ora-600 4194 错误一般是由于undo表空间损坏导致部分事务无法回滚，因而导致数据库无法启动或启动后又被PMON中止。

解决此问题关键是确定损坏的回滚段，一般报警日志中会记录损坏的回滚段，见红色字体。

如果数据库能够侥幸打开，我们也可以在打开的瞬间执行一个脚本来确定回滚段。

1、报警日志、重点关注红色字体

Mem# 0: /data/oradata/ecms/redo02.log

Block recovery completed at rba 11.69.16, scn 13.1485530026

Thu Mar 25 17:36:07 2010

Errors in file /u01/app/oracle/admin/ecms/bdump/ecms_smon_1722.trc:

ORA-01595: error freeing extent (75) of rollback segment (3))

ORA-00607: Internal error occurred while making a change to a data block

ORA-00600: internal error code, arguments: [4194], [19], [23], [], [], [], [], []

Thu Mar 25 17:37:07 2010

Errors in file /u01/app/oracle/admin/ecms/bdump/ecms_j000_2236.trc:

ORA-00600: internal error code, arguments: [4194], [19], [23], [], [], [], [], []

Thu Mar 25 17:37:08 2010

DEBUG: Replaying xcb 0x31621de4, pmd 0x3179c928 for failed op 8

Doing block recovery for file 2 block 58981

Block recovery from logseq 11, block 68 to scn 57320104895

Thu Mar 25 17:37:08 2010

Recovery of Online Redo Log: Thread 1 Group 2 Seq 11 Reading mem 0

  Mem# 0: /data/oradata/ecms/redo02.log

Block recovery completed at rba 11.84.16, scn 13.1485530049

Thu Mar 25 17:37:08 2010

Errors in file /u01/app/oracle/admin/ecms/bdump/ecms_j000_2236.trc:

ORA-00600: internal error code, arguments: [4194], [19], [23], [], [], [], [], []

ORA-00600: internal error code, arguments: [4194], [19], [23], [], [], [], [], []

Thu Mar 25 17:37:09 2010

Errors in file /u01/app/oracle/admin/ecms/bdump/ecms_j000_2236.trc:

ORA-00600: internal error code, arguments: [4194], [19], [23], [], [], [], [], []

ORA-00600: internal error code, arguments: [4194], [19], [23], [], [], [], [], []

Thu Mar 25 17:37:13 2010

DEBUG: Replaying xcb 0x31621de4, pmd 0x3179c928 for failed op 8

Doing block recovery for file 2 block 58981

Block recovery from logseq 11, block 68 to scn 57320104895

Thu Mar 25 17:37:13 2010

Recovery of Online Redo Log: Thread 1 Group 2 Seq 11 Reading mem 0

  Mem# 0: /data/oradata/ecms/redo02.log

Block recovery completed at rba 11.84.16, scn 13.1485530049

Thu Mar 25 17:37:13 2010

Errors in file /u01/app/oracle/admin/rcms/bdump/ecms_pmon_1710.trc:

ORA-00600: internal error code, arguments: [4194], [19], [23], [], [], [], [], []

Thu Mar 25 17:37:13 2010

Errors in file /u01/app/oracle/admin/rcms/bdump/ecms_pmon_1710.trc:

ORA-00600: internal error code, arguments: [4194], [19], [23], [], [], [], [], []

PMON: terminating instance due to error 472

Instance terminated by PMON, pid = 1710

2、确定损坏的回滚段

error freeing extent (75) of rollback segment (3))

根据该日志可以判断是第三个回滚段的第75个区损坏

Doing block recovery for file 2 block 58981

SELECT segment_name,segment_type,extent_id,block_id, blocks

from dba_extents t

where  file_id = 2

AND 58981 between block_id and (block_id + blocks - 1) ;

SEGMENT_NAME   SEGMENT_TYPE        EXTENT_ID   BLOCK_ID     BLOCKS

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

_SYSSMU3$      TYPE2 UNDO               74      58761       1024

根据那个错误日志基本可以判断是在读取file 2的58981数据块时出错的，可以执行上面sql确定损坏的回滚段

3、修改pfile如下

*.undo_management='MANUAL'

*._offline_rollback_segments='_SYSSMU57$'

*._corrupted_rollback_segments='_SYSSMU3$'

4、启动数据库，并重新创建undo表空间

$ !sql

sqlplus "/as sysdba"

SQL*Plus: Release 11.2.0.4.0 - Production on Fri Mar 26 13:46:28 2010

Copyright (c) 1982, 2007, Oracle.  All Rights Reserved.

Connected to an idle instance.

SQL> startup pfile='/tmp/a.txt';

ORACLE instance started.

Total System Global Area  314572800 bytes

Fixed Size                  1267212 bytes

Variable Size             188746228 bytes

Database Buffers          117440512 bytes

Redo Buffers                7118848 bytes

Database mounted.

Database opened.

SQL> create undo tablespace undotbs2 datafile '/data/oradata/ecms/undotbs02.dbf' size 2g;

Tablespace created.

SQL> drop tablespace undotbs1;

Tablespace dropped.

SQL> alter tablespace undotbs2 rename to undotbs1;

Tablespace altered.

SQL>

至此基本搞定