Mysql高可用MHA+Atlas读写分离

最新推荐文章于 2026-04-04 02:45:15 发布

原创最新推荐文章于 2026-04-04 02:45:15 发布 · 464 阅读

CC 4.0 BY-SA版权

本文详细介绍了MHA（Master High Availability）在MySQL高可用环境下的故障切换和主从提升功能，以及Atlas作为读写分离中间件的应用场景和配置步骤。MHA能够实现数据库故障的自动检测和快速切换，确保数据一致性；Atlas则简化了读写分离和从库负载均衡的复杂性，提高了数据库性能。

MHA简介：

MHA（Master High Availability）目前在MySQL高可用方面是一个相对成熟的解决方案，它由日本DeNA公司youshimaton（现就职于Facebook公司）开发，是一套优秀的作为MySQL高可用性环境下故障切换和主从提升的高可用软件。在MySQL故障切换过程中，MHA能做到在0~30秒之内自动完成数据库的故障切换操作，并且在进行故障切换的过程中，MHA能在最大程度上保证数据的一致性，以达到真正意义上的高可用。

该软件由两部分组成：MHA Manager（管理节点）和MHA Node（数据节点）。MHA Manager可以单独部署在一台独立的机器上管理多个master-slave集群，也可以部署在一台slave节点上。MHA Node运行在每台MySQL服务器上，MHA Manager会定时探测集群中的master节点，当master出现故障时，它可以自动将最新数据的slave提升为新的master，然后将所有其他的slave重新指向新的master。整个故障转移过程对应用程序完全透明。

在MHA自动故障切换过程中，MHA试图从宕机的主服务器上保存二进制日志，最大程度的保证数据的不丢失，但这并不总是可行的。例如，如果主服务器硬件故障或无法通过ssh访问，MHA没法保存二进制日志，只进行故障转移而丢失了最新的数据。使用MySQL 5.5的半同步复制，可以大大降低数据丢失的风险。MHA可以与半同步复制结合起来。如果只有一个slave已经收到了最新的二进制日志，MHA可以将最新的二进制日志应用于其他所有的slave服务器上，因此可以保证所有节点的数据一致性。

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。(出自 :《深入浅出MySQL(第二版)》 )

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: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保存的进制日志事件( binlog events) ;
(5)提升- -个slave为新的master ;
(6)使其他的slave连接新的masteri进行复制;

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

Manager工具包主要包括以下几个工具：

masterha_check_ssh #检查MHA的SSH配置状况
masterha_check_repl #检查MySQL复制状况
masterha_check_status #检测当前MHA运行状态
masterha_master_monitor #检测master是否宕机
masterha_manger #启动MHA
masterha_master_switch #控制故障转移（自动或者手动）
masterha_conf_host #添加或删除配置的server信息
masterha_secondary_check #试图建立TCP连接从远程服务器
masterha_stop #停止MHA

Node工具包主要包括以下几个工具：

save_binary_logs #保存和复制master的二进制日志
apply_diff_relay_logs #识别差异的中继日志事件
filter_mysqlbinlog #去除不必要的ROLLBACK事件
purge_relay_logs #清除中继日志

一、安装

环境：Centos7

主机名：c7-41 ip：10.0.0.41（master）

主机名：c7-42 ip：10.0.0.42（slave1）

主机名：c7-43 ip：10.0.0.43（slave2）

主机名：c7-44 ip：10.0.0.44（mha+atlas）

vip：10.0.0.49

1.1 环境准备

以下需要在四台机器上操作

时间同步

echo “*/5 * * * * /usr/sbin/ntpdate ntp1.aliyun.com >/dev/null 2>&1" >>/var/spool/cron/root

hosts解析

cat >>/etc/hosts<<EOF
> c7-41 10.0.0.41
> c7-42 10.0.0.42
> c7-43 10.0.0.43
> c7-44 10.0.0.44
> EOF

关闭防火墙和selinux

互相配置ssh免密码登录

vim ssh.sh

#!/bin/bash
yum -y install sshpass &> /dev/null
read -p "请输入服务器密码：" passwd
UserName=root
IP="10.0.0."
#创建密钥
ssh-keygen -t dsa -f ~/.ssh/id_dsa -P "" &>/dev/null
#分发公钥
for i in 41 42 43 44 #这里的要改成自己机子的ip
do
sshpass -p "$passwd" ssh-copy-id -i ~/.ssh/id_dsa.pub -p 22 -o StrictHostKeyChecking=no $UserName@$IP$i &>/dev/null
done

