Oracle数据库坏块修复

        昨天下午数据库负载突然升高，最高达每分钟40左右。正常情况下 最高也就每分钟2左右。经过常用的方法检查，也没有找到原因，后来自动好了。为了确保这样的情况不再莫名的发生，需要从更深的层次的找到原因。

        今天打开alert日志文件，发现有坏块错误。这时才怀疑是不是由于坏块导致的负载升高，这还有待进一步的确认。不过当前最主要的是把坏块给处理掉。

alert.log错误日志如下：

Hex dump of (file 34, block 43487) in trace file /opt/oracle/admin/mbar/udump/orcl_ora_18143.trc

Corrupt block relative dba: 0x0880a9df (file 34, block 43487)

Fractured block found during buffer read

Data in bad block:

 type: 6 format: 2 rdba: 0x0880a9df

 last change scn: 0x09c9.33765b70 seq: 0x2 flg: 0x06

 spare1: 0x0 spare2: 0x0 spare3: 0x0

 consistency value in tail: 0x44a90602

 check value in block header: 0xf70d

 computed block checksum: 0xe59a

Reread of rdba: 0x0880a9df (file 34, block 43487) found same corrupted data

Wed Jun 27 15:11:24 2012

Corrupt Block Found

         TSN = 10, TSNAME = BBS_DATA

         RFN = 34, BLK = 43487, RDBA = 142649823

         OBJN = 97543, BJD = 97543, BJECT = T_USER_LOGON, SUBOBJECT = 

         SEGMENT WNER = BBS, SEGMENT TYPE = Table Segment

从日志中可以清晰的看出是用户BBS下的BBS_DATA表空间的文件号为34存在坏块，并且表名也给出来了是T_USER_LOGON。

[oracle@db02 ~]$ sqlplus 用户名/密码

SQL> SELECT COUNT(*) FROM T_USER_LOGON;

执行命令后可以查出来数据，并没有报错。但是执行下面命令后就报错了。

SQL> SELECT COUNT(*) FROM T_USER_LOGON  T WHERE TO_CHAR(T.LOGON_TIME,'YYYY-MM-DD')='2012-06-28';

提示有坏块，为什么加上个where条件就不行，不知道是那里的问题。

以下是解决过程 ：

1、根据文件号查找对应的文件：

SQL> select file#,name from v$datafile where file#=34;

2、使用dbv检查： 

[oracle@db02 ~]$ dbv file=/opt/oradata/play03.dbf blocksize=8192

DBVERIFY: Release 10.2.0.1.0 - Production on 星期四 28 10:30:51 2012

Copyright (c) 1982, 2005, Oracle. All rights reserved.

DBVERIFY - Verification starting : FILE = /opt/oradata/play03.dbf

DBV-00200: Block, dba 30767628, already marked corrupted

DBVERIFY - Verification complete

Total Pages Examined : 4194302

Total Pages Processed (Data) : 3508146

Total Pages Failing (Data) : 0

Total Pages Processed (Index): 605887

Total Pages Failing (Index): 0

Total Pages Processed (Other): 75876

Total Pages Processed (Seg) : 0

Total Pages Failing (Seg) : 0

Total Pages Empty : 4393

Total Pages Marked Corrupt : 1

Total Pages Influx : 0

Highest block SCN : 3781089739 (0.3781089739)

3、其次要确认坏块类型是物理坏块还是逻辑坏块

physical corruption check: backup validate datafile 34;

logical corruption check: backup check logical validate datafile 34;

先进行物理坏块检测

RMAN> backup validate datafile 34;

Starting backup at 28-6月 -12

using target database control file instead of recovery catalog

allocated channel: ORA_DISK_1

channel ORA_DISK_1: sid=1035 devtype=DISK

channel ORA_DISK_1: starting full datafile backupset

channel ORA_DISK_1: specifying datafile(s) in backupset

input datafile fno=00034 name=/opt/oradata/play03.dbf

channel ORA_DISK_1: backup set complete, elapsed time: 00:01:25

Finished backup at 28-6月 -12

4、查看视图：

SQL> select * from V$DATABASE_BLOCK_CORRUPTION;

     FILE#     BLOCK#     BLOCKS CORRUPTION_CHANGE# CORRUPTIO

---------- ---------- ---------- ------------------ ---------

        34      43487          1                  0 FRACTURED

确认为物理坏块！

5、 最后也是最重要的一步，进行坏块恢复。注意这一步必须在有备份的情况下执行，否则恢复不了。

RMAN>  BLOCKRECOVER CORRUPTION LIST;

注： 由于我没有备份，所以也不能用这种方式恢复。哭呀！现在终于知道备份的重要性了。所以改用其他方法（exp/imp导出导入）。

6、 恢复步骤:

[oracle@db02 bdump]$ exp 用户名/密码 file='/opt/tlogon-0628.dmp' tables=(T_USER_LOGON)

Export: Release 10.2.0.1.0 - Production on 28 10:08:43 2012

Copyright (c) 1982, 2005, Oracle.  All rights reserved.

Connected to: Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production

With the Partitioning, OLAP and Data Mining options

Export done in ZHS16GBK character set and AL16UTF16 NCHAR character set

About to export specified tables via Conventional Path ...

. . exporting table               T_USER_LOGON

EXP-00056: ORACLE error 1578 encountered

ORA-01578: ORACLE data block corrupted (file # 34, block # 43487)

ORA-01110: data file 34: '/opt/oradata/play03.dbf'

Export terminated successfully with warnings.

[oracle@db02 bdump]$ 

[oracle@db02 bdump]$ sqlplus  用户名/密码

SQL*Plus: Release 10.2.0.1.0 - Production on 星期四 28 10:30:51 2012

Copyright (c) 1982, 2005, Oracle.  All rights reserved.

Connected to:

Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production

With the Partitioning, OLAP and Data Mining options

SQL> ALTER SYSTEM SET EVENTS='10231 trace name context forever,level 10';

System altered.

SQL> host

I注：

ALTER SYSTEM SET EVENTS='10231 trace name context forever,level 10' ;

内部事件，设置在全表扫描时跳过损坏的数据块.

[oracle@db02 bdump]$ exp  用户名/密码  file='/opt/tlogon-0628.dmp' tables=(T_USER_LOGON)

Export: Release 10.2.0.1.0 - Production on 星期四 28 10:31:11 2012

Copyright (c) 1982, 2005, Oracle.  All rights reserved.

Connected to: Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production

With the Partitioning, OLAP and Data Mining options

Export done in ZHS16GBK character set and AL16UTF16 NCHAR character set

About to export specified tables via Conventional Path ...

. . exporting table               T_USER_LOGON     768657 rows exported

Export terminated successfully without warnings.

注：这时候成功导出!

然后我们可以drop table，recreate，然后导入数据！

[oracle@db02 bdump]$ exit

SQL> ALTER SYSTEM SET EVENTS='10231 trace name context off'; 

System altered.

SQL> drop table T_USER_LOGON;

SQL>host

Disconnected from Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production

With the Partitioning, OLAP and Data Mining options

[oracle@db02 bdump]$ imp   用户名/密码  file='/opt/tlogon-0628.dmp' ignore=y full=y

7、 完成数据恢复！验证。

[oracle@db02 ~]$ sqlplus 用户名/密码

SQL>  SELECT COUNT(*) FROM T_USER_LOGON;

SQL>  SELECT COUNT(*) FROM T_USER_LOGON  T WHERE TO_CHAR(T.LOGON_TIME,'YYYY-MM-DD')='2012-06-28';

都没有报错，正常。

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/23009281/viewspace-734044/，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/23009281/viewspace-734044/