MySQL高可用集群MHA方案
MySQL高可用集群MHA方案
爱奇艺在用的数据库高可用方案
MHA 是目前比较成熟及流行的 MySQL 高可用解决方案，很多互联网公司正是直接使用或者基于 MHA 的架构进行改造实现 MySQL 的高可用。
MHA 能在 30 秒内对故障进行转移，并最大程度的保障数据的一致性。MHA 由两个模块组成：Manager 和 Node。
什么是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，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。
MHA作用是保证MySQL主从复制集群中的master高可用性，也就保证整个数据库集群业务不被故障影响。
- master故障时，MHA会在30s内实现故障自动检测+故障转移
- 选择一个最优的slave接替为新的master，并且保证new_master和其他slave继续保持数据一致性
高可用性HA、high availability
指的是一个经过设计的系统，能保证减少架构故障时的停工时间，保证业务程序的高度可用性
超哥也在各种运维业务场景下，接触过HA软件
无论是web、数据库、还是后端
MHA架构
整个MHA软件由两部分角色组成，即MHA Manager（管理节点）和MHA Node（数据节点）。
MHA Manager服务可以独立部署在一台linux机器，也可以部署在某一台主从复制从节点或者其他应用服务器节点上。
而MHA Node服务需要运行在每一个MySQL服务器上。
MHA Manager会定时通过主库上的MHA Node服务监测主库，当master出现故障时，它可以自动将最优slave（可以提前指定或由MHA判定）提升为新的master，然后让所有其他的从库与新的主库重新保持正常的复制状态。
故障的整个切换和转移的过程对客户以及应用程序几乎是完全透明的（也就是用户不会感知到有故障发生）
MHA工作原理
MHA主要功能
- master宕机、切换新的master，且保证其他slave和新的master保持一致复制
- 故障切换过程中，集群数据丢失量最小
一、选择新master
old_master宕机，在集群中选择一个新的slave作为new_master，这要根据MHA的配置，如根据其他slave的binlog位置点，选择最新的slave作为new_master
二、数据补全
进行故障切换、转移之前，必须要进行数据补全，否则即使故障切换了，数据丢了那也是不允许的
数据补全过程
- old_master数据库服务器还可以连接，MHA会SSH连接主库，保存主库所有的binlog
- 若ssh无法连接，放弃主库的binlog数据
- 以切换好的new_master主库的binlog位置点位基准点，通过
relay_log进行数据补全，使得其他所有slave和new_master数据一直 - 将宕机时从old_master上保存下来的binlog日志（如果存在的话）恢复到所有的数据库节点.
三、角色切换
- 已选择好的new_master正式提升为主库角色
- 其他的slave和new_master保持主从复制关系
四、有关master主库IP切换的问题，可以结合keepalived的VIP漂移来实现
MHA软件包介绍
MHA由2部分组成
Manager节点
Node节点
Manager节点命令
Node命令
MHA特点
- old_master宕机，slave快速切换为new_master
- 部署MHA与不会对现有的MySQL集群做大量改动
- MHA——manager功能强大，可以管理上百个节点、多套mysql集群
- 可以监控mysql状态，隔N秒向master发送ping包，性能不受影响
- 只要mySQL主从复制支持的存储引擎，MHA也都支持，不限于InnoDB
MHA部署
2台linux及以上
多个mysql实例之间实现复制
这里超哥准备四台linux机器
master1 10.211.55.12 MHA-node
Slave1 10.211.55.9 MHA-node
Slave2 10.211.55.11 MHA-node，MHA-Manager
Client 10.211.55.18 空
注意时间同步
ntpdate -u ntp.aliyun.com准备好一主两从-GTID
MHA需要支持一主多从架构，至少三台数据库，三台机器，基于GTID的主从复制，三个配置文件，仅有server-id不同
一个master
一个备用master
一个slave
额外配置my.cnf
relay_log_purge = 0 #<==不自动删除relay log，以便于宕机后修复数据。
log-bin=/mm_data/3306/mysql-bin #<==从库开启binlog，以便于宕机
后修复数据。
expire_logs_days = 7 #<==自动删除7天前的binlog。
log-slave-updates = 1 #<==从库开启Binlog，以便于宕机后修复数据。master基础配置
My.cnf
[client]
socket=/mm_data/3306/mysql.sock
[mysqld]
socket=/mm_data/3306/mysql.sock
basedir=/application/mysql-5.6.40-linux-glibc2.12-x86_64/
datadir=/mm_data/3306/data
log-bin=/mm_data/3306/mysql-bin
character-set-server=utf8
server-id=12
expire-logs-days=1
binlog_format=row
gtid_mode=on
enforce_gtid_consistency=1
log_slave_updates=1
[mysqld_safe]
log-error=/mm_data/3306/mysql_3306_error.log
pid-file=/mm_data/3306/mysqld_3306.pid
sql_mode=NO_ENGINE_SUBSTITUTION,STRTICT_TRANS_TABLES数据目录
[root@mysql-server56 tools]# /mm_data/3306/mysql_3306 restart
Restarting MySQL...