1.2 mysql主从配置

1.2.1安装yum repo

wget http://repo.mysql.com/mysql-community-release-el7-5.noarch.rpm
rpm -ivh mysql-community-release-el7-5.noarch.rpm
yum -y install mysql-server

1.2.2启动mysql

systemctl restart mysql

修改密码

update mysql.user set password=password('123456') where user='root' and host='localhost';

flush privileges;

1.2.3 配置主从

c7-41 主服务器操作

vim /etc/my.cnf

[mysqld]
server-id=1
log-bin=mysql-bin
#禁止mysql自动删除relaylog工能
relay_log_purge = 0
#mysql5.6已上的特性，开启gtid，必须主从全开
gtid_mode = on
enforce_gtid_consistency = 1
log_slave_updates = 1

#重启mysql

systemctl restart mysql

创建同步用户

mysql -p123456

mysql> grant replication slave on *.* to 'rep'@'10.0.0.%' identified by '123456';

mysql> flush privileges;

查看主库mysql的master状态

mysql> show master status \G;
*************************** 1. row ***************************
File: mysql-bin.000001
Position: 530
Binlog_Do_DB:
Binlog_Ignore_DB:
Executed_Gtid_Set: 4d4640d7-9c13-11ea-a1ee-000c29fda7cf:1-2
1 row in set (0.00 sec)

查看GTID的状态

mysql> show global variables like '%gtid%';
+-------------------------------------------------+---------------------------------------------------------+
| Variable_name | Value |
+------------------------------------------------+-----------------------------------------------------------+
| binlog_gtid_simple_recovery | OFF |
| enforce_gtid_consistency | ON |
| gtid_executed | 4d4640d7-9c13-11ea-a1ee-000c29fda7cf:1-2 |
| gtid_mode | ON |
| gtid_owned | |
| gtid_purged | |
| simplified_binlog_gtid_recovery | OFF |
+-----------------------------------------------+-------------------------------------------------------------+
7 rows in set (0.00 sec)

c7-42 从服务器操作

vim /etc/my.cnf

[mysqld]
server-id=2
log-bin=mysql-bin
#禁止mysql自动删除relaylog工能
relay_log_purge = 0
#mysql5.6已上的特性，开启gtid，必须主从全开
gtid_mode = on
enforce_gtid_consistency = 1
log_slave_updates = 1

#重启mysql

systemctl restart mysql

创建同步用户

mysql -p123456

mysql> grant replication slave on *.* to 'rep'@'10.0.0.%' identified by '123456';

mysql> flush privileges;

关闭从服务器的复制功能

mysql> stop slave;

配置从服务器指向master

change master to master_host='10.0.0.41',master_user='rep',master_password='123456',master_log_file='mysql-bin.000001',master_log_pos=530;

mysql> flush privileges;

开启从服务器的复制功能

mysql> start slave;

检查从服务器的复制功能状态

mysql> show slave status \G;
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.0.0.41
Master_User: rep
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000001
Read_Master_Log_Pos: 530
Relay_Log_File: mysqld-relay-bin.000002
Relay_Log_Pos: 314
Relay_Master_Log_File: mysql-bin.000001
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
是否双Yes

c7-43 从服务器操作

vim /etc/my.cnf

[mysqld]
server-id=3
log-bin=mysql-bin
#禁止mysql自动删除relaylog工能
relay_log_purge = 0
#mysql5.6已上的特性，开启gtid，必须主从全开
gtid_mode = on
enforce_gtid_consistency = 1
log_slave_updates = 1

#重启mysql

systemctl restart mysql

创建同步用户

mysql -p123456

mysql> grant replication slave on *.* to 'rep'@'10.0.0.%' identified by '123456';

mysql> flush privileges;

关闭从服务器的复制功能

mysql> stop slave;

配置从服务器指向master

mysql> change master to master_host='10.0.0.41',master_user='rep',master_password='123456',master_log_file='mysql-bin.000001',master_log_pos=530;

mysql> flush privileges;

开启从服务器的复制功能

mysql> start slave;

检查从服务器的复制功能状态

mysql> show slave status\G;
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.0.0.41
Master_User: rep
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000001
Read_Master_Log_Pos: 530
Relay_Log_File: mysqld-relay-bin.000002
Relay_Log_Pos: 314
Relay_Master_Log_File: mysql-bin.000001
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
是否为双Yes

