Instance terminated by PMON ORACLE自动宕掉

Instance terminated by PMON

QMNC started with pid=18, OS id=10455
Wed Nov  4 13:08:51 2009
Completed: ALTER DATABASE OPEN
Wed Nov  4 13:08:51 2009
Errors in file /usr/oracle/admin/ultraeos/bdump/ultraeos_mmon_10439.trc:
ORA-00600: Message 600 not found; No message file for product=RDBMS, facility=ORA; arguments: [4193] [37] [34]
Wed Nov  4 13:18:53 2009
Process 0x0xa75c5a48 appears to be hung while dumping
Current time = 1725304580, process death time = 1725244347 interval = 60000
Attempting to kill process 0x0xa75c5a48 with OS pid = 10439
OSD kill succeeded for process 0xa75c5a48
DEBUG: Replaying xcb 0xa6edb2e8, pmd 0xa54f0eb4 for failed op 8
Doing block recovery for file 2 block 4427
No block recovery was needed
Wed Nov  4 13:22:38 2009
Errors in file /usr/oracle/admin/ultraeos/bdump/ultraeos_pmon_10421.trc:
ORA-00600: Message 600 not found; No message file for product=RDBMS, facility=ORA; arguments: [4193] [37] [34]
Wed Nov  4 13:22:39 2009
DEBUG: Replaying xcb 0xa6edb2e8, pmd 0xa54f0eb4 for failed op 8
Doing block recovery for file 2 block 4427
No block recovery was needed
Wed Nov  4 13:22:39 2009
Errors in file /usr/oracle/admin/ultraeos/bdump/ultraeos_pmon_10421.trc:
ORA-00600: Message 600 not found; No message file for product=RDBMS, facility=ORA; arguments: [4193] [37] [34]
Wed Nov  4 13:22:39 2009
PMON: terminating instance due to error 472
Instance terminated by PMON, pid = 10421

启实例后，跑点SQL，实例就自动关闭。

以下是ultraeos_pmon_10421.trc

/usr/oracle/admin/ultraeos/bdump/ultraeos_pmon_10421.trc
Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production
With the Partitioning, OLAP and Data Mining options
ORACLE_HOME = /usr/oracle/product/10g
System name:    Linux
Node name:      linux5domain
Release:        2.6.18-8.el5xen
Version:        #1 SMP Thu Mar 15 21:02:53 EDT 2007
Machine:        i686
Instance name: ultraeos
Redo thread mounted by this instance: 1
Oracle process number: 2
Unix process pid: 10421, image: (PMON)

*** 2009-11-04 13:18:53.880
*** SERVICE NAMESYS$BACKGROUND) 2009-11-04 13:18:53.880
*** SESSION ID555.1) 2009-11-04 13:18:53.880
*** 2009-11-04 13:18:53.880
Process 0x0xa75c5a48 appears to be hung while dumping
Current time = 1725304580, process death time = 1725244347 interval = 60000
Attempting to kill process 0x0xa75c5a48 with OS pid = 10439
OSD kill succeeded for process 0xa75c5a48
DEBUG: Replaying xcb 0xa6edb2e8, pmd 0xa54f0eb4 for failed op 8
DEBUG: Reconstructing undo block 0x80114b for xcb 0xa6edb2e8
Doing block recovery for file 2 block 4427
Block header before block recovery:
buffer tsn: 1 rdba: 0x0080114b (2/4427)
scn: 0x0000.02d7d1e9 seq: 0x01 flg: 0x04 tail: 0xd1e90201
frmt: 0x02 chkval: 0x2bac type: 0x02=KTU UNDO BLOCK
No block recovery was needed
  lrda: 0.0 hscn: 48740923.0 hsub-scn: 65535
Block image after block recovery (no redo applied):
buffer tsn: 1 rdba: 0x0080114b (2/4427)
scn: 0x0000.02d7d1e9 seq: 0x01 flg: 0x04 tail: 0xd1e90201
frmt: 0x02 chkval: 0x2bac type: 0x02=KTU UNDO BLOCK
Hex dump of block: st=0, typ_found=1

另外，每次实例启动后马上就就会有个报错：

Completed: ALTER DATABASE OPEN
Wed Nov  4 13:08:51 2009
Errors in file /usr/oracle/admin/ultraeos/bdump/ultraeos_mmon_10439.trc:

文件内容：

/usr/oracle/admin/ultraeos/bdump/ultraeos_mmon_10439.trc
Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production
With the Partitioning, OLAP and Data Mining options
ORACLE_HOME = /usr/oracle/product/10g
System name:    Linux
Node name:      linux5domain
Release:        2.6.18-8.el5xen
Version:        #1 SMP Thu Mar 15 21:02:53 EDT 2007
Machine:        i686
Instance name: ultraeos
Redo thread mounted by this instance: 1
Oracle process number: 11
Unix process pid: 10439, image: (MMON)

*** SERVICE NAMESYS$BACKGROUND) 2009-11-04 13:08:51.403
*** SESSION ID546.1) 2009-11-04 13:08:51.403
*** ktuc_diag_dmp: dump of current change vector
ktudb big redo: cnt: 7 um: 0xa54f1f2c uba: 0x0080114b.0022.2e
*** ktuc_diag_dmp: dump of block after image in hex
Dump of memory from 0x760DE000 to 0x760E0000

解决方法：

SQL> create undo tablespace tbs_undo datafile '/usr/oracle/oradata/ultraeos/undotbs02.dbf' size 300M ;

Tablespace created.

SQL> alter system set undo_tablespace='tbs_undo' scope=both;

System altered.

SQL> show parameter undo

NAME                                 TYPE        VALUE
------------------------------------ ----------- ------------------------------
undo_management                      string      AUTO
undo_retention                       integer     900
undo_tablespace                      string      tbs_undo