.164:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Wed Aug 9 10:49:43 2017 - [info] GTID ON
Wed Aug 9 10:49:43 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Wed Aug 9 10:49:43 2017 - [info] Primary candidate for the new Master (candidate_master is set)
Wed Aug 9 10:49:43 2017 - [info] MHA-S2(10.180.2.165:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Wed Aug 9 10:49:43 2017 - [info] GTID ON
Wed Aug 9 10:49:43 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Wed Aug 9 10:49:43 2017 - [info] Current Alive Master: MHA-M1(10.180.2.163:3306)
Wed Aug 9 10:49:43 2017 - [info] Checking slave configurations..
Wed Aug 9 10:49:43 2017 - [info] Checking replication filtering settings..
Wed Aug 9 10:49:43 2017 - [info] binlog_do_db= , binlog_ignore_db= Wed Aug 9 10:49:43 2017 - [info] Replication filtering check ok.
Wed Aug 9 10:49:43 2017 - [info] GTID (with auto-pos) is supported. Skipping all SSH and Node package checking.
Wed Aug 9 10:49:43 2017 - [info] Checking SSH publickey authentication settings on the current master..
Wed Aug 9 10:49:43 2017 - [info] HealthCheck: SSH to MHA-M1 is reachable.
Wed Aug 9 10:49:43 2017 - [info]
MHA-M1(10.180.2.163:3306) (current master) +--MHA-S1(10.180.2.164:3306) +--MHA-S2(10.180.2.165:3306)
Wed Aug 9 10:49:43 2017 - [info] Checking replication health on MHA-S1..
Wed Aug 9 10:49:43 2017 - [info] ok.
Wed Aug 9 10:49:43 2017 - [info] Checking replication health on MHA-S2..
Wed Aug 9 10:49:43 2017 - [info] ok.
Wed Aug 9 10:49:43 2017 - [info] Checking master_ip_failover_script status:
Wed Aug 9 10:49:43 2017 - [info] /usr/local/bin/master_ip_failover --command=status --ssh_user=root --orig_master_host=MHA-M1 --orig_master_ip=10.180.2.163 --orig_master_port=3306
IN SCRIPT TEST====/sbin/ifconfig eth2:1 down==/sbin/ifconfig eth2:1 10.180.2.168/19===
Checking the Status of the script.. OK
Wed Aug 9 10:49:43 2017 – [info] OK.
Wed Aug 9 10:49:43 2017 – [warning] shutdown_script is not defined.
Wed Aug 9 10:49:43 2017 – [info] Got exit code 0 (Not master dead).
MySQL Replication Health is OK.
以上就是 MHA 安装配置的全过程，以下进行简单的测试。
（1）failover 测试
手动 kill 了 master 上面的 mysqld 进程，查看切换状态
[root@MHA-S2 tmp]# more manager.log
Wed Aug 9 17:47:11 2017 - [warning] Got error on MySQL select ping: 2006 (MySQL server has gone away)
Wed Aug 9 17:47:11 2017 - [info] Executing secondary network check script: /usr/local/bin/masterha_secondary_check -s MHA-S1 -s MHA-S2 --user=root --master_host=MHA-M1 --master_ip=10.180.2.163 --master_port=3306 --master_user=root --master_password=123456 --ping_type=SELECT
Wed Aug 9 17:47:11 2017 - [info] Executing SSH check script: exit 0Wed Aug 9 17:47:11 2017 - [info] HealthCheck: SSH to MHA-M1 is reachable.
Monitoring server MHA-S1 is reachable, Master is not reachable from MHA-S1. OK.
Monitoring server MHA-S2 is reachable, Master is not reachable from MHA-S2. OK.
Wed Aug 9 17:47:11 2017 - [info] Master is not reachable from all other monitoring servers. Failover should start.
Wed Aug 9 17:47:12 2017 - [warning] Got error on MySQL connect: 2013 (Lost connection to MySQL server at reading initial communication packet , system error: 111)
Wed Aug 9 17:47:12 2017 - [warning] Connection failed 2 time(s)..
Wed Aug 9 17:47:13 2017 - [warning] Got error on MySQL connect: 2013 (Lost connection to MySQL server at reading initial communication packet , system error: 111)
Wed Aug 9 17:47:13 2017 - [warning] Connection failed 3 time(s)..
Wed Aug 9 17:47:14 2017 - [warning] Got error on MySQL connect: 2013 (Lost connection to MySQL server at reading initial communication packet , system error: 111)
Wed Aug 9 17:47:14 2017 - [warning] Connection failed 4 time(s)..
Wed Aug 9 17:47:14 2017 - [warning] Master is not reachable from health checker!Wed Aug 9 17:47:14 2017 - [warning] Master MHA-M1(10.180.2.163:3306) is not reachable!Wed Aug 9 17:47:14 2017 - [warning] SSH is reachable.
Wed Aug 9 17:47:14 2017 - [info] Connecting to a master server failed. Reading configuration file /etc/masterha_default.cnf and /etc/mha/app1.cnf again, and tryin
g to connect to all servers to check server status..
Wed Aug 9 17:47:14 2017 - [warning] Global configuration file /etc/masterha_default.cnf not found. Skipping.
Wed Aug 9 17:47:14 2017 - [info] Reading application default configuration from /etc/mha/app1.cnf..
Wed Aug 9 17:47:14 2017 - [info] Reading server configuration from /etc/mha/app1.cnf..
Wed Aug 9 17:47:14 2017 - [info] GTID failover mode = 1Wed Aug 9 17:47:14 2017 - [info] Dead Servers:
Wed Aug 9 17:47:14 2017 - [info] MHA-M1(10.180.2.163:3306)
Wed Aug 9 17:47:14 2017 - [info] Alive Servers:
Wed Aug 9 17:47:14 2017 - [info] MHA-S1(10.180.2.164:3306)
Wed Aug 9 17:47:14 2017 - [info] MHA-S2(10.180.2.165:3306)
Wed Aug 9 17:47:14 2017 - [info] Alive Slaves:
Wed Aug 9 17:47:14 2017 - [info] MHA-S1(10.180.2.164:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Wed Aug 9 17:47:14 2017 - [info] GTID ON
Wed Aug 9 17:47:14 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Wed Aug 9 17:47:14 2017 - [info] Primary candidate for the new Master (candidate_master is set)
Wed Aug 9 17:47:14 2017 - [info] MHA-S2(10.180.2.165:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Wed Aug 9 17:47:14 2017 - [info] GTID ON
Wed Aug 9 17:47:14 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Wed Aug 9 17:47:14 2017 - [info] Checking slave configurations..
