查到mysqld的pid
[mysql@mvxl0782 binlog]$ netstat -nltp|grep mysqld
(Not all processes could be identified, non-owned process info
 will not be shown, you would have to be root to see it all.)
tcp        0      0 :::3306                     :::*                        LISTEN      18401/mysqld

這里18401

2.利用進程pid查看對應的文件描述符
[mysql@mvxl0782 binlog]$ ll /proc/18401/fd|egrep 'ib_|ibdata'
lrwx------ 1 mysql mysql 64 8月  11 10:33 10 -> /mysql/server/data/ib_logfile1
lrwx------ 1 mysql mysql 64 8月  11 10:33 11 -> /mysql/server/data/ib_logfile2
lrwx------ 1 mysql mysql 64 8月  11 10:33 4 -> /mysql/server/data/ibdata1
lrwx------ 1 mysql mysql 64 8月  11 10:33 9 -> /mysql/server/data/ib_logfile0

上面的10、11、4和9就是我們要恢復的文件

3、關閉前端業務或執行flush tables with read lock;  不讓有數據寫入

4、驗證有沒有數據寫入

a.先輸入下面命令，讓臟頁盡快寫入到磁盤里
set global innodb_max_dirty_pages_ptc=0;
b.查看binlog日志寫入情況，確保file和position沒有變化  show master status;
c.最后查看innodb狀態信息，確保臟頁已刷入到磁盤中
show engine inndodb status\G

TRANSACTIONS
...........
trx id counter A21837
Purge don for trx's n:o<A21837 undo n:o<0
##確保后臺purge進程把undo log全部清除掉，事務ID要一致

INSERT BUFFER AND ADAPTIVE HASH INDEX
Ibuf:size 1,free kust keb 65,seg size67,0 merges
##插入buffer合并插入緩存等于1

LOG
Log sequence number xxxx
Log flushed up to xxxx
Last checkpoint at xxx
##確保這3個值不在變化

BUFFER POOL AND MEMORY
............
Modified db pages 0
##確保臟頁數量為0

ROW OPERATIONS
..................
0.00 inserts/s,0.00 updates/s,0.00 deletes/s,0.00 reads/s
#確保插入、更新、刪除為0

5.恢復操作

[mysql@mvxl0782 binlog]$ ll /proc/18401/fd|egrep 'ib_|ibdata'
lrwx------ 1 mysql mysql 64 8月  11 10:33 10 -> /mysql/server/data/ib_logfile1
lrwx------ 1 mysql mysql 64 8月  11 10:33 11 -> /mysql/server/data/ib_logfile2
lrwx------ 1 mysql mysql 64 8月  11 10:33 4 -> /mysql/server/data/ibdata1
lrwx------ 1 mysql mysql 64 8月  11 10:33 9 -> /mysql/server/data/ib_logfile0

將上面文件復制并改名：

cp 10 /mysql/server/data/ib_logfile1
cp 11 /mysql/server/data/ib_logfile2
cp 4 /mysql/server/data/ibdata1
cp 9 /mysql/server/data/ib_logfile0

授權：
cd /mysql/server/data
chown -R mysql:dba ib*

最后重啟mysql