准备
1 1. 双主 master1 192.168.199.49 2 master2 192.168.199.50 3 VIP 192.168.199.52 //虚拟IP 4 2.环境 master:nginx + php + mysql + keepalived 5 VIP:只要和master在一个局域网内即可。 6 3. 服务器之间网络通畅，可以互相ping通。 7 4. 2个服务器的mysql版本要一致。数据库密码一致 8 5. 防火墙增加允许组播和允许VRRP（虚拟路由器冗余协）通信，这样主服务器在故障恢复后才能抢回资源 9 -A INPUT -s 182.148.15.0/24 -d 224.0.0.18 -j ACCEPT 10 -A INPUT -s 182.148.15.0/24 -p vrrp -j ACCEPT 11 重启生效：service iptables reload
配置服务器
1 服务器1： 2 3 在my.cnf文件的[mysqld]配置区域添加下面内容： 4 log-bin=mysql-bin #记录二进制文件 5 binlog_format=mixed #mysql默认采用的二进制格式 6 server-id = 2 #服务号，必须是唯一的，一般取IP的后8位 7 expire_logs_days = 10 #binlog过期清理时间 8 9 binlog-do-db = db1 #需要服务的数据库。 10 #replicate-ignore-db = db3,mysql,test,information_schema,performance_schema,transfer_new #不需要复制的数据库 11 #replicate-wild-ignore-table=garbage.% #有问题可以尝试使用此项配置 12 sync_binlog = 1 #表示binlog日志在每1次写入后与硬盘同步，设置1最安全也最慢 13 auto-increment-increment = 2 #服务器个数 14 auto-increment-offset = 2 #自增偏移量，如果有2台，则一台设置1，另一台设置2 15 slave-skip-errors = all #从库复制时跳过所有的错误 16 服务器2： 17 18 log-bin=mysql-bin 19 binlog_format=mixed 20 server-id = 1 21 expire_logs_days = 10 22 binlog-do-db = db1 23 #replicate-ignore-db = db3,mysql,test,information_schema,performance_schema,transfer_new 24 #replicate-wild-ignore-table=garbage.% #有问题可以尝试使用此项配置 25 sync_binlog = 1 26 auto-increment-increment = 2 27 auto-increment-offset = 1 28 slave-skip-errors = all
2台服务器都重启mysql服务：
1 service mysql restart
同步配置(2台服务器分别执行)
1 给对方授权复制权限 2 3 //对方服务器60.205.182.26允许复制本服务器数据库，用户名repl 密码ocnt-123 4 mysql> grant replication slave,replication client on *.* to repl@'60.205.182.26' identified by "ocnt-123"; 5 //生效 6 mysql> flush privileges; 7 //锁定表 8 mysql> flush tables with read lock;
同步数据
1 1.锁定表 2 mysql> flush tables with read lock; 3 2.同步数据，可使用navicat
同步操作(2台服务器分别执行)
1 //查看各自服务器当前二进制日志文件名称和记录位置 2 mysql> show master status; 3 4 <!-----------------------服务器1同步------------------------------------> 5 //解锁表 6 mysql> unlock tables; 7 //停止slave 8 mysql> slave stop; 9 //开始同步: IP、用户名和密码为服务器2刚才grant授权配置的,日志文件和pos值得位置通过服务器2 show master status查看得来 10 mysql> change master to master_host='101.200.63.35',master_user='repl',master_password='123456',master_log_file='mysql-bin.000016',master_log_pos=363; 11 //开启slave 12 mysql> start slave; 13 14 <!-----------------------服务器2同步------------------------------------> 15 //解锁表 16 mysql> unlock tables; 17 //停止slave 18 mysql> slave stop； 19 <!--开始同步--> 20 change master to master_host='60.205.182.26',master_user='repl',master_password='123456',master_log_file='mysql-bin.000028',master_log_pos=107; 21 <!--开启slave--> 22 mysql> start slave;
查看同步状态，如下出现两个“Yes”，表明同步成功！（Slave_IO_Running和Slave_SQL_Running）
1 mysql> show slave status G;
配置到此处就完成了双主热备。
keepalived故障自动切换配置