Stoping MySQL...
Starting MySQL...
[root@mysql-server56 tools]# ls /mm_data/3306/
data my.cnf mysql_3306 mysql_3306_error.log mysql-bin.000001 mysql-bin.000002 mysql-bin.index mysqld_3306.pid mysql.sock
[root@mysql-server56 tools]#
[root@mysql-server56 tools]# netstat -tunlp|grep mysql
tcp6 0 0 :::3306 :::* LISTEN 11040/mysqld
[root@mysql-server56 tools]#主库信息
mysql> show master status\G
*************************** 1. row ***************************
File: mysql-bin.000002
Position: 151
Binlog_Do_DB:
Binlog_Ignore_DB:
Executed_Gtid_Set:
1 row in set (0.00 sec)
mysql>
mysql> show slave status;
Empty set (0.00 sec)
mysql>创建复制账号
mysql> grant replication slave on *.* to 'repl_chaoge'@'10.211.55.%' identified by 'chaoge668';
Query OK, 0 rows affected (0.00 sec)
mysql> flush privileges;
Query OK, 0 rows affected (0.00 sec)
mysql>数据导出
[root@mysql-server56 3306]# mysqldump -S /mm_data/3306/mysql.sock -A -B -x --set-gtid-purged=OFF|gzip > /mm_data/m1_alldb_$(date +%F).sql.gz数据发给所有slave
[root@mysql-server56 tools]# scp -rp /mm_data/m1_alldb_2021-04-27.sql.gz root@10.211.55.9:/mm_data/
m1_alldb_2021-04-27.sql.gz 100% 177KB 28.6MB/s 00:00
[root@mysql-server56 tools]# scp -rp /mm_data/m1_alldb_2021-04-27.sql.gz root@10.211.55.11:/mm_data/
m1_alldb_2021-04-27.sql.gz 100% 177KB 38.5MB/s 00:00
[root@mysql-server56 tools]#查看GTID信息
mysql> show global variables like '%gtid%';
+---------------------------------+------------------------------------------+
| Variable_name | Value |
+---------------------------------+------------------------------------------+
| binlog_gtid_simple_recovery | OFF |
| enforce_gtid_consistency | ON |
| gtid_executed | 20bdbbd9-a5cd-11eb-a6af-001c4279bcf3:1-5 |
| gtid_mode | ON |
| gtid_owned | |
| gtid_purged | |
| simplified_binlog_gtid_recovery | OFF |
+---------------------------------+------------------------------------------+
7 rows in set (0.00 sec)
mysql>
mysql> show global variables like 'server%';
+----------------+--------------------------------------+
| Variable_name | Value |
+----------------+--------------------------------------+
| server_id | 12 |
| server_id_bits | 32 |
| server_uuid | 20bdbbd9-a5cd-11eb-a6af-001c4279bcf3 |
+----------------+--------------------------------------+
3 rows in set (0.00 sec)slave1~10.211.55.9~基础配置
my.cnf server-id=9
[client]
socket=/mm_data/3306/mysql.sock
[mysqld]
socket=/mm_data/3306/mysql.sock
basedir=/application/mysql-5.6.40-linux-glibc2.12-x86_64/
datadir=/mm_data/3306/data
log-bin=/mm_data/3306/mysql-bin
character-set-server=utf8
server-id=9
expire-logs-days=1
binlog_format=row
gtid_mode=on
enforce_gtid_consistency=1
log_slave_updates=1
[mysqld_safe]
log-error=/mm_data/3306/mysql_3306_error.log
pid-file=/mm_data/3306/mysqld_3306.pid
sql_mode=NO_ENGINE_SUBSTITUTION,STRTICT_TRANS_TABLES导入master数据
[root@chaoge_slave1 3306]# zcat /mm_data/m1_alldb_2021-04-27.sql.gz |mysql -S /mm_data/3306/mysql.sock
[root@chaoge_slave1 3306]# ./mysql_3306 restart
Restarting MySQL...
Stoping MySQL...
Starting MySQL...
