Oracle非归档模式Media Recovery错误之--ORA-26040

2021-11-22 04:35:34

系统环境：

操作系统：Linux RH55

Oracle：  Oracle 11gR2

模拟案例：

1、查看数据库模式

18:12:36 SYS@ prod>archive log list;

Database log mode              No Archive Mode

Automatic archival             Disabled

Archive destination            /dsk4/arch1

Oldest online log sequence     1

Current log sequence           3

2、创建新的Tablespace

1
2
3
18:13:19 SYS@ prod>create tablespace tbs2 
18:13:30   2  datafile '/u01/app/oracle/oradata/prod/tbs2.dbf' size 10m;
Tablespace created.

3、对数据库进行冷备

1
2
3
4
5
6
7
8
9
10
[oracle@rh6 ~]$ rman target /
Recovery Manager: Release 11.2.0.1.0 - Production on Thu Jul 24 18:14:30 2014
Copyright (c) 1982, 2009, Oracle and/or its affiliates.  All rights reserved.
connected to target database: PROD (DBID=239333010)
RMAN> run {
2> shutdown immediate;
3> startup mount;
4> backup database format '/dsk3/bak/%s_%d.bak';
5> alter database open;
6> }

4、备份完成在新的Tablesapce上创建Object

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
18:19:40 SYS@ prod>conn scott/tiger
Connected.
18:30:13 SCOTT@ prod>create table t1 tablespace tbs2 as select * from emp;
Table created.
Elapsed: 00:00:01.00
18:30:45 SCOTT@ prod>select count(*) from t1;
  COUNT(*)
----------
        14
Elapsed: 00:00:00.02
18:30:56 SCOTT@ prod>insert into t1 select * from t1 where rownum=1;
1 row created.
Elapsed: 00:00:00.02
18:31:16 SCOTT@ prod>commit;
Commit complete.
Elapsed: 00:00:00.04
18:31:18 SCOTT@ prod>insert into t1 select * from t1 where rownum=1;
1 row created.
Elapsed: 00:00:00.01
18:31:23 SCOTT@ prod>select count(*) from t1;
  COUNT(*)
----------
        16

5、查看Redo Log信息

1
2
3
4
5
6
7
18:32:14 SYS@ prod>select group#,sequence# ,status from v$log;
    GROUP#  SEQUENCE# STATUS
---------- ---------- ----------------
         1          4 CURRENT
         2          2 INACTIVE
         3          3 INACTIVE
Elapsed: 00:00:00.05

6、模拟数据文件被破坏

[oracle@rh6 ~]$ rm /u01/app/oracle/oradata/prod/tbs2.dbf

18:32:28 SYS@ prod>shutdown abort

ORACLE instance shut down.

重新启动Instance到mount，查看redo log，日志没有发生switch

1
2
3
4
5
6
7
8
9
10
11
12
13
14
18:33:06 SYS@ prod>startup mount;
ORACLE instance started.
Total System Global Area  835104768 bytes
Fixed Size                  2217952 bytes
Variable Size             775948320 bytes
Database Buffers           54525952 bytes
Redo Buffers                2412544 bytes
Database mounted.
18:33:42 SYS@ prod>select group#,sequence# ,status from v$log;
    GROUP#  SEQUENCE# STATUS
---------- ---------- ----------------
         1          4 CURRENT
         3          3 INACTIVE
         2          2 INACTIVE

Open database出现数据文件丢失错误：

18:34:17 SYS@ prod>alter database open;

alter database open

*

ERROR at line 1:

ORA-01157: cannot identify/lock data file 9 - see DBWR trace file

ORA-01110: data file 9: '/u01/app/oracle/oradata/prod/tbs2.dbf'

