如果在还原出来的数据文件中有坏块,而归档日志和联机日志是正常的,那么在应用日志恢复过程中,会出现什么情况,这里通过一个简单的测试给予其中一种情况的说明
创建测试表

SQL> select * from v$version;
BANNER
----------------------------------------------------------------
Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - Prod
PL/SQL Release 10.2.0.4.0 - Production
CORE    10.2.0.4.0      Production
TNS for Linux: Version 10.2.0.4.0 - Production
NLSRTL Version 10.2.0.4.0 - Production
SQL> create table t_xifenfei(object_id,object_name) tablespace xifenfei
  2  as
  3  select object_id,object_name from dba_objects
  4  where rownum<11;
Table created.
SQL> col object_name for a30
SQL> select   object_id,object_name,
  2   dbms_rowid.rowid_relative_fno(rowid)rel_fno,
  3  dbms_rowid.rowid_block_number(rowid)blockno
  4   from chf.t_xifenfei;
 OBJECT_ID OBJECT_NAME                       REL_FNO    BLOCKNO
---------- ------------------------------ ---------- ----------
        20 ICOL$                                   5         12
        44 I_USER1                                 5         12
        28 CON$                                    5         12
        15 UNDO$                                   5         12
        29 C_COBJ#                                 5         12
         3 I_OBJ#                                  5         12
        25 PROXY_ROLE_DATA$                        5         12
        39 I_IND1                                  5         12
        51 I_CDEF2                                 5         12
        26 I_PROXY_ROLE_DATA$_1                    5         12
10 rows selected.
SQL> select name from v$datafile where file#=5;
NAME
--------------------------------------------------------------
/u01/oracle/oradata/XFF/xifenfei01.dbf
SQL> update t_xifenfei set object_name='WWW.XIFENFEI.COM';
10 rows updated.
SQL> commit;
Commit complete.
SQL> create table t_xifenfei_new(object_id,object_name) tablespace xifenfei
  2  as
  3  select object_id,object_name from dba_objects
  4  where rownum<11;
Table created.
SQL> conn / as sysdba
Connected.
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.

备份数据文件

[oracle@xifenfei XFF]$ cp xifenfei01.dbf  ../tmp/
[oracle@xifenfei XFF]$ ll ../tmp/xifenfei01.dbf
-rw-r----- 1 oracle oinstall 10493952 Sep 28 19:05 ../tmp/xifenfei01.dbf
[oracle@xifenfei XFF]$ date
Fri Sep 28 19:05:42 CST 2012

bbed破坏备份文件

[oracle@xifenfei XFF]$ bbed password=blockedit
BBED: Release 2.0.0.0.0 - Limited Production on Fri Sep 28 19:05:59 2012
Copyright (c) 1982, 2007, Oracle.  All rights reserved.
************* !!! For Oracle Internal Use only !!! ***************
BBED> set filename '/u01/oracle/oradata/tmp/xifenfei01.dbf'
        FILENAME        /u01/oracle/oradata/tmp/xifenfei01.dbf
BBED> set block 12
        BLOCK#          12
BBED> set mode edit
        MODE            Edit
BBED> map
 File: /u01/oracle/oradata/tmp/xifenfei01.dbf (0)
 Block: 12                                    Dba:0x00000000
------------------------------------------------------------
 KTB Data Block (Table/Cluster)
 struct kcbh, 20 bytes                      @0
 struct ktbbh, 96 bytes                     @20
 struct kdbh, 14 bytes                      @124
 struct kdbt[1], 4 bytes                    @138
 sb2 kdbr[10]                               @142
 ub1 freespace[7666]                        @162
 ub1 rowdata[360]                           @7828
 ub4 tailchk                                @8188
BBED> p kcbh
struct kcbh, 20 bytes                       @0
   ub1 type_kcbh                            @0        0x06
   ub1 frmt_kcbh                            @1        0xa2
   ub1 spare1_kcbh                          @2        0x00
   ub1 spare2_kcbh                          @3        0x00
   ub4 rdba_kcbh                            @4        0x0140000c
   ub4 bas_kcbh                             @8        0x0004d7b0
   ub2 wrp_kcbh                             @12       0x000a
   ub1 seq_kcbh                             @14       0x01
   ub1 flg_kcbh                             @15       0x06 (KCBHFDLC, KCBHFCKV)
   ub2 chkval_kcbh                          @16       0xe573
   ub2 spare3_kcbh                          @18       0x0000
BBED> d offset 8188
 File: /u01/oracle/oradata/tmp/xifenfei01.dbf (0)
 Block: 12               Offsets: 8188 to 8191           Dba:0x00000000
------------------------------------------------------------------------
 0106b0d7
 <32 bytes per line>
BBED> m /x 11
Warning: contents of previous BIFILE will be lost. Proceed? (Y/N) y
 File: /u01/oracle/oradata/tmp/xifenfei01.dbf (0)
 Block: 12               Offsets: 8188 to 8191           Dba:0x00000000
------------------------------------------------------------------------
 1106b0d7
 <32 bytes per line>
BBED> sum apply
Check value for File 0, Block 12:
current = 0xe563, required = 0xe563
BBED> verify
DBVERIFY - Verification starting
FILE = /u01/oracle/oradata/tmp/xifenfei01.dbf
BLOCK = 12
Block 12 is corrupt
Corrupt block relative dba: 0x0140000c (file 0, block 12)
Fractured block found during verification
Data in bad block:
 type: 6 format: 2 rdba: 0x0140000c
 last change scn: 0x000a.0004d7b0 seq: 0x1 flg: 0x06
 spare1: 0x0 spare2: 0x0 spare3: 0x0
 consistency value in tail: 0xd7b00611
 check value in block header: 0xe563
 computed block checksum: 0x0
DBVERIFY - Verification complete
Total Blocks Examined         : 1
Total Blocks Processed (Data) : 0
Total Blocks Failing   (Data) : 0
Total Blocks Processed (Index): 0
Total Blocks Failing   (Index): 0
Total Blocks Empty            : 0
Total Blocks Marked Corrupt   : 1
Total Blocks Influx           : 2