[root@chaoge_slave1 3306]# ls
data my.cnf mysql_3306 mysql_3306_error.log mysql-bin.000001 mysql-bin.index mysqld_3306.pid mysql.sock
[root@chaoge_slave1 3306]#授权master-info
mysql> change master to
-> master_host='10.211.55.12',
-> master_port=3306,
-> master_user='repl_chaoge',
-> master_password='chaoge668',
-> master_auto_position=1;
Query OK, 0 rows affected, 2 warnings (0.00 sec)启动slave
mysql> start slave;
mysql> show slave status\G
mysql> show slave status\G
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.211.55.12
Master_User: repl_chaoge
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000002
Read_Master_Log_Pos: 644733
Relay_Log_File: mysqld_3306-relay-bin.000002
Relay_Log_Pos: 408
Relay_Master_Log_File: mysql-bin.000002
Slave_IO_Running: Yes
Slave_SQL_Running: Yesslave2~10.211.55.11~基础配置
my.cnf
server-id=11
[client]
socket=/mm_data/3306/mysql.sock
[mysqld]
socket=/mm_data/3306/mysql.sock
basedir=/application/mysql-5.6.40-linux-glibc2.12-x86_64/
datadir=/mm_data/3306/data
log-bin=/mm_data/3306/mysql-bin
character-set-server=utf8
server-id=11
expire-logs-days=1
binlog_format=row
gtid_mode=on
enforce_gtid_consistency=1
log_slave_updates=1
[mysqld_safe]
log-error=/mm_data/3306/mysql_3306_error.log
pid-file=/mm_data/3306/mysqld_3306.pid
sql_mode=NO_ENGINE_SUBSTITUTION,STRTICT_TRANS_TABLES导入master数据
[root@chaoge_slave2 3306]# zcat /mm_data/m1_alldb_2021-04-27.sql.gz |mysql -S /mm_data/3306/mysql.sock
[root@chaoge_slave2 3306]# ./mysql_3306 restart
Restarting MySQL...
Stoping MySQL...
Starting MySQL...
[root@chaoge_slave2 3306]# ls
data mysql_3306 mysql-bin.000001 mysql-bin.000003 mysqld_3306.pid
my.cnf mysql_3306_error.log mysql-bin.000002 mysql-bin.index mysql.sock
[root@chaoge_slave2 3306]#授权master-info
change master to
master_host='10.211.55.12',
master_port=3306,
master_user='repl_chaoge',
master_password='chaoge668',
master_auto_position=1;启动slave
mysql> start slave;
mysql> show slave status\G主从结果
主从结果是正常的
SSH免密登录
MHA_Manager管理节点是通过ssh服务连接其他node节点进行探测、以及获取数据，必须提前做好ssh免密连接
最终结果就是，四台机器，可以任意ssh免密互相登录，包括机器本身
# 生成公私钥
ssh-keygen -t dsa -P '' -f ~/.ssh/id_dsa > /dev/null 2>&1
# 传输公钥
ssh-copy-id -i ~/.ssh/id_dsa.pub 10.211.55.9
ssh-copy-id -i ~/.ssh/id_dsa.pub 10.211.55.11
ssh-copy-id -i ~/.ssh/id_dsa.pub 10.211.55.12
ssh-copy-id -i ~/.ssh/id_dsa.pub 10.211.55.18所有节点依赖安装
所有节点，安装MHA基础依赖
配置好yum源
yum install -y perl-DBD-MySQL \
perl-Config-Tiny \
perl-Log-Dispatch \
perl-Parallel-ForkManager \
perl-ExtUtils-CBuilder \
perl-ExtUtils-MakeMaker \
perl-CPAN所有mysql节点安装MHA-node
安装rpm包
wget --no-check-certificate https://qiniu.wsfnk.com/mha4mysql-node-0.58-0.el7.centos.noarch.rpm
rpm -ivh mha4mysql-node-0.58-0.el7.centos.noarch.rpm检查rpm安装出的命令
[root@chaoge_slave1 tools]# ls -l /usr/bin/*_*log*
-rwxr-xr-x 1 root root 17639 Mar 23 2018 /usr/bin/apply_diff_relay_logs
-rwxr-xr-x 1 root root 15704 Aug 9 2019 /usr/bin/db_log_verify
-rwxr-xr-x 1 root root 33032 Aug 9 2019 /usr/bin/db_printlog
-rwxr-xr-x 1 root root 4807 Mar 23 2018 /usr/bin/filter_mysqlbinlog
-rwxr-xr-x 1 root root 8337 Mar 23 2018 /usr/bin/purge_relay_logs
-rwxr-xr-x 1 root root 7525 Mar 23 2018 /usr/bin/save_binary_logs
-rwxr-xr-x. 1 root root 7910 Aug 4 2017 /usr/bin/scsi_logging_level
-rwxr-xr-x. 1 root root 94696 Aug 4 2017 /usr/bin/sg_logsMHA工具会检测mysql命令，这里还需要加一个软连接
# 三台mysql节点，都执行该命令
ln -s /application/mysql-5.6.40-linux-glibc2.12-x86_64/bin/mysqlbinlog /usr/bin/mysqlbinlog所有节点，创建MHA管理账号，三台机器都操作
mysql> grant all privileges on *.* to mha@'10.211.55.%' identified by 'mha_chaoge';
Query OK, 0 rows affected (0.00 sec)
mysql>MHA-Manager管理节点
MHA管理节点可以装在任何节点，超哥这里就给安装到了
slave02 节点因为Manager管理节点，通过ssh检测mysql集群，如果
master节点服务器宕机，或者网络故障，MHA也无法完成故障切换了。因此mha-manager不能装在
master节点
wget --no-check-certificate https://qiniu.wsfnk.com/mha4mysql-manager-0.58-0.el7.centos.noarch.rpm
[root@chaoge_slave2 tools]# ls mha4mysql-manager-0.58-0.el7.centos.noarch.rpm
mha4mysql-manager-0.58-0.el7.centos.noarch.rpm
[root@chaoge_slave2 tools]#安装管理节点的依赖
yum install -y perl-Config-Tiny epel-release perl-Log-Dispatch perl-Parallel-ForkManager perl-Time-HiRes安装MHA-Manager的包
[root@chaoge_slave2 tools]# rpm -ivh mha4mysql-manager-0.58-0.el7.centos.noarch.rpm
Preparing... ################################# [100%]
Updating / installing...
