MySQL高可用之MHA

MHA(Master High Availability)目前在MySQL高可用方面是一个相对成熟的解决方案,它由日本DeNA公司 youshimaton(现就职于Facebook公司)开发,是一套优秀的作为MySQL高可用性环境下故障切换和主从提升 的高可用软件。在MySQL故障切换过程中,MHA能做到在0~30秒之内自动完成数据库的故障切换操作,并且 在进行故障切换的过程中,MHA能在最大程度上保证数据的一致性,以达到真正意义上的高可用。 MHA里有 两个角色一个是MHA Node(数据节点)另一个是MHA Manager(管理节点)。 MHA Manager可以单独部署 在一台独立的机器上管理多个master-slave集群,也可以部署在一台slave节点上。MHA Node运行在每台 MySQL服务器上,MHA Manager会定时探测集群中的master节点,当master出现故障时,它可以自动将最新 数据的slave提升为新的master,然后将所有其他的slave重新指向新的master。整个故障转移过程对应用程序完 全透明。
MySQL高可用之MHA
MHA自动故障切换过程中,MHA试图从宕机的主服务器上保存二进制日志,最大程度的保证数据的不丢失,但 这并不总是可行的。例如,如果主服务器硬件故障或无法通过ssh访问,MHA没法保存二进制日志,只进行故 障转移而丢失了最新的数据。使用MySQL 5.5的半同步复制,可以大大降低数据丢失的风险。MHA可以与半同 步复制结合起来。如果只有一个slave已经收到了最新的二进制日志,MHA可以将最新的二进制日志应用于其他 所有的slave服务器上,因此可以保证所有节点的数据一致性。
注:从MySQL5.5开始,MySQL以插件的形式支持半同步复制。
如何理解半同步呢?

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

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

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

目前MHA主要支持一主多从的架构,要搭建MHA,要求一个复制集群中必须最少有三台数据库服务器,一主二 从,即一台充当master,一台充当备用master,另外一台充当从库,因为至少需要三台服务器。
MySQL高可用之MHA
其中master对外提供写服务,备选master(实际的slave,主机名slave1)提供读服务,slave也提供相关的读 服务,一旦master宕机,将会把备选master提升为新的master,slave指向新的master,manager作为管理服务 器。

一、开始部署
注:时间要同步

#在四台机器都配置epel源
[ ~]# wget -O /etc/yum.repos.d/epel.repo http://mirrors.aliyun.com/repo/epel-7.repo
#建立ssh无交互登录环境 
#Manager主机:
[ ~]# ssh-keygen -t rsa
Generating public/private rsa key pair.
Enter file in which to save the key (/root/.ssh/id_rsa): 
Created directory ‘/root/.ssh‘.
Enter passphrase (empty for no passphrase): 
Enter same passphrase again: 
Your identification has been saved in /root/.ssh/id_rsa.
Your public key has been saved in /root/.ssh/id_rsa.pub.
The key fingerprint is:
1c:cb:2d:4f:b1:80:ea:80:35:3b:89:48:5f:09:eb:2e 
The key‘s randomart image is:
+--[ RSA 2048]----+
|   .             |
|    o ..         |
| .o. o. o .      |
|o+o+.. o = o     |
|+ =o.   S +      |
|  .+     +       |
| E ..     .      |
|  .              |
|                 |
+-----------------+
[ ~]# for i in manager master slave1 slave2;do ssh-copy-id -i $i;done                    # 分发到其他主机
#master主机
[ ~]# ssh-keygen -t rsa
[ ~]# for i in manager master slave1 slave2;do ssh-copy-id -i $i;done
#slave1主机
[ ~]# ssh-keygen -t rsa
[ ~]# for i in manager master slave1 slave2;do ssh-copy-id -i $i;done
#slave2主机
[ ~]# ssh-keygen -t rsa
[ ~]# for i in manager master slave1 slave2;do ssh-copy-id -i $i;done