修改数据库记录

SQL> startup
ORACLE instance started.
Total System Global Area  318767104 bytes
Fixed Size                  1267236 bytes
Variable Size             109054428 bytes
Database Buffers          201326592 bytes
Redo Buffers                7118848 bytes
Database mounted.
Database opened.
SQL> conn chf/xifenfei
Connected.
SQL> update t_xifenfei set object_name='惜分飞';
10 rows updated.
SQL> update t_xifenfei_new set object_name='惜分飞';
10 rows updated.
SQL> commit;
Commit complete.
SQL> alter system switch logfile;
System altered.
SQL> alter system switch logfile;
System altered.
SQL> alter system switch logfile;
System altered.
SQL> conn / as sysdba
Connected.
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.

利用备份数据文件恢复数据库

[oracle@xifenfei XFF]$ cp xifenfei01.dbf xifenfei01.dbf_bak
[oracle@xifenfei XFF]$ cp ../tmp/xifenfei01.dbf  xifenfei01.dbf
[oracle@xifenfei XFF]$ sqlplus / as sysdba
SQL*Plus: Release 10.2.0.4.0 - Production on Fri Sep 28 19:13:59 2012
Copyright (c) 1982, 2007, Oracle.  All Rights Reserved.
Connected to an idle instance.
SQL> startup
ORACLE instance started.
Total System Global Area  318767104 bytes
Fixed Size                  1267236 bytes
Variable Size             109054428 bytes
Database Buffers          201326592 bytes
Redo Buffers                7118848 bytes
Database mounted.
ORA-01113: file 5 needs media recovery
ORA-01110: data file 5: '/u01/oracle/oradata/XFF/xifenfei01.dbf'
--提示数据需要恢复
SQL> recover datafile 5;
ORA-00279: change 42949990720 generated at 09/28/2012 19:04:10 needed for
thread 1
ORA-00289: suggestion : /u01/oracle/oradata/XFF/archivelog/1_24_792679299.dbf
ORA-00280: change 42949990720 for thread 1 is in sequence #24
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
auto
Log applied.
Media recovery complete.
SQL> alter database open;
Database altered.
--利用被破坏的数据文件+归档日志恢复数据库正常
SQL> col object_name for a30
SQL> select   object_id,object_name from t_xifenfei;
select   object_id,object_name from t_xifenfei
                                    *
ERROR at line 1:
ORA-01578: ORACLE data block corrupted (file # 5, block # 12)
ORA-01110: data file 5: '/u01/oracle/oradata/XFF/xifenfei01.dbf'
--提示被破坏的数据块,查询不能完成
--证明坏块之外的数据块还是被正常应用日志
SQL>  select   object_id,object_name from t_xifenfei_new;
 OBJECT_ID OBJECT_NAME
---------- ------------------------------
        20 惜分飞
        44 惜分飞
        28 惜分飞
        15 惜分飞
        29 惜分飞
         3 惜分飞
        25 惜分飞
        39 惜分飞
        51 惜分飞
        26 惜分飞
10 rows selected.
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.

dbv检查坏块

[oracle@xifenfei XFF]$ dbv file=xifenfei01.dbf
DBVERIFY: Release 10.2.0.4.0 - Production on Fri Sep 28 19:14:52 2012
Copyright (c) 1982, 2007, Oracle.  All rights reserved.
DBVERIFY - Verification starting : FILE = xifenfei01.dbf
DBV-00200: Block, DBA 20971532, already marked corrupt
--这里可以看出来，该数据块已经被标志为坏块
DBVERIFY - Verification complete
Total Pages Examined         : 1280
Total Pages Processed (Data) : 2
Total Pages Failing   (Data) : 0
Total Pages Processed (Index): 0
Total Pages Failing   (Index): 0
Total Pages Processed (Other): 14
Total Pages Processed (Seg)  : 0
Total Pages Failing   (Seg)  : 0
Total Pages Empty            : 1264
Total Pages Marked Corrupt   : 1
Total Pages Influx           : 0
Highest block SCN            : 318700 (10.318700)