7、对Database做Media Recovery（因为没有日志切换，做complete recover）

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
[oracle@rh6 ~]$ rman target /
Recovery Manager: Release 11.2.0.1.0 - Production on Thu Jul 24 18:34:35 2014
Copyright (c) 1982, 2009, Oracle and/or its affiliates.  All rights reserved.
connected to target database: PROD (DBID=239333010, not open)
RMAN> restore datafile 9;
Starting restore at 24-JUL-14
using target database control file instead of recovery catalog
allocated channel: ORA_DISK_1
channel ORA_DISK_1: SID=18 device type=DISK
channel ORA_DISK_1: starting datafile backup set restore
channel ORA_DISK_1: specifying datafile(s) to restore from backup set
channel ORA_DISK_1: restoring datafile 00009 to /u01/app/oracle/oradata/prod/tbs2.dbf
channel ORA_DISK_1: reading from backup piece /dsk3/bak/34_PROD.bak
channel ORA_DISK_1: piece handle=/dsk3/bak/34_PROD.bak tag=TAG20140724T181640
channel ORA_DISK_1: restored backup piece 1
channel ORA_DISK_1: restore complete, elapsed time: 00:00:01
Finished restore at 24-JUL-14
RMAN> recover datafile 9;
Starting recover at 24-JUL-14
using channel ORA_DISK_1
starting media recovery
media recovery complete, elapsed time: 00:00:01
Finished recover at 24-JUL-14

查看告警日志：

Thu Jul 24 18:34:49 2014

Full restore complete of datafile 9 /u01/app/oracle/oradata/prod/tbs2.dbf.  Elapsed time: 0:00:00 

  checkpoint is 2168258

Thu Jul 24 18:34:57 2014

alter database recover datafile list clear

Completed: alter database recover datafile list clear

alter database recover if needed

 datafile 9

Media Recovery Start

Serial Media Recovery started

Recovery of Online Redo Log: Thread 1 Group 3 Seq 3 Reading mem 0

  Mem# 0: /dsk1/oradata/prod/redo03a.log

  Mem# 1: /dsk2/oradata/prod/redo03b.log

Recovery of Online Redo Log: Thread 1 Group 1 Seq 4 Reading mem 0

  Mem# 0: /dsk1/oradata/prod/redo01a.log

  Mem# 1: /dsk2/oradata/prod/redo01b.log

Media Recovery Complete (prod)

Completed: alter database recover if needed

 datafile 9

通过group1和3，进行了media recovery ！


8、数据库open，发现T1无法访问（出现：ORA-26040）

18:35:36 SYS@ prod>alter database open;

Database altered.

18:35:49 SYS@ prod>select count(*) from scott.t1;

select count(*) from scott.t1

*

ERROR at line 1:

ORA-01578: ORACLE data block corrupted (file # 9, block # 131)

ORA-01110: data file 9: '/u01/app/oracle/oradata/prod/tbs2.dbf'

ORA-26040: Data block was loaded using the NOLOGGING option


查看告警日志：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
alter database open
Beginning crash recovery of 1 threads
Started redo scan
Completed redo scan
 read 83 KB redo, 26 data blocks need recovery
Started redo application at
 Thread 1: logseq 4, block 117
Recovery of Online Redo Log: Thread 1 Group 1 Seq 4 Reading mem 0
  Mem# 0: /dsk1/oradata/prod/redo01a.log
  Mem# 1: /dsk2/oradata/prod/redo01b.log
Completed redo application of 0.02MB
Completed crash recovery at
 Thread 1: logseq 4, block 284, scn 2208986
 26 data blocks read, 26 data blocks written, 83 redo k-bytes read
Thu Jul 24 18:35:45 2014
Thread 1 advanced to log sequence 5 (thread open)
Thread 1 opened at log sequence 5
  Current log# 2 seq# 5 mem# 0: /dsk1/oradata/prod/redo02a.log
  Current log# 2 seq# 5 mem# 1: /dsk2/oradata/prod/redo02b.log
Successful open of redo thread 1
MTTR advisory is disabled because FAST_START_MTTR_TARGET is not set
Thu Jul 24 18:35:45 2014
SMON: enabling cache recovery
Successfully onlined Undo Tablespace 5.
Verifying file header compatibility for 11g tablespace encryption..
Verifying 11g file header compatibility for tablespace encryption completed
SMON: enabling tx recovery
Database Characterset is ZHS16GBK
No Resource Manager plan active
replication_dependency_tracking turned off (no async multimaster replication found)
Starting background process QMNC
Thu Jul 24 18:35:48 2014
QMNC started with pid=20, OS id=13934 
Completed: alter database open
Thu Jul 24 18:35:54 2014
Starting background process CJQ0
Thu Jul 24 18:35:54 2014
CJQ0 started with pid=23, OS id=13946 
Thu Jul 24 18:36:06 2014
Errors in file /u01/app/oracle/diag/rdbms/prod/prod/trace/prod_ora_13932.trc  (incident=139383):
ORA-01578: ORACLE data block corrupted (file # 9, block # 131)
ORA-01110: data file 9: '/u01/app/oracle/oradata/prod/tbs2.dbf'
ORA-26040: Data block was loaded using the NOLOGGING option
Incident details in: /u01/app/oracle/diag/rdbms/prod/prod/incident/incdir_139383/prod_ora_13932_i139383.trc
Errors in file /u01/app/oracle/diag/rdbms/prod/prod/trace/prod_ora_13932.trc  (incident=139384):
ORA-01578: ORACLE data block corrupted (file # 9, block # 131)
ORA-01110: data file 9: '/u01/app/oracle/oradata/prod/tbs2.dbf'
ORA-26040: Data block was loaded using the NOLOGGING option
Incident details in: /u01/app/oracle/diag/rdbms/prod/prod/incident/incdir_139384/prod_ora_13932_i139384.trc
Thu Jul 24 18:36:13 2014
Trace dumping is performing id=[cdmp_20140724183613]
Thu Jul 24 18:36:14 2014
Sweep [inc][139384]: completed
Sweep [inc][139383]: completed
Sweep [inc2][139383]: completed
Trace dumping is performing id=[cdmp_20140724183617]
Thu Jul 24 18:37:14 2014
Sweep [inc2][139384]: completed

