EXP导出遇到ORA-00600 错误 kghstack_underflow_internal_1

EXP在导出表的过程中遇到 ORA-00600: 内部错误代码, 参数: [kghstack_underflow_internal_1], [0x7EB20A0], [kkocxj : predTailPP], [], [], [], [], []

问题重现如下：

 

SQL> host exp admin/admin file=c:\t.dmp tables=t

Export: Release 10.2.0.1.0 - Production on 星期三 11月 2 14:42:36 2011

Copyright (c) 1982, 2005, Oracle.  All rights reserved.

连接到: Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production
With the Partitioning, OLAP and Data Mining options
已导出 ZHS16GBK 字符集和 AL16UTF16 NCHAR 字符集

即将导出指定的表通过常规路径...
. . 正在导出表                               T
EXP-00008: 遇到 ORACLE 错误 600
ORA-00600: 内部错误代码, 参数: [kghstack_underflow_internal_1], [0x7EB20A0], [kk
ocxj : predTailPP], [], [], [], [], []
导出成功终止, 但出现警告。

查询了一下MOS，发现是Bug 5112856。

 

 

通过如下方法即可正常导出:

 

SQL> alter system set "_optimizer_join_sel_sanity_check"=true scope=memory;

 

系统已更改。

 

SQL> alter system set "_optimizer_cost_based_transformation"=on scope=memory;

 

系统已更改。

 

SQL> alter system set optimizer_mode='choose' scope=memory;

 

系统已更改。

 

SQL> host exp admin/admin file=c:\t.dmp tables=t

Export: Release 10.2.0.1.0 - Production on 星期三 11月 2 15:14:40 2011

Copyright (c) 1982, 2005, Oracle.  All rights reserved.

连接到: Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production
With the Partitioning, OLAP and Data Mining options
已导出 ZHS16GBK 字符集和 AL16UTF16 NCHAR 字符集

即将导出指定的表通过常规路径...
. . 正在导出表                               T导出了         151 行
成功终止导出, 没有出现警告。

 

ORACLE在11G中已经修复了这个BUG。

 

 

以下是此BUG的相关信息：

 

 

Bug 5112856: ORA-00600 [KGHSTACK_UNDERFLOW_INTERNAL_1], [XXX] [KKOCXJ : PREDTAILPP]

Bug 属性

类型	B - Defect	已在产品版本中修复	11.0
严重性	2 - Severe Loss of Service	产品版本	10.1.0.4.0
状态	80 - Development to Q/A	平台	59 - HP-UX PA-RISC (64-bit)
创建时间	22-Mar-2006	平台版本	-
更新时间	15-Sep-2011	基本 Bug	-
数据库版本	10.1.0.4.0
影响平台	Generic
产品源	Oracle

相关产品

产品线	Oracle Database Products	系列	Oracle Database
区域	Oracle Database	产品	5 - Oracle Server - Enterprise Edition

Hdr: 5112856 10.1.0.4.0 RDBMS 10.1.0.4.0 QRY OPTIMIZER PRODID-5 PORTID-59 ORA-600
Abstract: ORA-600 [KGHSTACK_UNDERFLOW_INTERNAL_1], [XXX] [KKOCXJ : PREDTAILPP]

*** 03/22/06 08:54 pm ***
TAR:
----

PROBLEM:
--------
ORA-600: internal error code, arguments: [kghstack_underflow_internal_1] ,
[0x8000000100292040], [kkocxj : predTailPP], [], [], [], [], []

Current SQL statement for this session:

select 'Attention : Le TBS '||tmax.TABLESPACE_NAME||' est proche de
maxextents!'
from
(select tablespace_name,max_extents v1 from dba_tablespaces
)tmax,
(select tablespace_name,sum(EXTENTS) v2 from dba_segments group by
tablespace_name
)tused
where tmax.TABLESPACE_NAME=tused.TABLESPACE_NAME
and tmax.TABLESPACE_NAME<>'TEMP'
and tmax.v1
DIAGNOSTIC ANALYSIS:
--------------------
Looks like heap corruption, but I see no other indication of such in ALERT
LOG. Unlike most kgh.c related issues, 
OERINM("kghstack_underflow_internal_1") seems to be more code related.

