通过全备+主从同步恢复被drop的库或表

MySQL 中drop 等高危误操作后恢复方法

实验目的:

• 本次实验以恢复drop操作为例,使用不同方法进行误操作的数据恢复.
  
  方法:
• 利用master同步(本文)]
• 伪master+Binlog+同步
• 利用binlog改为realy-log在临时slave更新
• 没有备份怎么恢复被drop的表(利用undrop-for-innodb)
  
  1.环境
• mysql 5.7.23-log
• master db210(192.168.99.210)
• slave db212(192.168.99.212)
• 空闲服务器:db211(192.168.99.211)
• 压力:在做本次实验过程用,使用sysbench同时做读写模拟压力(28thread+10*100000行)

2.本实验的前置条件:

• 在drop前至少有一次全量备份
• 从全备到drop操作时间点有全程binlog,并且binlog format='row'
  
  如果没有以上两个备份怎么办?见《无备份的误操作恢复方法》

3.模拟误操作

root@localhost [wenyz]>select count(*) from t2;
+----------+
| count(*) |
+----------+
|  163818 |
+----------+
1 row in set (0.12 sec)

root@localhost [wenyz]>drop table t2;
Query OK, 0 rows affected (0.02 sec)

root@localhost [wenyz]>show master status;
+------------------+-----------+--------------+------------------+------------------------------------------------+
| File            | Position  | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set                              |
+------------------+-----------+--------------+------------------+------------------------------------------------+
| mysql-bin.000020 | 136445791 |              |                  | 9e912d2b-92f6-11e8-9747-0050568be82c:1-9872111 |
+------------------+-----------+--------------+------------------+------------------------------------------------+
1 row in set (0.00 sec)

root@localhost [wenyz]>flush logs;
Query OK, 0 rows affected (1.41 sec)

4.在实验不同方法之前先完成以下步骤:

• 复制全备份和所有binlog文件到空闲服务器:db211上
• 以全备恢复到一个临时新实例(临时slave).

5.找到误操作的位置(GTID或binlog pos)

通过查找怀疑的binlog:
[root@db210_19:11:07 /data/57mysql/mysql3506/logs]
#ll
total 75564544
-rw-r----- 1 mysql mysql       177 Jul 29 14:14 mysql-bin.000001
...
-rw-r----- 1 mysql mysql  39336996 Jul 29 19:12 mysql-bin.000020
-rw-r----- 1 mysql mysql     13340 Jul 29 19:12 mysql-bin.index
[root@db210_19:12:48 /data/57mysql/mysql3506/logs] 

binlog文件导出命令:
#mysqlbinlog -v --base64-output=auto /data/57mysql/mysql3506/logs/mysql-bin.000020 >/data/backup/20.sql
binlog文件导出后查找:DROP TABLE `t2`:

在mysql-bin.000020中找到对应位置:
# at 97871478
#180731 19:33:29 server id 1853506  end_log_pos 97871543 CRC32 0x8cd4354d       GTID    last_committed=205567   sequence_number=205570  rbr_only=no     original_committed_timestamp=0  immediate_commit_timestamp=0    transaction_length=0
# original_commit_timestamp=0 (1970-01-01 08:00:00.000000 CST)
# immediate_commit_timestamp=0 (1970-01-01 08:00:00.000000 CST)
/*!80001 SET @@session.original_commit_timestamp=0*//*!*/;
SET @@SESSION.GTID_NEXT= '9e912d2b-92f6-11e8-9747-0050568be82c:9791090'/*!*/;
# at 97871543
#180731 19:33:29 server id 1853506  end_log_pos 97871660 CRC32 0x974a86d8       Query   thread_id=35    exec_time=0     error_code=0
use `wenyz`/*!*/;
SET TIMESTAMP=1533036809/*!*/;
/*!\C utf8 *//*!*/;
SET @@session.character_set_client=33,@@session.collation_connection=33,@@session.collation_server=33/*!*/;
DROP TABLE `t2` /* generated by server */
/*!*/;
# at 97871660
#180731 19:33:29 server id 1853506  end_log_pos 97871725 CRC32 0xb85257d6       GTID    last_committed=205567   sequence_number=205571  rbr_only=yes    original_committed_timestamp=0  immediate_commit_timestamp=0    transaction_length=0
/*!50718 SET TRANSACTION ISOLATION LEVEL READ COMMITTED*//*!*/;
# original_commit_timestamp=0 (1970-01-01 08:00:00.000000 CST)
# immediate_commit_timestamp=0 (1970-01-01 08:00:00.000000 CST)
/*!80001 SET @@session.original_commit_timestamp=0*//*!*/;
SET @@SESSION.GTID_NEXT= '9e912d2b-92f6-11e8-9747-0050568be82c:9791091'/*!*/;
# at 97871725