Wed Aug 9 17:47:14 2017 - [info] Checking replication filtering settings..
Wed Aug 9 17:47:14 2017 - [info] Replication filtering check ok.
Wed Aug 9 17:47:14 2017 - [info] Master is down!Wed Aug 9 17:47:14 2017 - [info] Terminating monitoring script.
Wed Aug 9 17:47:14 2017 - [info] Got exit code 20 (Master dead).
Wed Aug 9 17:47:14 2017 - [info] MHA::MasterFailover version 0.57.
Wed Aug 9 17:47:14 2017 - [info] Starting master failover.
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] * Phase 1: Configuration Check Phase..
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] GTID failover mode = 1Wed Aug 9 17:47:14 2017 - [info] Dead Servers:
Wed Aug 9 17:47:14 2017 - [info] MHA-M1(10.180.2.163:3306)
Wed Aug 9 17:47:14 2017 - [info] Checking master reachability via MySQL(double check)...
Wed Aug 9 17:47:14 2017 - [info] ok.
Wed Aug 9 17:47:14 2017 - [info] Alive Servers:
Wed Aug 9 17:47:14 2017 - [info] MHA-S1(10.180.2.164:3306)
Wed Aug 9 17:47:14 2017 - [info] MHA-S2(10.180.2.165:3306)
Wed Aug 9 17:47:14 2017 - [info] Alive Slaves:
Wed Aug 9 17:47:14 2017 - [info] MHA-S1(10.180.2.164:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Wed Aug 9 17:47:14 2017 - [info] GTID ON
Wed Aug 9 17:47:14 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Wed Aug 9 17:47:14 2017 - [info] Primary candidate for the new Master (candidate_master is set)
Wed Aug 9 17:47:14 2017 - [info] MHA-S2(10.180.2.165:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Wed Aug 9 17:47:14 2017 - [info] GTID ON
Wed Aug 9 17:47:14 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Wed Aug 9 17:47:14 2017 - [info] Starting GTID based failover.
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] ** Phase 1: Configuration Check Phase completed.
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] * Phase 2: Dead Master Shutdown Phase..
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] Forcing shutdown so that applications never connect to the current master..
Wed Aug 9 17:47:14 2017 - [info] Executing master IP deactivation script:
Wed Aug 9 17:47:14 2017 - [info] /usr/local/bin/master_ip_failover --orig_master_host=MHA-M1 --orig_master_ip=10.180.2.163 --orig_master_port=3306 --command=stopssh --ssh_user=root
IN SCRIPT TEST====/sbin/ifconfig eth2:1 down==/sbin/ifconfig eth2:1 10.180.2.168/24===Disabling the VIP on old master: MHA-M1
SIOCSIFFLAGS: Cannot assign requested address
Wed Aug 9 17:47:14 2017 - [info] done.
Wed Aug 9 17:47:14 2017 - [warning] shutdown_script is not set. Skipping explicit shutting down of the dead master.
Wed Aug 9 17:47:14 2017 - [info] * Phase 2: Dead Master Shutdown Phase completed.
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] * Phase 3: Master Recovery Phase..
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] * Phase 3.1: Getting Latest Slaves Phase..
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] The latest binary log file/position on all slaves is 3306-binlog.000003:194Wed Aug 9 17:47:14 2017 - [info] Retrieved Gtid Set: a5757eae-7981-11e7-82c7-005056b662d3:6-32210Wed Aug 9 17:47:14 2017 - [info] Latest slaves (Slaves that received relay log files to the latest):
Wed Aug 9 17:47:14 2017 - [info] MHA-S1(10.180.2.164:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Wed Aug 9 17:47:14 2017 - [info] GTID ON
Wed Aug 9 17:47:14 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Wed Aug 9 17:47:14 2017 - [info] Primary candidate for the new Master (candidate_master is set)
Wed Aug 9 17:47:14 2017 - [info] MHA-S2(10.180.2.165:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Wed Aug 9 17:47:14 2017 - [info] GTID ON
Wed Aug 9 17:47:14 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Wed Aug 9 17:47:14 2017 - [info] The oldest binary log file/position on all slaves is 3306-binlog.000003:194Wed Aug 9 17:47:14 2017 - [info] Retrieved Gtid Set: a5757eae-7981-11e7-82c7-005056b662d3:6-32210Wed Aug 9 17:47:14 2017 - [info] Oldest slaves:
Wed Aug 9 17:47:14 2017 - [info] MHA-S1(10.180.2.164:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Wed Aug 9 17:47:14 2017 - [info] GTID ON
Wed Aug 9 17:47:14 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Wed Aug 9 17:47:14 2017 - [info] Primary candidate for the new Master (candidate_master is set)
Wed Aug 9 17:47:14 2017 - [info] MHA-S2(10.180.2.165:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Wed Aug 9 17:47:14 2017 - [info] GTID ON
Wed Aug 9 17:47:14 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] * Phase 3.3: Determining New Master Phase..
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] Searching new master from slaves..
Wed Aug 9 17:47:14 2017 - [info] Candidate masters from the configuration file:
Wed Aug 9 17:47:14 2017 - [info] MHA-S1(10.180.2.164:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Wed Aug 9 17:47:14 2017 - [info] GTID ON
Wed Aug 9 17:47:14 2017 - [info] Replicating from MHA-M1(10.180.2.163:3306)
Wed Aug 9 17:47:14 2017 - [info] Primary candidate for the new Master (candidate_master is set)
Wed Aug 9 17:47:14 2017 - [info] Non-candidate masters:
Wed Aug 9 17:47:14 2017 - [info] Searching from candidate_master slaves which have received the latest relay log events..
Wed Aug 9 17:47:14 2017 - [info] New master is MHA-S1(10.180.2.164:3306)
Wed Aug 9 17:47:14 2017 - [info] Starting master failover..