9、通过DBV验证数据文件

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
[oracle@rh6 ~]$ dbv file=/u01/app/oracle/oradata/prod/tbs2.dbf blocksize=8192
DBVERIFY: Release 11.2.0.1.0 - Production on Thu Jul 24 18:45:59 2014
Copyright (c) 1982, 2009, Oracle and/or its affiliates.  All rights reserved.
DBVERIFY - Verification starting : FILE = /u01/app/oracle/oradata/prod/tbs2.dbf
DBV-00201: Block, DBA 37748867, marked corrupt for invalid redo application
DBVERIFY - Verification complete
Total Pages Examined         : 1280
Total Pages Processed (Data) : 4
Total Pages Failing   (Data) : 0
Total Pages Processed (Index): 0
Total Pages Failing   (Index): 0
Total Pages Processed (Other): 1274
Total Pages Processed (Seg)  : 0
Total Pages Failing   (Seg)  : 0
Total Pages Empty            : 2
Total Pages Marked Corrupt   : 1
Total Pages Influx           : 0
Total Pages Encrypted        : 0
Highest block SCN            : 2188961 (0.2188961)

在数据文件9上，出现了逻辑坏块 ！

10、转储数据块

14:35:48 SYS@ prod>alter system dump datafile 9 block 131;                   

System altered.

数据块转储信息：

[oracle@rh6 ~]$ cat /u01/app/oracle/diag/rdbms/prod/prod/trace/prod_ora_2822.trc|more

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Trace file /u01/app/oracle/diag/rdbms/prod/prod/trace/prod_ora_2822.trc
Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
ORACLE_HOME = /u01/app/oracle/product/11.2.0/db_1
System name:    Linux
Node name:      rh6
Release:        2.6.32-358.el6.x86_64
Version:        #1 SMP Tue Jan 29 11:47:41 EST 2013
Machine:        x86_64
Instance name: prod
Redo thread mounted by this instance: 1
Oracle process number: 29
Unix process pid: 2822, image: oracle@rh6 (TNS V1-V3)
*** 2014-07-25 14:37:54.634
*** SESSION ID:(28.16) 2014-07-25 14:37:54.634
*** CLIENT ID:() 2014-07-25 14:37:54.634
*** SERVICE NAME:(SYS$USERS) 2014-07-25 14:37:54.634
*** MODULE NAME:(sqlplus@rh6 (TNS V1-V3)) 2014-07-25 14:37:54.634
*** ACTION NAME:() 2014-07-25 14:37:54.634

Start dump data blocks tsn: 9 file#:9 minblk 131 maxblk 131

Block dump from cache:

Dump of buffer cache at level 4 for tsn=9, rdba=37748867

Block dump from disk:

buffer tsn: 9 rdba: 0x02400083 (9/131)

scn: 0x0000.00216666 seq: 0xff flg: 0x04 tail: 0x666600ff

frmt: 0x02 chkval: 0xa4e2 type: 0x00=unknown

Hex dump of block: st=0, typ_found=0

Dump of memory from 0x00007FDEF186CA00 to 0x00007FDEF186EA00

7FDEF186CA00 0000A200 02400083 00216666 04FF0000  [......@.ff!.....]