什么是GTID
●GTID(全球交易)全局事务标识符:是-一个唯一的标识符,它创建并与源服务器(主). 上提交的每个事务相关
联。此标识符不仅对其发起的服务器是唯一的,而且在给定复制设置中的所有服务器上都是唯一的。所有交易和
所有GTID之间都有1对1的映射。
●GTID实际上是由UUID + TID组成的。其中UUID是一个MySQL的实例的唯一标识TID代表了该实例上已经提交的
事务数量,并且随着事务提交单调递增。
GTID的新特性
(1)支持多线程复制:事实上是针对每个数据库开启相应的独立线程,即每个库有一个单独的( sql thread )
(2)支持启用GTID ,在配置主从复制,传统的方式里，你需要找到binlog和POS点,然后将master改为指向。在
mysq|5.6里,无须再知道binlog和POS点,只需要知道master的IP /端口/账号密码即可,因为同步复制是自动的，
MySQL的通过内部机制GTID自动找点同步。
(3)基于行复制只保存改变的列,大大节省磁盘空间，网络,内存等
(4)支持把主站和从站的相关信息记录在表中;原来是记录在文件里,现在则记录在表里,增强可用性
(5)支持延迟复制

1.3 安装MHA

所有节点操作

1.3.1 安装依赖

yum install perl-DBD-MySQL -y

yum install -y perl-Config-Tiny epel-release perl-Log-Dispatch perl-Parallel-ForkManager perl-Time-HiRes

1.3.2 授权MHA管理用户

grant all privileges on *.* to mha@'10.0.0.%' identified by 'mha';

flush privileges;

1.3.3 安装MHA node节点

上传 mha4mysql-node-0.58-0.el7.centos.noarch.rpm

rpm -ivh mha4mysql-node-0.58-0.el7.centos.noarch.rpm

1.3.4 安装MHA管理节点

注意:MHA管理节点不要装到mysql主库和切换的从库上(备用的主库),否则会在后面的vip无法漂移

[root@ c7-44 ~]# rpm -ivh mha4mysql-manager-0.58-0.el7.centos.noarch.rpm

附带源码安装方式:

yum -y instal1 per1-DBD-MySQL per1-Config-Tiny per1-Par ams -validate per1-CPAN per 1-deve1
per 1-ExtUtils-CBuilder per 1-ExtUtils-MakeMaker

node节点:
tar xf mha4mysq1-node-0. 58.tar .9z
cd mha4mysq1-node-0. 58
per1 Makefile.PL
make & make insta11

manager节点:
tar xf mha4mysq1 -manager-0.58. tar .gz
cd mha4mysq1-manager-0. 58
per1 Makefile.PL
make & make install

#安装完成后会在/usr/1oca1/bin目录下面生成以下脚本文件

1.3.5 配置MHA

[root@ c7-44 ~]# mkdir -p /etc/mha
[root@ c7-44 ~]# mkdir -p /var/log/mha/app1
[root@ c7-44 ~]# vim /etc/mha/app1.cnf

[server default]
manager_log=/var/log/mha/app1/manager.log
manager_workdir=/var/log/mha/app1
master_binlog_dir=/var/lib/mysql
user=mha
password=mha
ping_interval=2
repl_password=123456
repl_user=rep
ssh_user=root

[server1]
hostname=10.0.0.41
port=3306

[server2]
hostname=10.0.0.42
port=3306

[server3]
hostname=10.0.0.43
port=3306
ignore_fail=1
no_master=1
#candidate_master=1
#check_repl_delay=0