Wed Aug 9 17:47:14 2017 - [info]
From:
MHA-M1(10.180.2.163:3306) (current master) +--MHA-S1(10.180.2.164:3306) +--MHA-S2(10.180.2.165:3306)
MHA-S1(10.180.2.164:3306) (new master) +--MHA-S2(10.180.2.165:3306)
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] * Phase 3.3: New Master Recovery Phase..
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] Waiting all logs to be applied..
Wed Aug 9 17:47:14 2017 - [info] done.
Wed Aug 9 17:47:14 2017 - [info] Getting new master s binlog name and position..Wed Aug 9 17:47:14 2017 - [info] 3306-binlog.000003:61944788Wed Aug 9 17:47:14 2017 - [info] All other slaves should start replication from here. Statement should be: CHANGE MASTER TO MASTER_HOST= MHA-S1 or 10.180.2.164 , MASTER_PORT=3306, MASTER_AUTO_POSITION=1, MAS
TER_USER= repl , MASTER_PASSWORD= xxx
Wed Aug 9 17:47:14 2017 - [info] Master Recovery succeeded. File:Pos:Exec_Gtid_Set: 3306-binlog.000003, 61944788, 1c2dc99f-7b57-11e7-a280-005056b665cb:1-2,
a5757eae-7981-11e7-82c7-005056b662d3:1-32210Wed Aug 9 17:47:14 2017 - [info] Executing master IP activate script:
Wed Aug 9 17:47:14 2017 - [info] /usr/local/bin/master_ip_failover --command=start --ssh_user=root --orig_master_host=MHA-M1 --orig_master_ip=10.180.2.163 --orig_master_port=3306 --new_master_host=MHA-S1 --new_master_ip=10.180.2.164 --new_master_port=3306 --new_master_user= root --new_master_password=xxx
Unknown option: new_master_user
Unknown option: new_master_password
IN SCRIPT TEST====/sbin/ifconfig eth2:1 down==/sbin/ifconfig eth2:1 10.180.2.168/24===Enabling the VIP - 10.180.2.168/24 on the new master - MHA-S1
Wed Aug 9 17:47:14 2017 - [info] OK.
Wed Aug 9 17:47:14 2017 - [info] Setting read_only=0 on MHA-S1(10.180.2.164:3306)..
Wed Aug 9 17:47:14 2017 - [info] ok.
Wed Aug 9 17:47:14 2017 - [info] ** Finished master recovery successfully.
Wed Aug 9 17:47:14 2017 - [info] * Phase 3: Master Recovery Phase completed.
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] * Phase 4: Slaves Recovery Phase..
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] * Phase 4.1: Starting Slaves in parallel..
Wed Aug 9 17:47:14 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] -- Slave recovery on host MHA-S2(10.180.2.165:3306) started, pid: 18757. Check tmp log /var/log/masterha/app1/MHA-S2_3306_20170809174714.log if it takes time..
Wed Aug 9 17:47:15 2017 - [info]
Wed Aug 9 17:47:15 2017 - [info] Log messages from MHA-S2 ...
Wed Aug 9 17:47:15 2017 - [info]
Wed Aug 9 17:47:14 2017 - [info] Resetting slave MHA-S2(10.180.2.165:3306) and starting replication from the new master MHA-S1(10.180.2.164:3306)..
Wed Aug 9 17:47:14 2017 - [info] Executed CHANGE MASTER.
Wed Aug 9 17:47:15 2017 - [info] Slave started.
Wed Aug 9 17:47:15 2017 - [info] gtid_wait(1c2dc99f-7b57-11e7-a280-005056b665cb:1-2,
a5757eae-7981-11e7-82c7-005056b662d3:1-32210) completed on MHA-S2(10.180.2.165:3306). Executed 0 events.
Wed Aug 9 17:47:15 2017 - [info] End of log messages from MHA-S2.
Wed Aug 9 17:47:15 2017 - [info] -- Slave on host MHA-S2(10.180.2.165:3306) started.
Wed Aug 9 17:47:15 2017 - [info] All new slave servers recovered successfully.
Wed Aug 9 17:47:15 2017 - [info]
Wed Aug 9 17:47:15 2017 - [info] * Phase 5: New master cleanup phase..
Wed Aug 9 17:47:15 2017 - [info]
Wed Aug 9 17:47:15 2017 - [info] Resetting slave info on the new master..
Wed Aug 9 17:47:15 2017 - [info] MHA-S1: Resetting slave info succeeded.
Wed Aug 9 17:47:15 2017 - [info] Master failover to MHA-S1(10.180.2.164:3306) completed successfully.
Wed Aug 9 17:47:15 2017 - [info] Deleted server1 entry from /etc/mha/app1.cnf .
Wed Aug 9 17:47:15 2017 - [info]
----- Failover Report -----app1: MySQL Master failover MHA-M1(10.180.2.163:3306) to MHA-S1(10.180.2.164:3306) succeeded
Master MHA-M1(10.180.2.163:3306) is down!Check MHA Manager logs at MHA-S2:/var/log/masterha/app1/manager.log for details.
Started automated(non-interactive) failover.
Invalidated master IP address on MHA-M1(10.180.2.163:3306)
Selected MHA-S1(10.180.2.164:3306) as a new master.
MHA-S1(10.180.2.164:3306): OK: Applying all logs succeeded.
MHA-S1(10.180.2.164:3306): OK: Activated master IP address.
MHA-S2(10.180.2.165:3306): OK: Slave started, replicating from MHA-S1(10.180.2.164:3306)
MHA-S1(10.180.2.164:3306): Resetting slave info succeeded.
Master failover to MHA-S1(10.180.2.164:3306) completed successfully.
Wed Aug 9 17:47:15 2017 - [info] Sending mail..
