【MySQL】主从GTID复制修复

作者  董红禹

沃趣科技数据库工程师

导 读

---

GTID是5.6新增特性，减少DBA运维的工作。在以前一主两从架构下当主库M1发生故障我们需要选择一个从库S1作为新的主库，但是S2要重新change master 到新主库上 这时master_log_file和master_log_pos我们怎么获取到呢？在没有GTID时 MHA架构帮我们解决了这个问题，在有了GTID情况下，我们只需要在S2上面重新change master下设置MASTER_AUTO_POSITION = 1即可。 本文介绍下在GTID复制下遇到了错误如何解决。

场景描述

---

使用show slave status\G 查看主备复制状态时，错误如下:

[root@shadow1:/root 5.6.28-log_Instance1 root@localhost:test 12:56:57]>SHOW SLAVE STATUS\G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 10.10.30.101
                  Master_User: repl
                  Master_Port: 3306
                Connect_Retry: 10
              Master_Log_File: mysql-bin.000003
          Read_Master_Log_Pos: 28074558
               Relay_Log_File: mysql-relay-bin.000011
                Relay_Log_Pos: 100311
        Relay_Master_Log_File: mysql-bin.000003
             Slave_IO_Running: Yes
            Slave_SQL_Running: No
              Replicate_Do_DB:
          Replicate_Ignore_DB:
           Replicate_Do_Table:
       Replicate_Ignore_Table:
      Replicate_Wild_Do_Table:
  Replicate_Wild_Ignore_Table:
                   Last_Errno: 1007
                   Last_Error: Error 'Can't create database 'dhy2'; database exists' on query. Default database: 'dhy2'. Query: 'create database dhy2'
                 Skip_Counter: 0
          Exec_Master_Log_Pos: 28073671
              Relay_Log_Space: 287925
              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: NULL
Master_SSL_Verify_Server_Cert: No
                Last_IO_Errno: 0
                Last_IO_Error:
               Last_SQL_Errno: 1007
               Last_SQL_Error: Error 'Can't create database 'dhy2'; database exists' on query. Default database: 'dhy2'. Query: 'create database dhy2'
  Replicate_Ignore_Server_Ids:
             Master_Server_Id: 101
                  Master_UUID: dd2a02a3-f0be-11e5-af62-0050563a97cc
             Master_Info_File: mysql.slave_master_info
                    SQL_Delay: 0
          SQL_Remaining_Delay: NULL
      Slave_SQL_Running_State:
           Master_Retry_Count: 86400
                  Master_Bind:
      Last_IO_Error_Timestamp:
     Last_SQL_Error_Timestamp: 160408 12:57:02
               Master_SSL_Crl:
           Master_SSL_Crlpath:
           Retrieved_Gtid_Set: dd2a02a3-f0be-11e5-af62-0050563a97cc:70261-71462
            Executed_Gtid_Set: dd2a02a3-f0be-11e5-af62-0050563a97cc:1-71459,
e28420fb-f0be-11e5-af62-0050562edd64:1-81209
                Auto_Position: 0
1 row in set (0.00 sec)

根据报错描述:

Last_SQL_Error: Error 'Can't create database 'dhy2'; database exists' on query. Default database: 'dhy2'. Query: 'create database dhy2' 
dhy2 这个数据库已经在备库存在

我们在备库上面show database 查看dhy2这个数据库确实是存在的:

[root@shadow1:/root 5.6.28-log_Instance1 root@localhost:test 12:57:04]>show databases;
+--------------------+
| Database           |
+--------------------+
| information_schema |
| dhy2               |
| mysql              |
| performance_schema |
| test               |
+--------------------+
7 rows in set (0.00 sec)

修改过程

---

当我们使用GTID复制时，不能像传统复制一样跳过事务，只能注册一个空的事务骗过MySQL。具体操作步骤如下:

查看当前备份的GTID执行情况

           Retrieved_Gtid_Set: dd2a02a3-f0be-11e5-af62-0050563a97cc:70261-71462
            Executed_Gtid_Set: dd2a02a3-f0be-11e5-af62-0050563a97cc:1-71459,
e28420fb-f0be-11e5-af62-0050562edd64:1-81209
                Auto_Position: 0