查看恢复过程alert日志

Fri Sep 28 19:14:06 2012
Database mounted in Exclusive Mode
Completed: ALTER DATABASE   MOUNT
Fri Sep 28 19:14:06 2012
ALTER DATABASE OPEN
ORA-1113 signalled during: ALTER DATABASE OPEN...
Fri Sep 28 19:14:11 2012
ALTER DATABASE RECOVER  datafile 5
Media Recovery Start
 parallel recovery started with 2 processes
ORA-279 signalled during: ALTER DATABASE RECOVER  datafile 5  ...
Fri Sep 28 19:14:16 2012
ALTER DATABASE RECOVER    CONTINUE DEFAULT
Fri Sep 28 19:14:16 2012
--恢复数据库的时候,发现坏块
Media Recovery Log /u01/oracle/oradata/XFF/archivelog/1_24_792679299.dbf
Fri Sep 28 19:14:16 2012
Hex dump of (file 5, block 12) in trace file /u01/oracle/admin/XFF/bdump/xff_p001_23011.trc
Corrupt block relative dba: 0x0140000c (file 5, block 12)
Fractured block found during media recovery
Data in bad block:
 type: 6 format: 2 rdba: 0x0140000c
 last change scn: 0x000a.0004d7b0 seq: 0x1 flg: 0x06
 spare1: 0x0 spare2: 0x0 spare3: 0x0
 consistency value in tail: 0xd7b00611
 check value in block header: 0xe563
 computed block checksum: 0x0
Reread of rdba: 0x0140000c (file 5, block 12) found same corrupted data
--继续恢复
Fri Sep 28 19:14:16 2012
Recovery of Online Redo Log: Thread 1 Group 1 Seq 25 Reading mem 0
  Mem# 0: /u01/oracle/oradata/XFF/redo01.log
Fri Sep 28 19:14:16 2012
Recovery of Online Redo Log: Thread 1 Group 2 Seq 26 Reading mem 0
  Mem# 0: /u01/oracle/oradata/XFF/redo02.log
Fri Sep 28 19:14:16 2012
Recovery of Online Redo Log: Thread 1 Group 3 Seq 27 Reading mem 0
  Mem# 0: /u01/oracle/oradata/XFF/redo03.log
Fri Sep 28 19:14:16 2012
Media Recovery Complete (XFF)
Completed: ALTER DATABASE RECOVER    CONTINUE DEFAULT
Fri Sep 28 19:14:31 2012
alter database open

bbed查看修改相关信息

BBED> set filename '/u01/oracle/oradata/XFF/xifenfei01.dbf'
        FILENAME        /u01/oracle/oradata/XFF/xifenfei01.dbf
BBED> set block 12
        BLOCK#          12
BBED> map
 File: /u01/oracle/oradata/XFF/xifenfei01.dbf (0)
 Block: 12                                    Dba:0x00000000
------------------------------------------------------------
 KTB Data Block (Table/Cluster)
 struct kcbh, 20 bytes                      @0
 struct ktbbh, 96 bytes                     @20
 struct kdbh, 14 bytes                      @124
 struct kdbt[1], 4 bytes                    @138
 sb2 kdbr[10]                               @142
 ub1 freespace[7666]                        @162
 ub1 rowdata[360]                           @7828
 ub4 tailchk                                @8188
BBED> p kcbh
struct kcbh, 20 bytes                       @0
   ub1 type_kcbh                            @0        0x06
   ub1 frmt_kcbh                            @1        0xa2
   ub1 spare1_kcbh                          @2        0x00
   ub1 spare2_kcbh                          @3        0x00
   ub4 rdba_kcbh                            @4        0x0140000c
   ub4 bas_kcbh                             @8        0x00000000
   ub2 wrp_kcbh                             @12       0x0000
   ub1 seq_kcbh                             @14       0xff   <--因为被标记为坏块,所以为ff
   ub1 flg_kcbh                             @15       0x04 (KCBHFCKV)
   ub2 chkval_kcbh                          @16       0xe77d
   ub2 spare3_kcbh                          @18       0x0000
--查看数据块中记录
BBED> p *kdbr[5]
rowdata[69]
-----------
ub1 rowdata[69]                             @7897     0x2c
BBED> x /rnc
rowdata[69]                                 @7897
-----------
flag@7897: 0x2c (KDRHFL, KDRHFF, KDRHFH)
lock@7898: 0x02
cols@7899:    2
col    0[2] @7900: 3
col   1[16] @7903: WWW.XIFENFEI.COM  <--确实没有被恢复,而是直接被跳过
BBED> set mode edit
        MODE            Edit
BBED> m /x 01 offset 14
Warning: contents of previous BIFILE will be lost. Proceed? (Y/N) y
 File: /u01/oracle/oradata/XFF/xifenfei01.dbf (0)
 Block: 12               Offsets:   14 to  525           Dba:0x00000000
------------------------------------------------------------------------
 01047de7 00000100 00000dcc 000098d7 …………
 <32 bytes per line>
BBED> d offset 8188
 File: /u01/oracle/oradata/XFF/xifenfei01.dbf (0)
 Block: 12               Offsets: 8188 to 8191           Dba:0x00000000
------------------------------------------------------------------------
 ff060000    <--这么说明:数据块被标志为坏块的时候，同时会修改tailchk值
 <32 bytes per line>
BBED> m /x 01 offset 8188
 File: /u01/oracle/oradata/XFF/xifenfei01.dbf (0)
 Block: 12               Offsets: 8188 to 8191           Dba:0x00000000
------------------------------------------------------------------------
 01060000
 <32 bytes per line>
BBED> sum apply
Check value for File 0, Block 12:
current = 0xe77d, required = 0xe77d
--验证块已经标记为正常块
BBED> verify
DBVERIFY - Verification starting
FILE = /u01/oracle/oradata/XFF/xifenfei01.dbf
BLOCK = 12
DBVERIFY - Verification complete
Total Blocks Examined         : 1
Total Blocks Processed (Data) : 1
Total Blocks Failing   (Data) : 0
Total Blocks Processed (Index): 0
Total Blocks Failing   (Index): 0
Total Blocks Empty            : 0
Total Blocks Marked Corrupt   : 0
Total Blocks Influx           : 0