1.3.6 启动测试

ssh检查检测

[root@ c7-44 ~]# masterha_check_ssh --conf=/etc/mha/app1.cnf
Sat May 23 11:09:45 2020 - [warning] Global configuration file /etc/masterha_default.cnf not found. Skipping.
Sat May 23 11:09:45 2020 - [info] Reading application default configuration from /etc/mha/app1.cnf..
Sat May 23 11:09:45 2020 - [info] Reading server configuration from /etc/mha/app1.cnf..
Sat May 23 11:09:45 2020 - [info] Starting SSH connection tests..
Sat May 23 11:09:46 2020 - [debug]
Sat May 23 11:09:45 2020 - [debug] Connecting via SSH from root@10.0.0.41(10.0.0.41:22) to root@10.0.0.42(10.0.0.42:22)..
Sat May 23 11:09:46 2020 - [debug] ok.
Sat May 23 11:09:46 2020 - [debug] Connecting via SSH from root@10.0.0.41(10.0.0.41:22) to root@10.0.0.43(10.0.0.43:22)..
Sat May 23 11:09:46 2020 - [debug] ok.
Sat May 23 11:09:47 2020 - [debug]
Sat May 23 11:09:45 2020 - [debug] Connecting via SSH from root@10.0.0.42(10.0.0.42:22) to root@10.0.0.41(10.0.0.41:22)..
Sat May 23 11:09:46 2020 - [debug] ok.
Sat May 23 11:09:46 2020 - [debug] Connecting via SSH from root@10.0.0.42(10.0.0.42:22) to root@10.0.0.43(10.0.0.43:22)..
Sat May 23 11:09:47 2020 - [debug] ok.
Sat May 23 11:09:47 2020 - [debug]
Sat May 23 11:09:46 2020 - [debug] Connecting via SSH from root@10.0.0.43(10.0.0.43:22) to root@10.0.0.41(10.0.0.41:22)..
Sat May 23 11:09:46 2020 - [debug] ok.
Sat May 23 11:09:46 2020 - [debug] Connecting via SSH from root@10.0.0.43(10.0.0.43:22) to root@10.0.0.42(10.0.0.42:22)..
Sat May 23 11:09:47 2020 - [debug] ok.
Sat May 23 11:09:47 2020 - [info] All SSH connection tests passed successfully.

主从复制检测

[root@ c7-44 ~]# masterha_check_repl --conf=/etc/mha/app1.cnf

1.3.7 启动MHA

[root@ c7-44 ~]# nohup masterha_manager --conf=/etc/mha/app1.cnf --remove_dead_master_conf --ignore_last_failover < /dev/null > /var/log/mha/app1/manager.log 2>&1 &
[1] 19853

查看MHA状态

[root@ c7-44 ~]# masterha_check_status --conf=/etc/mha/app1.cnf
app1 (pid:19853) is running(0:PING_OK), master:10.0.0.41

关闭MHA

[root@ c7-44 ~]# masterha_stop --conf=/etc/mha/app1.cnf

从库重新加入新主

[root@ c7-44 ~]# grep -i "CHANGE MASTER TO MASTER" /var/log/mha/app1/manager.log | tail -1

1.4 测试MHA故障漂移

停掉c7-41主库的10.0.0.41

[root@ c7-41 ~]# systemctl stop mysql

查看c7-43的slave状态，发现Master_Host变成了10.0.0.42

mysql> show slave status\G;
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.0.0.42
Master_User: rep
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000002
Read_Master_Log_Pos: 1319
Relay_Log_File: mysqld-relay-bin.000003
Relay_Log_Pos: 871
Relay_Master_Log_File: mysql-bin.000002
Slave_IO_Running: Yes
Slave_SQL_Running: Yes

查看c7-42的master状态

[root@ c7-42 ~]# mysql -uroot -p123456 -e "show master status;"
Warning: Using a password on the command line interface can be insecure.
+-------------------------+----------+--------------+------------------+------------------------------------------------------------------------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+-------------------------+----------+--------------+------------------+------------------------------------------------------------------------------------+
| mysql-bin.000002 | 1319

查看c7-44的MHA配置文件

[root@ c7-44 ~]# cat /etc/mha/app1.cnf
[server default]
manager_log=/var/log/mha/app1/manager.log
manager_workdir=/var/log/mha/app1
master_binlog_dir=/var/lib/mysql
password=mha
ping_interval=2
repl_password=123456
repl_user=rep
ssh_user=root
user=mha

#少了一个[server1]

[server2]
hostname=10.0.0.42
port=3306

[server3]
hostname=10.0.0.43
ignore_fail=1
no_master=1
port=3306

发生故障时，MHA做了什么？

①当作为主库的c7-41上的MySQL宕机以后，mha通过检测发现c7-41的mysql宕机了，那么会将binlog日志最全的从库（c7-42）立刻提升为主库，而其他的从库会指向新的主库进行再次同步。