测试ssh无交互登录

[ ~]# for i in master manager slave1 slave2;do ssh $i hostname;done
master
manager
slave1
slave2
[ ~]# ssh master
Last failed login: Thu Jul  2 17:05:01 CST 2020 from 192.168.171.150 on ssh:notty
There was 1 failed login attempt since the last successful login.
Last login: Thu Jul  2 16:57:27 2020 from 192.168.171.1
[ ~]# ssh slave1
Last login: Thu Jul  2 17:02:59 2020 from 192.168.171.1
[ ~]# ssh slave2
Last login: Thu Jul  2 17:03:29 2020 from 192.168.171.1
#配置hosts环境
[ ~]# vim /etc/hosts
192.168.171.150 manager
192.168.171.151 master
192.168.171.152 slave1
192.168.171.153 slave2
[ ~]# scp /etc/hosts :/etc/
[ ~]# scp /etc/hosts :/etc/
[ ~]# scp /etc/hosts :/etc/

二、配置mysql半同步复制
为了尽可能的减少主库硬件损坏宕机造成的数据丢失,因此在配置MHA的同时建议 配置成MySQL的半同步复制。

**注:**mysql半同步插件是由谷歌提供,具体位置/usr/local/mysql/lib/plugin/下:
一个是master用的semisync_master.so;
一个是slave用的semisync_slave.so;

分别在主从节点上安装相关的插件(master, Candicate master,slave)

#检测数据库是否支持动态载入
mysql>show variables like ‘%have_dynamic%‘;
+----------------------+-------+
| Variable_name        | Value |
+----------------------+-------+
| have_dynamic_loading | YES   |           # 显示yes表示支持
+----------------------+-------+
1 row in set (0.00 sec)
#所有mysql数据库服务器,安装半同步插件(semisync_master.so,semisync_slave.so)
#要切记是所有mysql数据库服务器上都需要安装,这里我就在这写了一台该怎么安装
mysql>INSTALL PLUGIN rpl_semi_sync_master SONAME ‘semisync_master.so‘; 
Query OK, 0 rows affected (0.01 sec)
mysql>INSTALL PLUGIN rpl_semi_sync_slave SONAME ‘semisync_slave.so‘; 
Query OK, 0 rows affected (0.00 sec)
#检查Plugin是否已正确安装
mysql> show plugins;              # 可以再最后两行看到如下内容
| rpl_semi_sync_master       | ACTIVE   | REPLICATION        | semisync_master.so | GPL     |
| rpl_semi_sync_slave        | ACTIVE   | REPLICATION        | semisync_slave.so  | GPL     |
#也可以使用如下命令检查
mysql>select * from information_schema.plugins; 
#查看半同步相关信息
mysql>show variables like ‘%rpl_semi_sync%‘;
+-------------------------------------------+------------+
| Variable_name                             | Value      |
+-------------------------------------------+------------+
| rpl_semi_sync_master_enabled              | OFF        |
| rpl_semi_sync_master_timeout              | 10000      |
| rpl_semi_sync_master_trace_level          | 32         |
| rpl_semi_sync_master_wait_for_slave_count | 1          |
| rpl_semi_sync_master_wait_no_slave        | ON         |
| rpl_semi_sync_master_wait_point           | AFTER_SYNC |
| rpl_semi_sync_slave_enabled               | OFF        |
| rpl_semi_sync_slave_trace_level           | 32         |
+-------------------------------------------+------------+
8 rows in set (0.00 sec)
#上方所示内容可以看到半同复制插件已经成功安装,只是还没有启用,所以是off

