mysql replication cluster 双机热备集群搭建及常见错误排除

 一.参考网址
https://174.136.43.78/prox2/
http://www.clusterdb.com/mysql-cluster/setting-up-mysql-asynchronous-replication-for-high-availability/
http://oskb.blogbus.com/logs/53722298.html
http://www.howtoforge.com/mysql_master_master_replication

二.环境
 master主机        192.168.100.114
 slave主机机       192.168.100.15
 数据存放目录      /opt/cluster/mysql/var
 mysql安装目录     /opt/cluster/mysql
 mysql配置文件     /etc/my.cnf


三.安装mysql
 # cd mysql-5.1.44
 # ./configure --help
 # ./configure --prefix=/opt/cluster/mysql;make;makeinstall;
 # useradd mysql
 # cd /opt/cluster/mysql/bin/
 #./mysqladmin -u root password 123456
 # cp ./mysql.server /etc/init.d/mysql
 # chmod 755 /etc/init.d/mysql
 # chkconfig --level 345 mysql on
 # service mysql restart
 # cp support-files/my-medium.cnf /etc/my.cnf

四.master配置

1、远程同步用户授权
在master(192.168.100.114)上
[root@localhost bin]# pwd
/opt/cluster/mysql/bin
[root@localhost bin]# ./mysql -uroot -p123456
mysql>grant replication slave on *.* to [email protected] identified by 'jason';
mysql>GRANT FILE,SELECT,REPLICATION SLAVE ON *.* TO [email protected] IDENTIFIED BY 'jason';
mysql>\q

2、同步数据库
在master上
[root@localhost bin]# ./mysql -uroot -p123456
mysql>USE appdb;
mysql>FLUSH TABLES WITH READ LOCK;
注意：此时不能退出mysql的shell否则数据库的只读锁会自动取消。我们可以在新的窗口dump数据库，或者直接拷贝数据文件目录
手工同步完成后解锁
(打开一个新的远程窗口
# cd /opt/cluster/mysql/var
# scp -r appdb/ [email protected]:/opt/cluster/mysql/var
# 提示输入密码：15机器root用户密码
 开始传输文件，同步备份完成。)
返回刚才添加只读锁的窗口，对数据库表解锁
mysql>UNLOCK TABLES;
mysql>\q
起始数据库同步完成。

3、修改配置文件
打开master上/etc/my.cnf
在[mysqld] 标签下添加
server-id=1
binlog-do-db= appdb
重启master的mysql服务



---------------------------------------------------
五.配置slave(192.168.100.15)

1、修改配置文件
打开/etc/my.cnf
在[mysqld] 标签下添加
master-host=192.1668.100.114
master-user=jason
master-password=jason
master-port=3306
server-id=2
master-connect-retry=60
replicate-do-db=appdb
log-slave-updates


2、检验远程用户及权限
mysql -h 192.168.100.114 -u jason -pjason
mysql> SHOW GRANTS;
mysql>\q
重新启动mysql服务

3、更改master主机
[root@localhost bin]# pwd
/opt/cluster/mysql/bin
[root@localhost bin]# ./mysql -uroot -p123456
mysql> CHANGE MASTER TO MASTER_HOST='192.168.100.114', MASTER_PORT=3306, MASTER_USER='jason', MASTER_PASSWORD='jason', MASTER_LOG_FILE='', MASTER_LOG_POS=4;
Query OK, 0 rows affected (0.00 sec)
mysql> use appdb;
Database changed
mysql> show master status\G
*************************** 1. row ***************************
            File: mysql-bin.000031
        Position: 439
    Binlog_Do_DB:
Binlog_Ignore_DB:
1 row in set (0.00 sec)