启动数据库测试

SQL> startup
ORACLE instance started.
Total System Global Area  318767104 bytes
Fixed Size                  1267236 bytes
Variable Size             109054428 bytes
Database Buffers          201326592 bytes
Redo Buffers                7118848 bytes
Database mounted.
Database opened.
SQL> conn chf/xifenfei
Connected.
SQL> col object_name for a30
SQL> select   object_id,object_name from t_xifenfei;
 OBJECT_ID OBJECT_NAME
---------- ------------------------------
        20 WWW.XIFENFEI.COM
        44 WWW.XIFENFEI.COM
        28 WWW.XIFENFEI.COM
        15 WWW.XIFENFEI.COM
        29 WWW.XIFENFEI.COM
         3 WWW.XIFENFEI.COM
        25 WWW.XIFENFEI.COM
        39 WWW.XIFENFEI.COM
        51 WWW.XIFENFEI.COM
        26 WWW.XIFENFEI.COM
10 rows selected.
--通过修改数据块的seq_kcbh和tailchk,让这个块恢复正常,但是记录依然丢失，
--因为应用日志恢复之时标记为坏块跳过该块的日志应用

通过实验证明:
1.如果只有数据块异常,应用日志恢复,不一定会出现ORA-600[3020],而是直接把该块标记为坏块,继续应用日志
2.标记坏块其实就是修改seq_kcbh为ff,同时也修改tailchk值
3.经验值:如果在数据库应用日志恢复的时候,如果出现ORA-600[3020]错误,可以使用allow 2 corruption来跳过坏块处理,其实也是修改seq_kcbh为ff,然后让数据库跳过该块的恢复.

“在oracle出现ORA-8103/ORA-1578/ORA-376″等情况下抢救数据的争论没有停止过,很多人想到的是使用bbed,dul等工具来抢救,其实在很多时候我们使用pl/sql也可以完美的抢救数据.在这里我们通过模拟ORA-8103错误，然后使用plsql来找回数据.这中处理方法相对于bbed风险小,但是缺点是如果数据量大处理时间可能比较长,可能比dul有的一比,但是dul的工具不是任何人都有的.所以整体来说,在大部分情况下,这种方法处理某个数据块错误,抢救某个对象数据,还是很好的方法.
1.有非空列index情况