②MHA会自己结束自己的进程，还会将/etc/mha/app1.cnf配置文件中，坏掉的那台机器剔除。

1.4.1 MHA故障还原

1.先将宕机的mysql修复,然后加入到mysql一主两从集群

[root@ c7-44 ~]# grep "CHANGE MASTER TO MASTER" /var/log/mha/app1/manager.log | tail -1
Sat May 23 11:42:33 2020 - [info] All other slaves should start replication from here. Statement should be: CHANGE MASTER TO MASTER_HOST='10.0.0.42', MASTER_PORT=3306, MASTER_AUTO_POSITION=1, MASTER_USER='rep', MASTER_PASSWORD='xxx';

[root@ c7-41 ~]# systemctl restart mysql

[root@ c7-41 ~]# mysql -uroot -p123456 -e "CHANGE MASTER TO MASTER_HOST='10.0.0.42', MASTER_PORT=3306, MASTER_AUTO_POSITION=1, MASTER_USER='rep', MASTER_PASSWORD='123456';"
Warning: Using a password on the command line interface can be insecure.

[root@ c7-41 ~]# mysql -uroot -p123456 -e "start slave;"
Warning: Using a password on the command line interface can be insecure.

[root@ c7-41 ~]# mysql -uroot -p123456 -e "show slave status \G"
Warning: Using a password on the command line interface can be insecure.
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.0.0.42
Master_User: rep
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000002
Read_Master_Log_Pos: 1319
Relay_Log_File: mysqld-relay-bin.000003
Relay_Log_Pos: 871
Relay_Master_Log_File: mysql-bin.000002
Slave_IO_Running: Yes
Slave_SQL_Running: Yes

2.重新在将[server1]标签添加到MHA配置文件,并且启动MHA

[server1]
hostname=10.0.0.41
port=3306

3.重新启动MHA

[root@ c7-44 ~]# nohup masterha_manager --conf=/etc/mha/app1.cnf --remove_dead_master_conf --ignore_last_failover < /dev/null > /var/log/mha/app1/manager.log 2>&1 &
[1] 19798

[root@ c7-44 ~]# masterha_check_status --conf=/etc/mha/app1.cnf
app1 (pid:19798) is running(0:PING_OK), master:10.0.0.42

1.5 配置vip漂移

1.5.1IP漂移的两种方式

①通过keepalived的方式,管理虚拟IP的漂移。
②通过MHA自带脚本的方式,管理虚拟IP的漂移#用mha自带的一 -个VIP漂移的脚本,哪个提升为主,就飘到那个上
面,要根据binlog最新的slave提升。

1.5.2MHA脚本方式

修改MHA配置文件

[root@ c7-44 ~]# vim /etc/mha/app1.cnf

[server default]
master_ip_failover_script=/usr/bin/master_ip_failover

编写漂移脚本 master_ip_failover

注意:修改脚本中的网卡名和ip地址

[root@ c7-44 ~]# vim /usr/bin/master_ip_failover

#!/usr/bin/env perl

use strict;
use warnings FATAL => 'all';

use Getopt::Long;

my (
$command, $ssh_user, $orig_master_host, $orig_master_ip,
$orig_master_port, $new_master_host, $new_master_ip, $new_master_port
);

my $vip = '10.0.0.49/24';
my $key = '1';
my $ssh_start_vip = "/sbin/ifconfig ens33:$key $vip";
my $ssh_stop_vip = "/sbin/ifconfig ens33:$key down";

GetOptions(
'command=s' => \$command,
'ssh_user=s' => \$ssh_user,
'orig_master_host=s' => \$orig_master_host,
'orig_master_ip=s' => \$orig_master_ip,
'orig_master_port=i' => \$orig_master_port,
'new_master_host=s' => \$new_master_host,
'new_master_ip=s' => \$new_master_ip,
'new_master_port=i' => \$new_master_port,
);

exit &main();

sub main {

print "\n\nIN SCRIPT TEST====$ssh_stop_vip==$ssh_start_vip===\n\n";

if ( $command eq "stop" || $command eq "stopssh" ) {

my $exit_code = 1;
eval {
print "Disabling the VIP on old master: $orig_master_host \n";
&stop_vip();
$exit_code = 0;
};
if ($@) {
warn "Got Error: $@\n";
exit $exit_code;
}
exit $exit_code;
}
elsif ( $command eq "start" ) {

my $exit_code = 10;
eval {
print "Enabling the VIP - $vip on the new master - $new_master_host \n";
&start_vip();
$exit_code = 0;
};
if ($@) {
warn $@;
exit $exit_code;
}
exit $exit_code;
}
elsif ( $command eq "status" ) {
print "Checking the Status of the script.. OK \n";
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";
}