mysql> show slave status\G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 192.168.100.114
                  Master_User: jason
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000020
          Read_Master_Log_Pos: 439
               Relay_Log_File: localhost-relay-bin.000034
                Relay_Log_Pos: 584
        Relay_Master_Log_File: mysql-bin.000020
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes
              Replicate_Do_DB: appdb
          Replicate_Ignore_DB:
           Replicate_Do_Table:
       Replicate_Ignore_Table:
      Replicate_Wild_Do_Table:
  Replicate_Wild_Ignore_Table:
                   Last_Errno: 0
                   Last_Error:
                 Skip_Counter: 0
          Exec_Master_Log_Pos: 439
              Relay_Log_Space: 888
              Until_Condition: None
               Until_Log_File:
                Until_Log_Pos: 0
           Master_SSL_Allowed: No
           Master_SSL_CA_File:
           Master_SSL_CA_Path:
              Master_SSL_Cert:
            Master_SSL_Cipher:
               Master_SSL_Key:
        Seconds_Behind_Master: 0
Master_SSL_Verify_Server_Cert: No
                Last_IO_Errno: 0
                Last_IO_Error:
               Last_SQL_Errno: 0
               Last_SQL_Error:
1 row in set (0.00 sec)
mysql>\q
设置成功！

4、测试同步情况
当前情况下数据库两台服务器上的appdb数据库是相同的

(1)到master下修改数据库
[root@localhost bin]# ./mysql -uroot -p123456
mysql> USE appdb
Database changed
mysql> create table numbers (num1 int, num2 int);
Query OK, 0 rows affected (0.01 sec)

mysql> insert into numbers values (1,10),(2,20),(3,30);
Query OK, 3 rows affected (0.00 sec)
Records: 3  Duplicates: 0  Warnings: 0

mysql> select * from numbers;
+------+------+
| num1 | num2 |
+------+------+
|    1 |   10 |
|    2 |   20 |
|    3 |   30 |
+------+------+
3 rows in set (0.00 sec)
(2)到slave服务器上查看同步情况
[root@localhost bin]# ./mysql -uroot -p123456
mysql> show databases;
+--------------------+
| Database           |
+--------------------+
| information_schema |
| appdb            |
| mysql              |
| test               |
+--------------------+
4 rows in set (0.00 sec)

mysql> use appdb
Database changed
mysql> show tables;
+-------------------+
| Tables_in_cluster |
+-------------------+
| numbers           |
+-------------------+
1 row in set (0.00 sec)

mysql> select * from numbers;
+------+------+
| num1 | num2 |
+------+------+
|    1 |   10 |
|    2 |   20 |
|    3 |   30 |
+------+------+
3 rows in set (0.00 sec)

-------------------------------------------------------------------
以同样方法在 将两台机器角色互换一下，安装配置一次，双机高可用的集群配置完成
--------------------------------------------------------------

六.常见问题解决方案

(1) 如果你的第一个主数据库（master1）包含的状态（在撰写本文时），您就无法使用命令“从主服务器同步数据”。由于此错误：
你不得不要用现有系统的一些其他方法复制数据，例如通过mysqldump的复制和粘贴文件或数据流数据。需要打开数据库表的只读锁，方法参见上文(四、2章节)。

(2) 如果你的表里使用了自动增长的主键，而且你用的是双master的mysql集群，可能会遇到两台master添加了同一个id，从而出现crash。
那你就应该使用以下选项添加到my.cnf文件中,重启服务。
Make Master 1 only auto-increment odd numbers by adding this to my.cnf under [mysqld]:
auto_increment_increment= 2
auto_increment_offset   = 1

Make Master 2 only auto-increment even numbers by adding this to my.cnf under [mysqld]:
auto_increment_increment= 2
auto_increment_offset   = 2
此方法原理是通过设置是两个master在写入时自动增长的主键相互岔开,以避免出现crash。

(3)如果是在一个机子上做好后，将整套环境转移过去的话，最常见的就是文件权限问题。需要用chown来将文件的所属组更改正确。
  并使用chmod将新机器上对应的配置文件及相关目录的权限修改好。


(4) Slave_IO_Running: No  Slave_SQL_Running: No

   在master端msql登录后 show slave status\G后发现Slave_IO_Running: No  Slave_SQL_Running: No 而且在执行命令式出现此类错误报告
ERROR 1201 (HY000): Could not initialize master info structure; more error messages can be found in the MySQL error log
可能原因是你的master端已经有了前一个slave的日志文件，对更改新的slave没有日志文件。你需要重置一下你的slave
解决方案
登录到slave机器的mysql下   
[root@localhost bin]# pwd
/opt/cluster/mysql/bin
[root@localhost bin]# ./mysql -uroot -p123456
mysql> STOP SLAVE;
mysql> RESET SLAVE;
mysql> CHANGE MASTER TO MASTER_HOST='192.168.100.114(根据实际主机ip改变)', MASTER_PORT=3306, MASTER_USER='jason',
MASTER_PASSWORD='jason', MASTER_LOG_FILE='', MASTER_LOG_POS=4;
mysql> START SLVE;
mysql> show master status\G
mysql>\q