--创建测试表
SQL> create table xifenfei
  2  as
  3  select * from dba_objects;
Table created.
--修改某个项为非空值
SQL> alter table xifenfei modify object_id not null;
Table altered.
--创建一个唯一index
SQL> create unique index ind_xifenfei  on xifenfei(object_id);
Index created.
--表总记录
SQL> select count(*) from xifenfei;
  COUNT(*)
----------
     50088
--extent的分布情况
SQL> set pages 100
SQL>  select file_id,block_id,block_id+blocks-1
  2    from dba_extents
  3   where segment_name ='XIFENFEI' AND owner='CHF';
   FILE_ID   BLOCK_ID BLOCK_ID+BLOCKS-1
---------- ---------- -----------------
         9       1545              1552
         9       1553              1560
         9       1561              1568
         9       1569              1576
         9       1577              1584
         9       1585              1592
         9       1593              1600
         9       1601              1608
         9       1609              1616
         9       1617              1624
         9       1625              1632
         9       1633              1640
         9       1641              1648
         9       1649              1656
         9       1657              1664
         9       1665              1672
         9       1673              1800
         9       1801              1928
         9       1929              2056
         9       2057              2184
         9       2185              2312
21 rows selected.
--2200数据块包含记录
SQL> select   count(*)
  2  from chf.xifenfei where dbms_rowid.rowid_block_number(rowid)=2200;
  COUNT(*)
----------
        69
--关闭数据库
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
--破坏数据块
[oracle@xifenfei ~]$ dd if=/dev/zero of=/u01/oracle/oradata/XFF/users03.dbf  bs=8192  count=1 seek=2200 conv=notrunc
1+0 records in
1+0 records out
8192 bytes (8.2 kB) copied, 0.000151554 seconds, 54.1 MB/s
--启动数据库
SQL> startup
ORACLE instance started.
Total System Global Area  318767104 bytes
Fixed Size                  1267236 bytes
Variable Size             104860124 bytes
Database Buffers          205520896 bytes
Redo Buffers                7118848 bytes
Database mounted.
Database opened.
--查询结果
SQL>  select /*+ full(xifenfei) */ count(*) from chf.xifenfei;
 select /*+ full(xifenfei) */ count(*) from chf.xifenfei
                                                *
ERROR at line 1:
ORA-08103: object no longer exists
SQL> create table chf.xifenfei_new
  2  as
  3  select * from chf.xifenfei;
select * from chf.xifenfei
                  *
ERROR at line 3:
ORA-08103: object no longer exists
--创建备份表
SQL> create table chf.xifenfei_new
  2  as
  3  select * from chf.xifenfei where 1=0;
Table created.
--创建坏块相关rowid记录表
SQL> create table chf.bad_rows (row_id rowid, oracle_error_code number);
Table created.
--执行plsql脚本
DECLARE
 TYPE RowIDTab IS TABLE OF ROWID INDEX BY BINARY_INTEGER;
 CURSOR c1 IS  select /*+ index(xifenfei ind_xifenfei) */ rowid
 from chf.xifenfei
 where object_id is NOT NULL;
 r RowIDTab;
 rows  NATURAL := 20000;
 bad_rows number := 0 ;
 errors number;
 error_code number;
 myrowid rowid;
BEGIN
 OPEN c1;
 LOOP
   FETCH  c1 BULK COLLECT INTO r LIMIT rows;
   EXIT WHEN r.count=0;
   BEGIN
    FORALL i IN r.FIRST..r.LAST SAVE EXCEPTIONS
     insert into chf.xifenfei_new
     select /*+ ROWID(A) */ *
     from chf.xifenfei A where rowid = r(i);
   EXCEPTION
   when OTHERS then
    BEGIN
     errors := SQL%BULK_EXCEPTIONS.COUNT;
     FOR err1 IN 1..errors LOOP
       error_code := SQL%BULK_EXCEPTIONS(err1).ERROR_CODE;
       if error_code in (1410, 8103) then
         myrowid := r(SQL%BULK_EXCEPTIONS(err1).ERROR_INDEX);
         bad_rows := bad_rows + 1;
         insert into chf.bad_rows values(myrowid, error_code);
       else
         raise;
       end if;
     END LOOP;
     END;
   END;
  commit;
 END LOOP;
 commit;
 CLOSE c1;
 dbms_output.put_line('Total Bad Rows: '||bad_rows);
END;
/
--查询错误记录
SQL> select count(*) from chf.bad_rows ;
  COUNT(*)
----------
        69
SQL> select * from chf.bad_rows where rownum<10;
ROW_ID             ORACLE_ERROR_CODE
------------------ -----------------
AAAMugAAJAAAAiYAAA              8103
AAAMugAAJAAAAiYAAB              8103
AAAMugAAJAAAAiYAAC              8103
AAAMugAAJAAAAiYAAD              8103
AAAMugAAJAAAAiYAAE              8103
AAAMugAAJAAAAiYAAF              8103
AAAMugAAJAAAAiYAAG              8103
AAAMugAAJAAAAiYAAH              8103
AAAMugAAJAAAAiYAAI              8103
9 rows selected.
--查询备份表记录
SQL> select count(*) from chf.xifenfei_new;
  COUNT(*)
----------
     50019
50088-50019=69和被破坏块中记录一致,证明所有好块中记录全部被找回来

