keepalived+mysql实现双主高可用

环境：

DB1：centos6.8、mysql5.5、192.168.2.204  hostname：bogon

DB2：centos6.8、mysql5.5、192.168.2.205  hostname：localhost.localdomain

vip：192.168.2.33

一、先配置DB1和DB2的双主热备

1、分别在DB1和DB2上安装mysql，我这里是用的ansible自动部署

[root@www ansible]# ansible-playbook lnmp.yml 

PLAY [new] *********************************************************************

TASK [setup] *******************************************************************
ok: [192.168.2.205]
ok: [192.168.2.204]

TASK [mysql : Create backup folder] ********************************************
ok: [192.168.2.204]
ok: [192.168.2.205]

TASK [mysql : create log folder] ***********************************************
changed: [192.168.2.204]
changed: [192.168.2.205]

TASK [mysql : copy mysql_tar_gz to client] *************************************
changed: [192.168.2.204]
changed: [192.168.2.205]

TASK [mysql : copy install_script to client] ***********************************
changed: [192.168.2.204]
changed: [192.168.2.205]

TASK [mysql : copy my.cnf to /data/backup] *************************************
changed: [192.168.2.204]
changed: [192.168.2.205]

TASK [mysql : install mysql] ***************************************************
changed: [192.168.2.204]
changed: [192.168.2.205]

PLAY RECAP *********************************************************************
192.168.2.204              : ok=    changed=    unreachable=    failed=
192.168.2.205              : ok=    changed=    unreachable=    failed=   

2、修改mysql的配置文件

首先修改DB1主机的配置文件，在/etc/my.cnf文件中的[mysqld]段添加以下内容

[root@bogon ~]# vim /etc/my.cnf
server-id = 1　　　　#节点标示，主从节点不能相同，必须全局唯一
log-bin=mysql-bin　　#开启mysql的binlog日志功能
relay-log = mysql-relay-bin 　　#开启relay-log日志，relay-log日志记录的是从服务器I/O线程将主服务器的二进制日志读取过来记录到从服务器本地文件，然后SQL线程会读取relay-log日志的内容并应用到从服务器
replicate-wild-ignore-table=mysql.%　　#复制过滤选项
replicate-wild-ignore-table=test.%
replicate-wild-ignore-table=information_schema.%

然后修改DB2主机的配置文件，

[root@localhost ~]# vim /etc/my.cnf
server-id =
log-bin=mysql-bin
relay-log = mysql-relay-bin
replicate-wild-ignore-table=mysql.%
replicate-wild-ignore-table=test.%
replicate-wild-ignore-table=information_schema.%

最后分别重启DB1和DB2使配置生效

3、创建复制用户并授权

注：在执行主主互备之前要保证两台server上数据一致

首先在DB1的mysql库中创建复制用户

mysql> grant replication slave on *.* to 'repl_user'@'192.168.2.205' identified by 'repl_passwd';
Query OK,  rows affected (0.04 sec)

mysql> show master status;
+------------------+----------+--------------+------------------+
| File             | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+----------+--------------+------------------+
| mysql-bin. |       |              |                  |
+------------------+----------+--------------+------------------+
 row in set (0.00 sec)

然后在DB2的mysql库中将DB1设为自己的主服务器

mysql> change master to \
    -> master_host='192.168.2.204',
    -> master_user='repl_user',
    -> master_password='repl_passwd',
    -> master_log_file='mysql-bin.000004',
    -> master_log_pos=;
Query OK,  rows affected (0.07 sec)

这里需要注意master_log_file和master_log_pos两个选项，这两个选项的值是在DB1上通过“show master status” 查询到的结果

接着在DB2上启动slave服务

mysql> start slave;
Query OK,  rows affected (0.01 sec)

