linux rm误删除数据库文件的恢复方法
1.首先测试rm 误删除数据库文件$ ll
total 2153164
-rw-r----- 1 oracle dba 8536064 Nov 27 21:50 control01.ctl
-rw-r----- 1 oracle dba 8536064 Nov 27 21:50 control02.ctl
-rw-r----- 1 oracle dba 8536064 Nov 27 21:50 control03.ctl
-rw-r----- 1 oracle dba 104865792 Nov 27 21:32 ggs01.dbf
-rw-r----- 1 oracle dba 525344768 Nov 27 21:32 perfstat.dbf
-rw-r----- 1 oracle dba 52429312 Nov 27 21:10 redo01.log
-rw-r----- 1 oracle dba 52429312 Nov 27 21:49 redo02.log
-rw-r----- 1 oracle dba 52429312 Nov 27 21:10 redo03.log
-rw-r----- 1 oracle dba 10493952 Nov 27 21:32 shkz_data_01.dbf
-rw-r----- 1 oracle dba 524296192 Nov 27 21:49 sysaux01.dbf
-rw-r----- 1 oracle dba 524296192 Nov 27 21:49 system01.dbf
-rw-r----- 1 oracle dba 209723392 Nov 18 10:54 temp01.dbf
-rw-r----- 1 oracle dba 52436992 Nov 27 21:32 test01.dbf
-rw-r----- 1 oracle dba 251666432 Nov 27 21:49 undotbs01.dbf
-rw-r----- 1 oracle dba 26222592 Nov 27 21:32 users01.dbf
$ rm test01.dbf
2.找到数据库后台进程dbw0
oracle 3493 1 0 21:10 ? 00:00:00 ora_dbw0_dbwdn
进程号3493
3.找到被误删除的数据库文件
$ ll /proc/3493/fd
total 0
lr-x------ 1 oracle oinstall 64 Nov 27 21:52 0 -> /dev/null
lr-x------ 1 oracle oinstall 64 Nov 27 21:52 1 -> /dev/null
lr-x------ 1 oracle oinstall 64 Nov 27 21:52 10 -> /dev/zero
lr-x------ 1 oracle oinstall 64 Nov 27 21:52 11 -> /dev/zero
lr-x------ 1 oracle oinstall 64 Nov 27 21:52 12 -> /u01/app/oracle/product/10.2.0/db_1/rdbms/mesg/oraus.msb
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 13 -> /u01/app/oracle/product/10.2.0/db_1/dbs/hc_dbwdn.dat
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 14 -> /u01/app/oracle/product/10.2.0/db_1/dbs/lkDBWDN
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 15 -> /u01/app/oracle/oradata/dbwdn/control01.ctl
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 16 -> /u01/app/oracle/oradata/dbwdn/control02.ctl
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 17 -> /u01/app/oracle/oradata/dbwdn/control03.ctl
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 18 -> /u01/app/oracle/oradata/dbwdn/system01.dbf
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 19 -> /u01/app/oracle/oradata/dbwdn/undotbs01.dbf
lr-x------ 1 oracle oinstall 64 Nov 27 21:52 2 -> /dev/null
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 20 -> /u01/app/oracle/oradata/dbwdn/sysaux01.dbf
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 21 -> /u01/app/oracle/oradata/dbwdn/users01.dbf
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 22 -> /u01/app/oracle/oradata/dbwdn/test01.dbf (deleted)
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 23 -> /u01/app/oracle/oradata/dbwdn/ggs01.dbf
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 24 -> /u01/app/oracle/oradata/dbwdn/perfstat.dbf
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 25 -> /u01/app/oracle/oradata/dbwdn/shkz_data_01.dbf
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 26 -> /u01/app/oracle/oradata/dbwdn/temp01.dbf
lr-x------ 1 oracle oinstall 64 Nov 27 21:52 27 -> /u01/app/oracle/product/10.2.0/db_1/rdbms/mesg/oraus.msb
lr-x------ 1 oracle oinstall 64 Nov 27 21:52 3 -> /dev/null
lr-x------ 1 oracle oinstall 64 Nov 27 21:52 4 -> /dev/null
l-wx------ 1 oracle oinstall 64 Nov 27 21:52 5 -> /u01/app/oracle/admin/dbwdn/udump/dbwdn_ora_3485.trc
l-wx------ 1 oracle oinstall 64 Nov 27 21:52 6 -> /u01/app/oracle/admin/dbwdn/bdump/alert_dbwdn.log