Unknown option: conf
以上是切换的全日志过程，我们可以看到 MHA 切换主要经历以下步骤：
1. 配置文件检查阶段，这个阶段会检查整个集群配置文件配置
2. 宕机的 master 处理，这个阶段包括虚拟 ip 摘除操作，主机关机操作（这个我这里还没有实现，需要研究）
3. 复制 dead maste 和最新 slave 相差的 relay log，并保存到 MHA Manger 具体的目录下
4. 识别含有最新更新的 slave
5. 应用从 master 保存的二进制日志事件（binlog events）
6. 提升一个 slave 为新的 master 进行复制
7. 使其他的 slave 连接新的 master 进行复制
注意：
1. 切换完之后你会发现 MHA Manager 监控程序会自动死掉，官网有如下解释和解决方式：
Running MHA Manager from daemontoolsCurrently MHA Manager process does not run as a daemon. If failover completed successfully or the master process was killed by accident, the manager stops working. To run as a daemon, daemontool. or any external daemon program can be used. Here is an example to run from daemontools.这里我们用 shell 脚本的方式去执行就不会发生监控程序死掉的情况
[root@MHA-S2 bin]# more manager.sh #!/bin/shnohup masterha_manager --conf=/etc/mha/app1.cnf --remove_dead_master_conf --ignore_last_failover /dev/null /var/log/masterha/app1/manager.log 2 1 2. 当你修复完死掉的 master 想重新加入先有的两节点 MHA 也是可以的
旧 Master：
root@localhost:mysql3306.sock [tt] show master status\G*************************** 1. row ***************************
File: 3306-binlog.000004
Position: 194
Binlog_Do_DB:
Binlog_Ignore_DB:
Executed_Gtid_Set: a5757eae-7981-11e7-82c7-005056b662d3:1-322101 row in set (0.00 sec)
现有 master：
root@localhost:mysql3306.sock [tt] show master status\G*************************** 1. row ***************************
File: 3306-binlog.000003
Position: 61945043
Binlog_Do_DB:
Binlog_Ignore_DB:
Executed_Gtid_Set: 1c2dc99f-7b57-11e7-a280-005056b665cb:1-3,
a5757eae-7981-11e7-82c7-005056b662d3:1-322101 row in set (0.00 sec)
由于有 GTID，我们可以直接就 change master 切换过去，先对比一下数据：
旧 master：
root@localhost:mysql3306.sock [tt] select * from t1;+----+------+
| id | c1 |
+----+------+
| 1 | a1 |
| 2 | a2 |
| 3 | a3 |
| 4 | a4 |
+----+------+4 rows in set (0.02 sec)
新 master：
root@localhost:mysql3306.sock [tt] select * from t1;+----+------+
| id | c1 |
+----+------+
| 1 | a1 |
| 2 | a2 |
| 3 | a3 |
| 4 | a4 |
| 5 | a5 |+----+------+
旧 master 直接 change master to：
change master to master_host= MHA-S1 ,master_user= repl ,master_password= 123456 ,master_port=3306,master_auto_position=1;start slave 看输出：
root@localhost:mysql3306.sock [tt] show slave status\G
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: MHA-S1
Master_User: repl
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: 3306-binlog.000003
Read_Master_Log_Pos: 61945043
Relay_Log_File: MHA-M1-relay-bin.000004
Relay_Log_Pos: 715
Relay_Master_Log_File: 3306-binlog.000003
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
看是否会补全数据：
root@localhost:mysql3306.sock [tt] select * from t1;+----+------+
| id | c1 |
+----+------+
| 1 | a1 |
| 2 | a2 |
| 3 | a3 |
| 4 | a4 |
| 5 | a5 |
+----+------+
发现数据补全了，加入复制没问题。
最后还得修改 app1.cnf 把 server1 补上
[server1]hostname=MHA-M1
port=3306
重启监控程序并查看 MHA 状态
[root@MHA-S2 tmp]# masterha_check_repl --conf=/etc/mha/app1.cnf
Sat Aug 12 20:37:01 2017 - [info] Replication filtering check ok.
Sat Aug 12 20:37:01 2017 - [error][/usr/local/share/perl5/MHA/Server.pm, ln398] MHA-M1(10.180.2.163:3306): User repl does not exist or does not have REPLICATION SLAVE privilege! Other slaves can not start replication from this host.
Sat Aug 12 20:37:01 2017 - [error][/usr/local/share/perl5/MHA/MasterMonitor.pm, ln427] Error happened on checking configurations. at /usr/local/share/perl5/MHA/ServerManager.pm line 1403
Sat Aug 12 20:37:01 2017 - [error][/usr/local/share/perl5/MHA/MasterMonitor.pm, ln525] Error happened on monitoring servers.
Sat Aug 12 20:37:01 2017 - [info] Got exit code 1 (Not master dead).
发现权限有问题，赶紧修复一下：
MHA-M1：
set session sql_log_bin=OFF;
grant replication slave on *.* to repl@ % identified by 123456
set session sql_log_bin=ON;
再次执行 MHA 状态检查：
masterha_check_repl --conf=/etc/mha/app1.cnf
Sat Aug 12 20:41:14 2017 - [info] Checking replication health on MHA-M1..
Sat Aug 12 20:41:14 2017 - [info] ok.
Sat Aug 12 20:41:14 2017 - [info] Checking replication health on MHA-S2..
Sat Aug 12 20:41:14 2017 - [info] ok.
Sat Aug 12 20:41:14 2017 - [info] Checking master_ip_failover_script status:
Sat Aug 12 20:41:14 2017 - [info] /usr/local/bin/master_ip_failover --command=status --ssh_user=root --orig_master_host=MHA-S1 --orig_master_ip=10.180.2.164 --orig_master_port=3306 IN SCRIPT TEST====/sbin/ifconfig eth2:1 down==/sbin/ifconfig eth2:1 10.180.2.168/24===Checking the Status of the script.. OK
Sat Aug 12 20:41:15 2017 - [info] OK.
Sat Aug 12 20:41:15 2017 - [warning] shutdown_script is not defined.
Sat Aug 12 20:41:15 2017 - [info] Got exit code 0 (Not master dead).
MySQL Replication Health is OK.
