标签归档：oracle异常恢复

Oracle断电故障处理

发表于 2022 年 5 月 31 日由惜分飞

异常断电导致数据库异常恢复文件报ORA-00283 ORA-00742 ORA-00312

 
D:\check_db>sqlplus / as sysdba

SQL*Plus: Release 11.2.0.4.0 Production on 星期二 5月 31 00:38:42 2022

Copyright (c) 1982, 2013, Oracle.  All rights reserved.


连接到:
Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options

SQL> recover datafile 1;
ORA-00283: 恢复会话因错误而取消
ORA-00742: 日志读取在线程 %d 序列 %d 块 %d 中检测到写入丢失情况
ORA-00312: 联机日志 3 线程 1:
'D:\APP\ADMINISTRATOR\FAST_RECOVERY_AREA\ORCL\ONLINELOG\O1_MF_3_HJ32KJD5_.LOG'

这个错误比较明显是由于异常断电引起的写丢失导致.而且这种故障在没有备份的情况下，没有什么好处理方法,只能屏蔽一致性强制拉库,尝试强制拉库报错如下

SQL> startup mount pfile='d:/pfile.txt'
ORACLE 例程已经启动。

Total System Global Area 2.0310E+10 bytes
Fixed Size                  2290000 bytes
Variable Size            3690991280 bytes
Database Buffers         1.6576E+10 bytes
Redo Buffers               40837120 bytes
数据库装载完毕。
SQL> recover database until cancel;
ORA-00279: 更改 18755939194213 (在  生成) 对于线程 1 是必需的


指定日志: {<RET>=suggested | filename | AUTO | CANCEL}
D:\APP\ADMINISTRATOR\FAST_RECOVERY_AREA\ORCL\ONLINELOG\O1_MF_3_HJ32KJD5_.LOG
ORA-00600: internal error code, arguments: [3020], [2], [78824], [8467432], [],
[], [], [], [], [], [], []
ORA-10567: Redo is inconsistent with data block (file# 2, block# 78824, file
offset is 645726208 bytes)
ORA-10564: tablespace SYSAUX
ORA-01110: data file 2: 'D:\ORADATA\ORCL\SYSAUX01.DBF'
ORA-10561: block type 'TRANSACTION MANAGED INDEX BLOCK', data object# 80834


ORA-01112: 未启动介质恢复


SQL> alter database open resetlogs;
alter database open resetlogs
*
第 1 行出现错误:
ORA-00600: 内部错误代码, 参数: [krsi_al_hdr_update.15], [4294967295], [], [],[], [], [], [], [], [], [], []

ORA-600 krsi_al_hdr_update.15错误,主要是由于redo异常导致无法resetlogs成功,具体参考：Alter Database Open Resetlogs returns error ORA-00600: [krsi_al_hdr_update.15], (Doc ID 2026541.1)描述,处理这个问题之后,再次resetlogs库，报ORA-600 2662错误

SQL> alter database open resetlogs;
alter database open resetlogs
*
第 1 行出现错误:
ORA-00603: ORACLE server session terminated by fatal error
ORA-00600: internal error code, arguments: [2662], [4366], [4112122046],
[4366], [4112228996], [12583040], [], [], [], [], [], []
ORA-00600: internal error code, arguments: [2662], [4366], [4112122045],
[4366], [4112228996], [12583040], [], [], [], [], [], []
ORA-01092: ORACLE instance terminated. Disconnection forced
ORA-00600: internal error code, arguments: [2662], [4366], [4112122040],
[4366], [4112228996], [12583040], [], [], [], [], [], []
进程 ID: 4644
会话 ID: 1701 序列号: 3

这个问题比较简单,通过修改scn即可绕过去,之后数据库open报ORA-600 4194等错误

SQL> alter database open ;
alter database open 
*
第 1 行出现错误:
ORA-00600: 内部错误代码, 参数: [4194], [