1:mha4mysql-manager-0.58-0.el7.cent################################# [100%]
[root@chaoge_slave2 tools]#检查mha-manager的命令(slave2机器)
[root@chaoge_slave2 tools]# ls -l /usr/bin/masterha_*
-rwxr-xr-x 1 root root 1995 Mar 23 2018 /usr/bin/masterha_check_repl
-rwxr-xr-x 1 root root 1779 Mar 23 2018 /usr/bin/masterha_check_ssh
-rwxr-xr-x 1 root root 1865 Mar 23 2018 /usr/bin/masterha_check_status
-rwxr-xr-x 1 root root 3201 Mar 23 2018 /usr/bin/masterha_conf_host
-rwxr-xr-x 1 root root 2517 Mar 23 2018 /usr/bin/masterha_manager
-rwxr-xr-x 1 root root 2165 Mar 23 2018 /usr/bin/masterha_master_monitor
-rwxr-xr-x 1 root root 2373 Mar 23 2018 /usr/bin/masterha_master_switch
-rwxr-xr-x 1 root root 5172 Mar 23 2018 /usr/bin/masterha_secondary_check
-rwxr-xr-x 1 root root 1739 Mar 23 2018 /usr/bin/masterha_stop
[root@chaoge_slave2 tools]#创建MHA配置文件
基础信息如下，这里不需要操作
**主从账密**
mysql> grant replication slave on *.* to 'repl_chaoge'@'10.211.55.%' identified by 'chaoge668';
**MHA账密**
mysql> grant all privileges on *.* to mha@'10.211.55.%' identified by 'mha_chaoge';
Query OK, 0 rows affected (0.00 sec)
# 一定记住，刷新用户表
mysql> flush privileges;
Query OK, 0 rows affected (0.00 sec)
**主机host信息**，所有机器都做好/etc/hosts解析
10.211.55.11 chaoge_slave2 ,mha-manager安装在这里
10.211.55.9 chaoge_slave1
10.211.55.12 mysql-server56 ，如果master挂了，需要其他slave去检测注释版
[root@chaoge_slave2 tools]# mkdir -p /etc/mha #<==在/etc下创建mha目录。
[root@chaoge_slave2 tools]# mkdir -p /var/log/mha/app1 #<==在/etc下创建mha目录。
[root@chaoge_slave2 tools]# vim /etc/mha/app1.cnf #<==编辑mha配置文件，增加配置内容。
[server default] #<==默认模块标签。
manager_log=/var/log/mha/app1/manager.log #<==配置日志路径。
manager_workdir=/var/log/mha/app1.log #<==配置工作日志路径。
master_binlog_dir=/mm_data/3306/data/ #<==配置MHA保存主库binlog日志的路径。
user=mha #<==MySQL数据中授权的用户。
password=mha_chaoge #<==MySQL数据中授权的用户。
ping_interval=2 #<==设置监控主库发送ping数据包的时间间隔，
若尝试三次没有回应则自动进行failover。
repl_user=repl_chaoge #<==主从复制对应的用户。
repl_password=chaoge668 #<==主从复制用户对应的密码。
ssh_user=root #<==ssh远程连接服务器的用户。
report_script=/usr/local/send_report #<==设置故障发生切换后触发执行的脚本。
secondary_check_script=/usr/local/bin/masterha_secondary_check -s 10.211.55.9 -s 10.211.55.11 --user=root --master_host=mysql-server56 --master_ip=10.211.55.12 --master_port=3306
#<==当MHA Manager节点到MASTER节点（mysql-server56）的监控之间出现问题时，MHA Manager将会尝试从其他路径登录到MASTER（mysql-server56）节点。
#<==注：此配置在MHA Manager节点只有单独一台机器时起作用。意思就是，在Manager节点联系不上Master时，通过两个从节点（chaoge_slave1 、chaoge_slave2）去探视Master（mysql-server56）节点的状态。
shutdown_script="" #<==设置故障发生后执行主机脚本关闭故障机（防止故障机活过来发生脑裂）
[server1] #<==第一个mysql-master主机模块标签。
hostname=10.211.55.12 #<==第一个mysql-master主机IP。
port=3306 #<==第一个mysql主机端口。
[server2]
hostname=10.211.55.11
port=3306
candidate_master=1 #<==设定此参数后，server2标签的主机，将优先作为主库，宕机的候选服务器（切换主库优先选择）。
check_repl_delay=0 #<==设定此参数后，MHA会忽略主从复制延迟，将此服务器作为后选主机。
[server3]
hostname=10.211.55.9
port=3306最终版配置文件
[root@chaoge_slave2 tools]# mkdir -p /etc/mha
[root@chaoge_slave2 tools]# mkdir -p /var/log/mha/app1
[root@chaoge_slave2 tools]#