WORKAROUND:
-----------
unlnown

RELATED BUGS:
-------------
Bug.4600946
Bug.4208634/4148670
Bug.4144058
Bug.4098831
Bug.3946242
Bug.3940319/3803424
Bug.3935159/3803424
Bug.3897915
Bug.3803424
Bug.3341590/3277361
Bug.3274855
Bug.3232772/3241646

REPRODUCIBILITY:
----------------
Thought to be a one time

TEST CASE:
----------
n/a

STACK TRACE:
------------
ksedmp ksfdmp kgerinv kgeasnmierr kghstack_underflow_ internal
kghstack_free internal ksmfrs rpiswu2 kksLoadChild kkslod kglobld
kglobpn kglpim kglpin kksfbc kkspsc0 kksParseCursor opiosq0
kpooprx kpoal8 opiodr ttcpip opitsk opiino opiodr opidrv sou2o
main

SUPPORTING INFORMATION:
-----------------------
I will upload files

24 HOUR CONTACT INFORMATION FOR P1 BUGS:
----------------------------------------
n/a

DIAL-IN INFORMATION:
--------------------
n/a

IMPACT DATE:
------------
soon

*** 03/22/06 09:06 pm *** (CHG: Sta->16)
*** 03/22/06 09:06 pm ***
*** 03/23/06 04:14 am ***
*** 03/23/06 05:10 am ***
*** 03/23/06 10:18 pm *** (CHG: Sta->10 SubComp->QRY OPTIMIZER)
*** 03/23/06 10:18 pm ***
*** 03/24/06 05:18 am ***
ORA-600 can be reproduced with OPTIMIZER_COST_BASED_TRANSFORMATION = ON

1) OPTIMIZER_COST_BASED_TRANSFORMATION = "ON"

SQL> alter system flush shared_pool;

System  altered.

SQL> alter session set "_optimizer_cost_based_transformation" = on;

Session altered.

SQL> start /home/oracle/etc/sql/tbsl

from    dba_data_files
                        *
ERROR at line 14:
ORA-600: internal  error code, arguments: [kghstack_underflow_internal_1],
[0x800000010029A6A0],  [kkocxj :
predTailPP], [], [], [], [], []

2) OPTIMIZER_COST_BASED_TRANSFORMATION = "OFF"

SQL> alter session set "_optimizer_cost_based_transformation" = off;

Session altered.

SQL> start /home/oracle/etc/sql/tbsl;

TABLESPACE_N         BYTES_TOTAL      BYTES_FREE         MAXFREE  FREE_PCT   
 
      TO_ADD      MAX_NEXT   NB_NEXT
------------------------------------------------------------------------------
--
------------------------------------
PSDAT              2,146,435,072   2,145,894,400   2,145,894,400       100   

-2,145,759,232       204,800     10478
*** 03/24/06 06:07 am *** (CHG: Sta->16)
*** 03/24/06 06:07 am ***
Reproduced on


SQL> alter system set optimizer_mode='choose' scope=memory;
SQL> alter system set "_optimizer_join_sel_sanity_check"=false scope=memory;
SQL> alter system set "_optimizer_cost_based_transformation"=on scope=memory;

SQL> select 'Attention : Le TBS '||tmax.TABLESPACE_NAME||' est proche de
maxextents!'
  2  from
  3  (select tablespace_name,max_extents v1 from dba_tablespaces) tmax,
  4  (select tablespace_name,sum(EXTENTS) v2 from dba_segments group by
tablespace_name) tused
  5  where
  6  tmax.TABLESPACE_NAME=tused.TABLESPACE_NAME
  7  and tmax.TABLESPACE_NAME<>'TEMP'
  8* and tmax.v1SQL> /
(select tablespace_name,max_extents v1 from dba_tablespaces) tmax,
                                            *
ERROR at line 3:
ORA-600: internal error code, arguments: [kghstack_underflow_internal_1],
[0x800000010028C510], [kkocxj : predTailPP], [], [], [], [], []


Doesn't reproduce with

SQL> alter system set optimizer_mode='choose' scope=memory;
SQL> alter system set "_optimizer_join_sel_sanity_check"=true scope=memory;
SQL> alter system set "_optimizer_cost_based_transformation"=on scope=memory;