SMON: enabling tx recovery
Database Characterset is ZHS16GBK
Errors in file D:\APP\ADMINISTRATOR\diag\rdbms\orcl\orcl\trace\orcl_smon_5112.trc  (incident=322982):
ORA-00600: internal error code, arguments: [4137], [10.33.3070116], [0], [0], [], [], [], [], [], [], [], []
Incident details in: D:\APP\ADMINISTRATOR\diag\rdbms\orcl\orcl\incident\incdir_322982\orcl_smon_5112_i322982.trc
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
ARC3: Archival started
ARC0: STARTING ARCH PROCESSES COMPLETE
replication_dependency_tracking turned off (no async multimaster replication found)
LOGSTDBY: Validating controlfile with logical metadata
LOGSTDBY: Validation complete
Errors in file D:\APP\ADMINISTRATOR\diag\rdbms\orcl\orcl\trace\orcl_ora_3340.trc  (incident=323030):
ORA-00600: 内部错误代码, 参数: [4194], [
Incident details in: D:\APP\ADMINISTRATOR\diag\rdbms\orcl\orcl\incident\incdir_323030\orcl_ora_3340_i323030.trc
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
Tue May 31 09:05:04 2022
Sweep [inc][322982]: completed
ORACLE Instance orcl (pid = 13) - Error 600 encountered while recovering transaction (10, 33).
Errors in file D:\APP\ADMINISTRATOR\diag\rdbms\orcl\orcl\trace\orcl_smon_5112.trc:
ORA-00600: internal error code, arguments: [4137], [10.33.3070116], [0], [0], [], [], [], [], [], [], [], []
Checker run found 1 new persistent data failures
Tue May 31 09:05:05 2022
Sweep [inc][323030]: completed
Sweep [inc2][322982]: completed
Tue May 31 09:05:14 2022
Errors in file D:\APP\ADMINISTRATOR\diag\rdbms\orcl\orcl\trace\orcl_smon_5112.trc  (incident=322983):
ORA-00600: internal error code, arguments: [4193], [10.33.3070116], [0], [], [], [], [], [], [], [], [], []
Incident details in: D:\APP\ADMINISTRATOR\diag\rdbms\orcl\orcl\incident\incdir_322983\orcl_smon_5112_i322983.trc
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
Tue May 31 09:05:14 2022
ORA-600 signalled during: alter database open...
Block recovery stopped at EOT rba 2.61.16
Block recovery completed at rba 2.61.16, scn 4366.4112429058
Block recovery from logseq 2, block 60 to scn 18755939643393
Recovery of Online Redo Log: Thread 1 Group 2 Seq 2 Reading mem 0
  Mem# 0: D:\APP\ADMINISTRATOR\FAST_RECOVERY_AREA\ORCL\ONLINELOG\O1_MF_2_K9BSVC11_.LOG
Block recovery completed at rba 2.61.16, scn 4366.4112429058
Dumping diagnostic data in directory=[cdmp_2022053],requested by(instance=1,osid=5112(SMON)),summary=[incident=322983].
Errors in file D:\APP\ADMINISTRATOR\diag\rdbms\orcl\orcl\trace\orcl_smon_5112.trc:
ORA-01595: error freeing extent (3) of rollback segment (1))
ORA-00600: internal error code, arguments: [4193], [10.33.3070116], [3], [], [], [], [], [], [], [], [], []

对异常undo进行处理,数据库正常open成功

SQL> shutdown immediate;
ORA-00600: 内部错误代码, 参数: [4193], [


SQL> shutdown abort;
ORACLE 例程已经关闭。
SQL> startup mount
ORACLE 例程已经启动。

Total System Global Area 2.0310E+10 bytes
Fixed Size                  2290000 bytes
Variable Size            3690991280 bytes
Database Buffers         1.6576E+10 bytes
Redo Buffers               40837120 bytes
数据库装载完毕。
SQL> alter database open;

数据库已更改。

hcheck检测有一些字典不一致,建议客户逻辑导出,然后导入到新库中

HCheck Version 07MAY18 on 31-5月 -2022 09:12:22
----------------------------------------------
Catalog Version 11.2.0.4.0 (1102000400)
db_name: ORCL

                                   Catalog       Fixed
Procedure Name                     Version    Vs Release    Timestamp      Resul
t
------------------------------ ... ---------- -- ---------- -------------- -----
-
.- LobNotInObj                 ... 1102000400 <=  *All Rel* 05/31 09:12:22 PASS
.- MissingOIDOnObjCol          ... 1102000400 <=  *All Rel* 05/31 09:12:22 PASS
.- SourceNotInObj              ... 1102000400 <=  *All Rel* 05/31 09:12:22 PASS
.- OversizedFiles              ... 1102000400 <=  *All Rel* 05/31 09:12:22 PASS
.- PoorDefaultStorage          ... 1102000400 <=  *All Rel* 05/31 09:12:22 PASS
.- PoorStorage                 ... 1102000400 <=  *All Rel* 05/31 09:12:22 PASS
.- TabPartCountMismatch        ... 1102000400 <=  *All Rel* 05/31 09:12:22 PASS
.- OrphanedTabComPart          ... 1102000400 <=  *All Rel* 05/31 09:12:22 PASS
.- MissingSum$                 ... 1102000400 <=  *All Rel* 05/31 09:12:22 PASS
.- MissingDir$                 ... 1102000400 <=  *All Rel* 05/31 09:12:22 PASS
.- DuplicateDataobj            ... 1102000400 <=  *All Rel* 05/31 09:12:22 PASS
.- ObjSynMissing               ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- ObjSeqMissing               ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- OrphanedUndo                ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- OrphanedIndex               ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- OrphanedIndexPartition      ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- OrphanedIndexSubPartition   ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- OrphanedTable               ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- OrphanedTablePartition      ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- OrphanedTableSubPartition   ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- MissingPartCol              ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- OrphanedSeg$                ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- OrphanedIndPartObj#         ... 1102000400 <=  *All Rel* 05/31 09:12:23 FAIL

