Daily Archives: 2011 年 12 月 10 日

ORA-00600[4194]故障解决

Posted on by

朋友数据库因为断电，导致数据库正常启动片刻之后，自动down掉
一、alert日志

Starting ORACLE instance (normal)
LICENSE_MAX_SESSION = 0
LICENSE_SESSIONS_WARNING = 0
Picked latch-free SCN scheme 2
Autotune of undo retention is turned on. 
IMODE=BR
ILAT =18
LICENSE_MAX_USERS = 0
SYS auditing is disabled
ksdpec: called for event 13740 prior to event group initialization
Starting up ORACLE RDBMS Version: 10.2.0.1.0.
System parameters with non-default values:
  processes                = 150
  __shared_pool_size       = 58720256
  __large_pool_size        = 4194304
  __java_pool_size         = 4194304
  __streams_pool_size      = 4194304
  nls_date_format          = yyyy-mm-dd hh24:mi:ss
  sga_target               = 335544320
  control_files            = /u02/ezhou/control01.ctl
  db_block_size            = 8192
  compatible               = 10.2.0.1.0
  log_archive_dest         = /u02/arch
  log_archive_max_processes= 10
  db_file_multiblock_read_count= 16
  fast_start_mttr_target   = 300
  undo_management          = AUTO
  undo_tablespace          = UNDOTBS1
  remote_login_passwordfile= EXCLUSIVE
  db_domain                = 
  dispatchers              = (PROTOCOL=TCP) (SERVICE=ezhouXDB)
  job_queue_processes      = 10
  background_dump_dest     = /u01/pp/oracle/admin/ezhou/bdump
  user_dump_dest           = /u01/pp/oracle/admin/ezhou/udump
  core_dump_dest           = /u01/pp/oracle/admin/ezhou/cdump
  audit_file_dest          = /u01/pp/oracle/admin/ezhou/adump
  db_name                  = ezhou
  open_cursors             = 400
  sql_trace                = TRUE
  pga_aggregate_target     = 94371840
MMAN started with pid=4, OS id=5539
PMON started with pid=2, OS id=5535
DBW0 started with pid=5, OS id=5541
LGWR started with pid=6, OS id=5543
SMON started with pid=8, OS id=5547
CJQ0 started with pid=10, OS id=5577
RECO started with pid=9, OS id=5575
Sat Dec 10 17:15:40 2011
starting up 1 dispatcher(s) for network address '(ADDRESS=(PARTIAL=YES)(PROTOCOL=TCP))'
MMNL started with pid=12, OS id=5581
MMON started with pid=11, OS id=5579
Sat Dec 10 17:15:40 2011
starting up 1 shared server(s) ...
PSP0 started with pid=3, OS id=5537
CKPT started with pid=7, OS id=5545
Sat Dec 10 17:15:42 2011
ALTER DATABASE   MOUNT
Sat Dec 10 17:15:46 2011
Setting recovery target incarnation to 3
Sat Dec 10 17:15:47 2011
Successful mount of redo thread 1, with mount id 4055654398
Sat Dec 10 17:15:47 2011
Database mounted in Exclusive Mode
Completed: ALTER DATABASE   MOUNT
Sat Dec 10 17:15:47 2011
ALTER DATABASE OPEN
Sat Dec 10 17:15:47 2011
Beginning crash recovery of 1 threads
Sat Dec 10 17:15:47 2011
Started redo scan
Sat Dec 10 17:15:48 2011
Completed redo scan
 319 redo blocks read, 98 data blocks need recovery
Sat Dec 10 17:15:50 2011
Started redo application at
 Thread 1: logseq 24, block 3
Sat Dec 10 17:15:50 2011
Recovery of Online Redo Log: Thread 1 Group 3 Seq 24 Reading mem 0
  Mem# 0 errs 0: /u02/ezhou/redo03.log
Sat Dec 10 17:15:50 2011
Completed redo application
Sat Dec 10 17:15:51 2011
Completed crash recovery at
 Thread 1: logseq 24, block 322, scn 6168722
 98 data blocks read, 98 data blocks written, 319 redo blocks read
Sat Dec 10 17:15:51 2011
LGWR: STARTING ARCH PROCESSES
ARC1 started with pid=17, OS id=5645
ARC0 started with pid=16, OS id=5643
ARC3 started with pid=19, OS id=5649
ARC4 started with pid=20, OS id=5651
ARC2 started with pid=18, OS id=5647
ARC6 started with pid=22, OS id=5655
ARC7 started with pid=23, OS id=5657
ARC5 started with pid=21, OS id=5653
ARC8 started with pid=24, OS id=5659
Sat Dec 10 17:15:52 2011
ARC0: Archival started
ARC1: Archival started
ARC2: Archival started
ARC3: Archival started
ARC4: Archival started
ARC5: Archival started
ARC6: Archival started
ARC7: Archival started
ARC8: Archival started
ARC9: Archival started
LGWR: STARTING ARCH PROCESSES COMPLETE
ARC9 started with pid=25, OS id=5661
Sat Dec 10 17:15:52 2011
Thread 1 advanced to log sequence 25
Sat Dec 10 17:15:53 2011
ARC2: STARTING ARCH PROCESSES
Sat Dec 10 17:15:53 2011
ARC6: Becoming the 'no FAL' ARCH
ARC6: Becoming the 'no SRL' ARCH
Sat Dec 10 17:15:53 2011
ARC3: Becoming the heartbeat ARCH
Sat Dec 10 17:15:53 2011
Thread 1 opened at log sequence 25
  Current log# 1 seq# 25 mem# 0: /u02/ezhou/redo01.log
  Current log# 1 seq# 25 mem# 1: /u02/ezhou/redo01a.rdo
Successful open of redo thread 1
Sat Dec 10 17:15:53 2011
SMON: enabling cache recovery
Sat Dec 10 17:15:54 2011
ARCa: Archival started
ARC2: STARTING ARCH PROCESSES COMPLETE
ARCa started with pid=26, OS id=5663
Sat Dec 10 17:15:57 2011