lrwx------ 1 oracle oinstall 64 Nov 27 21:52 7 -> /u01/app/oracle/product/10.2.0/db_1/dbs/hc_dbwdn.dat
l-wx------ 1 oracle oinstall 64 Nov 27 21:52 8 -> /u01/app/oracle/admin/dbwdn/bdump/alert_dbwdn.log
误删除之后，数据库没有重启，系统也没用重启
4.拷贝数据库文件
$ cp /proc/3493/fd/22 /u01/app/oracle/oradata/dbwdn/test01.dbf
$ cd /u01/app/oracle/oradata/dbwdn
$ ls
control01.ctl control03.ctl perfstat.dbf redo02.log shkz_data_01.dbf system01.dbf test01.dbf users01.dbf
control02.ctl ggs01.dbf redo01.log redo03.log sysaux01.dbf temp01.dbf undotbs01.dbf
数据库文件找回来了。
5.恢复数据库文件
$ sqlplus "/as sysdba"
SQL*Plus: Release 10.2.0.5.0 - Production on Tue Nov 27 21:55:15 2012
Copyright (c) 1982, 2010, Oracle. All Rights Reserved.
Connected to:
Oracle Database 10g Enterprise Edition Release 10.2.0.5.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> alter database datafile 5 offline;
Database altered.
SQL> recover datafile 5;
Media recovery complete.
SQL> alter database datafile 5 online;
Database altered.
SQL> shutdown immediate;
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL> startup
ORACLE instance started.
Total System Global Area 285212672 bytes
Fixed Size 1273276 bytes
Variable Size 130024004 bytes
Database Buffers 150994944 bytes
Redo Buffers 2920448 bytes
Database mounted.
Database opened.
SQL>
成功恢复数据库
注：如果数据库重启之后，还能恢复吗，测试一下：
$ rm users01.dbf$ ps -ef |grep dbw0
oracle 3798 1 0 21:56 ? 00:00:00 ora_dbw0_dbwdn
oracle 4147 3439 0 22:07 pts/1 00:00:00 grep dbw0
$ ll /proc/3798/fd
total 0
lr-x------ 1 oracle oinstall 64 Nov 27 22:07 0 -> /dev/null
lr-x------ 1 oracle oinstall 64 Nov 27 22:07 1 -> /dev/null
lr-x------ 1 oracle oinstall 64 Nov 27 22:07 10 -> /dev/zero
lr-x------ 1 oracle oinstall 64 Nov 27 22:07 11 -> /dev/zero
lr-x------ 1 oracle oinstall 64 Nov 27 22:07 12 -> /u01/app/oracle/product/10.2.0/db_1/rdbms/mesg/oraus.msb
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 13 -> /u01/app/oracle/product/10.2.0/db_1/dbs/hc_dbwdn.dat
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 14 -> /u01/app/oracle/product/10.2.0/db_1/dbs/lkDBWDN
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 15 -> /u01/app/oracle/oradata/dbwdn/control01.ctl
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 16 -> /u01/app/oracle/oradata/dbwdn/control02.ctl
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 17 -> /u01/app/oracle/oradata/dbwdn/control03.ctl
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 18 -> /u01/app/oracle/oradata/dbwdn/system01.dbf
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 19 -> /u01/app/oracle/oradata/dbwdn/undotbs01.dbf
lr-x------ 1 oracle oinstall 64 Nov 27 22:07 2 -> /dev/null
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 20 -> /u01/app/oracle/oradata/dbwdn/sysaux01.dbf
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 21 -> /u01/app/oracle/oradata/dbwdn/users01.dbf (deleted)lrwx------ 1 oracle oinstall 64 Nov 27 22:07 22 -> /u01/app/oracle/oradata/dbwdn/test01.dbf
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 23 -> /u01/app/oracle/oradata/dbwdn/ggs01.dbf
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 24 -> /u01/app/oracle/oradata/dbwdn/perfstat.dbf
lrwx------ 1 oracle oinstall 64 Nov 27 22:07 25 -> /u01/app/oracle/oradata/dbwdn/shkz_data_01.dbf
数据库没有重启，这个数据文件还能恢复，如果我重启数据库之后，还能看见这个数据文件吗？
$ sqlplus "/as sysdba"
SQL*Plus: Release 10.2.0.5.0 - Production on Tue Nov 27 22:08:01 2012
Copyright (c) 1982, 2010, Oracle. All Rights Reserved.