1 master1:192.168.199.49 主机1 2 master2:192.168.199.50 主机2 备份 3 VIP:192.168.199.52 虚拟IP 4 ### 2台服务器安装keepalived 5 6 [root@localhost ]#yum install -y openssl-devel 7 [root@localhost src]#cd /usr/local/src/ 8 [root@localhost src]# wget http://www.keepalived.org/software/keepalived-1.3.5.tar.gz 9 [root@localhost src]# tar -zvxf keepalived-1.3.5.tar.gz 10 [root@localhost src]# cd keepalived-1.3.5 11 [root@localhost keepalived-1.3.5]# ./configure --prefix=/usr/local/keepalived 12 [root@localhost keepalived-1.3.5]# make && make install 13 14 [root@localhost keepalived-1.3.5]# cp /usr/local/src/keepalived-1.3.5/keepalived/etc/init.d/keepalived /etc/rc.d/init.d/ 15 [root@localhost keepalived-1.3.5]# cp /usr/local/keepalived/etc/sysconfig/keepalived /etc/sysconfig/ 16 [root@localhost keepalived-1.3.5]# mkdir /etc/keepalived/ 17 [root@localhost keepalived-1.3.5]# cp /usr/local/keepalived/etc/keepalived/keepalived.conf /etc/keepalived/ 18 [root@localhost keepalived-1.3.5]# cp /usr/local/keepalived/sbin/keepalived /usr/sbin/ 19 [root@localhost keepalived-1.3.5]# echo "/etc/init.d/keepalived start" >> /etc/rc.local
master1服务器配置
1 [root@master1 ~]# cp /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf.bak 2 [root@master1 ~]# vim /etc/keepalived/keepalived.conf ＃删除所有内容，重新配置如下 3 4 ! Configuration File for keepalived 5 global_defs { 6 notification_email { 7 412140451@qq.com 8 } 9 10 notification_email_from 412140451@qq.com 11 smtp_server 127.0.0.1 12 smtp_connect_timeout 30 13 router_id MASTER-HA 14 } 15 16 vrrp_script chk_mysql_port { #检测mysql服务是否在运行 17 script "/opt/chk_mysql.sh" #这里通过脚本监测 18 interval 3 #脚本执行间隔，每3s检测一次 19 weight -5 #脚本结果导致的优先级变更，检测失败（脚本返回非0）则优先级 -5 20 fall 2 #检测连续2次失败才算确定是真失败。会用weight减少优先级（1-255之间） 21 rise 1 #检测1次成功就算成功。但不修改优先级 22 } 23 24 vrrp_instance VI_1 { 25 state MASTER 26 interface eth0 #指定虚拟ip的网卡接口 27 mcast_src_ip 192.168.199.49 #本机master的 IP地址 28 virtual_router_id 100 #路由器标识，MASTER和BACKUP必须是一致的 29 priority 100 #定义优先级，数字越大，优先级越高，在同一个vrrp_instance下，MASTER的优先级必须大于BACKUP的优先级。这样MASTER故障恢复后，就可以将VIP资源再次抢回来 30 advert_int 1 31 authentication { #加密协议 32 auth_type PASS 33 auth_pass 1111 34 } 35 virtual_ipaddress { 36 192.168.199.52 #虚拟IP地址 VIP 37 } 38 39 track_script { 40 chk_mysql_port 41 } 42 }
master2服务器配置 修改 mcast_src_ip priority
1 global_defs { 2 notification_email { 3 412140451@qq.com 4 } 5 6 notification_email_from 412140451@qq.com 7 smtp_server 127.0.0.1 8 smtp_connect_timeout 30 9 router_id MASTER-HA 10 } 11 12 vrrp_script chk_mysql_port { 13 script "/opt/chk_mysql.sh" 14 interval 3 15 weight -5 16 fall 2 17 rise 1 18 } 19 20 vrrp_instance VI_1 { 21 state BACKUP 22 interface eth0 23 mcast_src_ip 192.168.199.50 24 virtual_router_id 100 25 priority 90 26 advert_int 1 27 authentication { 28 auth_type PASS 29 auth_pass 1111 30 } 31 virtual_ipaddress { 32 192.168.199.52 33 } 34 35 track_script { 36 chk_mysql_port 37 } 38 }
2个服务器都编写检测切换mysql脚本