Successfully onlined Undo Tablespace 1.
Sat Dec 10 17:15:57 2011
SMON: enabling tx recovery
Sat Dec 10 17:15:57 2011
Database Characterset is AL32UTF8
replication_dependency_tracking turned off (no async multimaster replication found)
Starting background process QMNC
QMNC started with pid=27, OS id=5666
Sat Dec 10 17:16:13 2011
Errors in file /u01/pp/oracle/admin/ezhou/bdump/ezhou_smon_5547.trc:
ORA-00600: internal error code, arguments: [4194], [30], [27], [], [], [], [], []
Sat Dec 10 17:16:17 2011
Completed: ALTER DATABASE OPEN
Sat Dec 10 17:16:27 2011
Doing block recovery for file 2 block 4124
Block recovery from logseq 25, block 68 to scn 6168829
Sat Dec 10 17:16:27 2011
Recovery of Online Redo Log: Thread 1 Group 1 Seq 25 Reading mem 0
  Mem# 0 errs 0: /u02/ezhou/redo01.log
  Mem# 1 errs 0: /u02/ezhou/redo01a.rdo
Block recovery stopped at EOT rba 25.126.16
Block recovery completed at rba 25.126.16, scn 0.6168829
Doing block recovery for file 2 block 73
Block recovery from logseq 25, block 68 to scn 6168786
Sat Dec 10 17:16:28 2011
Recovery of Online Redo Log: Thread 1 Group 1 Seq 25 Reading mem 0
  Mem# 0 errs 0: /u02/ezhou/redo01.log
  Mem# 1 errs 0: /u02/ezhou/redo01a.rdo
Block recovery completed at rba 25.69.16, scn 0.6168789
Sat Dec 10 17:16:28 2011
Errors in file /u01/pp/oracle/admin/ezhou/bdump/ezhou_smon_5547.trc:
ORA-01595: error freeing extent (2) of rollback segment (5))
ORA-00607: Internal error occurred while making a change to a data block
ORA-00600: internal error code, arguments: [4194], [30], [27], [], [], [], [], []
Sat Dec 10 17:16:30 2011
Errors in file /u01/pp/oracle/admin/ezhou/bdump/ezhou_j002_5690.trc:
ORA-00600: internal error code, arguments: [4194], [30], [27], [], [], [], [], []
Sat Dec 10 17:16:37 2011
Doing block recovery for file 2 block 4124
Block recovery from logseq 25, block 68 to scn 6168829
Sat Dec 10 17:16:37 2011
Recovery of Online Redo Log: Thread 1 Group 1 Seq 25 Reading mem 0
  Mem# 0 errs 0: /u02/ezhou/redo01.log
  Mem# 1 errs 0: /u02/ezhou/redo01a.rdo
Block recovery completed at rba 25.126.16, scn 0.6168830
Doing block recovery for file 2 block 73
Block recovery from logseq 25, block 68 to scn 6168841
Sat Dec 10 17:16:37 2011
Recovery of Online Redo Log: Thread 1 Group 1 Seq 25 Reading mem 0
  Mem# 0 errs 0: /u02/ezhou/redo01.log
  Mem# 1 errs 0: /u02/ezhou/redo01a.rdo
Block recovery completed at rba 25.149.16, scn 0.6168843
Sat Dec 10 17:16:37 2011
Errors in file /u01/pp/oracle/admin/ezhou/bdump/ezhou_j002_5690.trc:
ORA-12012: error on auto execute of job 8886
ORA-00607: Internal error occurred while making a change to a data block
Sat Dec 10 17:16:41 2011
Errors in file /u01/pp/oracle/admin/ezhou/bdump/ezhou_j003_5692.trc:
ORA-00600: internal error code, arguments: [4194], [30], [27], [], [], [], [], []
Sat Dec 10 17:16:42 2011
DEBUG: Replaying xcb 0x32a2b17c, pmd 0x32bdbd24 for failed op 8
Doing block recovery for file 2 block 4124
Block recovery from logseq 25, block 68 to scn 6168829
Sat Dec 10 17:16:42 2011
Recovery of Online Redo Log: Thread 1 Group 1 Seq 25 Reading mem 0
  Mem# 0 errs 0: /u02/ezhou/redo01.log
  Mem# 1 errs 0: /u02/ezhou/redo01a.rdo
Block recovery completed at rba 25.126.16, scn 0.6168830
Sat Dec 10 17:16:43 2011
Errors in file /u01/pp/oracle/admin/ezhou/bdump/ezhou_j003_5692.trc:
ORA-00600: internal error code, arguments: [4194], [30], [27], [], [], [], [], []
ORA-00600: internal error code, arguments: [4194], [30], [27], [], [], [], [], []
Sat Dec 10 17:16:46 2011
Errors in file /u01/pp/oracle/admin/ezhou/bdump/ezhou_j003_5692.trc:
ORA-00600: internal error code, arguments: [4194], [30], [27], [], [], [], [], []
ORA-00600: internal error code, arguments: [4194], [30], [27], [], [], [], [], []
Sat Dec 10 17:17:46 2011
DEBUG: Replaying xcb 0x32a2b17c, pmd 0x32bdbd24 for failed op 8
Doing block recovery for file 2 block 4124
Block recovery from logseq 25, block 68 to scn 6168829
Sat Dec 10 17:17:46 2011
Recovery of Online Redo Log: Thread 1 Group 1 Seq 25 Reading mem 0
  Mem# 0 errs 0: /u02/ezhou/redo01.log
  Mem# 1 errs 0: /u02/ezhou/redo01a.rdo
Block recovery completed at rba 25.126.16, scn 0.6168830
Sat Dec 10 17:17:48 2011
Errors in file /u01/pp/oracle/admin/ezhou/bdump/ezhou_pmon_5535.trc:
ORA-00600: internal error code, arguments: [4194], [30], [27], [], [], [], [], []
Sat Dec 10 17:17:49 2011
Errors in file /u01/pp/oracle/admin/ezhou/bdump/ezhou_pmon_5535.trc:
ORA-00600: internal error code, arguments: [4194], [30], [27], [], [], [], [], []
PMON: terminating instance due to error 472
Instance terminated by PMON, pid = 5535