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

#master主机
[ ~]# vim /etc/my.cnf
server-id = 1
log-bin = mysql-bin
binlog_format = mixed
log-bin-index = mysql-bin.index
rpl_semi_sync_master_enabled = 1                 # 1表示启用,0表示关闭,slave同样
rpl_semi_sync_master_timeout = 1000         # 毫秒单位,主服务器等待确认消息10秒后,不在等待,变为异步方式
rpl_semi_sync_slave_enabled = 1
relay_log_purge = 0
relay-log = relay-bin
relay-log-index = slave-relay-bin.index
#candicate master主机
[ ~]# vim /etc/my.cnf
server-id = 2
log-bin = mysql-bin
binlog_format = mixed
log-bin-index = mysql-bin.index
relay_log_purge = 0           # 0表示禁止 SQL 线程在执行完一个 relay log 后自动将其删除,对于MHA场景下,对 于某些滞后从库的恢复依赖于其他从库的relay log,因此采取禁用自动删除功能 
relay-log = relay-bin
relay-log-index = slave-relay-bin.index
rpl_semi_sync_master_enabled = 1
rpl_semi_sync_master_timeout = 10000
rpl_semi_sync_slave_enabled = 1
#salve2主机
Server-id = 3
log-bin = mysql-bin
relay-log = relay-bin
relay-log-index = slave-relay-bin.index
read_only = 1
rpl_semi_sync_slave_enabled = 1

#查看半同步相关信息
mysql> show variables like ‘%rpl_semi_sync%‘;
+-------------------------------------------+------------+
| Variable_name                             | Value      |
+-------------------------------------------+------------+
| rpl_semi_sync_master_enabled              | ON         |
| rpl_semi_sync_master_timeout              | 10000      |
| rpl_semi_sync_master_trace_level          | 32         |
| rpl_semi_sync_master_wait_for_slave_count | 1          |
| rpl_semi_sync_master_wait_no_slave        | ON         |
| rpl_semi_sync_master_wait_point           | AFTER_SYNC |
| rpl_semi_sync_slave_enabled               | ON         |
| rpl_semi_sync_slave_trace_level           | 32         |
+-------------------------------------------+------------+
8 rows in set (0.00 sec)
#查看半同步状态
mysql> show status like ‘%rpl_semi_sync%‘;
+--------------------------------------------+-------+
| Variable_name                              | Value |
+--------------------------------------------+-------+
| Rpl_semi_sync_master_clients               | 0     |
| Rpl_semi_sync_master_net_avg_wait_time     | 0     |
| Rpl_semi_sync_master_net_wait_time         | 0     |
| Rpl_semi_sync_master_net_waits             | 0     |
| Rpl_semi_sync_master_no_times              | 0     |
| Rpl_semi_sync_master_no_tx                 | 0     |
| Rpl_semi_sync_master_status                | ON    |
| Rpl_semi_sync_master_timefunc_failures     | 0     |
| Rpl_semi_sync_master_tx_avg_wait_time      | 0     |
| Rpl_semi_sync_master_tx_wait_time          | 0     |
| Rpl_semi_sync_master_tx_waits              | 0     |
| Rpl_semi_sync_master_wait_pos_backtraverse | 0     |
| Rpl_semi_sync_master_wait_sessions         | 0     |
| Rpl_semi_sync_master_yes_tx                | 0     |
| Rpl_semi_sync_slave_status                 | OFF   |
+--------------------------------------------+-------+
15 rows in set (0.00 sec)
#有几个状态参数值得关注的: 
<< rpl_semi_sync_master_status :显示主服务是异步复制模式还是半同步复制模式 
<< rpl_semi_sync_master_clients :显示有多少个从服务器配置为半同步复制模式 
<< rpl_semi_sync_master_yes_tx :显示从服务器确认成功提交的数量 
<< rpl_semi_sync_master_no_tx :显示从服务器确认不成功提交的数量
<< rpl_semi_sync_master_tx_avg_wait_time :事务因开启 semi_sync ,平均需要额外等待的时间 
<< rpl_semi_sync_master_net_avg_wait_time :事务进入等待队列后,到网络平均等待时间 
#master主机:

mysql> grant replication slave on *.* to ‘192.168.171.%‘ identified by ‘123‘;                    # 创建一个用于主从复制的帐号,在master和candicate master的主机上创建即可
mysql> grant all privileges on *.* to ‘192.168.171.%‘ identified by ‘123‘;                   # 创建MHA管理账号,所有mysql服务器上都需要
mysql> show master status;
+------------------+----------+--------------+------------------+-------------------+
| File             | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+------------------+----------+--------------+------------------+-------------------+
| mysql-bin.000002 |      746 |              |                  |                   |
+------------------+----------+--------------+------------------+-------------------+
1 row in set (0.00 sec)
#Candicate master主机:
mysql> grant replication slave on *.* to ‘192.168.171.%‘ identified by ‘123‘;
mysql> grant all privileges on *.* to ‘192.168.171.%‘ identified by ‘123‘;
mysql> change master to master_host=‘192.168.171.151‘,master_port=3306,master_user=‘mharep‘,master_password=‘123‘,master_log_file=‘mysql-bin.000001‘,master_log_pos=746;
mysql> start slave;
#slave主机:
mysql> grant all privileges on *.* to ‘192.168.171.%‘ identified by ‘123‘; 
mysql> change master to master_host=‘192.168.171.151‘,master_port=3306,master_user=‘mharep‘,master_password=‘123‘,master_log_file=‘mysql-bin.000001‘,master_log_pos=746;
mysql> start slave;

查看Candicate master和slave2 两个从服务器的状态
MySQL高可用之MHA
MySQL高可用之MHA

#查看master服务器的半同步状态
mysql> show status like ‘%rpl_semi_sync%‘;
+--------------------------------------------+-------+
| Variable_name                              | Value |
+--------------------------------------------+-------+
| Rpl_semi_sync_master_clients               | 2     |
| Rpl_semi_sync_master_net_avg_wait_time     | 0     |
| Rpl_semi_sync_master_net_wait_time         | 0     |
| Rpl_semi_sync_master_net_waits             | 0     |
| Rpl_semi_sync_master_no_times              | 0     |
| Rpl_semi_sync_master_no_tx                 | 0     |
| Rpl_semi_sync_master_status                | ON    |
| Rpl_semi_sync_master_timefunc_failures     | 0     |
| Rpl_semi_sync_master_tx_avg_wait_time      | 0     |
| Rpl_semi_sync_master_tx_wait_time          | 0     |
| Rpl_semi_sync_master_tx_waits              | 0     |
| Rpl_semi_sync_master_wait_pos_backtraverse | 0     |
| Rpl_semi_sync_master_wait_sessions         | 0     |
| Rpl_semi_sync_master_yes_tx                | 0     |
| Rpl_semi_sync_slave_status                 | OFF   |
+--------------------------------------------+-------+
15 rows in set (0.00 sec)

三、配置mysql-mha mha包括manager节点和data节点

#在所有主机上安装mha所依赖的软 件包(需要系统自带的yum源并联网)
[ ~]# yum -y install perl-DBD-MySQL perl-Config-Tiny perl-Log-Dispatch perl-ParallelForkManager perl-Config-IniFiles ncftp perl-Params-Validate perl-CPAN perl-TestMock-LWP.noarch perl-LWP-Authen-Negotiate.noarch perl-devel perl-ExtUtils-CBuilder perl-ExtUtils-MakeMaker