(5) Slave_IO_Running: Yes  Slave_SQL_Running: No

   1、首先定位同步不成功的原因，查数据库日志。
   2、若是从主机重启，事物回滚，则
登录到slave机器的mysql下   
[root@localhost bin]# pwd
/opt/cluster/mysql/bin
[root@localhost bin]# ./mysql -uroot -p123456
mysql> slave stop;
mysql> set GLOBAL SQL_SLAVE_SKIP_COUNTER=1;
mysql> slave start;
mysql> show master status\G
mysql>\q

   3、其他网络故障，则登录先到master下的mysql，查看主机的日志情况
[root@localhost bin]# pwd
/opt/cluster/mysql/bin
[root@localhost bin]# ./mysql -uroot -p123456
mysql> show master status;
+------------------+-----------+--------------+------------------+
| File             | Position  | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+-----------+--------------+------------------+
| mysql-bin.000004 | 244274056 | appdb        |                  |
+------------------+-----------+--------------+------------------+
1 row in set (0.00 sec)
mysql>\q

 现在登录到slave的mysql下
[root@localhost bin]# pwd
/opt/cluster/mysql/bin
[root@localhost bin]# ./mysql -uroot -p123456
mysql> slave stop;
CHANGE MASTER TO MASTER_HOST='192.168.100.114(根据实际主机ip改变)', MASTER_PORT=3306, MASTER_USER='jason', MASTER_PASSWORD='jason', MASTER_LOG_FILE='mysql-bin.000004', MASTER_LOG_POS=244274056;
(注意将MASTER_LOG_FILE改为上文看到的master的日志表的file下文件名称，同时将MASTER_LOG_POS改为上文看到的master的日志表的Position下文件名称)
mysql> slave start;
mysql> show master status\G
mysql>\q

(6) Slave_IO_Running: No  Slave_SQL_Running: Yes
    
两种办法，重置slave，表解锁同步数据
  
    1、重置slave
登录到master下的mysql，查看主机的日志情况
[root@localhost bin]# pwd
/opt/cluster/mysql/bin
[root@localhost bin]# ./mysql -uroot -p123456
mysql> show master status;
+------------------+-----------+--------------+------------------+
| File             | Position  | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+-----------+--------------+------------------+
| mysql-bin.000005 | 244274178 | appdb        |                  |
+------------------+-----------+--------------+------------------+
1 row in set (0.00 sec)
mysql>\q
 现在登录到slave的mysql下
[root@localhost bin]# pwd
/opt/cluster/mysql/bin
[root@localhost bin]# ./mysql -uroot -p123456
mysql> slave stop;
mysql> reset slave;
mysql> CHANGE MASTER TO MASTER_HOST='192.168.100.114(根据实际主机ip改变)', MASTER_PORT=3306, MASTER_USER='jason', MASTER_PASSWORD='jason', MASTER_LOG_FILE='mysql-bin.000005', MASTER_LOG_POS=244274178;
(注意将MASTER_LOG_FILE改为上文看到的master的日志表的file下文件名称，同时将MASTER_LOG_POS改为上文看到的master的日志表的Position下文件名称)
mysql> slave start;
mysql> show master status\G
mysql>\q

  2、表解锁同步数据
[root@localhost bin]# pwd
/opt/cluster/mysql/bin
[root@localhost bin]# ./mysql -uroot -p123456
mysql> slave stop;
mysql> FLUSH TABLES WITH READ LOCK;
mysql> UNLOCK TABLES;
mysql> CHANGE MASTER TO MASTER_LOG_FILE='mysqld-bin.00000*';
mysql> CHANGE MASTER TO MASTER_LOG_POS=*;
mysql> slave start;
mysql> show master status\G
mysql>\q