HCKE-0024: Orphaned Index Partition Obj# (no OBJ$) (Doc ID 1360935.1)
ORPHAN INDPART$: OBJ#=149167 BO#=6378 - no OBJ$ row
ORPHAN INDPART$: OBJ#=149168 BO#=6378 - no OBJ$ row

.- DuplicateBlockUse           ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- FetUet                      ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- Uet0Check                   ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- SeglessUET                  ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- BadInd$                     ... 1102000400 <=  *All Rel* 05/31 09:12:23 FAIL

HCKE-0030: OBJ$ INDEX entry has no IND$ or INDPART$/INDSUBPART$ entry (Doc ID 13
60528.1)
OBJ$ INDEX PARTITION has no INDPART$ entry: Obj#=148278 SYS Name=WRH$_FILESTATXS
_PK PARTITION=WRH$_FILEST_1572571104_16462
OBJ$ INDEX PARTITION has no INDPART$ entry: Obj#=148920 SYS Name=WRH$_FILESTATXS
_PK PARTITION=WRH$_FILEST_1572571104_16678

.- BadTab$                     ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- BadIcolDepCnt               ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- ObjIndDobj                  ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- TrgAfterUpgrade             ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- ObjType0                    ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- BadOwner                    ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- StmtAuditOnCommit           ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- BadPublicObjects            ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- BadSegFreelist              ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- BadDepends                  ... 1102000400 <=  *All Rel* 05/31 09:12:23 WARN

HCKW-0016: Dependency$ p_timestamp mismatch for VALID objects (Doc ID 1361045.1)

[E] - P_OBJ#=6376 D_OBJ#=6765

.- CheckDual                   ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- ObjectNames                 ... 1102000400 <=  *All Rel* 05/31 09:12:23 PASS
.- BadCboHiLo                  ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
.- ChkIotTs                    ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
.- NoSegmentIndex              ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
.- BadNextObject               ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
.- DroppedROTS                 ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
.- FilBlkZero                  ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
.- DbmsSchemaCopy              ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
.- OrphanedObjError            ... 1102000400 >  1102000000 05/31 09:12:24 PASS
.- ObjNotLob                   ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
.- MaxControlfSeq              ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
.- SegNotInDeferredStg         ... 1102000400 >  1102000000 05/31 09:12:24 PASS
.- SystemNotRfile1             ... 1102000400 >   902000000 05/31 09:12:24 PASS
.- DictOwnNonDefaultSYSTEM     ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
.- OrphanTrigger               ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
.- ObjNotTrigger               ... 1102000400 <=  *All Rel* 05/31 09:12:24 PASS
---------------------------------------
31-5月 -2022 09:12:24  Elapsed: 2 secs
---------------------------------------
Found 4 potential problem(s) and 1 warning(s)
Contact Oracle Support with the output and trace file
to check if the above needs attention or not

PL/SQL 过程已成功完成。

发表在 Oracle备份恢复 | 标签为 krsi_al_hdr_update.15, ORA-00742, ORA-600 2662, ORA-600 4194, oracle异常恢复, oracle断电恢复 | 评论关闭

ORA-600 kccpb_sanity_check_2故障恢复

发表于 2015 年 7 月 29 日由惜分飞

今天是有人在淘宝旺旺上找我，需要oracle数据库恢复支持
wangwang

远程登录上去一看发现数据库mount的时候报ORA-600[kccpb_sanity_check_2]错误

C:\Documents and Settings\Administrator>sqlplus / as sysdba

SQL*Plus: Release 10.2.0.3.0 - Production on Wed Jul 29 16:23:18 2015

Copyright (c) 1982, 2006, Oracle.  All Rights Reserved.