最后启动监控程序
[root@MHA-S2 bin]# nohup monitor.sh [root@MHA-S2 bin]# masterha_check_status --conf=/etc/mha/app1.cnf
app1 (pid:32084) is running(0:PING_OK), master:MHA-S1
（2) 手动在线切换测试
在许多情况下，需要将现有的主服务器迁移到另外一台服务器上。比如主服务器硬件故障，RAID 控制卡需要重建，将主服务器移到性能更好的服务器上等等。维护主服务器引起性能下降，导致停机时间至少无法写入数据。另外，阻塞或杀掉当前运行的会话会导致主主之间数据不一致的问题发生。MHA 提供快速切换和优雅的阻塞写入，这个切换过程只需要 0.5-2s 的时间，这段时间内数据是无法写入的。在很多情况下，0.5-2s 的阻塞写入是可以接受的。因此切换主服务器不需要计划分配维护时间窗口。
MHA 在线切换的大概过程：
1. 检测复制设置和确定当前主服务器
2. 确定新的主服务器
3. 阻塞写入到当前主服务器
4. 等待所有从服务器赶上复制
5. 授予写入到新的主服务器
6. 重新设置从服务器
注意，在线切换的时候应用架构需要考虑以下两个问题：
1. 自动识别 master 和 slave 的问题（master 的机器可能会切换），如果采用了 vip 的方式，基本可以解决这个问题。
2. 负载均衡的问题（可以定义大概的读写比例，每台机器可承担的负载比例，当有机器离开集群时，需要考虑这个问题）
为了保证数据完全一致性，在最快的时间内完成切换，MHA 的在线切换必须满足以下条件才会切换成功，否则会切换失败。
1. 所有 slave 的 IO 线程都在运行
2. 所有 slave 的 SQL 线程都在运行
3. 所有的 show slave status 的输出中 Seconds_Behind_Master 参数小于或者等于 running_updates_limit 秒，如果在切换过程中不指定 running_updates_limit, 那么默认情况下 running_updates_limit 为 1 秒。
4. 在 master 端，通过 show processlist 输出，没有一个更新花费的时间大于 running_updates_limit 秒。
在线切换步骤如下：
先停止监控程序
[root@MHA-S2 app1]# masterha_stop --conf=/etc/mha/app1.cnf
Stopped app1 successfully.修改 master_ip_online_change 脚本如下：
[root@MHA-S2 bin]# more master_ip_online_change
#!/usr/bin/env perl# Copyright (C) 2011 DeNA Co.,Ltd.
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc.,
# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
## Note: This is a sample script and is not complete. Modify the script based on your environment.
use strict;
use warnings FATAL = all
use Getopt::Long;
use MHA::DBHelper;
use MHA::NodeUtil;
use Time::HiRes qw( sleep gettimeofday tv_interval );
use Data::Dumper;
my $_tstart;
my $_running_interval = 0.1;
my (
$command, $orig_master_host, $orig_master_ip,
$orig_master_port, $orig_master_user,
$new_master_host, $new_master_ip, $new_master_port,
$new_master_user,
my $vip = 10.180.2.168/19 # Virtual IP
my $key = 1
my $ssh_start_vip = /sbin/ifconfig eth2:$key $vip
my $ssh_stop_vip = /sbin/ifconfig eth2:$key down
my $ssh_user = root
my $new_master_password= 123456
my $orig_master_password= 123456
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, orig_master_user=s = \$orig_master_user,
# orig_master_password=s = \$orig_master_password, 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,
exit main();
sub current_time_us { my ( $sec, $microsec ) = gettimeofday();
my $curdate = localtime($sec);
return $curdate . . sprintf( %06d , $microsec );
sub sleep_until { my $elapsed = tv_interval($_tstart); if ( $_running_interval $elapsed ) { sleep( $_running_interval - $elapsed );
}
sub get_threads_util {
my $dbh = shift;
my $my_connection_id = shift;
my $running_time_threshold = shift;
my $type = shift;
$running_time_threshold = 0 unless ($running_time_threshold);
$type = 0 unless ($type);
my @threads;
my $sth = $dbh- prepare( SHOW PROCESSLIST
$sth- execute(); while ( my $ref = $sth- fetchrow_hashref() ) { my $id = $ref- {Id};
my $user = $ref- {User};
my $host = $ref- {Host};
my $command = $ref- {Command};
my $state = $ref- {State};
my $query_time = $ref- {Time};
my $info = $ref- {Info};
$info =~ s/^\s*(.*?)\s*$/$1/ if defined($info);
next if ( $my_connection_id == $id );
next if ( defined($query_time) $query_time $running_time_threshold );
next if ( defined($command) $command eq Binlog Dump );
next if ( defined($user) $user eq system user );
next if ( defined($command) $command eq Sleep
defined($query_time) $query_time = 1 ); if ( $type = 1 ) { next if ( defined($command) $command eq Sleep );
next if ( defined($command) $command eq Connect );
} if ( $type = 2 ) { next if ( defined($info) $info =~ m/^select/i );
next if ( defined($info) $info =~ m/^show/i );
}
push @threads, $ref;
}
return @threads;
sub main { if ( $command eq stop ) {
## Gracefully killing connections on the current master
# 1. Set read_only= 1 on the new master
# 2. DROP USER so that no app user can establish new connections
# 3. Set read_only= 1 on the current master
# 4. Kill current queries
# * Any database access failure will result in script die.
my $exit_code = 1;
eval { ## Setting read_only=1 on the new master (to avoid accident)
my $new_master_handler = new MHA::DBHelper();
# args: hostname, port, user, password, raise_error(die_on_error)_or_not
$new_master_handler- connect( $new_master_ip, $new_master_port,
$new_master_user, $new_master_password, 1 );
print current_time_us() . Set read_only on the new master..