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

#在所有数据库节点上安装node节点
[ ~]# tar zxf mha4mysql-node-0.58.tar.gz 
[ ~]# cd mha4mysql-node-0.58/
[ mha4mysql-node-0.58]# perl Makefile.PL 
*** Module::AutoInstall version 1.06
*** Checking for Perl dependencies...
[Core Features]
- DBI        ...loaded. (1.627)
- DBD::mysql ...loaded. (4.023)
*** Module::AutoInstall configuration finished.
Checking if your kit is complete...
Looks good
Writing Makefile for mha4mysql::node
[ mha4mysql-node-0.58]# make && make install 
#这里我就忽略其他两个数据库安装node了,步骤都一样
#管理节点两个都需要安装
[ ~]# tar zxf mha4mysql-node-0.58.tar.gz 
[ ~]# cd mha4mysql-node-0.58/
[ mha4mysql-node-0.58]# perl Makefile.PL 
[ mha4mysql-node-0.58]# make && make install 
#安装manager包
[ ~]# tar zxf mha4mysql-manager-0.58.tar.gz 
[ ~]# cd mha4mysql-manager-0.58/
[ mha4mysql-manager-0.58]# perl Makefile.PL 
*** Module::AutoInstall version 1.06
*** Checking for Perl dependencies...
[Core Features]
- DBI                   ...loaded. (1.627)
- DBD::mysql            ...loaded. (4.023)
- Time::HiRes           ...loaded. (1.9725)
- Config::Tiny          ...loaded. (2.14)
- Log::Dispatch         ...loaded. (2.41)
- Parallel::ForkManager ...missing.
- MHA::NodeConst        ...loaded. (0.58)
==> Auto-install the 1 mandatory module(s) from CPAN? [y] y
*** Dependencies will be installed the next time you type ‘make‘.
*** Module::AutoInstall configuration finished.
Checking if your kit is complete...
Looks good
Warning: prerequisite Parallel::ForkManager 0 not found.
Writing Makefile for mha4mysql::manager
[ mha4mysql-manager-0.58]# make && make install                  # 根据提示输入
[ ~]# mkdir /etc/masterha
[ ~]# mkdir -p /masterha/app1
[ ~]# mkdir /scripts
[ ~]# cd mha4mysql-manager-0.58/
[ mha4mysql-manager-0.58]# cp samples/conf/* /etc/masterha/
[ mha4mysql-manager-0.58]# cp samples/scripts/* /scripts/

编辑MHA配置文件

[ ~]# vim /etc/masterha/app1.cnf 
[server default]
manager_workdir=/masterha/app1               # 设置manager的工作目录 
manager_log=/masterha/app1/manager.log         # 设置manager的日志
user=manager               # 设置监控用户
password=123             # 监控用户的密码
ssh_user=root            # ssh连接用户
repl_user=mharep         # 主从复制用户 
repl_password=123         # 主从复制用户密码 
ping_interval=1        # 设置监控主库,发送ping包的时间间隔,默认 是3秒

[server1]
hostname=192.168.171.151
port=3306
master_binlog_dir=/usr/local/mysql/data
candidate_master=1               # 设置为候选master,如果设置该参数以后,发生主从切换以后将会将此从库提升为主库

[server2]
hostname=192.168.171.152
port=3306
master_binlog_dir=/usr/local/mysql/data
candidate_master=1

[server3]
hostname=192.168.171.153
port=3306
master_binlog_dir=/usr/local/mysql/data
no_master=1

SSH有效性验证

[ ~]# masterha_check_ssh --global_conf=/etc/masterha/masterha_default.cnf --conf=/etc/masterha/app1.cnf

MySQL高可用之MHA
集群复制的有效性验证(mysql必须都启动)