Connected to:
Oracle Database 10g Enterprise Edition Release 10.2.0.3.0 - Production
With the Partitioning, OLAP and Data Mining options

SQL> alter database mount;
alter database mount
*
ERROR at line 1:
ORA-00600: internal error code, arguments: [kccpb_sanity_check_2], [14169],
[14160], [0x0], [], [], [], []

尝试重建控制文件

SQL> shutdown immediate;
ORA-01507: database not mounted


ORACLE instance shut down.
SQL> startup pfile='D:\database\m104\pfile\init.ora' nomount
ORACLE instance started.

Total System Global Area  444596224 bytes
Fixed Size                  1291072 bytes
Variable Size             155192512 bytes
Database Buffers          281018368 bytes
Redo Buffers                7094272 bytes

SQL> SHOW PARAMETER CONT;

NAME                                 TYPE        VALUE
------------------------------------ ----------- ------------------------------
control_file_record_keep_time        integer     7
control_files                        string      D:\DATABASE\M104\CTRL\CONTROL0
                                                 2.CTL
global_context_pool_size             string
SQL> ALTER DATABASE MOUNT;
ALTER DATABASE MOUNT
*
ERROR at line 1:
ORA-00600: internal error code, arguments: [kccpb_sanity_check_2], [14169],
[14160], [0x0], [], [], [], []


SQL>
SQL> CREATE CONTROLFILE REUSE DATABASE "m104_db" NORESETLOGS  FORCE LOGGING NOAR
CHIVELOG
  2      MAXLOGFILES 16
  3      MAXLOGMEMBERS 3
  4      MAXDATAFILES 100
  5      MAXINSTANCES 8
  6      MAXLOGHISTORY 2921
  7  LOGFILE
  8    GROUP 1 'D:\database\m104\log\redo01.log'  SIZE 51200K,
  9    GROUP 2 'D:\database\m104\log\redo02.log'  SIZE 51200K,
 10    GROUP 3 'D:\database\m104\log\redo03.log'  SIZE 51200K
 11  DATAFILE
 12    'd:\database\m104\data\system01.dbf',
 13    'd:\database\m104\data\sysaux01.dbf',
 14    'd:\database\m104\data\USERS01.DBF',
 15    'd:\database\m104\data\UNDOTBS01.DBF',
 16    'd:\database\m104\data\INDX01.DBF'
 17  CHARACTER SET WE8ISO8859P1
 18  ;
CREATE CONTROLFILE REUSE DATABASE "m104_db" NORESETLOGS  FORCE LOGGING NOARCHIVE
LOG
*
ERROR at line 1:
ORA-01503: CREATE CONTROLFILE failed
ORA-00600: internal error code, arguments: [kccsga_update_ckpt_4], [1], [8],
[], [], [], [], []

SQL>
SQL> CREATE CONTROLFILE REUSE DATABASE "m104_db" RESETLOGS  FORCE LOGGING NOARCH
IVELOG
  2      MAXLOGFILES 16
  3      MAXLOGMEMBERS 3
  4      MAXDATAFILES 100
  5      MAXINSTANCES 8
  6      MAXLOGHISTORY 2921
  7  LOGFILE
  8    GROUP 1 'D:\database\m104\log\redo01.log'  SIZE 51200K,
  9    GROUP 2 'D:\database\m104\log\redo02.log'  SIZE 51200K,
 10    GROUP 3 'D:\database\m104\log\redo03.log'  SIZE 51200K
 11  DATAFILE
 12    'd:\database\m104\data\system01.dbf',
 13    'd:\database\m104\data\sysaux01.dbf',
 14    'd:\database\m104\data\USERS01.DBF',
 15    'd:\database\m104\data\UNDOTBS01.DBF',
 16    'd:\database\m104\data\INDX01.DBF'
 17  CHARACTER SET WE8ISO8859P1
 18  ;
CREATE CONTROLFILE REUSE DATABASE "m104_db" RESETLOGS  FORCE LOGGING NOARCHIVELO
G
*
ERROR at line 1:
ORA-01503: CREATE CONTROLFILE failed
ORA-00600: internal error code, arguments: [kccsga_update_ckpt_4], [1], [8],
[], [], [], [], []

无论是使用noresetlogs还是resetlogs，重建控制文件都报ORA-600[kccsga_update_ckpt_4]错误.比较奇怪，无解指定控制文件新名称重建试试看

修改控制文件路径

SQL> SHUTDOWN ABORT
ORACLE instance shut down.
SQL> startup pfile='D:\database\m104\pfile\init.ora' nomount
ORACLE instance started.