2.无非空列index情况

--创建表
SQL> CONN CHF/XIFENFEI
Connected.
SQL> create table t_xifenfei
  2  as
  3  select * from dba_objects;
Table created.
--表中记录总数
SQL> select count(*) from t_xifenfei;
  COUNT(*)
----------
     50086
--extent分布
SQL> SET PAGES 100
SQL>  select file_id,block_id,block_id+blocks-1
  2    from dba_extents
  3   where segment_name ='T_XIFENFEI' AND owner='CHF';
   FILE_ID   BLOCK_ID BLOCK_ID+BLOCKS-1
---------- ---------- -----------------
         9          9                16
         9         17                24
         9         25                32
         9         33                40
         9         41                48
         9         49                56
         9         57                64
         9         65                72
         9         73                80
         9         81                88
         9         89                96
         9         97               104
         9        105               112
         9        113               120
         9        121               128
         9        129               136
         9        137               264
         9        265               392
         9        393               520
         9        521               648
         9        649               776
21 rows selected.
--700数据块中记录数
SQL> select   count(*)
  2  from chf.t_xifenfei where dbms_rowid.rowid_block_number(rowid)=700;
  COUNT(*)
----------
        73
--关闭数据库
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
--破坏block 700的数据块
[oracle@xifenfei ~]$ dd if=/dev/zero of=/u01/oracle/oradata/XFF/users03.dbf  bs=8192  count=1 seek=700 conv=notrunc
1+0 records in
1+0 records out
8192 bytes (8.2 kB) copied, 0.000156576 seconds, 52.3 MB/s
--启动数据库
SQL> startup
ORACLE instance started.
Total System Global Area  318767104 bytes
Fixed Size                  1267236 bytes
Variable Size             104860124 bytes
Database Buffers          205520896 bytes
Redo Buffers                7118848 bytes
Database mounted.
Database opened.
--查询报错
SQL> select count(*) from chf.t_xifenfei;
select count(*) from chf.t_xifenfei
                         *
ERROR at line 1:
ORA-08103: object no longer exists
--创建备份表
SQL> CREATE TABLE T_XIFENFEI_NEW
  2  AS
  3  SELECT * FROM T_XIFENFEI WHERE 1=0;
--找回记录
set serveroutput on
set concat off
DECLARE
 nrows number;
 rid rowid;
 dobj number;
 ROWSPERBLOCK number;
BEGIN
 ROWSPERBLOCK:=1000;  --估算最大的一个块中记录条数
 nrows:=0;
 select data_object_id  into dobj
 from dba_objects
 where owner = 'CHF'
 and object_name = 'T_XIFENFEI'
-- and subobject_name = '<table partition>'  Add this condition if table is partitioned
 ;
 for i in (select relative_fno, block_id, block_id+blocks-1 totblocks
           from dba_extents
           where owner = 'CHF'
             and segment_name = 'T_XIFENFEI'
-- and partition_name = '<table partition>' Add this condition if table is partitioned
-- and file_id != <OFFLINED DATAFILE> This condition is only used if a datafile needs to be skipped due to ORA-376 (A)
          order by extent_id)
 loop
   for br in i.block_id..i.totblocks loop
    for j in 1..ROWSPERBLOCK loop
    begin
      rid := dbms_rowid.ROWID_CREATE(1,dobj,i.relative_fno, br , j-1);
      insert into CHF.T_XIFENFEI_NEW
      select /*+ ROWID(A) */ *
      from CHF.T_XIFENFEI A
      where rowid = rid;
      if sql%rowcount = 1 then nrows:=nrows+1; end if;
      if (mod(nrows,10000)=0) then commit; end if;
    exception when others then null;
    end;
    end loop;
  end loop;
 end loop;
 COMMIT;
 dbms_output.put_line('Total rows: '||to_char(nrows));
END;
/
--找回记录数
SQL> SELECT COUNT(*) FROM CHF.T_XIFENFEI_NEW;
  COUNT(*)
----------
     50013
50086-50013=73  证明非坏块中的数据都被完全寻找回来

参考：
ORA-8103 Troubleshooting, Diagnostic and Solution [ID 268302.1]
Extract rows from a CORRUPT table creating ROWID from DBA_EXTENTS [ID 422547.1]

当我们对数据库进行异常恢复,很多时候要选择屏蔽回滚段,但是我们有没有办法来评估屏蔽回滚段到底会对我们的数据库的数据产生多大影响呢？其实我们可以通过TXChecker工具来评估数据库undo异常时候受到影响的数据对象有哪些,从而进一步确定是否真的需要对其undo下手处理.
模拟数据异常事务为提交情况

[oracle@xifenfei ~]$ sqlplus / as sysdba
SQL*Plus: Release 10.2.0.4.0 - Production on Thu Aug 30 20:32:27 2012
Copyright (c) 1982, 2007, Oracle.  All Rights Reserved.
Connected to an idle instance.
SQL> startup;
ORACLE instance started.
Total System Global Area  318767104 bytes
Fixed Size                  1267236 bytes
Variable Size             100665820 bytes
Database Buffers          209715200 bytes
Redo Buffers                7118848 bytes
Database mounted.
Database opened.
SQL> create table t_xifenfei tablespace system
  2  as
  3  select * from dba_objects;
Table created.
SQL> create table chf.t_xifenfei tablespace users
  2  as
  3  select * from dba_objects;
Table created.
SQL> delete from t_xifenfei where rownum<10;
9 rows deleted.
SQL> delete from chf.t_xifenfei where rownum<100;
99 rows deleted.
SQL> alter system checkpoint;
System altered.
SQL> shutdow abort;
ORACLE instance shut down.