$new_master_handler- enable_read_only(); if ( $new_master_handler- is_read_only() ) {
print ok.\n
} else {
die Failed!\n
}
$new_master_handler- disconnect();
# Connecting to the orig master, die if any database error happens
my $orig_master_handler = new MHA::DBHelper();
$orig_master_handler- connect( $orig_master_ip, $orig_master_port,
$orig_master_user, $orig_master_password, 1 );
## Drop application user so that nobody can connect. Disabling per-session binlog beforehand
#$orig_master_handler- disable_log_bin_local();
#print current_time_us() . Drpping app user on the orig master..\n
#FIXME_xxx_drop_app_user($orig_master_handler);
## Waiting for N * 100 milliseconds so that current connections can exit
my $time_until_read_only = 15;
$_tstart = [gettimeofday];
my @threads = get_threads_util( $orig_master_handler- {dbh},
$orig_master_handler- {connection_id} ); while ( $time_until_read_only 0 $#threads = 0 ) { if ( $time_until_read_only % 5 == 0 ) { printf %s Waiting all running %d threads are disconnected.. (max %d milliseconds)\n ,
current_time_us(), $#threads + 1, $time_until_read_only * 100; if ( $#threads 5 ) { print Data::Dumper- new( [$_] )- Indent(0)- Terse(1)- Dump . \n
foreach (@threads);
}
}
sleep_until();
$_tstart = [gettimeofday];
$time_until_read_only--;
@threads = get_threads_util( $orig_master_handler- {dbh},
$orig_master_handler- {connection_id} );
}
## Setting read_only=1 on the current master so that nobody(except SUPER) can write
print current_time_us() . Set read_only=1 on the orig master..
$orig_master_handler- enable_read_only(); if ( $orig_master_handler- is_read_only() ) {
print ok.\n
} else {
die Failed!\n
}
## Waiting for M * 100 milliseconds so that current update queries can complete
my $time_until_kill_threads = 5;
@threads = get_threads_util( $orig_master_handler- {dbh},
$orig_master_handler- {connection_id} ); while ( $time_until_kill_threads 0 $#threads = 0 ) { if ( $time_until_kill_threads % 5 == 0 ) { printf %s Waiting all running %d queries are disconnected.. (max %d milliseconds)\n ,
current_time_us(), $#threads + 1, $time_until_kill_threads * 100; if ( $#threads 5 ) { print Data::Dumper- new( [$_] )- Indent(0)- Terse(1)- Dump . \n
foreach (@threads);
}
}
sleep_until();
$_tstart = [gettimeofday];
$time_until_kill_threads--;
@threads = get_threads_util( $orig_master_handler- {dbh},
$orig_master_handler- {connection_id} );
}
print Disabling the VIP on old master: $orig_master_host \n stop_vip();
## Terminating all threads
print current_time_us() . Killing all application threads..\n
$orig_master_handler- kill_threads(@threads) if ( $#threads = 0 );
print current_time_us() . done.\n
#$orig_master_handler- enable_log_bin_local();
$orig_master_handler- disconnect();
## After finishing the script, MHA executes FLUSH TABLES WITH READ LOCK
$exit_code = 0;
}; if ($@) {
warn Got Error: $@\n
exit $exit_code;
}
exit $exit_code;
}
elsif ( $command eq start ) {
## Activating master ip on the new master
# 1. Create app user with write privileges
# 2. Moving backup script if needed
# 3. Register new master s ip to the catalog database# We don t return error even though activating updatable accounts/ip failed so that we don t interrupt slaves recovery.# If exit code is 0 or 10, MHA does not abort
my $exit_code = 10;
eval { my $new_master_handler = new MHA::DBHelper();
# args: hostname, port, user, password, raise_error_or_not
$new_master_handler- connect( $new_master_ip, $new_master_port,
$new_master_user, $new_master_password, 1 );
## Set read_only=0 on the new master
#$new_master_handler- disable_log_bin_local();
print current_time_us() . Set read_only=0 on the new master.\n
$new_master_handler- disable_read_only();
## Creating an app user on the new master
#print current_time_us() . Creating app user on the new master..\n
#FIXME_xxx_create_app_user($new_master_handler);
#$new_master_handler- enable_log_bin_local();
$new_master_handler- disconnect();
## Update master ip on the catalog database, etc
print Enabling the VIP - $vip on the new master - $new_master_host \n start_vip();
$exit_code = 0;
}; if ($@) {
warn Got Error: $@\n
exit $exit_code;
}
exit $exit_code;
}
elsif ( $command eq status ) {
# do nothing
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_online_change --command=start|stop|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
die;
}
执行切换
[root@MHA-S2 tmp]# masterha_master_switch --conf=/etc/mha/app1.cnf --master_state=alive --new_master_host=MHA-M1 --new_master_port=3306 --orig_master_is_new_slave --running_updates_limit=10000其中参数的意思：
–orig_master_is_new_slave 切换时加上此参数是将原 master 变为 slave 节点，如果不加此参数，原来的 master 将不启动
–running_updates_limit=10000, 故障切换时, 候选 master 如果有延迟的话，mha 切换不能成功，加上此参数表示延迟在此时间范围内都可切换（单位为 s），但是切换的时间长短是由 recover 时 relay 日志的大小决定
查看切换后各机器的状态：
S2：
root@localhost:mysql3306.sock [tt] show slave status\G
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: MHA-M1
Master_User: repl
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: 3306-binlog.000004
Read_Master_Log_Pos: 748
Relay_Log_File: MHA-S2-relay-bin.000002
Relay_Log_Pos: 420
Relay_Master_Log_File: 3306-binlog.000004
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
S1：
root@localhost:mysql3306.sock [tt] show slave status\G*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: MHA-M1
Master_User: repl
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: 3306-binlog.000004
Read_Master_Log_Pos: 748
Relay_Log_File: MHA-S1-relay-bin.000002
Relay_Log_Pos: 420
Relay_Master_Log_File: 3306-binlog.000004
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
M1：
root@localhost:mysql3306.sock [tt] show slave status\G
Empty set (0.00 sec)
在线切换的日志：
[root@MHA-S2 tmp]# more sw.log
[root@MHA-S2 bin]# masterha_master_switch --conf=/etc/mha/app1.cnf --master_state=alive --new_master_host=MHA-M1 --new_master_port=3306 --orig_master_is_new_slave --running_updates_limit=10000Sat Aug 12 21:34:54 2017 - [info] MHA::MasterRotate version 0.57.
Sat Aug 12 21:34:54 2017 - [info] Starting online master switch..
Sat Aug 12 21:34:54 2017 - [info]
Sat Aug 12 21:34:54 2017 - [info] * Phase 1: Configuration Check Phase..
Sat Aug 12 21:34:54 2017 - [info]
Sat Aug 12 21:34:54 2017 - [warning] Global configuration file /etc/masterha_default.cnf not found. Skipping.