# 配置文件如下
[server default]
manager_log=/var/log/mha/app1/manager.log
manager_workdir=/var/log/mha/app1.log
master_binlog_dir=/mm_data/3306/data/
# 该脚本暂时先注释
#master_ip_failover_script=/usr/local/bin/master_ip_failover
user=mha
password=mha_chaoge
ping_interval=2
repl_user=repl_chaoge
repl_password=chaoge668
ssh_user=root
shutdown_script=""
[server1]
candidate_master=1
check_repl_delay=0
hostname=10.211.55.12
port=3306
[server2]
hostname=10.211.55.11
port=3306
candidate_master=1
check_repl_delay=0
[server3]
hostname=10.211.55.9
port=3306自愈检测脚本
#!/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 = '';
my $key = '1';
my $ssh_start_vip = "/sbin/ifconfig eth1:$key $vip";
my $ssh_stop_vip = "/sbin/ifconfig eth1:$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";
}预备启动MHA
检测如下MHA运行条件
- SSH免密登录
- MySQL主从复制
检测SSH登录
如下结果表明服务器之间的SSH免密登录没有问题
[root@chaoge_slave2 tools]# masterha_check_ssh --conf=/etc/mha/app1.conf
Tue Apr 27 15:06:51 2021 - [warning] Global configuration file /etc/masterha_default.cnf not found. Skipping.
Tue Apr 27 15:06:51 2021 - [info] Reading application default configuration from /etc/mha/app1.conf..
Tue Apr 27 15:06:51 2021 - [info] Reading server configuration from /etc/mha/app1.conf..
Tue Apr 27 15:06:51 2021 - [info] Starting SSH connection tests..
Tue Apr 27 15:06:52 2021 - [debug]
Tue Apr 27 15:06:51 2021 - [debug] Connecting via SSH from root@10.211.55.12(10.211.55.12:22) to root@10.211.55.11(10.211.55.11:22)..
Tue Apr 27 15:06:51 2021 - [debug] ok.
Tue Apr 27 15:06:51 2021 - [debug] Connecting via SSH from root@10.211.55.12(10.211.55.12:22) to root@10.211.55.9(10.211.55.9:22)..
Tue Apr 27 15:06:51 2021 - [debug] ok.
Tue Apr 27 15:06:52 2021 - [debug]
Tue Apr 27 15:06:51 2021 - [debug] Connecting via SSH from root@10.211.55.11(10.211.55.11:22) to root@10.211.55.12(10.211.55.12:22)..
Tue Apr 27 15:06:51 2021 - [debug] ok.
Tue Apr 27 15:06:51 2021 - [debug] Connecting via SSH from root@10.211.55.11(10.211.55.11:22) to root@10.211.55.9(10.211.55.9:22)..
Tue Apr 27 15:06:52 2021 - [debug] ok.
Tue Apr 27 15:06:53 2021 - [debug]
Tue Apr 27 15:06:52 2021 - [debug] Connecting via SSH from root@10.211.55.9(10.211.55.9:22) to root@10.211.55.12(10.211.55.12:22)..
Tue Apr 27 15:06:52 2021 - [debug] ok.
Tue Apr 27 15:06:52 2021 - [debug] Connecting via SSH from root@10.211.55.9(10.211.55.9:22) to root@10.211.55.11(10.211.55.11:22)..
Tue Apr 27 15:06:52 2021 - [debug] ok.
Tue Apr 27 15:06:53 2021 - [info] All SSH connection tests passed successfully.
[root@chaoge_slave2 tools]#检测主从复制情况
MHA也提供了主从复制检测
[root@chaoge_slave2 3306]# masterha_check_repl --conf=/etc/mha/app1.conf
Tue Apr 27 16:54:50 2021 - [warning] Global configuration file /etc/masterha_default.cnf not found. Skipping.
Tue Apr 27 16:54:50 2021 - [info] Reading application default configuration from /etc/mha/app1.conf..
Tue Apr 27 16:54:50 2021 - [info] Reading server configuration from /etc/mha/app1.conf..
Tue Apr 27 16:54:50 2021 - [info] MHA::MasterMonitor version 0.58.