SQL> select 'Attention : Le TBS '||tmax.TABLESPACE_NAME||' est proche de
maxextents!'
  2  from
  3  (select tablespace_name,max_extents v1 from dba_tablespaces) tmax,
  4  (select tablespace_name,sum(EXTENTS) v2 from dba_segments group by
tablespace_name) tused
  5  where
  6  tmax.TABLESPACE_NAME=tused.TABLESPACE_NAME
  7  and tmax.TABLESPACE_NAME<>'TEMP'
  8  and tmax.v1
no rows selected


And doesn't reproduce either with

SQL> alter system set optimizer_mode='choose' scope=memory;
SQL> alter system set "_optimizer_join_sel_sanity_check"=false scope=memory;
SQL> alter system set "_optimizer_cost_based_transformation"=off
scope=memory;

SQL> select 'Attention : Le TBS '||tmax.TABLESPACE_NAME||' est proche de
maxextents!'
  2  from
  3  (select tablespace_name,max_extents v1 from dba_tablespaces) tmax,
  4  (select tablespace_name,sum(EXTENTS) v2 from dba_segments group by
tablespace_name) tused
  5  where
  6  tmax.TABLESPACE_NAME=tused.TABLESPACE_NAME
  7  and tmax.TABLESPACE_NAME<>'TEMP'
  8  and tmax.v1
no rows selected



So, it seems to be caused by a combination of
_OPTIMIZER_JOIN_SEL_SANITY_CHECK
and _OPTIMIZER_COST_BASED_TRANSFORMATION
*** 03/27/06 01:27 am *** (CHG: Sta->11)
*** 03/27/06 01:27 am ***
*** 03/27/06 03:27 am ***
*** 03/27/06 03:27 am ***
*** 03/27/06 07:21 am ***
*** 03/27/06 07:24 am ***
*** 03/27/06 07:53 am ***
*** 03/27/06 07:53 am ***
*** 03/30/06 08:04 am ***
REDISCOVERY INFORMATION:
If you get an error with similar symptoms (same function in
the error message and same call stack) then you are likely
to have the same problem.
WORKAROUND:
None
RELEASE NOTES:
]]None
*** 03/30/06 08:08 am ***
*** 03/30/06 08:08 am *** (CHG: Sta->80)
*** 03/31/06 03:20 am ***
RDBMS_MAIN_LINUX_060331 contains changes from mzait_bug-5112856
*** 04/03/06 05:16 am ***
RDBMS_MAIN_LINUXAMD64_060331 contains changes from mzait_bug-5112856
*** 04/24/06 12:05 pm ***
RDBMS_MAIN_LINUXIA64_060417.UIP contains changes from mzait_bug-5112856
*** 05/11/06 07:31 am ***
RDBMS_MAIN_LINUXIA64_060427 contains changes from mzait_bug-5112856
*** 05/19/06 04:58 pm ***
RDBMS_SAGE_LINUX_060520 contains changes from mzait_bug-5112856
*** 05/21/06 05:49 pm ***
RDBMS_SAGE_LINUX_060521 contains changes from mzait_bug-5112856
*** 12/05/06 07:58 am ***
*** 12/05/06 07:58 am ***
*** 12/28/06 11:20 pm ***
*** 06/18/07 03:15 am ***
*** 06/18/07 03:16 am ***
*** 06/30/07 11:41 am ***
*** 06/30/07 04:09 pm ***
*** 07/03/07 06:49 am ***
*** 07/03/07 06:51 am ***
*** 04/21/08 02:46 pm ***
Application Demo Services team filed a bug#6969364 and were told by apps
development to apply rdbms patch#5112856. I checked an ARU and it looks like
the backport was created for rdbms 10.2.0.3 only for AIX platform.

Is it possible to request a backport for Linux x86 for rdbms 10.2.0.3?
*** 04/28/08 05:19 pm ***
Could you please let us know who should file a backport request?
*** 05/08/08 09:25 am ***
*** 05/08/08 09:25 am ***
*** 08/12/08 03:07 pm ***
*** 08/12/08 03:15 pm ***
*** 08/19/08 01:47 am ***
*** 01/03/11 10:51 pm ***
*** 01/04/11 01:04 am ***
*** 09/09/11 08:43 am ***
*** 09/09/11 10:38 pm ***
*** 09/10/11 08:24 am ***
*** 09/11/11 07:09 am ***
*** 09/13/11 10:14 am ***
*** 09/14/11 07:23 am ***
*** 09/15/11 07:13 am ***