6.恢复方法一：

• 构建主的从库,并通过:start slave sql_thread until sql_before_gtids='9e912d2b-92f6-11e8-9747-0050568be82c:9791090' 试设置SQL回放位置停在drop操作前一个事务.
• Executed_Gtid_Set应该停在: 9e912d2b-92f6-11e8-9747-0050568be82c:1-9791089

change master to master_host='db210',master_port=3506,master_user='wyz',master_password='xxxx',master_auto_position=1;
root@localhost [(none)]>start slave io_thread ;
Query OK, 0 rows affected (0.01 sec)

root@localhost [(none)]>show slave status\G;
*************************** 1. row ***************************
               Slave_IO_State: Queueing master event to the relay log
                  Master_Host: db210
                  Master_User: wyz
                  Master_Port: 3506
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000015
          Read_Master_Log_Pos: 10476294
               Relay_Log_File: relay-bin.000001
                Relay_Log_Pos: 4
        Relay_Master_Log_File:
             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: 0
                   Last_Error:
                 Skip_Counter: 0
          Exec_Master_Log_Pos: 0
              Relay_Log_Space: 27669048
              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: 0
               Last_SQL_Error:
  Replicate_Ignore_Server_Ids:
             Master_Server_Id: 1853506
                  Master_UUID: 9e912d2b-92f6-11e8-9747-0050568be82c
             Master_Info_File: /data/57mysql/mysql3506/data/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:
               Master_SSL_Crl:
           Master_SSL_Crlpath:
           Retrieved_Gtid_Set: 9e912d2b-92f6-11e8-9747-0050568be82c:6730274-6788383
            Executed_Gtid_Set: 9e912d2b-92f6-11e8-9747-0050568be82c:1-6730273
                Auto_Position: 1
         Replicate_Rewrite_DB:
                 Channel_Name:
           Master_TLS_Version:
1 row in set (0.00 sec)

ERROR:
No query specified

root@localhost [(none)]>
root@localhost [(none)]>start slave sql_thread until sql_before_gtids='9e912d2b-92f6-11e8-9747-0050568be82c:9791090';
Query OK, 0 rows affected (0.00 sec)

• 查看sql_thread 执行位置停在9791089

root@localhost [(none)]>show slave status\G;
*************************** 1. row ***************************
               Slave_IO_State: Queueing master event to the relay log
                  Master_Host: db210
                  Master_User: wyz
                  Master_Port: 3506
                Connect_Retry: 60
              Master_Log_File: mysql-bin.000079
          Read_Master_Log_Pos: 15969193
               Relay_Log_File: relay-bin.000024
                Relay_Log_Pos: 97871691
        Relay_Master_Log_File: mysql-bin.000020
             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: 0
                   Last_Error:
                 Skip_Counter: 0
          Exec_Master_Log_Pos: 97871478
              Relay_Log_Space: 15602695666
              Until_Condition: SQL_BEFORE_GTIDS
               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: 0
               Last_SQL_Error:
  Replicate_Ignore_Server_Ids:
             Master_Server_Id: 1853506
                  Master_UUID: 9e912d2b-92f6-11e8-9747-0050568be82c
             Master_Info_File: /data/57mysql/mysql3506/data/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:
               Master_SSL_Crl:
           Master_SSL_Crlpath:
           Retrieved_Gtid_Set: 9e912d2b-92f6-11e8-9747-0050568be82c:6730274-42357224
            Executed_Gtid_Set: 9e912d2b-92f6-11e8-9747-0050568be82c:1-9791089
                Auto_Position: 1
         Replicate_Rewrite_DB:
                 Channel_Name:
           Master_TLS_Version:
1 row in set (0.00 sec)

ERROR:
No query specified

root@localhost [(none)]>

• 查看数据是否和drop相同:

root@localhost [(none)]>select count(*) from wenyz.t2;
+----------+
| count(*) |
+----------+
|   163818 |
+----------+
1 row in set (0.10 sec)

root@localhost [(none)]>

• 将数据表导出

/usr/local/mysql57/bin/mysqldump -S /tmp/mysql3506.sock -pxxxx --master-data=2 --single-transaction wenyz t2 >/data/backup/t2.sql

• 将t2.sql传回master 并导入

[root@db211_20:49:35 /root]
#scp /data/backup/t2.sql db210:/data/backup/
t2.sql

[root@db211_20:50:10 /root]
#/usr/local/mysql57/bin/mysql -S /tmp/mysql3506.sock -pxxxx <t2.sql
mysql: [Warning] Using a password on the command line interface can be insecure.
[root@db210_21:01:36 /data/backup]
#fg
/usr/local/mysql57/bin/mysql --login-path=p3506	(wd: /tmp)
root@localhost [wenyz]>select count(*) from t2;
+----------+
| count(*) |
+----------+
|   163818 |
+----------+
1 row in set (0.40 sec)