[ ~]# masterha_check_repl -global_conf=/etc/masterha/masterha_default.cnf --conf=/etc/masterha/app1.cnf
Thu Jul  2 19:20:58 2020 - [info] Reading default configuration from /etc/masterha/masterha_default.cnf..
Thu Jul  2 19:20:58 2020 - [info] Reading application default configuration from /etc/masterha/app1.cnf..
Thu Jul  2 19:20:58 2020 - [info] Reading server configuration from /etc/masterha/app1.cnf..
Thu Jul  2 19:20:58 2020 - [info] MHA::MasterMonitor version 0.58.
Thu Jul  2 19:20:59 2020 - [info] GTID failover mode = 0
Thu Jul  2 19:20:59 2020 - [info] Dead Servers:
Thu Jul  2 19:20:59 2020 - [info] Alive Servers:
Thu Jul  2 19:20:59 2020 - [info]   192.168.171.151(192.168.171.151:3306)
Thu Jul  2 19:20:59 2020 - [info]   192.168.171.152(192.168.171.152:3306)
Thu Jul  2 19:20:59 2020 - [info]   192.168.171.153(192.168.171.153:3306)
Thu Jul  2 19:20:59 2020 - [info] Alive Slaves:
Thu Jul  2 19:20:59 2020 - [info]   192.168.171.152(192.168.171.152:3306)  Version=5.7.22-log (oldest major version between slaves) log-bin:enabled
Thu Jul  2 19:20:59 2020 - [info]     Replicating from 192.168.171.151(192.168.171.151:3306)
Thu Jul  2 19:20:59 2020 - [info]     Primary candidate for the new Master (candidate_master is set)
Thu Jul  2 19:20:59 2020 - [info]   192.168.171.153(192.168.171.153:3306)  Version=5.7.22-log (oldest major version between slaves) log-bin:enabled
Thu Jul  2 19:20:59 2020 - [info]     Replicating from 192.168.171.151(192.168.171.151:3306)
Thu Jul  2 19:20:59 2020 - [info]     Not candidate for the new Master (no_master is set)
Thu Jul  2 19:20:59 2020 - [info] Current Alive Master: 192.168.171.151(192.168.171.151:3306)
Thu Jul  2 19:20:59 2020 - [info] Checking slave configurations..
Thu Jul  2 19:20:59 2020 - [info]  read_only=1 is not set on slave 192.168.171.152(192.168.171.152:3306).
Thu Jul  2 19:20:59 2020 - [warning]  relay_log_purge=0 is not set on slave 192.168.171.153(192.168.171.153:3306).
Thu Jul  2 19:20:59 2020 - [info] Checking replication filtering settings..
Thu Jul  2 19:20:59 2020 - [info]  binlog_do_db= , binlog_ignore_db= 
Thu Jul  2 19:20:59 2020 - [info]  Replication filtering check ok.
Thu Jul  2 19:20:59 2020 - [info] GTID (with auto-pos) is not supported
Thu Jul  2 19:20:59 2020 - [info] Starting SSH connection tests..
Thu Jul  2 19:21:00 2020 - [info] All SSH connection tests passed successfully.
Thu Jul  2 19:21:00 2020 - [info] Checking MHA Node version..
Thu Jul  2 19:21:01 2020 - [info]  Version check ok.
Thu Jul  2 19:21:01 2020 - [info] Checking SSH publickey authentication settings on the current master..
Thu Jul  2 19:21:01 2020 - [info] HealthCheck: SSH to 192.168.171.151 is reachable.
Thu Jul  2 19:21:01 2020 - [info] Master MHA Node version is 0.58.
Thu Jul  2 19:21:01 2020 - [info] Checking recovery script configurations on 192.168.171.151(192.168.171.151:3306)..
Thu Jul  2 19:21:01 2020 - [info]   Executing command: save_binary_logs --command=test --start_pos=4 --binlog_dir=/usr/local/mysql/data --output_file=/var/tmp/save_binary_logs_test --manager_version=0.58 --start_file=mysql-bin.000003 
Thu Jul  2 19:21:01 2020 - [info]   Connecting to (192.168.171.151:22).. 
  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.000003
Thu Jul  2 19:21:01 2020 - [info] Binlog setting check done.
Thu Jul  2 19:21:01 2020 - [info] Checking SSH publickey authentication and checking recovery script configurations on all alive slave servers..
Thu Jul  2 19:21:01 2020 - [info]   Executing command : apply_diff_relay_logs --command=test --slave_user=‘manager‘ --slave_host=192.168.171.152 --slave_ip=192.168.171.152 --slave_port=3306 --workdir=/var/tmp --target_version=5.7.22-log --manager_version=0.58 --relay_log_info=/usr/local/mysql/data/relay-log.info  --relay_dir=/usr/local/mysql/data/  --slave_pass=xxx
Thu Jul  2 19:21:01 2020 - [info]   Connecting to (192.168.171.152:22).. 
  Checking slave recovery environment settings..
    Opening /usr/local/mysql/data/relay-log.info ... ok.
    Relay log found at /usr/local/mysql/data, up to relay-bin.000002
    Temporary relay log file is /usr/local/mysql/data/relay-bin.000002
    Checking if super_read_only is defined and turned on.. not present or turned off, ignoring.
    Testing mysql connection and privileges..
mysql: [Warning] Using a password on the command line interface can be insecure.
 done.
    Testing mysqlbinlog output.. done.
    Cleaning up test file(s).. done.
Thu Jul  2 19:21:01 2020 - [info]   Executing command : apply_diff_relay_logs --command=test --slave_user=‘manager‘ --slave_host=192.168.171.153 --slave_ip=192.168.171.153 --slave_port=3306 --workdir=/var/tmp --target_version=5.7.22-log --manager_version=0.58 --relay_log_info=/usr/local/mysql/data/relay-log.info  --relay_dir=/usr/local/mysql/data/  --slave_pass=xxx
Thu Jul  2 19:21:01 2020 - [info]   Connecting to (192.168.171.153:22).. 
  Checking slave recovery environment settings..
    Opening /usr/local/mysql/data/relay-log.info ... ok.
    Relay log found at /usr/local/mysql/data, up to relay-bin.000002
    Temporary relay log file is /usr/local/mysql/data/relay-bin.000002
    Checking if super_read_only is defined and turned on.. not present or turned off, ignoring.
    Testing mysql connection and privileges..
mysql: [Warning] Using a password on the command line interface can be insecure.
 done.
    Testing mysqlbinlog output.. done.
    Cleaning up test file(s).. done.
Thu Jul  2 19:21:02 2020 - [info] Slaves settings check done.
Thu Jul  2 19:21:02 2020 - [info] 
192.168.171.151(192.168.171.151:3306) (current master)
 +--192.168.171.152(192.168.171.152:3306)
 +--192.168.171.153(192.168.171.153:3306)