二、MOS记录

ERROR:              
  ORA-600 [4194] [a] [b]
 
VERSIONS:           
  versions 6.0 to 10.1 

DESCRIPTION:

  A mismatch has been detected between Redo records and rollback (Undo) 
  records.

  We are validating the Undo record number relating to the change being 
  applied against the maximum undo record number recorded in the undo block.

  This error is reported when the validation fails.

ARGUMENTS:
  Arg [a] Maximum Undo record number in Undo block
  Arg [b] Undo record number from Redo block

三、解决办法
1、修改参数
undo_management= MANUAL
undo_tablespace= SYSTEM
2、打开数据库，删除当前undo空间，重建新undo空间
3、修改参数
undo_management= AUTO
undo_tablespace= UNDOTBSNEW
4、重新启动数据库

in/exists和not in/not exists语意探讨

Posted on by

本篇只讨论in/exists和not in/not exists语意，不涉及这些写法的执行效率问题，至于效率问题请见:in/exists和not in/not exists执行效率
1、准备实验环境

C:\Users\XIFENFEI>sqlplus chf_xff/xifenfei

SQL*Plus: Release 11.2.0.1.0 Production on 星期六 12月 10 14:55:14 2011

Copyright (c) 1982, 2010, Oracle.  All rights reserved.


连接到:
Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - Production
With the Partitioning, Oracle Label Security, OLAP, Data Mining,
Oracle Database Vault and Real Application Testing options

SQL> create table t1 (c1 number,c2 number);

表已创建。

SQL> create table t2 (c1 number,c2 number);

表已创建。

SQL> insert into t1 values (1,2);

已创建 1 行。

SQL> insert into t1 values (1,3);

已创建 1 行。

SQL> insert into t2 values (1,2);

已创建 1 行。

SQL> insert into t2 values (1,null);

已创建 1 行。

SQL> insert into t1 values (1,null);

已创建 1 行。

SQL> commit;

提交完成。

SQL> select * from t1;

        C1         C2
---------- ----------
         1          2
         1          3
         1

SQL> select * from t2;