Connected to:
Oracle Database 10g Enterprise Edition Release 10.2.0.5.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> shutdown immediate;
ORA-01116: error in opening database file 4
ORA-01110: data file 4: '/u01/app/oracle/oradata/dbwdn/users01.dbf'
ORA-27041: unable to open file
Linux Error: 2: No such file or directory
Additional information: 3
SQL> shutdown abort;ORACLE instance shut down.
SQL> exit
Disconnected from Oracle Database 10g Enterprise Edition Release 10.2.0.5.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
$ sqlplus "/as sysdba"
SQL*Plus: Release 10.2.0.5.0 - Production on Tue Nov 27 22:08:19 2012
Copyright (c) 1982, 2010, Oracle. All Rights Reserved.
Connected to an idle instance.
SQL> startup
ORACLE instance started.
Total System Global Area 285212672 bytes
Fixed Size 1273276 bytes
Variable Size 130024004 bytes
Database Buffers 150994944 bytes
Redo Buffers 2920448 bytes
Database mounted.
ORA-01157: cannot identify/lock data file 4 - see DBWR trace file
ORA-01110: data file 4: '/u01/app/oracle/oradata/dbwdn/users01.dbf'
SQL> exit
Disconnected from Oracle Database 10g Enterprise Edition Release 10.2.0.5.0 - Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
$ ll /proc/3798/fd
ls: /proc/3798/fd: No such file or directory
$ ps -ef |grep dbw0
oracle 4160 1 0 22:08 ? 00:00:00 ora_dbw0_dbwdn
oracle 4193 3439 0 22:09 pts/1 00:00:00 grep dbw0
$ ll /proc/4160/fd
total 0
lr-x------ 1 oracle oinstall 64 Nov 27 22:09 0 -> /dev/null
lr-x------ 1 oracle oinstall 64 Nov 27 22:09 1 -> /dev/null
lr-x------ 1 oracle oinstall 64 Nov 27 22:09 10 -> /dev/zero
lr-x------ 1 oracle oinstall 64 Nov 27 22:09 11 -> /dev/zero
lr-x------ 1 oracle oinstall 64 Nov 27 22:09 12 -> /u01/app/oracle/product/10.2.0/db_1/rdbms/mesg/oraus.msb
lrwx------ 1 oracle oinstall 64 Nov 27 22:09 13 -> /u01/app/oracle/product/10.2.0/db_1/dbs/hc_dbwdn.dat
lrwx------ 1 oracle oinstall 64 Nov 27 22:09 14 -> /u01/app/oracle/product/10.2.0/db_1/dbs/lkDBWDN
lrwx------ 1 oracle oinstall 64 Nov 27 22:09 15 -> /u01/app/oracle/oradata/dbwdn/control01.ctl
lrwx------ 1 oracle oinstall 64 Nov 27 22:09 16 -> /u01/app/oracle/oradata/dbwdn/control02.ctl
lrwx------ 1 oracle oinstall 64 Nov 27 22:09 17 -> /u01/app/oracle/oradata/dbwdn/control03.ctl
l-wx------ 1 oracle oinstall 64 Nov 27 22:09 2 -> /u01/app/oracle/admin/dbwdn/bdump/dbwdn_dbw0_4160.trc
lr-x------ 1 oracle oinstall 64 Nov 27 22:09 3 -> /dev/null
lr-x------ 1 oracle oinstall 64 Nov 27 22:09 4 -> /dev/null
l-wx------ 1 oracle oinstall 64 Nov 27 22:09 5 -> /u01/app/oracle/admin/dbwdn/udump/dbwdn_ora_4152.trc
l-wx------ 1 oracle oinstall 64 Nov 27 22:09 6 -> /u01/app/oracle/admin/dbwdn/bdump/alert_dbwdn.log
lrwx------ 1 oracle oinstall 64 Nov 27 22:09 7 -> /u01/app/oracle/product/10.2.0/db_1/dbs/hc_dbwdn.dat
l-wx------ 1 oracle oinstall 64 Nov 27 22:09 8 -> /u01/app/oracle/admin/dbwdn/bdump/alert_dbwdn.log
lrwx------ 1 oracle oinstall 64 Nov 27 22:09 9 -> /u01/app/oracle/product/10.2.0/db_1/dbs/lkinstdbwdn (deleted)
$
已经没法恢复了。
说明需要注意的是，rm误删除文件之后，千万不要慌张，稳住神，不要重启数据库，不要重启动系统。
AIX系统，我没有测试过。
使用备份集恢复上面的数据库文件：
$ rman target /
Recovery Manager: Release 10.2.0.5.0 - Production on Tue Nov 27 22:13:45 2012
Copyright (c) 1982, 2007, Oracle. All rights reserved.