下面查看DB2上slave的运行状态，

mysql> show slave status\G
*************************** . row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.2.204
                  Master_User: repl_user
                  Master_Port:
                Connect_Retry:
              Master_Log_File: mysql-bin.
          Read_Master_Log_Pos:
               Relay_Log_File: mysql-relay-bin.
                Relay_Log_Pos:
        Relay_Master_Log_File: mysql-bin.
             Slave_IO_Running: Yes　　　　#重点
            Slave_SQL_Running: Yes　　　　#重点
              Replicate_Do_DB:
          Replicate_Ignore_DB:
           Replicate_Do_Table:
       Replicate_Ignore_Table:
      Replicate_Wild_Do_Table:
  Replicate_Wild_Ignore_Table: mysql.%,test.%,information_schema.%　　#跳过的表
                   Last_Errno:
                   Last_Error:
                 Skip_Counter:
          Exec_Master_Log_Pos:
              Relay_Log_Space:
              Until_Condition: None
               Until_Log_File:
                Until_Log_Pos:
           Master_SSL_Allowed: No
           Master_SSL_CA_File:
           Master_SSL_CA_Path:
              Master_SSL_Cert:
            Master_SSL_Cipher:
               Master_SSL_Key:
        Seconds_Behind_Master:
Master_SSL_Verify_Server_Cert: No
                Last_IO_Errno:
                Last_IO_Error:
               Last_SQL_Errno:
               Last_SQL_Error:
  Replicate_Ignore_Server_Ids:
             Master_Server_Id:
 row in set (0.00 sec)

到这里，从DB1到DB2的mysql主从复制已经完成。接下来开始配置从DB2到DB1的mysql主从复制

在DB2的mysql库中创建复制用户

mysql> grant replication slave on *.* to 'repl_user'@'192.168.2.204' identified by 'repl_passwd';
Query OK,  rows affected (0.00 sec)

mysql> show master status;
+------------------+----------+--------------+------------------+
| File             | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+----------+--------------+------------------+
| mysql-bin. |       |              |                  |
+------------------+----------+--------------+------------------+
 row in set (0.00 sec)

然后在DB1的mysql库中将DB2设为自己的主服务器

mysql> change master to \
    -> master_host='192.168.2.205',
    -> master_user='repl_user',
    -> master_password='repl_passwd',
    -> master_log_file='mysql-bin.000005',
    -> master_log_pos=;
Query OK,  rows affected (0.07 sec)

最后，在DB1上启动slave服务

mysql> start slave;
Query OK,  rows affected (0.01 sec)

查看DB1上slave的运行状态

mysql> show slave status\G
*************************** . row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.2.205
                  Master_User: repl_user
                  Master_Port:
                Connect_Retry:
              Master_Log_File: mysql-bin.
          Read_Master_Log_Pos:
               Relay_Log_File: mysql-relay-bin.
                Relay_Log_Pos:
        Relay_Master_Log_File: mysql-bin.
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes
              Replicate_Do_DB:
          Replicate_Ignore_DB:
           Replicate_Do_Table:
       Replicate_Ignore_Table:
      Replicate_Wild_Do_Table:
  Replicate_Wild_Ignore_Table: mysql.%,test.%,information_schema.%
                   Last_Errno:
                   Last_Error:
                 Skip_Counter:
          Exec_Master_Log_Pos:
              Relay_Log_Space:
              Until_Condition: None
               Until_Log_File:
                Until_Log_Pos:
           Master_SSL_Allowed: No
           Master_SSL_CA_File:
           Master_SSL_CA_Path:
              Master_SSL_Cert:
            Master_SSL_Cipher:
               Master_SSL_Key:
        Seconds_Behind_Master:
Master_SSL_Verify_Server_Cert: No
                Last_IO_Errno:
                Last_IO_Error:
               Last_SQL_Errno:
               Last_SQL_Error:
  Replicate_Ignore_Server_Ids:
             Master_Server_Id:
 row in set (0.00 sec)

二、配置keepalived实现mysql双主高可用

1、安装keepalived

[root@bogon src]# tar zxf keepalived-1.2..tar.gz
[root@bogon src]# cd keepalived-1.2.
[root@bogon keepalived-1.2.]# ./configure --sysconf=/etc --with-kernel-dir=/lib/modules/2.6.-642.3..el6.x86_64/
[root@bogon keepalived-1.2.]# make && make install
[root@bogon keepalived-1.2.]# ln -s /usr/local/sbin/keepalived /sbin/
[root@bogon keepalived-1.2.]# chkconfig --add keepalived
[root@bogon keepalived-1.2.]# chkconfig --level  keepalived on
[root@bogon keepalived-1.2.24]# yum  -y install ipvsadm  ####之前没安装ipvsadm，导致 keepalived配置中lvs配置部分不生效，其中定义的notify_down 字段死活不生效，查了好久在发现是没安装ipvsadm导致的，泪奔！！！
[root@bogon keepalived-1.2.24]# ipvsadm