[root@ c7-44 ~]# chmod +x /usr/bin/master_ip_failover

1.5.3 手动绑定vip

[root@ c7-42 ~]# ifconfig ens33:1 10.0.0.49/24
[root@ c7-42 ~]# ip a show ens33
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether 00:0c:29:76:cc:ce brd ff:ff:ff:ff:ff:ff
inet 10.0.0.42/24 brd 10.0.0.255 scope global ens33
valid_lft forever preferred_lft forever
inet 10.0.0.49/24 brd 10.0.0.255 scope global secondary ens33:1
valid_lft forever preferred_lft forever
inet6 fe80::20c:29ff:fe76:ccce/64 scope link
valid_lft forever preferred_lft forever

1.5.4 重启MHA

[root@ c7-44 ~]# masterha_check_status --conf=/etc/mha/app1.cnf
app1 (pid:19798) is running(0:PING_OK), master:10.0.0.42
[root@ c7-44 ~]# masterha_stop --conf=/etc/mha/app1.cnf
Stopped app1 successfully.
[1]+ Exit 1 nohup masterha_manager --conf=/etc/mha/app1.cnf --remove_dead_master_conf --ignore_last_failover < /dev/null > /var/log/mha/app1/manager.log 2>&1
[root@ c7-44 ~]# nohup masterha_manager --conf=/etc/mha/app1.cnf --remove_dead_master_conf --ignore_last_failover < /dev/null > /var/log/mha/app1/manager.log 2>&1 &
[1] 20194

[root@ c7-44 ~]# masterha_check_status --conf=/etc/mha/app1.cnf
app1 (pid:20194) is running(0:PING_OK), master:10.0.0.42

1.5.5模拟主库宕机vip漂移

关闭c7-42主库的mysql

[root@ c7-42 ~]# systemctl stop mysql
[root@ c7-42 ~]# ip a show ens33
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether 00:0c:29:76:cc:ce brd ff:ff:ff:ff:ff:ff
inet 10.0.0.42/24 brd 10.0.0.255 scope global ens33
valid_lft forever preferred_lft forever
inet6 fe80::20c:29ff:fe76:ccce/64 scope link
valid_lft forever preferred_lft forever

现在查看c7-41上mysql的master状态和vip

[root@ c7-41 ~]# ip a show ens33
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether 00:0c:29:fd:a7:cf brd ff:ff:ff:ff:ff:ff
inet 10.0.0.41/24 brd 10.0.0.255 scope global ens33
valid_lft forever preferred_lft forever
inet 10.0.0.49/24 brd 10.0.0.255 scope global secondary ens33:1
valid_lft forever preferred_lft forever
inet6 fe80::20c:29ff:fefd:a7cf/64 scope link
valid_lft forever preferred_lft forever
[root@ c7-41 ~]# mysql -uroot -p123456 -e 'show master status\G'
Warning: Using a password on the command line interface can be insecure.
*************************** 1. row ***************************
File: mysql-bin.000003
Position: 1449
Binlog_Do_DB:
Binlog_Ignore_DB:
Executed_Gtid_Set: 4d4640d7-9c13-11ea-a1ee-000c29fda7cf:1-6,
ff6afc95-9c11-11ea-a1e6-000c2976ccce:1-7

二. Mysql之Atlas(读写分离)

数据库中间件Atlas与Mycat比较分库分表压测报告https://blog.csdn.net/izhitao/article/details/71680714

Atlas是由Qihoo 360公司Web平台部基础架构团队开发维护的一-个基于MySQL协议的数据中间层项目。它在MySQL言
方推出的MySQL-Proxy 0.8.2版本的基础上,修改了大量bug ,添加了很多功能特性。目前该项目在360公司内部得到了广
泛应用,很多MySQL业务已经接入了Atlas平台,每天承载的读写请求数达几十亿条。
源码Github : https://github.com/Qihoo360/Atlas

2.1什么是读写分离

读写分离,基本的原理是让主数据库处理事务性增、删、改操作( INSERT. DELETE、 UPDATE ) , 而从数据库处理
SELECT查询操作。数据库复制把主数据库操作的变更同步到集群中的从数据库。

2.2为什么读写分离.

