请教:ORA-00603: ORACLE server session terminated by fatal error

曾经醒悟

oracle10g

磁盘空间100%
oracle宕机!
清除文件(与oracle无关的文件)
启动oracle提示:ORA-00603: ORACLE server session terminated by fatal error
下面是:alert_orcl.log错误

Fri Nov 10 12:02:35 2006
PMON: terminating instance due to error 472
Instance terminated by PMON, pid = 4005
Mon Nov 13 15:45:23 2006
Starting ORACLE instance (normal)
Cannot determine all dependent dynamic libraries for /proc/self/exe
Unable to find dynamic library libocr10.so in search paths
RPATH = /ade/aime1_build2101/oracle/has/lib/:/ade/aime1_build2101/oracle/lib/:/ade/aime1_build2101/oracle/has/lib/:
LD_LIBRARY_PATH = :ORACLE_HOME/lib
The default library directories are /lib and /usr/lib
Unable to find dynamic library libocrb10.so in search paths
Unable to find dynamic library libocrutl10.so in search paths
Unable to find dynamic library libocrutl10.so in search paths
LICENSE_MAX_SESSION = 0
LICENSE_SESSIONS_WARNING = 0
Picked latch-free SCN scheme 2
Using LOG_ARCHIVE_DEST_10 parameter default value as USE_DB_RECOVERY_FILE_DEST
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       = 75497472
  __large_pool_size        = 4194304
  __java_pool_size         = 4194304
  __streams_pool_size      = 0
  sga_target               = 167772160
  control_files            = /usr/u01/app/oracle/oradata/orcl/control01.ctl, /usr/u01/app/oracle/oradata/orcl/control02.ctl, /usr/u01/app/oracle/oradata/orcl/control03.ctl
  db_block_size            = 8192
  __db_cache_size          = 79691776
  compatible               = 10.2.0.1.0
  db_file_multiblock_read_count= 16
  db_recovery_file_dest    = /usr/u01/app/oracle/flash_recovery_area
  db_recovery_file_dest_size= 2147483648
  undo_management          = AUTO
  undo_tablespace          = UNDOTBS1
  remote_login_passwordfile= EXCLUSIVE
  db_domain                =
  dispatchers              = (PROTOCOL=TCP) (SERVICE=orclXDB)
  job_queue_processes      = 10
  background_dump_dest     = /usr/u01/app/oracle/admin/orcl/bdump
  user_dump_dest           = /usr/u01/app/oracle/admin/orcl/udump
  core_dump_dest           = /usr/u01/app/oracle/admin/orcl/cdump
  audit_file_dest          = /usr/u01/app/oracle/admin/orcl/adump
  db_name                  = orcl
  open_cursors             = 300
  pga_aggregate_target     = 16777216
PMON started with pid=2, OS id=13191
PSP0 started with pid=3, OS id=13193
MMAN started with pid=4, OS id=13195
DBW0 started with pid=5, OS id=13197
LGWR started with pid=6, OS id=13199
CKPT started with pid=7, OS id=13201
SMON started with pid=8, OS id=13203
RECO started with pid=9, OS id=13205
CJQ0 started with pid=10, OS id=13207
MMON started with pid=11, OS id=13209
MMNL started with pid=12, OS id=13211
Mon Nov 13 15:45:24 2006
starting up 1 dispatcher(s) for network address '(ADDRESS=(PARTIAL=YES)(PROTOCOL=TCP))'...
starting up 1 shared server(s) ...
Mon Nov 13 15:45:24 2006
ALTER DATABASE   MOUNT
Mon Nov 13 15:45:28 2006
Setting recovery target incarnation to 2
Mon Nov 13 15:45:28 2006
Successful mount of redo thread 1, with mount id 1134203924
Mon Nov 13 15:45:28 2006
Database mounted in Exclusive Mode
Completed: ALTER DATABASE   MOUNT
Mon Nov 13 15:45:28 2006
ALTER DATABASE OPEN
Mon Nov 13 15:45:28 2006
Beginning crash recovery of 1 threads
Mon Nov 13 15:45:28 2006
Started redo scan
Mon Nov 13 15:45:29 2006
Completed redo scan
72 redo blocks read, 23 data blocks need recovery
Mon Nov 13 15:45:29 2006
Started redo application at
Thread 1: logseq 126, block 3
Mon Nov 13 15:45:29 2006
Recovery of Online Redo Log: Thread 1 Group 2 Seq 126 Reading mem 0
  Mem# 0 errs 0: /usr/u01/app/oracle/oradata/orcl/redo02.log
Mon Nov 13 15:45:29 2006
Completed redo application
Mon Nov 13 15:45:29 2006
Completed crash recovery at
Thread 1: logseq 126, block 75, scn 4269705
23 data blocks read, 23 data blocks written, 72 redo blocks read
Mon Nov 13 15:45:29 2006
Thread 1 advanced to log sequence 127
Thread 1 opened at log sequence 127
  Current log# 3 seq# 127 mem# 0: /usr/u01/app/oracle/oradata/orcl/redo03.log
Successful open of redo thread 1
Mon Nov 13 15:45:29 2006
MTTR advisory is disabled because FAST_START_MTTR_TARGET is not set
Mon Nov 13 15:45:29 2006
SMON: enabling cache recovery
Mon Nov 13 15:45:31 2006
Successfully onlined Undo Tablespace 1.
Mon Nov 13 15:45:31 2006
SMON: enabling tx recovery
Mon Nov 13 15:45:31 2006
Database Characterset is ZHS16GBK
Mon Nov 13 15:45:33 2006
Errors in file /usr/u01/app/oracle/admin/orcl/udump/orcl_ora_13217.trc:
ORA-00600: internal error code, arguments: [4194], [46], [45], [], [], [], [], []
Doing block recovery for file 2 block 579
Block recovery from logseq 127, block 63 to scn 4269861
Mon Nov 13 15:45:37 2006
Recovery of Online Redo Log: Thread 1 Group 3 Seq 127 Reading mem 0
  Mem# 0 errs 0: /usr/u01/app/oracle/oradata/orcl/redo03.log
Block recovery stopped at EOT rba 127.65.16
Block recovery completed at rba 127.65.16, scn 0.4269859
Doing block recovery for file 2 block 153
Block recovery from logseq 127, block 63 to scn 4269858
Mon Nov 13 15:45:37 2006
Recovery of Online Redo Log: Thread 1 Group 3 Seq 127 Reading mem 0
  Mem# 0 errs 0: /usr/u01/app/oracle/oradata/orcl/redo03.log
Block recovery completed at rba 127.65.16, scn 0.4269859
Mon Nov 13 15:45:37 2006
Errors in file /usr/u01/app/oracle/admin/orcl/udump/orcl_ora_13217.trc:
ORA-00600: internal error code, arguments: [4194], [24], [23], [], [], [], [], []
DEBUG: Replaying xcb 0x288d5904, pmd 0x28a54c7c for failed op 8
Doing block recovery for file 2 block 934
No block recovery was needed
Mon Nov 13 15:45:38 2006
Errors in file /usr/u01/app/oracle/admin/orcl/bdump/orcl_smon_13203.trc:
ORA-00600: internal error code, arguments: [4097], [], [], [], [], [], [], []
Mon Nov 13 15:45:39 2006
Non-fatal internal error happenned while SMON was doing logging scn->time mapping.
SMON encountered 1 out of maximum 100 non-fatal internal errors.
Mon Nov 13 15:46:29 2006
Errors in file /usr/u01/app/oracle/admin/orcl/udump/orcl_ora_13217.trc:
ORA-00600: internal error code, arguments: [4194], [24], [23], [], [], [], [], []
ORA-00600: internal error code, arguments: [4194], [24], [23], [], [], [], [], []
Mon Nov 13 15:46:29 2006
DEBUG: Replaying xcb 0x288d5904, pmd 0x28a54c7c for failed op 8
Doing block recovery for file 2 block 934
No block recovery was needed
Mon Nov 13 15:46:30 2006
Errors in file /usr/u01/app/oracle/admin/orcl/udump/orcl_ora_13217.trc:
ORA-00600: internal error code, arguments: [4194], [24], [23], [], [], [], [], []
ORA-00600: internal error code, arguments: [4194], [24], [23], [], [], [], [], []