2、配置keepalived

DB1上keepalived.conf配置为

[root@bogon keepalived-1.2.]# cat /etc/keepalived/keepalived.conf
! Configuration File for keepalived

global_defs {
   notification_email {
     acassen@firewall.loc
     failover@firewall.loc
     sysadmin@firewall.loc
   }
   notification_email_from Alexandre.Cassen@firewall.loc
   smtp_server 192.168.200.1
   smtp_connect_timeout
   router_id LVS_DEVEL
   vrrp_skip_check_adv_addr
   vrrp_strict
   vrrp_garp_interval
   vrrp_gna_interval
}

vrrp_instance HA_1 {
    state BACKUP　　　　#在DB1和DB2上均配置为BACKUP
    interface eth1
    virtual_router_id
    priority
    advert_int
    nopreempt　　　　#不抢占模式，只有优先级高的机器上设置即可，优先级低的机器可不设置
    authentication {
        auth_type PASS
        auth_pass
    }
    virtual_ipaddress {
    192.168.2.33
    }
}

virtual_server 192.168.2.33  {
     delay_loop
     lb_algo wrr
     lb_kind DR
     persistence_timeout 60　　#会话保持时间 
     protocol TCP
     real_server 192.168.2.204  {
         weight
         notify_down /root/shutdown.sh　　#检测到服务down后执行的脚本 
         TCP_CHECK {
             connect_timeout 10　　#连接超时时间
             nb_get_retry 3　　　　#重连次数
             delay_before_retry 3　　 #重连间隔时间  
             connect_port 3306　　　　 #健康检查端口
         }
     }
}

DB2上keepalived.conf配置为

[root@localhost keepalived-1.2.]# cat /etc/keepalived/keepalived.conf
! Configuration File for keepalived

global_defs {
   notification_email {
     acassen@firewall.loc
     failover@firewall.loc
     sysadmin@firewall.loc
   }
   notification_email_from Alexandre.Cassen@firewall.loc
   smtp_server 192.168.200.1
   smtp_connect_timeout
   router_id LVS_DEVEL
   vrrp_skip_check_adv_addr
   vrrp_strict
   vrrp_garp_interval
   vrrp_gna_interval
}

vrrp_instance HA_1 {
    state BACKUP
    interface eth1
    virtual_router_id
    priority
    advert_int
    authentication {
        auth_type PASS
        auth_pass
    }
    virtual_ipaddress {
    192.168.2.33
    }
}

virtual_server 192.168.2.33  {
     delay_loop
     lb_algo wrr
     lb_kind DR
     persistence_timeout
     protocol TCP
     real_server 192.168.2.205  {
         weight
         notify_down /root/shutdown.sh
         TCP_CHECK {
             connect_timeout
             nb_get_retry
             delay_before_retry
             connect_port
         }
     }
}

编写检测服务down后所要执行的脚本shutdown.sh

[root@bogon ~]# cat /root/shtdown.sh
#!/bin/bash
killall keepalived

注：此脚本是上面配置文件notify_down选项所用到的，keepalived使用notify_down选项来检查real_server的服务状态，当发现real_server服务故障时，便触发此脚本；我们可以看到，脚本就一个命令，通过killall keepalived强制杀死keepalived进程，从而实现了MySQL故障自动转移。另外，我们不用担心两个MySQL会同时提供数据更新操作，因为每台MySQL上的keepalived的配置里面只有本机MySQL的IP+VIP，而不是两台MySQL的IP+VIP

启动keepalived并查看日志

[root@bogon keepalived-1.2.]# chmod  /etc/init.d/keepalived
[root@bogon keepalived-1.2.]# service keepalived start
正在启动 keepalived：                                      [确定]
[root@bogon keepalived-1.2.]# tail -f /var/log/messages
Oct  :: bogon Keepalived_vrrp[]: Sending gratuitous ARP on eth1 for 192.168.2.33
Oct  :: bogon Keepalived_vrrp[]: Sending gratuitous ARP on eth1 for 192.168.2.33
Oct  :: bogon Keepalived_vrrp[]: Sending gratuitous ARP on eth1 for 192.168.2.33
Oct  :: bogon Keepalived_vrrp[]: Sending gratuitous ARP on eth1 for 192.168.2.33
Oct  :: bogon Keepalived_vrrp[]: Sending gratuitous ARP on eth1 for 192.168.2.33
Oct  :: bogon Keepalived_vrrp[]: VRRP_Instance(HA_1) Sending/queueing gratuitous ARPs on eth1 for 192.168.2.33
Oct  :: bogon Keepalived_vrrp[]: Sending gratuitous ARP on eth1 for 192.168.2.33
Oct  :: bogon Keepalived_vrrp[]: Sending gratuitous ARP on eth1 for 192.168.2.33
Oct  :: bogon Keepalived_vrrp[]: Sending gratuitous ARP on eth1 for 192.168.2.33
Oct  :: bogon Keepalived_vrrp[]: Sending gratuitous ARP on eth1 for 192.168.2.33