因为数据库的“写" (写10000条数据到oracle可能要3分钟)操作是比较耗时的。但是数据库的“读" (从oracle读10000
条数据可能只要5秒钟)。所以读写分离,解决的是，数据库的写入，影响了查询的效率。

2.3什么时候要读写分离

数据库不一定要读写分离,如果程序使用数据库较多时,而更新少,查询多的情况下会考虑使用,利用数据库主从同步
可以减少数据库压力,提高性能。当然,数据库也有其它优化方案。memcache 或是表折分,或是搜索引学。都是解
决方法。

2.4 Atlas的功用与应用场景

Atlas的功能有:
读写分离、从库负载均衡、自动分表、IP过滤、 SQL语句黑白名单、DBA可平滑上下线DB、自动摘除宕机的DB。

Atlas的使用场景:
Atlas是一个位于前端应用与后端MySQL数据库之间的中间件,它使得应用程序员无需再关心读写分离、分表等与
MySQL相关的细节,可以专注于编写业务逻辑,同时使得DBA的运维工作对前端应用透明，上下线DB前端应用无感知。

2.5 Atlas的安装过程

注意:
1.Atlas只能安装运行在64位的系统上
2.Centos 5.X安装Atlas-XX.el5.x86_ _64.rpm , Centos 6.X安装Atlas-XX.el6.x86 _64.rpm(经过测试centos7也可以使用6
的版本)
3、后端mysq|版本应大于5.1 ,建议使用Mysql 5.6以上

1.安装Atlas
[root@ c7-44 ~]# wget https://github.com/Qihoo360/Atlas/releases/download/2.2.1/Atlas-2.2.1.el6.x86_64.rpm
[root@ c7-44 ~]# rpm -ivh Atlas-2.2.1.el6.x86_64.rpm

2.修改配置文件

[root@ c7-44 ~]# egrep -v '^#|^$' /usr/local/mysql-proxy/conf/test.cnf
[mysql-proxy]
admin-username = user
admin-password = pwd
proxy-backend-addresses = 10.0.0.49:3306 #设置写入主库vip的地址
proxy-read-only-backend-addresses = 10.0.0.42:3306,10.0.0.43:3306 #设置只读的从库地址
pwds = rep:/iZxz+0GRoA=,mha:O2jBXONX098=
daemon = true
keepalive = true
event-threads = 8
log-level = message
log-path = /usr/local/mysql-proxy/log
sql-log =ON
proxy-address = 0.0.0.0:1234
admin-address = 0.0.0.0:2345
charset = utf8

#加密方式 /usr/local/mysql-proxy/bin/encrypt 密码
[root@c704 ~]# /usr/local/mysql-proxy/bin/encrypt 123456
/iZxz+0GRoA=
[root@c704 ~]# /usr/local/mysql-proxy/bin/encrypt mha
O2jBXONX098=

3.启动Atlas

[root@ c7-44 ~]# /usr/local/mysql-proxy/bin/mysql-proxyd test start
OK: MySQL-Proxy of test is started
#Atlas可以通过不同配置文件，同时启动多个Atlas代理多套MHA节点;

#/usr/local/mysql-proxy/bin/mysql-proxyd test start
#/usr/local/mysql-proxy/bin/mysql-proxyd test stop
#/usr/local/mysql-proxy/bin/mysql-proxyd test restart

4.查看Atlas

[root@ c7-44 ~]# ps -ef|grep mysql-proxy
root 20400 1 0 12:34 ? 00:00:00 /usr/local/mysql-proxy/bin/mysql-proxy --defaults-file=/usr/local/mysql-proxy/conf/test.cnf
root 20401 20400 0 12:34 ? 00:00:00 /usr/local/mysql-proxy/bin/mysql-proxy --defaults-file=/usr/local/mysql-proxy/conf/test.cnf
root 20415 19433 0 12:36 pts/1 00:00:00 grep --color=auto mysql-proxy

2.6 Atlas读写分离测试

读测试:

[root@ c7-44 ~]# mysql -umha -pmha -P1234 -h10.0.0.44

mysql> select @@server_id;
+-------------------+
| @@server_id |
+-------------------+
| 3 |
+-------------------+
1 row in set (0.00 sec)

mysql> select @@server_id;
+-------------------+
| @@server_id |
+-------------------+
| 2 |
+-------------------+
1 row in set (0.00 sec)

#注:发现上面'server. jid每次的结果都不-样，分别是2台从库的server_id ,并且每执行一次命令 , server_id就会变换一次，这是因为默认读操作的权重都是1，两台从DB默认就是负载均衡。

