在数据文件下的innodb ibdata包括表空间:ibdata1,ibdata2,回滚日志ib_logfile0,ib_logfile1,ib_logfile2.
[root@localhost data]# ls
5k72 db-bin.000015 dbjijin-bin.000001 ibdata1 ib_logfile1 localhost-slow.log mysql.err test
auto.cnf db-bin.000016 dbjijin-bin.000002 ibdata2 ib_logfile2 monitor performance_schema test333
db-bin.000014 db-bin.index dbjijin-bin.index ib_logfile0 localhost.localdomain.pid mysql rrr ttt
[root@localhost data]#

现在删除这几个文件:

[root@localhost data]# rm -f ib*
[root@localhost data]# ls
5k72 db-bin.000015 dbjijin-bin.000001 localhost-slow.log mysql.err test
auto.cnf db-bin.000016 dbjijin-bin.000002 monitor performance_schema test333
db-bin.000014 db-bin.index dbjijin-bin.index localhost.localdomain.pid mysql rrr ttt
[root@localhost data]#

删除之后,数据库还可以正常工作,切记不要停止mysql服务,不然神仙也难救.....

先找到mysql的pid.
[root@localhost data]# netstat -ntlp|grep mysqld
tcp 0 0 :::3307 :::* LISTEN 4863/mysqld
tcp 0 0 :::3306 :::* LISTEN 2169/mysqld


第一个是我们需要的pid号,4863
通过pid号,找到相关的文件,红色标记的是我们需要的文件:
[root@localhost data]# ll /proc/4863/fd
total 0
lr-x------. 1 root root 64 Apr 24 10:58 0 -> /dev/null
l-wx------. 1 root root 64 Apr 24 10:58 1 -> /home/mysql3307/data/mysql.err
lrwx------. 1 root root 64 Apr 24 10:58 10 -> /home/mysql3307/data/ib_logfile0 (deleted)
lrwx------. 1 root root 64 Apr 24 10:58 11 -> /home/mysql3307/data/ib_logfile1 (deleted)
lrwx------. 1 root root 64 Apr 24 10:58 12 -> /home/mysql3307/data/ib_logfile2 (deleted)
.
..
...
lrwx------. 1 root root 64 Apr 24 10:58 4 -> /home/mysql3307/data/ibdata1 (deleted)
..
...
....
lrwx------. 1 root root 64 Apr 24 10:58 9 -> /home/mysql3307/data/ibdata2 (deleted)

这时,需要暂停前端的业务,也就是需要停止数据的写操作.
关闭业务,或者: flush tables with read lock;
然后把脏页尽快刷入到磁盘里,
mysql>set global innodb_max_dirty_pages_pct=0;
然后查看binlog日志写入情况,确保file和position的值没有变化
Your MySQL connection id is 7
Server version: 5.6.20-log Source distribution
Copyright (c) 2000, 2014, 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> show master status;
+---------------+----------+--------------+------------------+-------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+---------------+----------+--------------+------------------+-------------------+
| db-bin.000016 | 3415 | | | |
+---------------+----------+--------------+------------------+-------------------+
1 row in set (0.00 sec)
mysql> show master status;
+---------------+----------+--------------+------------------+-------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+---------------+----------+--------------+------------------+-------------------+
| db-bin.000016 | 3415 | | | |
+---------------+----------+--------------+------------------+-------------------+
1 row in set (0.00 sec)
mysql> show master status;
+---------------+----------+--------------+------------------+-------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+---------------+----------+--------------+------------------+-------------------+
| db-bin.000016 | 3415 | | | |
+---------------+----------+--------------+------------------+-------------------+
1 row in set (0.00 sec)
mysql> show master status;
+---------------+----------+--------------+------------------+-------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+---------------+----------+--------------+------------------+-------------------+
| db-bin.000016 | 3415 | | | |
+---------------+----------+--------------+------------------+-------------------+
1 row in set (0.00 sec)

然后查看innodb状态信息,确保脏页已经刷入磁盘.
mysql> show engine innodb status \G
************************** 1. row ***************************
Type: InnoDB
Name:
Status:
=====================================
2015-04-24 11:51:52 7f7038202700 INNODB MONITOR OUTPUT
------------
TRANSACTIONS
------------
Trx id counter 18696
Purge done for trx's n:o < 18696 undo n:o < 0 state: running but idle
##确保后台purge进程吧 undo log 全部清除掉,事务id要一致.

-------------------------------------
INSERT BUFFER AND ADAPTIVE HASH INDEX
-------------------------------------
Ibuf: size 1, free list len 0, seg size 2, 0 merges
## insert buffer 合并插入缓存等于1
---
LOG
---
Log sequence number 5196495
Log flushed up to 5196495
Pages flushed up to 5196495
Last checkpoint at 5196495
## 确保这4个值不会变

----------------------
BUFFER POOL AND MEMORY
----------------------
Total memory allocated 1098907648; in additional pool allocated 0
Dictionary memory allocated 74136
Buffer pool size 65528
Free buffers 65121
Database pages 406
Old database pages 0
Modified db pages 0
## 确保脏页数据为0

--------------
ROW OPERATIONS
--------------
0 queries inside InnoDB, 0 queries in queue
0 read views open inside InnoDB
Main thread process no. 1964, id 140119885477632, state: sleeping
Number of rows inserted 1, updated 0, deleted 0, read 4
0.00 inserts/s, 0.00 updates/s, 0.00 deletes/s, 0.00 reads/s
## 确保插入,更新,删除为0


经过上面确认后,可以进行恢复操作了.复制上面表示部分数据到数据库目录:
[root@localhost fd]# cp 4 /home/mysql3307/data/ibdata1
[root@localhost fd]# cp 10 /home/mysql3307/data/ib_logfile0
[root@localhost fd]# cp 11 /home/mysql3307/data/ib_logfile1
[root@localhost fd]# cp 12 /home/mysql3307/data/ib_logfile2
给文件授权:
[root@localhost fd]# chown mysql:mysql /home/mysql3307/data/ib*