Sat Aug 12 21:34:54 2017 - [info] Reading application default configuration from /etc/mha/app1.cnf..
Sat Aug 12 21:34:54 2017 - [info] Reading server configuration from /etc/mha/app1.cnf..
Sat Aug 12 21:34:54 2017 - [info] GTID failover mode = 1Sat Aug 12 21:34:54 2017 - [info] Current Alive Master: MHA-S1(10.180.2.164:3306)
Sat Aug 12 21:34:54 2017 - [info] Alive Slaves:
Sat Aug 12 21:34:54 2017 - [info] MHA-M1(10.180.2.163:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Sat Aug 12 21:34:54 2017 - [info] GTID ON
Sat Aug 12 21:34:54 2017 - [info] Replicating from MHA-S1(10.180.2.164:3306)
Sat Aug 12 21:34:54 2017 - [info] MHA-S2(10.180.2.165:3306) Version=5.7.18-log (oldest major version between slaves) log-bin:enabled
Sat Aug 12 21:34:54 2017 - [info] GTID ON
Sat Aug 12 21:34:54 2017 - [info] Replicating from MHA-S1(10.180.2.164:3306)
It is better to execute FLUSH NO_WRITE_TO_BINLOG TABLES on the master before switching. Is it ok to execute on MHA-S1(10.180.2.164:3306)? (YES/no): yes
Sat Aug 12 21:35:07 2017 - [info] Executing FLUSH NO_WRITE_TO_BINLOG TABLES. This may take long time..
Sat Aug 12 21:35:07 2017 - [info] ok.
Sat Aug 12 21:35:07 2017 - [info] Checking MHA is not monitoring or doing failover..
Sat Aug 12 21:35:07 2017 - [info] Checking replication health on MHA-M1..
Sat Aug 12 21:35:07 2017 - [info] ok.
Sat Aug 12 21:35:07 2017 - [info] Checking replication health on MHA-S2..
Sat Aug 12 21:35:07 2017 - [info] ok.
Sat Aug 12 21:35:07 2017 - [info] MHA-M1 can be new master.
Sat Aug 12 21:35:07 2017 - [info]
From:
MHA-S1(10.180.2.164:3306) (current master) +--MHA-M1(10.180.2.163:3306) +--MHA-S2(10.180.2.165:3306)
MHA-M1(10.180.2.163:3306) (new master) +--MHA-S2(10.180.2.165:3306) +--MHA-S1(10.180.2.164:3306)
Starting master switch from MHA-S1(10.180.2.164:3306) to MHA-M1(10.180.2.163:3306)? (yes/NO): yes
Sat Aug 12 21:35:15 2017 - [info] Checking whether MHA-M1(10.180.2.163:3306) is ok for the new master..
Sat Aug 12 21:35:15 2017 - [info] ok.
Sat Aug 12 21:35:15 2017 - [info] MHA-S1(10.180.2.164:3306): SHOW SLAVE STATUS returned empty result. To check replication filtering rules, temporarily executing CHANGE MASTER to a dummy host.
Sat Aug 12 21:35:15 2017 - [info] MHA-S1(10.180.2.164:3306): Resetting slave pointing to the dummy host.
Sat Aug 12 21:35:15 2017 - [info] ** Phase 1: Configuration Check Phase completed.
Sat Aug 12 21:35:15 2017 - [info]
Sat Aug 12 21:35:15 2017 - [info] * Phase 2: Rejecting updates Phase..
Sat Aug 12 21:35:15 2017 - [info]
Sat Aug 12 21:35:15 2017 - [info] Executing master ip online change script to disable write on the current master:
Sat Aug 12 21:35:15 2017 - [info] /usr/local/bin/master_ip_online_change --command=stop --orig_master_host=MHA-S1 --orig_master_ip=10.180.2.164 --orig_master_port=3306 --orig_master_user= root --new_master_
host=MHA-M1 --new_master_ip=10.180.2.163 --new_master_port=3306 --new_master_user= root --orig_master_ssh_user=root --new_master_ssh_user=root --orig_master_is_new_slave --orig_master_password=xxx --new_mas
ter_password=xxx
Unknown option: orig_master_ssh_user
Unknown option: new_master_ssh_user
Unknown option: orig_master_is_new_slave
Unknown option: orig_master_password
Unknown option: new_master_password
Sat Aug 12 21:35:15 2017 568580 Set read_only on the new master.. ok.
Sat Aug 12 21:35:15 2017 573508 Waiting all running 2 threads are disconnected.. (max 1500 milliseconds)
{Time = 272878 , Command = Binlog Dump GTID , db = undef, Id = 40 , Info = undef, User = repl , State = Master has sent all binlog to slave; waiting for more updates , Host = MHA-S2:46970}{Time = 3738 , Command = Binlog Dump GTID , db = undef, Id = 55 , Info = undef, User = repl , State = Master has sent all binlog to slave; waiting for more updates , Host = MHA-M1:51506}
Sat Aug 12 21:35:16 2017 075020 Waiting all running 2 threads are disconnected.. (max 1000 milliseconds)
{Time = 272879 , Command = Binlog Dump GTID , db = undef, Id = 40 , Info = undef, User = repl , State = Master has sent all binlog to slave; waiting for more updates , Host = MHA-S2:46970}{Time = 3739 , Command = Binlog Dump GTID , db = undef, Id = 55 , Info = undef, User = repl , State = Master has sent all binlog to slave; waiting for more updates , Host = MHA-M1:51506}
Sat Aug 12 21:35:16 2017 576059 Waiting all running 2 threads are disconnected.. (max 500 milliseconds)
{Time = 272879 , Command = Binlog Dump GTID , db = undef, Id = 40 , Info = undef, User = repl , State = Master has sent all binlog to slave; waiting for more updates , Host = MHA-S2:46970}{Time = 3739 , Command = Binlog Dump GTID , db = undef, Id = 55 , Info = undef, User = repl , State = Master has sent all binlog to slave; waiting for more updates , Host = MHA-M1:51506}
Sat Aug 12 21:35:17 2017 076940 Set read_only=1 on the orig master.. ok.
