一台服務器昨天晚上的exp備份出現ora-24324錯誤﹐

并且從alert_prm.log中發現許多ora-00600和錯誤ORA-07445﹕

ORA-07445: exception encountered: core dump [kpofcr()+645] [SIGSEGV] [Address not mapped to object] [0x3A38343E] [] []
Fri Jan 12 10:25:28 2007
Errors in file /oracle/admin/prm/udump/prm_ora_24806.trc:
ORA-07445: exception encountered: core dump [kpofcr()+645] [SIGSEGV] [Address not mapped to object] [0x3A38343E] [] []
Fri Jan 12 10:26:49 2007
Errors in file /oracle/admin/prm/udump/prm_ora_24810.trc:
ORA-07445: exception encountered: core dump [kpofcr()+645] [SIGSEGV] [Address not mapped to object] [0x3A38343E] [] []
Fri Jan 12 10:37:08 2007
Errors in file /oracle/admin/prm/udump/prm_ora_24860.trc:
ORA-00600: internal error code, arguments: [17147], [0x9774DA40], [], [], [], [], [], []
Fri Jan 12 10:37:09 2007
Errors in file /oracle/admin/prm/udump/prm_ora_24860.trc:
ORA-00600: internal error code, arguments: [17147], [0x9774DA40], [], [], [], [], [], []
ORA-00600: internal error code, arguments: [17147], [0x9774DA40], [], [], [], [], [], []

通過trace文件﹐出現ORA-07445錯誤﹐ 發現是因exp中的SELECT /*+NESTED_TABLE_GET_REFS+*/ "PMS"."MPTM".* FROM "PMS"."MPTM"引起的。

而ORA-00600 [17147]錯誤﹕

Note: For additional ORA-600 related information please read Note 146580.1

PURPOSE:
This article discusses the internal error "ORA-600 [17147]", what
it means and possible actions. The information here is only applicable
to the versions listed and is provided only for guidance.

ERROR:
ORA-600 [17147] [a] [b] [c] [d] [e]

VERSIONS:
versions 7.0 to 9.2

DESCRIPTION:

Oracle has detected that the magic number in a memory chunk header has
been overwritten.

This is a heap (in memory) corruption and there is no underlying data
corruption.

The error may occur in the one of the process specific heaps
(the Call heap, PGA heap, or session heap) or in the shared heap (SGA).

FUNCTIONALITY:
HEAP MANAGEMENT

IMPACT:
PROCESS FAILURE
MEMORY CORRUPTION, POSSIBLE INSTANCE FAILURE.
NON DATA CORRUPTIVE - No underlying data corruption.

SUGGESTIONS:

Many are one time occurrances.

Please see if the problem is reproducable.

Have the hardware vendor check all system memory for possible errors.

For repeatable occurrances, please submit the trace
files and alert.log to Oracle Support Services for further analysis.

并且當我建立mptm_t1表﹐并單獨exp時沒問題。

create table mptm_t1 as select * from mptm where rownum<10000;

[oracle@paclx pms]$ exp pms/pms file=/backup/mptm.dmp tables=(mptm_t1)

Export: Release 9.2.0.8.0 - Production on 星期五 1月 12 13:30:15 2007

Copyright (c) 1982, 2002, Oracle Corporation. All rights reserved.

Connected to: Oracle9i Enterprise Edition Release 9.2.0.8.0 - Production
With the Partitioning, OLAP and Oracle Data Mining options
JServer Release 9.2.0.8.0 - Production
Export done in ZHT16BIG5 character set and AL16UTF16 NCHAR character set

About to export specified tables via Conventional Path ...
. . exporting table MPTM_T1 9999 rows exported
Export terminated successfully without warnings.
[oracle@paclx pms]$

但是當資料增加到100000時﹐腳報錯﹕

create table mptm_t1 as select * from mptm where rownum<100000;

. . exporting table MPTM_T1
EXP-00056: ORACLE error 600 encountered
ORA-00600: internal error code, arguments: [%s], [%s], [%s], [%s], [%s], [%s], [%s], [%s]
EXP-00056: ORACLE error 24324 encountered
ORA-24324: service handle not initialized
EXP-00056: ORACLE error 24324 encountered
ORA-24324: service handle not initialized
EXP-00056: ORACLE error 24324 encountered
ORA-24324: service handle not initialized
EXP-00056: ORACLE error 24324 encountered
ORA-24324: service handle not initialized
EXP-00000: Export terminated unsuccessfully

并且前端應用人員也反應﹐只要涉及到MPTM這個表的查詢﹐都會出錯﹐但如果在詢時對某些欄位加上trim后就不會出錯了。

從ora-00600[17147]錯誤來看﹐是內存出了問題。我首先想到的是Flushing the buffer cache﹕

ALTER SYSTEM SET EVENTS 'immediate trace name flush_cache';

然后叫前端人員前面出錯誤的語句﹐這時發現可以執了﹐并沒有出現錯誤信息。

接著我又對MPTM_T1表進行導出﹐OK﹐可以正常exp了。