        C1         C2
---------- ----------
         1          2
         1

Note:t1和t2表都有null，且t1比t2多一条记录

2、t2做内部表

SQL> select * from t1 where c2 in (select c2 from t2 );

        C1         C2
---------- ----------
         1          2

SQL> select * from t1 where exists (select c2 from t2 where t1.c2=t2.c2);

        C1         C2
---------- ----------
         1          2

SQL> select * from t1 where c2 not in (select c2 from t2 );

未选定行

SQL> select * from t1 where not exists (select 1 from t2 where t1.c2=t2.c2);

        C1         C2
---------- ----------
         1          3
         1

3、t1为内部表

SQL> select * from t2 where c2 in (select c2 from t1 );

        C1         C2
---------- ----------
         1          2

SQL> select * from t2 where exists (select c2 from t1 where t1.c2=t2.c2);

        C1         C2
---------- ----------
         1          2

SQL> select * from t2 where c2 not in (select c2 from t1 );

未选定行

SQL> select * from t2 where not exists (select 1 from t1 where t1.c2=t2.c2);

        C1         C2
---------- ----------
         1

3、结论
in和exists结果相同(都会排除掉null,无论内部表中有无null)
not in会过滤掉外部表中的null(即使内部表中无null)
not exists不会过滤掉外部表的null(即使内部表有null)
由于篇幅关系,括号中的部分实验过程未展现出来

回收dba中alter system处理方法

Posted on by

今天在pub上看到一个问题，一个朋友想回收dba的alter system权限，直接回收这个系统权限从dba的做法是不推荐使用，因为修改了系统默认的dba角色所具有的系统权限，可能会导致未知的后果。好的做法是创建新的角色，使其有dba中除alter system之外的所有权限。
1、数据库版本

SQL> select * from v$version;

BANNER
-------------------------------------------------------------------------

Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - Production
PL/SQL Release 11.2.0.1.0 - Production
CORE    11.2.0.1.0      Production
TNS for 32-bit Windows: Version 11.2.0.1.0 - Production
NLSRTL Version 11.2.0.1.0 - Production

2、创建角色并授权

SQL> create role xifenfei;

角色已创建。

SQL> declare
  2  cursor cur is select privilege from role_sys_privs where role='DBA'
  3  AND PRIVILEGE NOT IN ('ALTER SYSTEM','ADMINISTER RESOURCE MANAGER');
  4  BEGIN
  5    FOR C in cur loop
  6    EXECUTE IMMEDIATE  'grant '||c.privilege||' to xifenfei';
  7    END loop;
  8  END;
  9  /

PL/SQL 过程已成功完成。

SQL> SELECT PRIVILEGE FROM ROLE_SYS_PRIVS
  2  WHERE ROLE ='DBA' AND PRIVILEGE NOT IN(
  3  SELECT PRIVILEGE FROM ROLE_SYS_PRIVS
  4  WHERE ROLE ='XIFENFEI'
  5  );

PRIVILEGE
----------------------------------------
ALTER SYSTEM
ADMINISTER RESOURCE MANAGER

说明：授予创建角色出ALTER SYSTEM和ADMINISTER RESOURCE MANAGER系统权限之外的所有权限

3、创建用户并授权角色

SQL> create user chf_xff identified by xifenfei;

用户已创建。

SQL> grant xifenfei to chf_xff;

授权成功。

4、单独授予ADMINISTER RESOURCE MANAGER权限

SQL> exec dbms_resource_manager_privs.grant_system_privilege(
   2 grantee_name => 'CHF_XFF',admin_option => false);

PL/SQL 过程已成功完成。

SQL> CONN chf_xff/xifenfei
已连接。

SQL> SELECT * FROM SESSION_PRIVS WHERE
  2  PRIVILEGE ='ADMINISTER RESOURCE MANAGER';

PRIVILEGE
----------------------------------------
ADMINISTER RESOURCE MANAGER

说明:
1)通过授权xifenfei角色和ADMINISTER RESOURCE MANAGER权限，完成回收dba中的alter system权限要求。
2)如果只有个别用户有这样的需求，那么可以直接生成批量授权语句实现，而不用建立类此xifenfei这样的角色。

5、为何单独授予ADMINISTER RESOURCE MANAGER权限

SQL> CONN / AS SYSDBA
已连接。
SQL> GRANT ADMINISTER RESOURCE MANAGER TO CHF_XFF;
GRANT ADMINISTER RESOURCE MANAGER TO CHF_XFF
      *
第 1 行出现错误:
ORA-00990: 权限缺失或无效

说明：ADMINISTER RESOURCE MANAGER这个系统权限在10g及其以后版本中，就不能直接使用GRANT/REVOKE直接授权/回收权限,而必面使用dbms_resource_manager_privs.grant_system_privilege和revoke_system_privilege过程进行处理。