Sat Aug 12 21:35:17 2017 079645 Waiting all running 2 queries are disconnected.. (max 500 milliseconds)
{Time = 272880 , Command = Binlog Dump GTID , db = undef, Id = 40 , Info = undef, User = repl , State = Master has sent all binlog to slave; waiting for more updates , Host = MHA-S2:46970}{Time = 3740 , Command = Binlog Dump GTID , db = undef, Id = 55 , Info = undef, User = repl , State = Master has sent all binlog to slave; waiting for more updates , Host = MHA-M1:51506}
Disabling the VIP on old master: MHA-S1
Sat Aug 12 21:35:17 2017 683769 Killing all application threads..
Sat Aug 12 21:35:17 2017 686090 done.
Sat Aug 12 21:35:17 2017 - [info] ok.
Sat Aug 12 21:35:17 2017 - [info] Locking all tables on the orig master to reject updates from everybody (including root):
Sat Aug 12 21:35:17 2017 - [info] Executing FLUSH TABLES WITH READ LOCK..
Sat Aug 12 21:35:17 2017 - [info] ok.
Sat Aug 12 21:35:17 2017 - [info] Orig master binlog:pos is 3306-binlog.000003:61945043.
Sat Aug 12 21:35:17 2017 - [info] Waiting to execute all relay logs on MHA-M1(10.180.2.163:3306)..
Sat Aug 12 21:35:17 2017 - [info] master_pos_wait(3306-binlog.000003:61945043) completed on MHA-M1(10.180.2.163:3306). Executed 0 events.
Sat Aug 12 21:35:17 2017 - [info] done.
Sat Aug 12 21:35:17 2017 - [info] Getting new master s binlog name and position..Sat Aug 12 21:35:17 2017 - [info] 3306-binlog.000004:748Sat Aug 12 21:35:17 2017 - [info] All other slaves should start replication from here. Statement should be: CHANGE MASTER TO MASTER_HOST= MHA-M1 or 10.180.2.163 , MASTER_PORT=3306, MASTER_AUTO_POSITION=1, MAS
TER_USER= repl , MASTER_PASSWORD= xxx
Sat Aug 12 21:35:17 2017 - [info] Executing master ip online change script to allow write on the new master:
Sat Aug 12 21:35:17 2017 - [info] /usr/local/bin/master_ip_online_change --command=start --orig_master_host=MHA-S1 --orig_master_ip=10.180.2.164 --orig_master_port=3306 --orig_master_user= root --new_master
_host=MHA-M1 --new_master_ip=10.180.2.163 --new_master_port=3306 --new_master_user= root --orig_master_ssh_user=root --new_master_ssh_user=root --orig_master_is_new_slave --orig_master_password=xxx --new_ma
ster_password=xxx
Unknown option: orig_master_ssh_user
Unknown option: new_master_ssh_user
Unknown option: orig_master_is_new_slave
Unknown option: orig_master_password
Unknown option: new_master_password
Sat Aug 12 21:35:17 2017 865209 Set read_only=0 on the new master.
Enabling the VIP - 10.180.2.168/19 on the new master - MHA-M1
Sat Aug 12 21:35:17 2017 - [info] ok.
Sat Aug 12 21:35:17 2017 - [info]
Sat Aug 12 21:35:17 2017 - [info] * Switching slaves in parallel..
Sat Aug 12 21:35:17 2017 - [info]
Sat Aug 12 21:35:17 2017 - [info] -- Slave switch on host MHA-S2(10.180.2.165:3306) started, pid: 2327Sat Aug 12 21:35:17 2017 - [info]
Sat Aug 12 21:35:18 2017 - [info] Log messages from MHA-S2 ...
Sat Aug 12 21:35:18 2017 - [info]
Sat Aug 12 21:35:18 2017 - [info] Waiting to execute all relay logs on MHA-S2(10.180.2.165:3306)..
Sat Aug 12 21:35:18 2017 - [info] master_pos_wait(3306-binlog.000003:61945043) completed on MHA-S2(10.180.2.165:3306). Executed 0 events.
Sat Aug 12 21:35:18 2017 - [info] done.
Sat Aug 12 21:35:18 2017 - [info] Resetting slave MHA-S2(10.180.2.165:3306) and starting replication from the new master MHA-M1(10.180.2.163:3306)..
Sat Aug 12 21:35:18 2017 - [info] Executed CHANGE MASTER.
Sat Aug 12 21:35:18 2017 - [info] Slave started.
Sat Aug 12 21:35:18 2017 - [info] End of log messages from MHA-S2 ...
Sat Aug 12 21:35:18 2017 - [info]
Sat Aug 12 21:35:18 2017 - [info] -- Slave switch on host MHA-S2(10.180.2.165:3306) succeeded.
Sat Aug 12 21:35:18 2017 - [info] Unlocking all tables on the orig master:
Sat Aug 12 21:35:18 2017 - [info] Executing UNLOCK TABLES..
Sat Aug 12 21:35:18 2017 - [info] ok.
Sat Aug 12 21:35:18 2017 - [info] Starting orig master as a new slave..
Sat Aug 12 21:35:18 2017 - [info] Resetting slave MHA-S1(10.180.2.164:3306) and starting replication from the new master MHA-M1(10.180.2.163:3306)..
Sat Aug 12 21:35:18 2017 - [info] Executed CHANGE MASTER.
Sat Aug 12 21:35:19 2017 - [info] Slave started.
Sat Aug 12 21:35:19 2017 - [info] All new slave servers switched successfully.
Sat Aug 12 21:35:19 2017 - [info]
Sat Aug 12 21:35:19 2017 - [info] * Phase 5: New master cleanup phase..
Sat Aug 12 21:35:19 2017 - [info]
Sat Aug 12 21:35:19 2017 - [info] MHA-M1: Resetting slave info succeeded.
Sat Aug 12 21:35:19 2017 - [info] Switching master to MHA-M1(10.180.2.163:3306) completed successfully.
感谢你能够认真阅读完这篇文章，希望丸趣 TV 小编分享的“MySQL 高可用架构之 MHA 的原理分析”这篇文章对大家有帮助，同时也希望大家多多支持丸趣 TV，关注丸趣 TV 行业资讯频道，更多相关知识等着你来学习!
正文完