Thu Jul  2 19:21:02 2020 - [info] Checking replication health on 192.168.171.152..
Thu Jul  2 19:21:02 2020 - [info]  ok.
Thu Jul  2 19:21:02 2020 - [info] Checking replication health on 192.168.171.153..
Thu Jul  2 19:21:02 2020 - [info]  ok.
Thu Jul  2 19:21:02 2020 - [warning] master_ip_failover_script is not defined.
Thu Jul  2 19:21:02 2020 - [warning] shutdown_script is not defined.
Thu Jul  2 19:21:02 2020 - [info] Got exit code 0 (Not master dead).

MySQL Replication Health is OK.

**验证成功的话会自动识别出所有服务器和主从状况

注:验证成功的话会自动识别出所有服务器和主从状况 在验 证时,若遇到这个错误:Can‘t exec "mysqlbinlog" ...... 解决方法是在所有服务器上执行:**

ln -s /usr/local/mysql/bin/* /usr/local/bin/ 
#启动 manager
[ ~]# nohup masterha_manager --conf=/etc/masterha/app1.cnf &>/tmp/mha_manager.log &
[1] 49094
注:在应用Unix/Linux时,我们一般想让某个程序在后台运行,于是我们将常会用 & 在程序结尾来让程 序自动运行。比如我们要运行mysql在后台: /usr/local/mysql/bin/mysqld_safe –user=mysql &。可是 有很多程序并不想mysqld一样,这样我们就需要nohup命令
#状态检查
[ ~]# masterha_check_status --conf=/etc/masterha/app1.cnf 
app1 (pid:49094) is running(0:PING_OK), master:192.168.171.151