Total System Global Area  444596224 bytes
Fixed Size                  1291072 bytes
Variable Size             155192512 bytes
Database Buffers          281018368 bytes
Redo Buffers                7094272 bytes
SQL> SHOW PARAMETER CONT;

NAME                                 TYPE        VALUE
------------------------------------ ----------- ------------------------------
control_file_record_keep_time        integer     7
control_files                        string      D:\DATABASE\M104\CTRL\CONTROL0
                                                 4.CTL
global_context_pool_size             string
SQL> CREATE CONTROLFILE REUSE DATABASE "m104_db" RESETLOGS  FORCE LOGGING NOARCH
IVELOG
  2      MAXLOGFILES 16
  3      MAXLOGMEMBERS 3
  4      MAXDATAFILES 100
  5      MAXINSTANCES 8
  6      MAXLOGHISTORY 2921
  7  LOGFILE
  8    GROUP 1 'D:\database\m104\log\redo01.log'  SIZE 51200K,
  9    GROUP 2 'D:\database\m104\log\redo02.log'  SIZE 51200K,
 10    GROUP 3 'D:\database\m104\log\redo03.log'  SIZE 51200K
 11  DATAFILE
 12    'd:\database\m104\data\system01.dbf',
 13    'd:\database\m104\data\sysaux01.dbf',
 14    'd:\database\m104\data\USERS01.DBF',
 15    'd:\database\m104\data\UNDOTBS01.DBF',
 16    'd:\database\m104\data\INDX01.DBF'
 17  CHARACTER SET WE8ISO8859P1
 18  ;

Control file created.

使用新的控制文件位置，这次终于数据库重建控制文件成功
尝试指定redo进行恢复，数据库正常打开

SQL> RECOVER DATABASE USING BACKUP CONTROLFILE UNTIL CANCEL;
ORA-00279: change 3643108240801 generated at 07/26/2015 20:15:22 needed for
thread 1
ORA-00289: suggestion :
D:\ORACLE\PRODUCT\10.2.0\DB_1\RDBMS\ARC00567_0866390669.001
ORA-00280: change 3643108240801 for thread 1 is in sequence #567


Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
D:\database\m104\log\redo01.log
ORA-00310: archived log contains sequence 566; sequence 567 required
ORA-00334: archived log: 'D:\DATABASE\M104\LOG\REDO01.LOG'


ORA-01547: warning: RECOVER succeeded but OPEN RESETLOGS would get error below
ORA-01194: file 1 needs more recovery to be consistent
ORA-01110: data file 1: 'D:\DATABASE\M104\DATA\SYSTEM01.DBF'


SQL> RECOVER DATABASE USING BACKUP CONTROLFILE UNTIL CANCEL;
ORA-00279: change 3643108240801 generated at 07/26/2015 20:15:22 needed for
thread 1
ORA-00289: suggestion :
D:\ORACLE\PRODUCT\10.2.0\DB_1\RDBMS\ARC00567_0866390669.001
ORA-00280: change 3643108240801 for thread 1 is in sequence #567


Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
D:\database\m104\log\redo02.log
Log applied.
Media recovery complete.
SQL> ALTER DATABASE OPEN RESETLOGS;

Database altered.

数据库恢复完成。整个数据库恢复比较简单，但是注意这里的ORA-600[kccsga_update_ckpt_4]通过修改控制文件路径规避，具体原因待查。

知识点补充：ORA-600 [kccpb_sanity_check_2] [a] [b] {c}

VERSIONS:
  Versions 10.2 to 11.2

DESCRIPTION:

  This internal error is raised when the sequence number (seq#) of the
  current block of the controlfile is greater than the seq# in the controlfile header.

  The header value should always be equal to, or greater than the value
  held in the control file block(s).

  This extra check was introduced in Oracle 10gR2 to detect lost writes
  or stale reads to the header.

ARGUMENTS:
  Arg [a] seq# in control block header.
  Arg [b] seq# in the control file header.
  Arg {c}

发表在 ORA-xxxxx, Oracle备份恢复 | 标签为 kccpb_sanity_check_2, kccsga_update_ckpt_4, ORA-600 kccpb_sanity_check_2, ORA-600 kccsga_update_ckpt_4, oracle异常恢复 | 评论关闭

Oracle异常恢复前备份保护现场建议—ASM环境

发表于 2015 年 5 月 15 日由惜分飞

在上一篇中写道了文件系统的库,在进行异常恢复前的备份方法（Oracle异常恢复前备份保护现场建议—FileSystem环境）,对于asm库,因为asm 里面的数据文件无法直接dd文件头,因此备份方式也有所改变.对于asm是mount,但是数据库不能打开,使用rman或者asm的cp命令全部备份数据文件也来不及或者空间不足,这样的情况下,你可以考虑使用rman或者cp命令备份控制文件和system表空间文件,cp命令备份redo,dd命令备份文件头,来完成asm情况下数据库异常恢复前备份

