-
要求和原理:
(1)需要奇数个节点
(2)当一个master节点挂掉之后投票选出一个新的master,偶数个节点会出现票数相同的状况,但奇数个就不会
(3)所有数据节点的数据一致,每一台数据节点都有可能作为master
(4)每个节点都要安装master和slave插件; -
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。
-
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
-
实验环境:
主机名 | IP | 角色 |
---|---|---|
server1 | 172.25.19.1 | master |
server2 | 172.25.19.2 | slave(备master) |
server3 | 172.25.19.3 | slave |
server4 | 172.25.19.4 | MHA |
- 配置过程:
1.(1)重新配置server1的mysql服务
[root@server1 mysql]# systemctl stop mysqld
[root@server1 mysql]# vim /etc/my.cnf
[root@server1 mysql]# ls
29 server-id=1
30 gtid_mode=ON
31 enforce_gtid_consistency=ON
32 log_bin=binlog
33 log_slave_updates=ON
[root@server1 mysql]# rm -fr *
[root@server1 mysql]# ls
[root@server1 mysql]# systemctl start mysqld
[root@server1 mysql]# grep password /var/log/mysqld.log
[root@server1 mysql]# mysql -p
mysql> alter user root@localhost identified by 'Yjy+123+mm';
mysql> show databases;
mysql> GRANT REPLICATION SLAVE ON *.* TO repl@'172.25.19.%' IDENTIFIED BY 'Yjy+123+mm';
mysql> SHOW MASTER STATUS;
(2)配置server2的mysql服务
[root@server2 ~]# systemctl stop mysqld
[root@server2 ~]# cd /var/lib/mysql
[root@server2 mysql]# ls
server-id = 2
gtid_mode=ON
enforce_gtid_consistency=ON
log_slave_updates=ON
log_bin=binlog
[root@server2 mysql]# rm -fr *
[root@server2 mysql]# ls
[root@server2 mysql]# vim /etc/my.cnf
[root@server2 mysql]# systemctl start mysqld
[root@server2 mysql]# grep password /var/log/mysqld.log
[root@server2 mysql]# mysql -p
mysql> alter user root@localhost identified by 'Yjy+123+mm';
mysql> show databases;
mysql> CHANGE MASTER TO MASTER_HOST='172.25.19.1',MASTER_USER='repl',
MASTER_PASSWORD='Yjy+123+mm',MASTER_AUTO_POSITION=1;
mysql> start slave;
mysql> show slave status\G
(3)配置server3的mysql服务
[root@server3 ~]# systemctl stop mysqld
[root@server3 ~]# cd /var/lib
[root@server3 lib]# cd mysql
[root@server3 mysql]# ls
server-id=3
gtid_mode=ON
enforce_gtid_consistency=ON
log_slave_updates=ON
log_bin=binlog
[root@server3 mysql]# rm -fr *
[root@server3 mysql]# ls
[root@server3 mysql]# vim /etc/my.cnf
[root@server3 mysql]# systemctl start mysqld
[root@server3 mysql]# grep password /var/log/mysqld.log
[root@server3 mysql]# mysql -p
mysql> alter user root@localhost identified by 'Yjy+123+mm';
mysql> show databases;
mysql> CHANGE MASTER TO MASTER_HOST='172.25.19.1',MASTER_USER='repl',
MASTER_PASSWORD='Yjy+123+mm',MASTER_AUTO_POSITION=1;
mysql> start slave;
mysql> show slave status\G
2.测试主从复制:
server1:
mysql> create database westos;
mysql> use westos;
mysql> create table usertb (
-> username varchar(10) not null,
-> password varchar(15) not null);
mysql> insert into usertb values ('user1','123');
mysql> select * from usertb;
server2:
mysql> select * from westos.usertb
-> ;
server3:
mysql> select * from westos.usertb;
备份数据库 :
[root@server1 ~]# mysqldump --all-databases --single-transaction --triggers
--routines --events --host=127.0.0.1 --port=3306 --user=root
--password=Yjy+123+mm > yjy.sql
一、手动同步
1.安装需要的软件
[root@server4 MHA-7]# yum install mha4mysql-manager-0.58-0.el7.centos.noarch.rpm
mha4mysql-node-0.58-0.el7.centos.noarch.rpm perl-*
2.生成ssh密钥,管理节点是不需要输入密码,将生成的公钥和私钥传给数据节点
[root@server4 ~]# ssh-keygen
[root@server4 ~]# ssh-copy-id server1
[root@server4 ~]# ssh-copy-id server2
[root@server4 ~]# ssh-copy-id server3
[root@server4 ~]# scp -r .ssh server1:
[root@server4 ~]# scp -r .ssh server2:
[root@server4 ~]# scp -r .ssh server3:
测试:
[root@server4 ~]# ssh server1
[root@server4 ~]# ssh server2
[root@server4 ~]# ssh server3
3.将节点的安装包传给数据节点
[root@server4 ~]# ls
MHA-7
[root@server4 ~]# cd MHA-7/
[root@server4 MHA-7]# ls
[root@server4 MHA-7]# scp -r mha4mysql-node-0.58-0.el7.centos.noarch.rpm server1:/root/
[root@server4 MHA-7]# scp -r mha4mysql-node-0.58-0.el7.centos.noarch.rpm server2:/root/
[root@server4 MHA-7]# scp -r mha4mysql-node-0.58-0.el7.centos.noarch.rpm server3:/root/
4.数据节点安装
[root@server1 ~]# yum install -y mha4mysql-node-0.58-0.el7.centos.noarch.rpm
[root@server2 ~]# yum install -y mha4mysql-node-0.58-0.el7.centos.noarch.rpm
[root@server3 ~]# yum install -y mha4mysql-node-0.58-0.el7.centos.noarch.rpm
5.管理节点创建一个目录,编辑配置文件
[root@server4 ~]# mkdir -p /etc/masterha
[root@server4 ~]# cd /etc/masterha/
[root@server4 masterha]# ls
[root@server4 masterha]#
[root@server4 masterha]# vim app1.cnf
[root@server4 masterha]#
[root@server4 masterha]# cd
[server default]
manager_workdir=/etc/masterha
manager_log=/var/log/masterha.log
master_binlog_dir=/etc/masterha
#master_ip_failover_script= /usr/local/bin/master_ip_failover
#master_ip_online_change_script= /usr/local/bin/master_ip_online_change
password=Yjy+123+mm
user=root
ping_interval=1
remote_workdir=/tmp
repl_password=Yjy+123+mm
repl_user=repl
#report_script=/usr/local/send_report
#secondary_check_script= /usr/local/bin/masterha_secondary_check -s server03 -s server02
#shutdown_script=""
ssh_user=root
[server1]
hostname=172.25.19.1
port=3306
[server2]
hostname=172.25.19.2
port=3306
candidate_master=1
check_repl_delay=0
[server3]
hostname=172.25.19.3
port=3306
no_master=1
6.管理节点ssh检查
[root@server4 ~]# masterha_check_ssh --conf=/etc/masterha/app1.cnf
Fri May 3 11:47:19 2019 - [debug] ok.
Fri May 3 11:47:19 2019 - [info] All SSH connection tests passed successfully.
7.数据节点授权
server1: 所有权限
mysql> grant all on *.* to root@'%' identified by 'Yjy+123+mm';
server2 和 server3: 只读
mysql> set global read_only=1;
8.管理节点环境健康检查
[root@server4 ~]# masterha_check_repl --conf=/etc/masterha/app1.cnf
MySQL Replication Health is OK.
9. 手动替换master
(1)关闭当前的master节点的mysql服务
[root@server1 ~]# systemctl stop mysqld
(2)手动替换master节点
[root@server4 ~]# masterha_master_switch --master_state=dead
--conf=/etc/masterha/app1.cnf --dead_master_host=172.25.19.1
--dead_master_port=3306 --new_master_host=172.25.19.2 --new_master_port=3306
Master 172.25.19.1(172.25.19.1:3306) is dead. Proceed? (yes/NO): yes
Starting master switch from 172.25.19.1(172.25.19.1:3306) to 172.25.19.2
(172.25.19.2:3306)? (yes/NO): yes
(3)开启server1的master服务,将他的master改为server2
[root@server1 ~]# systemctl start mysqld
[root@server1 ~]# mysql -p
mysql> CHANGE MASTER TO MASTER_HOST='172.25.19.2',MASTER_USER='repl',
MASTER_PASSWORD='Yjy+123+mm',MASTER_AUTO_POSITION=1;
## 打开slave
mysql> start slave;
测试:
server1:
mysql> show slave status\G
server2:
mysql> show slave status\G
查看不出来
server3:
mysql> show slave status\G
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 172.25.19.2
二、自动切换
1.删除故障文件
[root@server4 ~]# cd /etc/masterha/
[root@server4 masterha]# ls
app1.cnf app1.failover.complete
[root@server4 masterha]#
[root@server4 masterha]# rm -fr app1.failover.complete
[root@server4 masterha]# ls
2.输入命令自动替换
masterha_master_switch --conf=/etc/masterha/app1.cnf --master_state=alive --
new_master_host=172.25.19.1 --new_master_port=3306 --orig_master_is_new_slave --