Tue Apr 27 16:54:51 2021 - [info] GTID failover mode = 1
Tue Apr 27 16:54:51 2021 - [info] Dead Servers:
Tue Apr 27 16:54:51 2021 - [info] Alive Servers:
Tue Apr 27 16:54:51 2021 - [info] 10.211.55.12(10.211.55.12:3306)
Tue Apr 27 16:54:51 2021 - [info] 10.211.55.11(10.211.55.11:3306)
Tue Apr 27 16:54:51 2021 - [info] 10.211.55.9(10.211.55.9:3306)
Tue Apr 27 16:54:51 2021 - [info] Alive Slaves:
Tue Apr 27 16:54:51 2021 - [info] 10.211.55.11(10.211.55.11:3306) Version=5.6.40-log (oldest major version between slaves) log-bin:enabled
Tue Apr 27 16:54:51 2021 - [info] GTID ON
Tue Apr 27 16:54:51 2021 - [info] Replicating from 10.211.55.12(10.211.55.12:3306)
Tue Apr 27 16:54:51 2021 - [info] Primary candidate for the new Master (candidate_master is set)
Tue Apr 27 16:54:51 2021 - [info] 10.211.55.9(10.211.55.9:3306) Version=5.6.40-log (oldest major version between slaves) log-bin:enabled
Tue Apr 27 16:54:51 2021 - [info] GTID ON
Tue Apr 27 16:54:51 2021 - [info] Replicating from 10.211.55.12(10.211.55.12:3306)
Tue Apr 27 16:54:51 2021 - [info] Current Alive Master: 10.211.55.12(10.211.55.12:3306)
Tue Apr 27 16:54:51 2021 - [info] Checking slave configurations..
Tue Apr 27 16:54:51 2021 - [info] read_only=1 is not set on slave 10.211.55.11(10.211.55.11:3306).
Tue Apr 27 16:54:51 2021 - [info] read_only=1 is not set on slave 10.211.55.9(10.211.55.9:3306).
Tue Apr 27 16:54:51 2021 - [info] Checking replication filtering settings..
Tue Apr 27 16:54:51 2021 - [info] binlog_do_db= , binlog_ignore_db=
Tue Apr 27 16:54:51 2021 - [info] Replication filtering check ok.
Tue Apr 27 16:54:51 2021 - [info] GTID (with auto-pos) is supported. Skipping all SSH and Node package checking.
Tue Apr 27 16:54:51 2021 - [info] Checking SSH publickey authentication settings on the current master..
Tue Apr 27 16:54:51 2021 - [info] HealthCheck: SSH to 10.211.55.12 is reachable.
Tue Apr 27 16:54:51 2021 - [info]
10.211.55.12(10.211.55.12:3306) (current master)
+--10.211.55.11(10.211.55.11:3306)
+--10.211.55.9(10.211.55.9:3306)
Tue Apr 27 16:54:51 2021 - [info] Checking replication health on 10.211.55.11..
Tue Apr 27 16:54:51 2021 - [info] ok.
Tue Apr 27 16:54:51 2021 - [info] Checking replication health on 10.211.55.9..
Tue Apr 27 16:54:51 2021 - [info] ok.
Tue Apr 27 16:54:51 2021 - [warning] master_ip_failover_script is not defined.
Tue Apr 27 16:54:51 2021 - [warning] shutdown_script is not defined.
Tue Apr 27 16:54:51 2021 - [info] Got exit code 0 (Not master dead).
MySQL Replication Health is OK.
[root@chaoge_slave2 3306]#结果总结
结果必须和超哥一样，全都是info级别的日志信息，而不得有error日志
并且提示mysql replication health is OK，表示复制检查正常。
配置VIP漂移
上面超哥是临时关闭了VIP的漂移脚本
# 该脚本暂时先注释
#master_ip_failover_script=/usr/local/bin/master_ip_failover这个作用是当master发生故障，迁移后数据库IP发生变化，解决这个问题，因此得使用VIP进行漂移
MHA已经提供好了perl脚本
我们只需要创建VIP即可，首先在master节点上创建
# 创建
[root@mysql-server56 3306]# ifconfig eth0:1 10.211.55.77/24
# 删除
[root@mysql-server56 3306]# ifconfig eth0:1 del 10.211.55.77/24
# 关闭
[root@mysql-server56 3306]# ifconfig eth0:1 down使用vip脚本
脚本代码
master_ip_failover_script=/usr/local/bin/master_ip_failover
如下
vim /usr/local/bin/master_ip_failover
给与执行权限
chmod +x /usr/local/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 = '10.211.55.77/24';
my $key = '1';
my $ssh_start_vip = "/sbin/ifconfig eth0:$key $vip";
my $ssh_stop_vip = "/sbin/ifconfig eth0:$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";
}再次检测MHA程序
[root@chaoge_slave2 3306]# masterha_check_repl --conf=/etc/mha/app1.conf
Tue Apr 27 17:29:19 2021 - [warning] Global configuration file /etc/masterha_default.cnf not found. Skipping.