控制文件备份
11.2及其以后版本使用asmcmd cp命令处理

select 'asmcmd cp '||name||' &&backup_dir/' from v$datafile where ts#=0
union all
select 'asmcmd cp '||name||' &&backup_dir/crontrofile_'||rownum||'.ctl' from v$controlfile
union all
select 'asmcmd cp '||member||' &&backup_dir/'||thread#||'_'||a.group#||'_'||sequence#||'_'||substr(member,
instr(member,'/',-1)+1)  FROM v$log a, v$logfile b WHERE a.group# = B.GROUP#;

其他版本使用rman命令处理

--rman备份控制文件(/tmp目录自己修改)
copy current controlfile to '/tmp/ctl.ctl';

--rman备份system表空间
select 'copy datafile '||file#||' to ''&backup_dir/system_'||file#||'.dbf'';' 
from v$datafile where ts#=0;

--redo无法直接备份

备份文件头

[grid@xifenfei ~]$ ss

SQL*Plus: Release 11.2.0.4.0 Production on Fri May 1 04:15:18 2015

Copyright (c) 1982, 2013, Oracle.  All rights reserved.


Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Automatic Storage Management option

SQL> set lines 150
SQL> select 'dd if='||c.PATH_KFDSK||' of=&&backup_path/'||a.GROUP_KFFXP||'_'||a.disk_kffxp||'_'||
  2  b.NUMBER_KFFIL||'.asm count=1 bs='|| d.AUSIZE_KFGRP||' skip='||a.au_kffxp backup_dd_cmd
  3   FROM x$kffxp a, X$KFFIL  b,X$KFDSK c,X$KFGRP d  WHERE
  4  a.GROUP_KFFXP=b.GROUP_KFFIL
  5  and a.NUMBER_KFFXP=b.NUMBER_KFFIL
  6  and b.FTYPE_KFFIL in(2,12)
  7  and b.NUMBER_KFFIL>255
  8  and a.xnum_kffxp=0
  9  and a.GROUP_KFFXP=c.GRPNUM_KFDSK
 10  and a.disk_kffxp=c.NUMBER_KFDSK
 11  and a.GROUP_KFFXP=d.NUMBER_KFGRP;
Enter value for backup_path: /tmp
old   1: select 'dd if='||c.PATH_KFDSK||' of=&&backup_path/'||a.GROUP_KFFXP||'_'||a.disk_kffxp||'_'||
new   1: select 'dd if='||c.PATH_KFDSK||' of=/tmp/'||a.GROUP_KFFXP||'_'||a.disk_kffxp||'_'||

BACKUP_DD_CMD
------------------------------------------------------------------------------------------------------------------
dd if=/dev/asm-disk1 of=/tmp/1_0_256.asm count=1 bs=1048576 skip=29
dd if=/dev/asm-disk2 of=/tmp/1_1_257.asm count=1 bs=1048576 skip=404
dd if=/dev/asm-disk2 of=/tmp/1_1_258.asm count=1 bs=1048576 skip=641
dd if=/dev/asm-disk1 of=/tmp/1_0_259.asm count=1 bs=1048576 skip=648
dd if=/dev/asm-disk3 of=/tmp/2_0_256.asm count=1 bs=1048576 skip=51

还原文件头

SQL> set lines 150
SQL> select 'dd of='||c.PATH_KFDSK||' if=&&backup_path/'||a.GROUP_KFFXP||'_'||a.disk_kffxp||
  2  '_'||b.NUMBER_KFFIL||'.asm count=1 conv=notrunc bs='|| d.AUSIZE_KFGRP||' seek='||a.au_kffxp restore_dd_cmd 
  3   FROM x$kffxp a, X$KFFIL  b,X$KFDSK c,X$KFGRP d  WHERE
  4  a.GROUP_KFFXP=b.GROUP_KFFIL
  5  and a.NUMBER_KFFXP=b.NUMBER_KFFIL
  6  and b.FTYPE_KFFIL in(2,12)
  7  and b.NUMBER_KFFIL>255
  8  and a.xnum_kffxp=0
  9  and a.GROUP_KFFXP=c.GRPNUM_KFDSK
 10  and a.disk_kffxp=c.NUMBER_KFDSK
 11  and a.GROUP_KFFXP=d.NUMBER_KFGRP;
old   1: select 'dd of='||c.PATH_KFDSK||' if=&&backup_path/'||a.GROUP_KFFXP||'_'||a.disk_kffxp||
new   1: select 'dd of='||c.PATH_KFDSK||' if=/tmp/'||a.GROUP_KFFXP||'_'||a.disk_kffxp||