Retrieved_Gtid_Set 代表已经接受到的GTID集合 
Executed_Gtid_Set 代码已经执行的GTID集合

针对上面GTID执行情况 我们可以看到:

Retrieved_Gtid_Set: dd2a02a3-f0be-11e5-af62-0050563a97cc:70261-71462
 Executed_Gtid_Set: dd2a02a3-f0be-11e5-af62-0050563a97cc:1-71459

接收到了71462 执行到了71459就报错了。

手工注册事务

[root@shadow1:/root 5.6.28-log_Instance1 root@localhost:test 12:59:41]>stop slave;
Query OK, 0 rows affected (0.00 sec)
[root@shadow1:/root 5.6.28-log_Instance1 root@localhost:test 13:04:57]>set gtid_next='dd2a02a3-f0be-11e5-af62-0050563a97cc:71460';
Query OK, 0 rows affected (0.00 sec)
[root@shadow1:/root 5.6.28-log_Instance1 root@localhost:test 13:05:23]>begin;
Query OK, 0 rows affected (0.00 sec)
[root@shadow1:/root 5.6.28-log_Instance1 root@localhost:test 13:05:29]>commit;
Query OK, 0 rows affected (0.02 sec)
[root@shadow1:/root 5.6.28-log_Instance1 root@localhost:test 13:05:31]>SET GTID_NEXT='AUTOMATIC';
Query OK, 0 rows affected (0.00 sec)
[root@shadow1:/root 5.6.28-log_Instance1 root@localhost:test 13:05:36]>start slave;
Query OK, 0 rows affected (0.00 sec)

这里需要注意的是

set gtid_next='dd2a02a3-f0be-11e5-af62-0050563a97cc:71460'; 这个值是
Executed_Gtid_Set: dd2a02a3-f0be-11e5-af62-0050563a97cc:1-71459  71459这个GTID+1
并且 uuid一定是Retrieved_Gtid_Set: dd2a02a3-f0be-11e5-af62-0050563a97cc:70261-71462 接收到这里显示的uuid

再次查看show slave status\G 已经恢复正常

[root@shadow1:/root 5.6.28-log_Instance1 root@localhost:test 13:05:39]>show slave status\G
*************************** 1. row ***************************
               Slave_IO_State: Waiting for master to send event
                  Master_Host: 10.10.30.101
                  Master_User: repl
                  Master_Port: 3306
                Connect_Retry: 10
              Master_Log_File: mysql-bin.000003
          Read_Master_Log_Pos: 28275288
               Relay_Log_File: mysql-relay-bin.000012
                Relay_Log_Pos: 21794
        Relay_Master_Log_File: mysql-bin.000003
             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:
                   Last_Errno: 0
                   Last_Error:
                 Skip_Counter: 0
          Exec_Master_Log_Pos: 28275288
              Relay_Log_Space: 302335
              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:
  Replicate_Ignore_Server_Ids:
             Master_Server_Id: 101
                  Master_UUID: dd2a02a3-f0be-11e5-af62-0050563a97cc
             Master_Info_File: mysql.slave_master_info
                    SQL_Delay: 0
          SQL_Remaining_Delay: NULL
      Slave_SQL_Running_State: Slave has read all relay log; waiting for the slave I/O thread to update it
           Master_Retry_Count: 86400
                  Master_Bind:
      Last_IO_Error_Timestamp:
     Last_SQL_Error_Timestamp:
               Master_SSL_Crl:
           Master_SSL_Crlpath:
           Retrieved_Gtid_Set: dd2a02a3-f0be-11e5-af62-0050563a97cc:70261-72005
            Executed_Gtid_Set: dd2a02a3-f0be-11e5-af62-0050563a97cc:1-72005,
e28420fb-f0be-11e5-af62-0050562edd64:1-81209
                Auto_Position: 0

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/28218939/viewspace-2137231/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/28218939/viewspace-2137231/