请帮忙分析一下
先谢谢了!

[ 本帖最后由 曾经醒悟 于 2006-11-13 16:03 编辑 ]

managergh

我到metalink上帮你查了一下 上面写的东西 我给你贴出来

主题:  OERR: ORA 603 "ORACLE server session terminated by fatal error"
  文档 ID:  注释:18488.1 类型:  REFERENCE
  上次修订日期:  10-JUN-1999 状态:  PUBLISHED




Error:  ORA 603
Text:   ORACLE server session terminated by fatal error
-------------------------------------------------------------------------------
Cause:  An ORACLE server session is in an unrecoverable state.
Action: Login to ORACLE again so a new server session will be created

*** Important: The notes below are for experienced users - See Note 22080.1

Explanation:
        A severe error occurred so this session died.
        This is a fairly generic error that can be raised in many places
        so the error itself is of little use.       

Diagnosis:
        See the server side trace file and alert log. There is usually another
        error on the error stack to show why the session died.
        Eg: ORA 1092 - Instance terminated.
        The stack trace should show what we were doing when we died but this is
        often a side effect of some other session dying .

曾经醒悟

上去

numenhuang

这是因为redo log信息和undo segments信息不一致造成的，一般通过删除原有的undo tablespace，创建新的undo tablespace来解决。

1. 先做好备份
2. 在init_<sid>.ora中设置_corrupted_rollback_segments=...
    要获得自动回滚段的名称：
           可以通过 ： strings system01.dbf | grep _SYSSMU
           可能的结果是：
_SYSSMU1
_SYSSMU2
_SYSSMU3
_SYSSMU4
_SYSSMU5
_SYSSMU6
_SYSSMU7
_SYSSMU8
_SYSSMU9
          则_corrupted_rollback_segments=('_SYSSMU1','_SYSSMU2',...);－－－>写上所有的自动回滚段名称。

3. startup database
4. create new undo tablespace
5. alter system set undo_tablespace=your_new_undo_tablespace
6. drop old undo tablespace
7. shutdown database and restart database
8. make full database backup