Tue Apr 27 17:29:19 2021 - [info] Reading application default configuration from /etc/mha/app1.conf..
Tue Apr 27 17:29:19 2021 - [info] Reading server configuration from /etc/mha/app1.conf..
Tue Apr 27 17:29:19 2021 - [info] MHA::MasterMonitor version 0.58.
Tue Apr 27 17:29:20 2021 - [info] GTID failover mode = 1
Tue Apr 27 17:29:20 2021 - [info] Dead Servers:
Tue Apr 27 17:29:20 2021 - [info] Alive Servers:
Tue Apr 27 17:29:20 2021 - [info] 10.211.55.12(10.211.55.12:3306)
Tue Apr 27 17:29:20 2021 - [info] 10.211.55.11(10.211.55.11:3306)
Tue Apr 27 17:29:20 2021 - [info] 10.211.55.9(10.211.55.9:3306)
Tue Apr 27 17:29:20 2021 - [info] Alive Slaves:
Tue Apr 27 17:29:20 2021 - [info] 10.211.55.11(10.211.55.11:3306) Version=5.6.40-log (oldest major version between slaves) log-bin:enabled
Tue Apr 27 17:29:20 2021 - [info] GTID ON
Tue Apr 27 17:29:20 2021 - [info] Replicating from 10.211.55.12(10.211.55.12:3306)
Tue Apr 27 17:29:20 2021 - [info] Primary candidate for the new Master (candidate_master is set)
Tue Apr 27 17:29:20 2021 - [info] 10.211.55.9(10.211.55.9:3306) Version=5.6.40-log (oldest major version between slaves) log-bin:enabled
Tue Apr 27 17:29:20 2021 - [info] GTID ON
Tue Apr 27 17:29:20 2021 - [info] Replicating from 10.211.55.12(10.211.55.12:3306)
Tue Apr 27 17:29:20 2021 - [info] Current Alive Master: 10.211.55.12(10.211.55.12:3306)
Tue Apr 27 17:29:20 2021 - [info] Checking slave configurations..
Tue Apr 27 17:29:20 2021 - [info] read_only=1 is not set on slave 10.211.55.11(10.211.55.11:3306).
Tue Apr 27 17:29:20 2021 - [info] read_only=1 is not set on slave 10.211.55.9(10.211.55.9:3306).
Tue Apr 27 17:29:20 2021 - [info] Checking replication filtering settings..
Tue Apr 27 17:29:20 2021 - [info] binlog_do_db= , binlog_ignore_db=
Tue Apr 27 17:29:20 2021 - [info] Replication filtering check ok.
Tue Apr 27 17:29:20 2021 - [info] GTID (with auto-pos) is supported. Skipping all SSH and Node package checking.
Tue Apr 27 17:29:20 2021 - [info] Checking SSH publickey authentication settings on the current master..
Tue Apr 27 17:29:20 2021 - [info] HealthCheck: SSH to 10.211.55.12 is reachable.
Tue Apr 27 17:29:20 2021 - [info]
10.211.55.12(10.211.55.12:3306) (current master)
+--10.211.55.11(10.211.55.11:3306)
+--10.211.55.9(10.211.55.9:3306)
Tue Apr 27 17:29:20 2021 - [info] Checking replication health on 10.211.55.11..
Tue Apr 27 17:29:20 2021 - [info] ok.
Tue Apr 27 17:29:20 2021 - [info] Checking replication health on 10.211.55.9..
Tue Apr 27 17:29:20 2021 - [info] ok.
Tue Apr 27 17:29:20 2021 - [info] Checking master_ip_failover_script status:
Tue Apr 27 17:29:20 2021 - [info] /usr/local/bin/master_ip_failover --command=status --ssh_user=root --orig_master_host=10.211.55.12 --orig_master_ip=10.211.55.12 --orig_master_port=3306
IN SCRIPT TEST====/sbin/ifconfig eth1:1 down==/sbin/ifconfig eth1:1 10.211.55.77/24===
Checking the Status of the script.. OK
Tue Apr 27 17:29:20 2021 - [info] OK.
Tue Apr 27 17:29:20 2021 - [warning] shutdown_script is not defined.
Tue Apr 27 17:29:20 2021 - [info] Got exit code 0 (Not master dead).