connected to target database (not started)
RMAN> startup nomount;
Oracle instance started
Total System Global Area 285212672 bytes
Fixed Size 1273276 bytes
Variable Size 130024004 bytes
Database Buffers 150994944 bytes
Redo Buffers 2920448 bytes
RMAN> alter database mount;
using target database control file instead of recovery catalog
database mounted
RMAN> run
2> {
3> allocate channel t1 type to disk;
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-00558: error encountered while parsing input commands
RMAN-01009: syntax error: found "to": expecting one of: "double-quoted-string, disk, equal, identifier, single-quoted-string"
RMAN-01007: at line 3 column 26 file: standard input
RMAN> run
2> {
3> allocate channel t1 device type disk;
4> restore datafile 4;
5> recover datafile 4;
6> release channel t1;
7> }
allocated channel: t1
channel t1: sid=156 devtype=DISK
Starting restore at 27-NOV-2012 22:16:26
channel t1: starting datafile backupset restore
channel t1: specifying datafile(s) to restore from backup set
restoring datafile 00004 to /u01/app/oracle/oradata/dbwdn/users01.dbf
channel t1: reading from backup piece /u01/bak/03nrctnp_1_1.bak
channel t1: restored backup piece 1
piece handle=/u01/bak/03nrctnp_1_1.bak tag=TAG20121127T213241
channel t1: restore complete, elapsed time: 00:00:04
Finished restore at 27-NOV-2012 22:16:30
Starting recover at 27-NOV-2012 22:16:30
starting media recovery
media recovery complete, elapsed time: 00:00:01
Finished recover at 27-NOV-2012 22:16:31
released channel: t1
RMAN> alter database open;
database opened
RMAN>
Windows下数据恢复（把回收站清空后的恢复方法）：
步骤：
1、单击“开始——运行，然后输入regedit （打开注册表）
2、依次展开：HEKEY——LOCAL——MACHIME/SOFTWARE/microsoft/WINDOWS/ CURRENTVERSION/EXPLORER/DESKTOP/NAMESPACE 在左边空白外点击“新建”
，选择：“主键”，把它命名为“645FFO40——5081——101B——9F08——00AA002F954E”
再把右边的“默认”的主键的键值设为“回收站”，然后退出注册表。就OK啦。
3、要重启计算机。
只要机器没有运行过磁盘整理。系统完好.任何时候的文件都可以找回来。 好详细的教程啊！必须收藏备用！
之前公司有个新入职的DBA差点酿成大祸，就是不知道这招，差点冲动重启数据库。还好被老员工及时拦住了。
补充几点经验：
1. 预防永远比补救重要。建议在生产环境给重要数据文件所在目录加个保护，比如用chattr +i保护，或者干脆把数据文件目录单独挂载一个分区并设置为只读。
2. 养成好习惯：rm -i 强制提示确认，虽然平时嫌麻烦，但关键时候能救命。或者直接alias rm='rm -i'。
3. 关于Windows那个注册表方法，确实有用，不过现在很多Ghost系统或者精简版系统可能没有效果，还是得靠专业的数据恢复软件。
4. 如果有RMAN备份的话，恢复确实是最后的安全保障，但作者也展示了proc方法在数据库还在运行时的强大之处。
总之记住一点：误删文件后，第一时间要做的就是停止一切可能覆盖磁盘的操作，能不关机就不关机，能不重启数据库就不重启。
感谢楼主的无私分享！👍
頁:
[1]