故障转移验证

#模拟master宕机,查看备用master是否会接手
[ ~]# systemctl stop mysqld  
mysql> show slave status\G             # 在slave从上查看到已经更改IP
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.171.152       # 为备用master的IP
                  Master_User: mharep
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000002
          Read_Master_Log_Pos: 746
               Relay_Log_File: relay-bin.000002
                Relay_Log_Pos: 320
        Relay_Master_Log_File: mysql-bin.000002
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes
#查看MHA日志
[ ~]# cat /masterha/app1/manager.log 
Master 192.168.171.151(192.168.171.151:3306) is down!       # down机

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

Started automated(non-interactive) failover.
The latest slave 192.168.171.152(192.168.171.152:3306) has all relay logs for recovery.
Selected 192.168.171.152(192.168.171.152:3306) as a new master.
192.168.171.152(192.168.171.152:3306): OK: Applying all logs succeeded.
192.168.171.153(192.168.171.153:3306): This host has the latest relay log events.
Generating relay diff files from the latest slave succeeded.
192.168.171.153(192.168.171.153:3306): OK: Applying all logs succeeded. Slave started, replicating from 192.168.171.152(192.168.171.152:3306)
192.168.171.152(192.168.171.152:3306): Resetting slave info succeeded.
Master failover to 192.168.171.152(192.168.171.152:3306) completed successfully.      # 更改成功

至此,MHA成功搭建
MHA Manager 端日常主要操作步骤
1)检查是否有下列文件,有则删除

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

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

2)检查MHA复制检查:(需要把master设置成candicatade的从服务器)

mysql> CHANGE MASTER TO MASTER_HOST=‘192.168.18.6‘, MASTER_PORT=3306, MASTER_LOG_FILE=‘mysql-bin.000002‘, MASTER_LOG_POS=154, MASTER_USER=‘mharep‘, MASTER_PASSWORD=‘123‘; Query OK, 0 rows affected, 2 warnings (0.01 sec) 
# masterha_check_repl  --conf=/etc/masterha/app1.cnf

3)停止MHA

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

4)启动MHA

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

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

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

5) 检查状态

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

6) 检查日志

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

7)主从切换后续工作重构

重构就是你的主挂了,切换到Candicate master上,Candicate master变成了主, 因此重构的一种方案原主库修复成一个新的slave 主库切换后,把原主库修复成新从库,然后重新执行以上5 步。原主库数据文件完整的情况下,可通过以下方式找出最后执行的CHANGE MASTER命令:

[ ~]# grep "CHANGE MASTER TO Tue Feb 18 13:14:02 2020 - [info]  All other slaves should start replication from here. Statement should be: CHANGE MASTER TO MASTER_HOST=‘192.168.18.6‘, MASTER_PORT=3306, MASTER_LOG_FILE=‘mysqlbin.000002‘, MASTER_LOG_POS=154, MASTER_USER=‘mharep‘, MASTER_PASSWORD=‘xxx‘; Tue Feb 18 13:14:03 2020 - [info]  Executed CHANGE MASTER. 
mysql> CHANGE MASTER TO MASTER_HOST=‘192.168.18.6‘, MASTER_PORT=3306, MASTER_LOG_FILE=‘mysql-bin.000002‘, MASTER_LOG_POS=154, MASTER_USER=‘mharep‘, MASTER_PASSWORD=‘123‘; 
mysql> start slave;
mysql> show slave status\G 
#启动manager
     # nohup masterha_manager --conf=/etc/masterha/app1.cnf &>/tmp/mha_manager.log &
         # masterha_check_status --conf=/etc/masterha/app1.cnf 
#注意:如果正常,会显示"PING_OK",否则会显示"NOT_RUNNING",这代表MHA监控没有开启。

定期删除中继日志 在配置主从复制中,slave上设置了参数relay_log_purge=0,所以slave节点需要定期删除中继日志,建议每个slave节点删除中继日志的时间错开。

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

MHA后传主要写的是添加了Keepalived和脚本切换VIP

相关推荐