RESTORE_DD_CMD
-----------------------------------------------------------------------------------------------------------------
dd of=/dev/asm-disk1 if=/tmp/1_0_256.asm count=1 conv=notrunc bs=1048576 seek=29
dd of=/dev/asm-disk2 if=/tmp/1_1_257.asm count=1 conv=notrunc bs=1048576 seek=404
dd of=/dev/asm-disk2 if=/tmp/1_1_258.asm count=1 conv=notrunc bs=1048576 seek=641
dd of=/dev/asm-disk1 if=/tmp/1_0_259.asm count=1 conv=notrunc bs=1048576 seek=648
dd of=/dev/asm-disk3 if=/tmp/2_0_256.asm count=1 conv=notrunc bs=1048576 seek=51

SQL>

备份还原文件头测试–通过测试证明该方法备份文件头是ok的
关闭数据库,使用dd备份文件头

[oracle@xifenfei ~]$ sqlplus / as sysdba

SQL*Plus: Release 11.2.0.4.0 Production on Fri May 1 04:21:49 2015

Copyright (c) 1982, 2013, Oracle.  All rights reserved.


Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Partitioning, Automatic Storage Management, OLAP, Data Mining
and Real Application Testing options

SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.

dul查看当前dbname值为XIFENFEI

[oracle@xifenfei dul]$ ./dul

Data UnLoader: 10.2.0.6.5 - Internal Only - on Fri May  1 04:37:43 2015
with 64-bit io functions

Copyright (c) 1994 2015 Bernard van Duijnen All rights reserved.

 Strictly Oracle Internal Use Only


Disk group DATA, dul group_cid 0
Discovered disk /dev/asm-disk1 as diskgroup DATA, disk number 0 size 3922 Mb File1 starts at 2, dul_disk_cid 0
Discovered disk /dev/asm-disk2 as diskgroup DATA, disk number 1 size 3922 Mb without File1 meta data, dul_disk_cid 1
Disk group XIFENFEI, dul group_cid 1
Discovered disk /dev/asm-disk3 as diskgroup XIFENFEI, disk number 0 size 4439 Mb File1 starts at 2, dul_disk_cid 2

DUL: Warning: Dictionary cache DC_ASM_EXTENTS is empty
Probing for attributes in File9, the attribute directory, for disk group DATA
attribute name "_extent_sizes", value "1 4 16"
attribute name "_extent_counts", value "20000 20000 2147483647"
Oracle data file size 775954432 bytes, block size 8192
Found db_id = 1495013434
Found db_name = XIFENFEI   <-----db name
DUL: Error: Filedir block not allocated, file does not exist
DUL: Error: Could not load asm meta data for group XIFENFEI file 9
Probing for filenames in File6, the alias directory, for disk group XIFENFEI
+XIFENFEI/XIFENFEI/DATAFILE/XIFENFEI.256.878397315
Probing for database datafiles in File1, the file directory,  for disk group XIFENFEI
File 256 datafile size 104865792, block size 8192
Disk group XIFENFEI has one file of type datafile

使用dd备份1文件头

[oracle@xifenfei tmp]$ dd if=/dev/asm-disk1 of=/tmp/1_0_256.asm count=1 bs=1048576 skip=29
1+0 records in
1+0 records out
1048576 bytes (1.0 MB) copied, 0.0168209 seconds, 62.3 MB/s

尝试把dbname从XIFENFEI修改为ORCL

SQL> select dump('XIFENFEI',16) from dual;

DUMP('XIFENFEI',16)
-------------------------------------
Typ=96 Len=8: 58,49,46,45,4e,46,45,49

SQL> SELECT DUMP('ORCL',16) FROM DUAL; 

DUMP('ORCL',16)
-------------------------
Typ=96 Len=4: 4f,52,43,4c

SQL>

bbed修改XIFENFEI为ORCL

[oracle@xifenfei tmp]$ bbed filename='/tmp/1_0_256.asm' mode=edit 
Password: 

BBED: Release 2.0.0.0.0 - Limited Production on Fri May 1 04:24:06 2015

Copyright (c) 1982, 2011, Oracle and/or its affiliates.  All rights reserved.