写测试：

[root@ c7-44 ~]# mysql -umha -pmha -P1234 -h10.0.0.44

mysql> begin;select @@server_id;commit;
ERROR 2013 (HY000): Lost connection to MySQL server during query
ERROR 2006 (HY000): MySQL server has gone away
No connection. Trying to reconnect...
Connection id: 16
Current database: *** NONE ***
#此报错是，连不上写的数据库，注意vip是否已经绑定到机器上

mysql> begin;select @@server_id;commit;
Query OK, 0 rows affected (0.00 sec)

+--------------------+
| @@server_id |
+--------------------+
| 1 |
+--------------------+
1 row in set (0.00 sec)

Query OK, 0 rows affected (0.00 sec)

mysql> create database www;
Query OK, 1 row affected (0.10 sec)

mysql> show databases;
+--------------------+
| Database |
+--------------------+
| information_schema |
| mysql |
| performance_schema |
| www |
+--------------------+
4 rows in set (0.12 sec)

2.7 Atlas管理操作

登入管理接口

[root@ c7-44 ~]# mysql -uuser -ppwd -P2345 -h 10.0.0.44

查看帮助信息

mysql> SELECT * FROM help;
+--------------------------------------+-----------------------------------------------------------------+
| command | description |
+--------------------------------------+------------------------------------------------------------------+
| SELECT * FROM help | shows this help |
| SELECT * FROM backends | lists the backends and their state |
| SET OFFLINE $backend_id | offline backend server, $backend_id is backend_ndx's id |
| SET ONLINE $backend_id | online backend server, ... |
| ADD MASTER $backend | example: "add master 127.0.0.1:3306", ... |
| ADD SLAVE $backend | example: "add slave 127.0.0.1:3306", ... |
| REMOVE BACKEND $backend_id | example: "remove backend 1", ... |
| SELECT * FROM clients | lists the clients |
| ADD CLIENT $client | example: "add client 192.168.1.2", ... |
| REMOVE CLIENT $client | example: "remove client 192.168.1.2", ... |
| SELECT * FROM pwds | lists the pwds |
| ADD PWD $pwd | example: "add pwd user:raw_password", ... |
| ADD ENPWD $pwd | example: "add enpwd user:encrypted_password", ... |
| REMOVE PWD $pwd | example: "remove pwd user", ... |
| SAVE CONFIG | save the backends to config file |
| SELECT VERSION | display the version of Atlas |
+----------------------------------+--------------------------------------------------------------------+
16 rows in set (0.00 sec)

查看后端的代理库

16 rows in set (0.00 sec)

mysql> SELECT * FROM backends;
+-------------+----------------+-------+----------+
| backend_ndx | address | state | type |
+-------------+----------------+-----------+------+
| 1 | 10.0.0.49:3306 | up | rw |
| 2 | 10.0.0.42:3306 | down | ro |
| 3 | 10.0.0.43:3306 | up | ro |
+-------------+----------------+-------+-----------+
3 rows in set (0.00 sec)

下线后端节点

mysql> set offline 3;
+-------------+----------------+---------+------+
| backend_ndx | address | state | type |
+-------------+----------------+---------+------+
| 3 | 10.0.0.43:3306 | offline | ro |
+-------------+----------------+---------+------+
1 row in set (0.00 sec)

mysql> SELECT * FROM backends;
+-------------+----------------+---------+------+
| backend_ndx | address | state | type |
+-------------+----------------+---------+------+
| 1 | 10.0.0.49:3306 | up | rw |
| 2 | 10.0.0.42:3306 | up | ro |
| 3 | 10.0.0.43:3306 | offline | ro |
+-------------+----------------+---------+------+
3 rows in set (0.00 sec)

上线后端节点

mysql> set online 3;
+-------------+----------------+---------+------+
| backend_ndx | address | state | type |
+-------------+----------------+---------+------+
| 3 | 10.0.0.43:3306 | unknown | ro |
+-------------+----------------+---------+------+
1 row in set (0.00 sec)

mysql> SELECT * FROM backends;
+-------------+----------------+-------+------+
| backend_ndx | address | state | type |
+-------------+----------------+-------+------+
| 1 | 10.0.0.49:3306 | up | rw |
| 2 | 10.0.0.42:3306 | up | ro |
| 3 | 10.0.0.43:3306 | up | ro |
+-------------+----------------+-------+------+
3 rows in set (0.00 sec)