在system和users表空间分别模拟了sys和chf的t_xifenfei表含事务未提交情况

TXChecker用法说明

[oracle@xifenfei txchecker]$ ./TXChecker
TXChecker - v1.4 by Center Of Expertise (COE), Oracle Corporation (build 10/16/07)
Usage is: TXChecker [options]
Options:
  -a                     When scanning datafiles (with -d/-f/-l/-t options) report objects using any of the undo
                         segments (not just those with errors) (OPTIONAL)
  -b                     For objects found, print the datablock addresses. See readme for further details (OPTIONAL)
  -c<controlfile_name>   Fully qualified controlfile name to read (MANDATORY)
  -d                     Scan database for active TXs (use when undo not available) (OPTIONAL)
  -f<filename>           Scan the named datafile for active TXs (OPTIONAL)
  -g                     Indicates you want to find all blocks taking part in transactions with
                         a USN > than the USN supplied in -x<XID> parameter (same constraints as -w) (OPTIONAL)
  -l<listfile>           Scan all the datafiles listed in the listfile for active TXs (OPTIONAL)
  -m<minutes>            Number of minutes used to consider a TX as active (1-120) (DEFAULTS TO 60 MINUTES)
  -p                     Show the names and last known status of the UNDO segments (OPTIONAL)
  -s                     Skip read-only or offline normal datafiles (OPTIONAL)
  -t<tablespace>         Scan all the datafiles for this tablespace (OPTIONAL)
  -u                     Report ITL entries active if marked with an upper bound ('U' flag) fast commit SCN
                         instead if active transactions (OPTIONAL)
  -w<wrap#>              Wrap# for XID in ITL entry to report blocks where wrap# > this one (OPTIONAL)
                         Must use -x with this option. See the readme for details
  -x<XID>                XID for transaction wanting to search for (OPTIONAL)
                         Use format rrrr.ssss.wwwwwwww using Hexadecimal numbers
                         See the readme for full instructions on using -x, -w and -g options
NOTE: Options -d/-f/-l/-t are exclusive, and only one should be specified.
[/sql]
<strong>TXChecker初始</strong>