7FDEF186CA10 0000A4E2 FFFFFFFF FFFFFFFF FFFFFFFF  [................]

7FDEF186CA20 FFFFFFFF FFFFFFFF FFFFFFFF FFFFFFFF  [................]

        Repeat 508 times

7FDEF186E9F0 FFFFFFFF FFFFFFFF FFFFFFFF 666600FF  [..............ff]

End dump data blocks tsn: 9 file#: 9 minblk 131 maxblk 131

----可以看出block 131除了头部信息外，剩下的都是‘FFFFFFFF’，应该是一个逻辑坏块！

11、转储对应的logfile

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
14:35:48 SYS@ prod>alter system dump logfile '/dsk1/oradata/prod/redo01a.log'；
System altered.
14:35:53 SYS@ prod>alter system dump logfile '/dsk1/oradata/prod/redo03a.log'；
System altered.
[oracle@rh6 ~]$ cat /u01/app/oracle/diag/rdbms/prod/prod/trace/prod_ora_2406.trc
DUMP OF REDO FROM FILE '/dsk1/oradata/prod/redo01a.log'
 Opcodes *.*
 RBAs: 0x000000.00000000.0000 thru 0xffffffff.ffffffff.ffff
 SCNs: scn: 0x0000.00000000 thru scn: 0xffff.ffffffff
 Times: creation thru eternity
 FILE HEADER:
        Compatibility Vsn = 186646528=0xb200000
        Db ID=239333010=0xe43ee92, Db Name='PROD'
        Activation ID=264808982=0xfc8aa16
        Control Seq=2747=0xabb, File size=102400=0x19000
        File Number=1, Blksiz=512, File Type=2 LOG
 descrip:"Thread 0001, Seq# 0000000004, SCN 0x000000216562-0x00000021b4dc"
 thread: 1 nab: 0x11c seq: 0x00000004 hws: 0x6 eot: 0 dis: 0
 resetlogs count: 0x32d7d50e scn: 0x0000.00206c24 (2124836)
 prev resetlogs count: 0x32d7cc14 scn: 0x0000.0020680f (2123791)
 Low  scn: 0x0000.00216562 (2188642) 07/24/2014 18:27:42
 Next scn: 0x0000.0021b4dc (2208988) 07/24/2014 18:35:45
 Enabled scn: 0x0000.00206c24 (2124836) 07/15/2014 17:59:42
 Thread closed scn: 0x0000.0021b4da (2208986) 07/24/2014 18:33:04
 Disk cksum: 0x1af0 Calc cksum: 0x1af0
 Terminal recovery stop scn: 0x0000.00000000
 Terminal recovery  01/01/1988 00:00:00
 Most recent redo scn: 0x0000.00000000
 Largest LWN: 0 blocks
 End-of-redo stream : No
 Unprotected mode
 Miscellaneous flags: 0x800000
 Thread internal enable indicator: thr: 0, seq: 0 scn: 0x0000.00000000

REDO RECORD - Thread:1 RBA: 0x000004.000000c6.0198 LEN: 0x0034 VLD: 0x01

SCN: 0x0000.00216666 SUBSCN:  1 07/24/2014 18:30:45

CHANGE #1 INVLD AFN:9 DBA:0x02400083 BLKS:0x0001 OBJ:75139 SCN:0x0000.00216666 SEQ:1 OP:19.2 ENC:0

Direct Loader invalidate block range redo entry

REDO RECORD - Thread:1 RBA: 0x000006.000007b6.0110 LEN: 0x0034 VLD: 0x01

SCN: 0x0000.00220ad0 SUBSCN:  1 07/25/2014 14:22:51

CHANGE #1 INVLD AFN:2 DBA:0x0080f09c BLKS:0x0001 OBJ:6214 SCN:0x0000.00220ad0 SEQ:1 OP:19.2 ENC:0

Direct Loader invalidate block range redo entry

-在redo日志里记录了“Direct Loader”的动作（OBJ:75139，OBJ:6214）


查看对应的对象：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
14:29:23 SYS@ prod>select owner,object_id,object_name from dba_objects
15:18:47   2  where object_id=75139;
OWNER                           OBJECT_ID OBJECT_NAME
------------------------------ ---------- --------------------
SCOTT                               75139 T1
Elapsed: 00:00:00.04
 