三、测试功能

1、在远程客户端通过vip登陆测试

[root@www ansible]# mysql -h 192.168.2.33 -uroot -p
Enter password:
Welcome to the MySQL monitor.  Commands end with ; or \g.
Your MySQL connection id is
Server version: 5.5.-log Source distribution

Copyright (c) , , 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>

mysql> show variables like "%hostname%"
  -> ;
  +---------------+-------+
  | Variable_name | Value |
  +---------------+-------+
  | hostname | bogon |
  +---------------+-------+
  1 row in set (0.00 sec)

从sql输出结果看，可以通过vip登陆，并且登陆了DB1服务器

2、创建一个数据库，然后在这个库重创建一个表，并插入数据

mysql> create database repldb;
Query OK,  row affected (0.02 sec)

mysql> show databases;
+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| performance_schema |
| repldb             |
| test               |
+--------------------+
 rows in set (0.06 sec)

mysql> use repldb;
Database changed
mysql> create table repl_table(id int,email varchar(),password varchar() not null);
Query OK,  rows affected (0.03 sec)

mysql> show tables;
+------------------+
| Tables_in_repldb |
+------------------+
| repl_table       |
+------------------+
 row in set (0.01 sec)

mysql> insert into repl_table(id,email,password) values(,"master@163.com","qweasd");
Query OK,  row affected (0.00 sec)

登陆DB2主机的mysql，可数据是否复制成功

mysql> show variables like "%hostname%";
+---------------+-----------------------+
| Variable_name | Value                 |
+---------------+-----------------------+
| hostname      | localhost.localdomain |
+---------------+-----------------------+
 row in set (0.01 sec)

mysql> show databases;
+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| performance_schema |
| repldb             |
| test               |
+--------------------+
 rows in set (0.05 sec)

mysql> use repldb;
Database changed
mysql> show tables;
+------------------+
| Tables_in_repldb |
+------------------+
| repl_table       |
+------------------+
 row in set (0.00 sec)

mysql> select * from repl_table;
+------+----------------+----------+
| id   | email          | password |
+------+----------------+----------+
|     | master@.com | qweasd   |
+------+----------------+----------+
 row in set (0.08 sec)

3、停止DB1主机上的mysql，查看故障是否自动转移

[root@bogon ~]# service mysqld stop
Shutting down MySQL.. SUCCESS! 

登陆192.168.2.33查看：

mysql> show variables like "%hostname%";
ERROR  (HY000): MySQL server has gone away
No connection. Trying to reconnect...
Connection id:
Current database: repldb

+---------------+-----------------------+
| Variable_name | Value                 |
+---------------+-----------------------+
| hostname      | localhost.localdomain |
+---------------+-----------------------+
 row in set (0.01 sec)

可以看到现在登陆的是DB2 故障自动切换成功

接着，插入数据看DB1是否能复制

mysql> insert into repl_table(id,email,password) values(,"slave@163.com","qweasd");
Query OK,  row affected (0.06 sec)

mysql> use repldb;
Database changed
mysql> select * from repl_table;
+------+----------------+----------+
| id   | email          | password |
+------+----------------+----------+
|     | master@.com | qweasd   |
|     | slave@.com  | qweasd   |
+------+----------------+----------+
 rows in set (0.00 sec)

登陆DB1查看表数据

[root@bogon ~]# service mysqld start
Starting MySQL. SUCCESS!
[root@bogon ~]# mysql -uroot -p
Enter password:
Welcome to the MySQL monitor.  Commands end with ; or \g.
Your MySQL connection id is
Server version: 5.5.-log Source distribution

Copyright (c) , , 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> use repldb;
Database changed
mysql> select * from repl_table;
+------+----------------+----------+
| id   | email          | password |
+------+----------------+----------+
|     | master@.com | qweasd   |
|     | slave@.com  | qweasd   |
+------+----------------+----------+
 rows in set (0.02 sec)

复制成功！

到此全部完成！！！