[oracle@xifenfei txchecker]$ ./TXChecker -c/u01/oracle/oradata/XFF/control01.ctl -d -a
TXChecker - v1.4 by Center Of Expertise (COE), Oracle Corporation (build 10/16/07)
Database Name: XFF   Version: 10.2.0
*** Database last checkpointed at 08/30/2012 20:34:43 (SCN: 0xa.0x1e142)
*** Using 60 minutes to find most active transactions (-m60)
*** Undo Segments:
USN:   0  Name: SYSTEM          TBS#:  0 File:   1  Block:      9  Instance:  0  SMU: N  Status:  3 - Online
                                SCN: 0000.00000000   XactSQN: 0x00000000   UndoSQN: 0x00000000
USN:   1  Name: _SYSSMU1$       TBS#:  1 File:   2  Block:      9  Instance:  0  SMU: Y  Status:  3 - Online
                                SCN: 000a.000191c3   XactSQN: 0x000000d2   UndoSQN: 0x0000013c
…………省略
USN:   9  Name: _SYSSMU9$       TBS#:  1 File:   2  Block:    137  Instance:  0  SMU: Y  Status:  3 - Online
                                SCN: 000a.000191d8   XactSQN: 0x0000011a   UndoSQN: 0x000000dc
USN:  10  Name: _SYSSMU10$      TBS#:  1 File:   2  Block:    153  Instance:  0  SMU: Y  Status:  3 - Online
                                SCN: 000a.000191d7   XactSQN: 0x000000c1   UndoSQN: 0x000000d9
…………省略
USN:  20  Name: _SYSSMU20$      TBS#:  5 File:   5  Block:    153  Instance:  0  SMU: Y  Status:  1 - Invalid / Dropped
                                SCN: 0000.00070ec6   XactSQN: 0x00000002   UndoSQN: 0x00000001
*** Active Transactions:
USN: 6  Name: _SYSSMU6$  File: 2  Block: 89  Instance: 0  Status: 3 - Online
* Active TX at slot 6  #undo blocks: 3  Last bk: 2.90
Obj#:  51938  Name: CHF.T_XIFENFEI                                    Type: TABLE               Undo recs: 99
              Used undo segment IDs: 6
Obj#:  51937  Name: SYS.T_XIFENFEI                                    Type: TABLE               Undo recs: 9
              Used undo segment IDs: 6
         File (validate_objects.sql) being created for object validattion already exists
         and will be overwritten!!!
         Do you want to continue overwriting [Y/N]?y
*** Undo segments (headers) that encountered errors preventing Active TX scan:
USN:  11  Name: _SYSSMU11$      File:   5  Block:       9  Instance:  0  Error: 27 - Undo segment was dropped
…………省略
USN:  20  Name: _SYSSMU20$      File:   5  Block:     153  Instance:  0  Error: 27 - Undo segment was dropped
WARNING: Analyzing the full database for active transactions will take some time!
         Are you sure you want to analyze the full database [Y/N]?
         Are you sure you want to analyze the full database [Y/N]?y
*** Scanning database for datablocks that may require undo (PLEASE WAIT...):
*** Asterisk ('*') denotes blocks being updated since 08/08/2012 03:30:32 (SCN: 0x0.0x79607)
Scanning datafile:  3 - /u01/oracle/oradata/XFF/sysaux01.dbf (SYSAUX)                           - Active TX blocks: 147 *
Scanning datafile:  1 - /u01/oracle/oradata/XFF/system01.dbf (SYSTEM)                           - Active TX blocks: 11597 *
--undo表空间跳过
Undo datafile (/u01/oracle/oradata/XFF/undotbs01.dbf) - SKIPPING
Scanning datafile:  4 - /u01/oracle/oradata/XFF/users01.dbf (USERS)                             - Active TX blocks: 2 *
--因为数据文件丢失控制文件中offline的跳过(其实只要数据文件丢失就会跳过)
Cannot access datafile (/u01/oracle/oradata/XFF/xifenfei01.dbf) (error 2 - No such file or directory) - SKIPPING
Cannot access datafile (/u01/oracle/oradata/XFF/xifenfei02.dbf) (error 2 - No such file or directory) - SKIPPING
Scanning datafile:  7 - /u01/oracle/oradata/XFF/xifenfei03.dbf (XIFENFEI2)                      - Active TX blocks: 0
*** Objects that may require undo data:
*** Asterisk ('*') denotes blocks being updated since 08/08/2012 03:30:32 (SCN: 0x0.0x79607)
DataObj#:  51938  Name: CHF.T_XIFENFEI                                    Type: TABLE               Datablocks: 2 *
Used undo segment IDs: 6
DataObj#:  46434  Name: MDSYS.SYS_IL0000046432C00006$$                    Type: INDEX               Datablocks: 4
Used undo segment IDs: 2
DataObj#:    124  Name: SYS.I_ACCESS1                                     Type: INDEX               Datablocks: 27
Used undo segment IDs: 1 2 3 4 5 6 7 8 9 10
…………省略
DataObj#:   8824  Name: SYS.SYS_IL0000008822C00008$$                      Type: INDEX               Datablocks: 1 *
Used undo segment IDs: 4
DataObj#:  51937  Name: SYS.T_XIFENFEI                                    Type: TABLE               Datablocks: 1 *
Used undo segment IDs: 6
DataObj#:  51557  Name: SYS.UTL_RECOMP_COMPILED                           Type: TABLE               Datablocks: 1
Used undo segment IDs: 8
…………省略
DataObj#:  42131  Name: XDB.XDB$ELEMENT_PROPNUMBER                        Type: INDEX               Datablocks: 1
Used undo segment IDs: 2
*** Use validate_objects.sql script file to validate the structure of possibly corrupt objects
    if the undo required is not available.
Undo Segment Usage Summary
**************************
*** Undo segments identified in use by active transaction datablocks AFTER 08/08/2012 03:30:32 (SCN: 0x0.0x79607):
USN:   1  Name: _SYSSMU1$
USN:   2  Name: _SYSSMU2$
USN:   3  Name: _SYSSMU3$
USN:   4  Name: _SYSSMU4$
USN:   5  Name: _SYSSMU5$
USN:   6  Name: _SYSSMU6$
USN:   9  Name: _SYSSMU9$
*** Undo segments identified in use by active transacation datablocks BEFORE 08/08/2012 03:30:32 (SCN: 0x0.0x79607):
USN:   1  Name: _SYSSMU1$
USN:   2  Name: _SYSSMU2$
USN:   3  Name: _SYSSMU3$
USN:   4  Name: _SYSSMU4$
USN:   5  Name: _SYSSMU5$
USN:   7  Name: _SYSSMU7$
USN:   8  Name: _SYSSMU8$
USN:   9  Name: _SYSSMU9$
USN:  10  Name: _SYSSMU10$
NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!NOTE!
*************************************************************************************
Upload the logfile (TXChecker_083012_2042_XFF.log) to Oracle Support Services for analysis.
Do NOT attempt to force the database open until the logfile has been analyzed.

验证对象脚本

[oracle@xifenfei txchecker]$ more validate_objects.sql
rem validate_objects.sql - checks strcuture of objects needing unavailable undo data
rem                        Created by findtxns program, Oracle Corporation
set echo on
ANALYZE TABLE CHF.T_XIFENFEI VALIDATE STRUCTURE CASCADE;
ANALYZE INDEX MDSYS.SYS_IL0000046432C00006$$ VALIDATE STRUCTURE;
ANALYZE INDEX SYS.I_ACCESS1 VALIDATE STRUCTURE;
…………省略
ANALYZE INDEX XDB.SYS_C003167 VALIDATE STRUCTURE;
ANALYZE INDEX XDB.XDB$ELEMENT_PROPNAME VALIDATE STRUCTURE;
ANALYZE INDEX XDB.XDB$ELEMENT_PROPNUMBER VALIDATE STRUCTURE;