15:18:51 SYS@ prod>select owner,object_id,object_name from dba_objects
15:18:58   2  where object_id=6214;
OWNER                           OBJECT_ID OBJECT_NAME
------------------------------ ---------- --------------------
SYS                                  6214 SYS_LOB0000006213C00
                                          038$$
Elapsed: 00:00:00.05
15:19:07 SYS@ prod>

   从以上，查看信息可以推断，在Media Recover阶段，create table T1 as select * from 采用了“Direct Loader”，在非归档下应该是采用了‘nologging’方式，以致后续的Insert的数据通过redo log无法被恢复，导致出现逻辑坏块！


备注：（借鉴Maclean Liu博客观点）

http://www.askmaclean.com/archives/nologging-unrecoverable.html

【数据恢复】NOLOGGING UNRECOVERABLE ORA-26040解析

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
SQL> select count(*) from abc;select count(*) from abc*第 1 行出现错误:ORA-01578: ORACLE 数据块损坏 (文件号 17, 块号 131)ORA-01110: 数据文件 17:‘C:\APP\XIANGBLI\ORADATA\MACLEAN\DATAFILE\O1_MF_NLOGGING_9475OCS5_.DBF’ORA-26040: 数据块是使用 NOLOGGING 选项加载的
 SQL> select UNRECOVERABLE_CHANGE# , UNRECOVERABLE_time from v$datafile where file#=17;
UNRECOVERABLE_CHANGE# UNRECOVERABLE_——————— ————–6486756 26-9月 -13
  
把 (文件号 17, 块号 131) dump出来看一下
  
  
*** 2013-09-26 10:07:46.584Start dump data blocks tsn: 17 file#:17 minblk 131 maxblk 131Block dump from cache:Dump of buffer cache at level 4 for pdb=0 tsn=17 rdba=71303299Block dump from disk:buffer tsn: 17 rdba: 0×04400083 (17/131)scn: 0×0.62faac seq: 0xff flg: 0×04 tail: 0xfaac00fffrmt: 0×02 chkval: 0xa2a1 type: 0×00=unknownHex dump of block: st=0, typ_found=0Dump of memory from 0x000000000BFF2200 to 0x000000000BFF420000BFF2200 0000A200 04400083 0062FAAC 04FF0000 [......@...b.....]00BFF2210 0000A2A1 FFFFFFFF FFFFFFFF FFFFFFFF [................]00BFF2220 FFFFFFFF FFFFFFFF FFFFFFFF FFFFFFFF [................]Repeat 508 times00BFF41F0 FFFFFFFF FFFFFFFF FFFFFFFF FAAC00FF [................]End dump data blocks tsn: 17 file#: 17 minblk 131 maxblk 131
  
scn: 0×0.62faac seq: 0xff
==》 对应的SCN为6486700，可以看到内容除了头部一点外 全是0XFF
  
dump 对应redo logfile 可以看到
REDO RECORD – Thread:1 RBA: 0×000074.00015418.0078 LEN: 0x003c VLD: 0×01 CON_UID: 0SCN: 0×0000.0062faac SUBSCN: 1 09/26/2013 10:04:39CHANGE #1 INVLD CON_ID:0 AFN:17 DBA:0×04400083 BLKS:0x000d OBJ:123054 SCN:0×0000.0062faac SEQ:1 OP:19.2 ENC:0Direct Loader invalidate block range redo entry
  
OP:19.2=》Layer 19 : Direct Loader Log Blocks – KCOCODLB Opcode 2 : Invalidate range – KCBLCOIR
==》这里在redo里标记了 直接路径加载造成块失效的范围，在redo logfile dump中可以看到大量类似数据
  
即当recover时读取redo，读到“Direct Loader invalidate block range redo entry”信息时，则将对应的数据块的内容除了kcbh头部外全部记录为0XFF
当Oracle读取到这些块时就会知道这些块是SOFT Corrupt ，原因是nologging造成的。
Block is marked as SOFT Corrupt and has 0xff along the block. This is theformat used by Oracle to mark a block as corrupt due to redo invalidation(NOLOGGING).NOLOGGING OPERATION in redo:









本文转自 客居天涯 51CTO博客，原文链接：http://blog.51cto.com/tiany/1530094，如需转载请自行联系原作者
上一篇：C#发送邮件三种方法(Localhost，SMTP，SSL-SMTP)


下一篇：【Oracle】RMAN-03009 ORA-01400 错误解决方法