1 KeepAlived做心跳检测，如果Master的MySQL服务挂了(3306端口挂了),那么它就会选择自杀。Slave的KeepAlived通过心跳检测发现这个情况，就会将VIP的请求接管 2 vim /opt/chk_mysql.sh 3 4 #!/bin/bash 5 CHECK_TIME=3 6 7 #Mysql 8 host='localhost' 9 port='3306' 10 user='root' 11 password='ocnt-123' 12 mysqlclient="mysql" 13 14 #mysql is working MYSQL_OK is 1 , mysql down MYSQL_OK is 0 15 function check_mysql_helth (){ 16 #NEW=$MYSQL -h $MYSQL_HOST -u $MYSQL_USER -p${MYSQL_PASSWORD} -e "show status;" >/dev/null 2>&1 17 $mysqlclient --host=$host --port=$port --user=$user --password=$password -e "show databases;" > /dev/null 2>&1 18 19 if [ $? == 0 ] 20 then 21 MYSQL_OK=1 22 else 23 MYSQL_OK=0 24 fi 25 # echo $MYSQL_OK 26 return $MYSQL_OK 27 } 28 29 while [ $CHECK_TIME -ne 0 ] 30 do 31 let "CHECK_TIME -= 1" 32 check_mysql_helth 33 if [ $MYSQL_OK == 1 ] ; then 34 CHECK_TIME=0 35 exit 0 36 fi 37 if [ $MYSQL_OK -eq 0 ] && [ $CHECK_TIME -eq 0 ] 38 then 39 pkill keepalived 40 exit 1 41 fi 42 sleep 1 43 done
赋权限
1 chmod 755 /opt/chk_mysql.sh
启动服务
1 关闭防火墙 2 service iptables stop 3 启动服务 4 /etc/init.d/keepalived start
数据库测试
1 使用局域网内的客户端连接VIP 2 mysql -h192.168.199.52 -uroot -pocnt-123 3 测试是否可用
故障切换检测
1 默认情况下，vip是在master1上的。使用"ip addr"命令查看vip切换情况 2 ip addr 3 查找 4 inet 192.168.199.49/32 scope global eth0 //这个32位子网掩码的vip地址表示该资源目前还在master1机器上，而master2机器上没有 5 同时查看 tail -f /var/log/message 会发现 6 Dec 13 17:52:38 localhost Keepalived_vrrp[65545]: VRRP_Instance(VI_1) Entering MASTER STATE 7 Dec 13 17:52:38 localhost Keepalived_vrrp[65545]: VRRP_Instance(VI_1) setting protocol VIPs. 8 Dec 13 17:52:38 localhost Keepalived_vrrp[65545]: Sending gratuitous ARP on eth0 for 192.168.199.51 9 10 关掉master1的mysql服务 11 service mysql stop 12 查看 master1的日志 tail -f /var/log/message 会发现 13 Dec 13 18:52:44 localhost Keepalived_vrrp[65545]: Stopped 14 Dec 13 18:52:44 localhost Keepalived[65542]: Stopped Keepalived v1.3.5 (03/19,2017), git commit v1.3.5-6-g6fa32f2 15 16 查看master2的日志 17 Dec 13 18:52:45 localhost Keepalived_vrrp[28451]: VRRP_Instance(VI_1) Transition to MASTER STATE 18 Dec 13 18:52:46 localhost Keepalived_vrrp[28451]: VRRP_Instance(VI_1) Entering MASTER STATE 19 Dec 13 18:52:46 localhost Keepalived_vrrp[28451]: VRRP_Instance(VI_1) setting protocol VIPs. 20 Dec 13 18:52:46 localhost Keepalived_vrrp[28451]: Sending gratuitous ARP on eth0 for 192.168.199.51 21 22 此时就已经转移到master2上， 23 恢复master1 的mysql服务 24 service mysql start 25 恢复master1 的keepalive 26 /etc/init.d/keepalived start 27 28 查看master1的日志 29 Dec 13 18:54:41 localhost Keepalived_vrrp[71002]: VRRP_Instance(VI_1) Transition to MASTER STATE 30 Dec 13 18:54:42 localhost Keepalived_vrrp[71002]: VRRP_Instance(VI_1) Entering MASTER STATE 31 Dec 13 18:54:42 localhost Keepalived_vrrp[71002]: VRRP_Instance(VI_1) setting protocol VIPs. 32 Dec 13 18:54:42 localhost Keepalived_vrrp[71002]: Sending gratuitous ARP on eth0 for 192.168.199.51 33 34 查看master2的日志 35 Dec 13 18:54:41 localhost Keepalived_vrrp[28451]: VRRP_Instance(VI_1) Received advert with higher priority 100, ours 90 36 Dec 13 18:54:41 localhost Keepalived_vrrp[28451]: VRRP_Instance(VI_1) Entering BACKUP STATE 37 Dec 13 18:54:41 localhost Keepalived_vrrp[28451]: VRRP_Instance(VI_1) removing protocol VIPs. 38 39 说明服务已经转移到master2上了。 40 至此完成
完成
其它：删除主从复制
1 mysql> stop slave; 2 mysql> slave reset; 3 mysql> reset master;
keepalived
1 启动 2 /etc/init.d/keepalived start
异常解决办法
1 如果不能正常复制，查看状态 2 sql> show slave status G; 3 然后可以看到错误信息，一般情况下，重启一下slave就好了 4 sql> stop slave;