************* !!! For Oracle Internal Use only !!! ***************

BBED> set blocksize 8192
        BLOCKSIZE       8192

BBED> set block 1
        BLOCK#          1

BBED> map
 File: /tmp/1_0_256.asm (0)
 Block: 1                                     Dba:0x00000000
------------------------------------------------------------
 Data File Header

 struct kcvfh, 860 bytes                    @0       

 ub4 tailchk                                @8188    

BBED> p kcvfhhdr
struct kcvfhhdr, 76 bytes                   @20      
   ub4 kccfhswv                             @20       0x00000000
   ub4 kccfhcvn                             @24       0x0b200400
   ub4 kccfhdbi                             @28       0x591c183a
   text kccfhdbn[0]                         @32      X
   text kccfhdbn[1]                         @33      I
   text kccfhdbn[2]                         @34      F
   text kccfhdbn[3]                         @35      E
   text kccfhdbn[4]                         @36      N
   text kccfhdbn[5]                         @37      F
   text kccfhdbn[6]                         @38      E
   text kccfhdbn[7]                         @39      I

BBED> d seek 32
 File: /tmp/1_0_256.asm (0)
 Block: 1                seeks:   32 to   63           Dba:0x00000000
------------------------------------------------------------------------
 58494645 4e464549 12040000 00720100 00200000 01000300 00000000 00000000 

 <32 bytes per line>

dd把修改的block还原到asm中

[oracle@xifenfei dul]$ dd of=/dev/asm-disk1 if=/tmp/1_0_256.asm count=1 conv=notrunc bs=1048576 seek=29
1+0 records in
1+0 records out
1048576 bytes (1.0 MB) copied, 0.00253244 seconds, 414 MB/s

dul验证dbname 修改为ORCL成功

[oracle@xifenfei dul]$ ./dul

Data UnLoader: 10.2.0.6.5 - Internal Only - on Fri May  1 04:41:33 2015
with 64-bit io functions

Copyright (c) 1994 2015 Bernard van Duijnen All rights reserved.

 Strictly Oracle Internal Use Only


Disk group DATA, dul group_cid 0
Discovered disk /dev/asm-disk1 as diskgroup DATA, disk number 0 size 3922 Mb File1 starts at 2, dul_disk_cid 0
Discovered disk /dev/asm-disk2 as diskgroup DATA, disk number 1 size 3922 Mb without File1 meta data, dul_disk_cid 1
Disk group XIFENFEI, dul group_cid 1
Discovered disk /dev/asm-disk3 as diskgroup XIFENFEI, disk number 0 size 4439 Mb File1 starts at 2, dul_disk_cid 2

DUL: Warning: Dictionary cache DC_ASM_EXTENTS is empty
Probing for attributes in File9, the attribute directory, for disk group DATA
attribute name "_extent_sizes", value "1 4 16"
attribute name "_extent_counts", value "20000 20000 2147483647"
Oracle data file size 775954432 bytes, block size 8192
Found db_id = 1495013434
Found db_name = ORCL   <----修改后的dbname
DUL: Error: Filedir block not allocated, file does not exist
DUL: Error: Could not load asm meta data for group XIFENFEI file 9
Probing for filenames in File6, the alias directory, for disk group XIFENFEI
+XIFENFEI/XIFENFEI/DATAFILE/XIFENFEI.256.878397315
Probing for database datafiles in File1, the file directory,  for disk group XIFENFEI
File 256 datafile size 104865792, block size 8192
Disk group XIFENFEI has one file of type datafile

对于asm无法mount情况下备份asm disk header
asm磁盘的备份主要是备份磁盘头100M空间,使用dd命令直接备份

set lines 150
set pages 1000
select 'dd if='||path||' of=&asmbackup_dir/'||group_number||'_'||disk_number||'.asm bs=1048576 
count=100' from v$asm_disk;

set lines 150
set pages 1000
select 'dd of='||path||' if=&asmbackup_dir/'||group_number||'_'||disk_number||'.asm bs=1048576 
count=100 conv=notrunc' from v$asm_disk;

asmlib需要注意把ORCL:替换为/dev/oracleasm/disks/对应目录.

发表在 Oracle备份恢复 | 标签为 oracle异常恢复, oracle现场保护, 异常恢复备份 | 评论关闭

17813235971

QQ 咨询

标签归档：oracle异常恢复

Oracle断电故障处理

ORA-600 kccpb_sanity_check_2故障恢复

Oracle异常恢复前备份保护现场建议—ASM环境

文章分类

最近发表

友情链接

标签归档：oracle异常恢复

Oracle断电故障处理

ORA-600 kccpb_sanity_check_2故障恢复

Oracle异常恢复前备份保护现场建议—ASM环境

标签云

文章分类

最近发表

友情链接