MySQL Replication Health is OK.一切OK，启动MHA
启动MHA
在MHA的manager节点，启动MHA进程
nohup masterha_manager --conf=/etc/mha/app1.conf --ignore_last_failover /var/log/mha/app1/manager.log 2>&1 &
命令参数：
--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。
--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，也就是上面设置的manager_workdir目录中产生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。
# 停止命令
masterha_stop --conf=/etc/mha/app1.conf运行命令，检测日志
[root@chaoge_slave2 3306]# nohup masterha_manager --conf=/etc/mha/app1.conf --ignore_last_failover /var/log/mha/app1/manager.log 2>&1 &
[1] 31466
[root@chaoge_slave2 3306]# nohup: ignoring input and appending output to ‘nohup.out’
[root@chaoge_slave2 3306]# ps -ef|grep master
root 31466 25629 2 17:33 pts/0 00:00:00 perl /usr/bin/masterha_manager --conf=/etc/mha/app1.conf --ignore_last_failover /var/log/mha/app1/manager.log
root 31488 25629 0 17:33 pts/0 00:00:00 grep --color=auto master
[root@chaoge_slave2 3306]# tail -f /var/log/mha/app1/manager.log
IN SCRIPT TEST====/sbin/ifconfig eth1:1 down==/sbin/ifconfig eth1:1 10.211.55.77/24===
Checking the Status of the script.. OK
Tue Apr 27 17:33:36 2021 - [info] OK.
Tue Apr 27 17:33:36 2021 - [warning] shutdown_script is not defined.
Tue Apr 27 17:33:36 2021 - [info] Set master ping interval 2 seconds.
Tue Apr 27 17:33:36 2021 - [warning] secondary_check_script is not defined. It is highly recommended setting it to check master reachability from two or more routes.
Tue Apr 27 17:33:36 2021 - [info] Starting ping health check on 10.211.55.12(10.211.55.12:3306)..
Tue Apr 27 17:33:36 2021 - [info] Ping(SELECT) succeeded, waiting until MySQL doesn't respond..检测MHA状态
检测mysql主从集群的状态
[root@chaoge_slave2 ~]# masterha_check_status --conf=/etc/mha/app1.conf
app1 (pid:31466) is running(0:PING_OK), master:10.211.55.12检查VIP当前在哪，在当前的mysql-master机器
[root@mysql-server56 3306]# ifconfig eth0:1
eth0:1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 10.211.55.77 netmask 255.255.255.0 broadcast 10.211.55.255
ether 00:1c:42:79:bc:f3 txqueuelen 1000 (Ethernet)
[root@mysql-server56 3306]#第一次进行VIP脚本自动切换，VIP必须在master机器上
停止master主库
查看结果
- master切换
- VIP切换
[root@mysql-server56 3306]# /mm_data/3306/mysql_3306 stop
Stoping MySQL...见证MHA漂移结果
最终发生了如下变化
- MHA软件在切换后会自动停止进程
- VIP发生漂移
- 主从复制关系发生变化
slave01机器，主从角色发生变化
[root@chaoge_slave1 3306]# mysql -S /mm_data/3306/mysql.sock
Welcome to the MySQL monitor. Commands end with ; or \g.
Your MySQL connection id is 28
Server version: 5.6.40-log MySQL Community Server (GPL)
Copyright (c) 2000, 2018, Oracle and/or its affiliates. All rights reserved.
Oracle is a registered trademark of Oracle Corporation and/or its
affiliates. Other names may be trademarks of their respective
owners.
Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
mysql> show slave status\G
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.211.55.11
Master_User: repl_chaoge
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: mysql-bin.000003
Read_Master_Log_Pos: 1929
Relay_Log_File: mysqld_3306-relay-bin.000003
Relay_Log_Pos: 1405
Relay_Master_Log_File: mysql-bin.000003
Slave_IO_Running: Yes
Slave_SQL_Running: Yesslave01机器，已经没有slave的角色，成为了主库
[root@chaoge_slave2 ~]# mysql -S /mm_data/3306/mysql.sock
Welcome to the MySQL monitor. Commands end with ; or \g.
Your MySQL connection id is 29
Server version: 5.6.40-log MySQL Community Server (GPL)
Copyright (c) 2000, 2018, Oracle and/or its affiliates. All rights reserved.
Oracle is a registered trademark of Oracle Corporation and/or its
affiliates. Other names may be trademarks of their respective
owners.
Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
mysql> show slave status\G
Empty set (0.00 sec)
mysql>
【推荐】国内首个AI IDE，深度理解中文开发场景，立即下载体验Trae
【推荐】编程新体验，更懂你的AI，立即体验豆包MarsCode编程助手
【推荐】抖音旗下AI助手豆包，你的智能百科全书，全免费不限次数
【推荐】轻量又高性能的 SSH 工具 IShell：AI 加持，快人一步
· 阿里最新开源QwQ-32B，效果媲美deepseek-r1满血版，部署成本又又又降低了！
· AI编程工具终极对决：字节Trae VS Cursor，谁才是开发者新宠？
· 开源Multi-agent AI智能体框架aevatar.ai，欢迎大家贡献代码
· Manus重磅发布：全球首款通用AI代理技术深度解析与实战指南
· 被坑几百块钱后，我竟然真的恢复了删除的微信聊天记录！