running_updates_limit=10000
测试
server1: 查看不到
mysql> mysql> show slave status\G
Empty set (0.00 sec)
server2和server3: master是server1
mysql> mysql> show slave status\G
三、自动切换
1.server4创建一个检测进程
[root@server4 masterha]# nohup masterha_manager
--conf=/etc/masterha/app1.cnf &> /dev/null &
[1] 11841
[root@server4 masterha]# ps a
2.server1(当前master关闭myslq服务)
[root@server1 ~]# systemctl stop mysqld
[root@server1 ~]#
3.server4上监控进程停止
[root@server4 masterha]# ps a
PID TTY STAT TIME COMMAND
640 tty1 Ss+ 0:00 /sbin/agetty --noclear tty1 linux
2046 pts/0 Ss 0:00 -bash
11914 pts/0 R+ 0:00 ps a
[1]+ Done nohup masterha_manager
--conf=/etc/masterha/app1.cnf &>/dev/null
4.server2上查看不到
mysql> show slave status\G
Empty set (0.00 sec)
server3上查看到master是server2
mysql> show slave status\G
5.server1重新开启master
[root@server1 ~]# systemctl start mysqld
配置主从复制
[root@server1 ~]# mysql -p
mysql> CHANGE MASTER TO MASTER_HOST='172.25.19.2',MASTER_USER='repl',
MASTER_PASSWORD='Yjy+123+mm',MASTER_AUTO_POSITION=1;
打开slave
mysql> start slave;
查看master是server2
mysql> show slave status\G
四、脚本控制(通过vip的漂移查看)
1.编辑配置文件,将以下两行的注释打开
master_ip_failover_script= /usr/local/bin/master_ip_failover
master_ip_online_change_script= /usr/local/bin/master_ip_online_change
2.编写两个脚本,主要内容如下:
my $ssh_start_vip = "/sbin/ip addr add $vip dev eth0";
my $ssh_stop_vip = "/sbin/ip addr del $vip dev eth0";
my $vip = '172.25.19.100/24'; # Virtual IP
my $key = "1";
my $ssh_start_vip = "/sbin/ip addr add $vip dev eth0";
my $ssh_stop_vip = "/sbin/ip addr del $vip dev eth0";
my $exit_code = 0;
3.将这两个脚本复制到/usr/local/bin下
[root@server4 ~]# cp master_ip_* /usr/local/bin/
4.给两个脚本加上可执行权限
[root@server4 ~]# chmod +x /usr/local/bin/master_ip_*
[root@server4 ~]#
[root@server4 ~]# ll /usr/local/bin/master_ip_*
-rwxr-xr-x 1 root root 2172 May 3 14:36 /usr/local/bin/master_ip_failover
-rwxr-xr-x 1 root root 3847 May 3 14:36 /usr/local/bin/master_ip_online_change
5.给server2(当前master)上添加一个vip
[root@server2 mysql]# cd
[root@server2 ~]# ip addr add 172.25.19.100/24 dev eth0
[root@server2 ~]# ip addr
6.server4 手动切换master到server1
[root@server4 masterha]# masterha_master_switch --conf=/etc/masterha/app1.cnf
--master_state=alive --new_master_host=172.25.19.1 --new_master_port=3306
--orig_master_is_new_slave --running_updates_limit=10000
7.测试:
[root@server1 ~]# ip addr
server1可以查看到vip
[root@server2 ~]# ip addr
server2查看不到vip
server3查看现在的master是server1
mysql> show slave status\G
五、自动切换
1.删除故障文件
[root@server4 masterha]# ls
app1.cnf app1.failover.complete
[root@server4 masterha]# rm -fr app1.failover.complete
[root@server4 masterha]# ls
app1.cnf
2.创建监控master的进程
[root@server4 masterha]# nohup masterha_manager --conf=/etc/masterha/app1.cnf &> /dev/null &
[1] 11957
[root@server4 masterha]# ps a
3.server1关闭mysql服务
[root@server1 ~]# systemctl stop mysqld
4.监控master的进程挂掉
[root@server4 masterha]#
[1]+ Done nohup masterha_manager --conf=/etc/masterha/app1.cnf &>/dev/null
[root@server4 masterha]#
5.测试:
server2可以查看到vip
[root@server2 ~]# ip a
server3查看到master是server2
mysql> show slave status\G
[root@server1 ~]# ip a
注意:每一次切换都会生成一个app1.failover.complete文件(简称故障文件),要将这个文件删除,否则会出错
创建一个监控master的进程,这个进程会实时监控master节点的状态,一旦master节点出现故障(宕机),server4(管理节点)就会自动执行/usr/local/bin下的两个脚本,在正常运行的节点中选择一个最新的来作为新的master节点,同时当前的进程更会挂掉,因为他监控的master已经不再当前集群
[root@server1 ~]# systemctl stop mysqld
[root@server2 ~]# systemctl stop mysqld
[root@server3 mysql]# systemctl stop mysqld