Oracle Database Discussions

My db was crash as GATHER_STATS_JOB was trying to gather some status from db instance, rise the error ora-07445 [kglophup+66] at 6:00 am.

I start it when I found it was down and works well till now.

-------------------------------------------------------------------------------------------------------------------------------------------
Sat Dec 22 06:00:31 中国标准时间 2012
Errors in file e:\oracle\product\10.2.0\admin\cdw\bdump\cdw_j001_24836.trc:
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [kglophup+66] [PC:0x58D4E34] [ADDR:0x127F873133] [UNABLE_TO_WRITE] []

Sat Dec 22 06:00:43 中国标准时间 2012
Errors in file e:\oracle\product\10.2.0\admin\cdw\bdump\cdw_pmon_3696.trc:
ORA-07445: 出现异常错误: 核心转储 [ACCESS_VIOLATION] [kglophup+66] [PC:0x58D4E34] [ADDR:0x127F873133] [UNABLE_TO_WRITE] []
---------------------------------------------------------------------------------------------------------------------------------------------

1.I don't know whether the schedule job of oracle(GATHER_STATS_JOB) could cause the crash of db instance, but it seems true from the trace file which means oracle kills himself..

2.I don't know whether there is something wrong do with the memory in hardware, I 'll check the status of hardware later

3.the db was just create one month ago, it was migrate from oracle10201 to oracle10205 by user exp and imp and without error in exp(imp) logs.
I don't know whether exp and imp method of migration was correct or not?
(1)"exp user/paswd file=" from oracle 10201
(2)"imp user/paswd file=" to oracle 10205

4.I do opatch 13460968 which is the 10205 bundle patch 14 for windows x64 based on 10205 to make the db less bug,but I don't know whether the bundle patch cause the crash of instance?

some information:
5.oracle10205 with "Patch 13460968: ORACLE 10G 10.2.0.5 PATCH 14 BUG FOR WINDOWS (64-BIT AMD64 AND INTEL EM64"
6.oracle 10205 on win2008 x64 r2
7.128GB RAM(70G for sga and 30G for pga)

trace file:
-----------------------------------------------------------------------------------------------------------
Dump file e:\oracle\product\10.2.0\admin\cdw\bdump\cdw_j001_24836.trc

racle Database 10g Enterprise Edition Release 10.2.0.5.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
Windows NT Version V6.1 Service Pack 1
CPU : 48 - type 8664, 24 Physical Cores
Process Affinity : 0x0000000000000000
Memory (Avail/Total): Ph:55481M/131058M, Ph+PgF:186456M/262115M
Instance name: cdw

Redo thread mounted by this instance: 1

Oracle process number: 43

Windows thread id: 24836, image: ORACLE.EXE (J001)


*** ACTION NAME:(GATHER_STATS_JOB) 2012-12-22 06:00:31.438
*** MODULE NAME:(DBMS_SCHEDULER) 2012-12-22 06:00:31.438
*** SERVICE NAME:(SYS$USERS) 2012-12-22 06:00:31.438
*** SESSION ID:(1041.23881) 2012-12-22 06:00:31.438
*** 2012-12-22 06:00:31.438
ksedmp: internal or fatal error
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [kglophup+66] [PC:0x58D4E34] [ADDR:0x127F873133] [UNABLE_TO_WRITE] []
Current SQL statement for this session:
select substrb(dump(val,16,0,32),1,120) ep, cnt from (select /*+ index_rs(t) no_parallel(t) no_parallel_index(t) dbms_stats cursor_sharing_exact use_weak_name_resl dynamic_sampling(0) no_monitoring */"INSTANCE_NUMBER" val,count(*) cnt from "SYS"."WRH$_DB_CACHE_ADVICE" t where TBL$OR$IDX$PART$NUM("SYS"."WRH$_DB_CACHE_ADVICE",0,4,0,"ROWID") = :objn and "INSTANCE_NUMBER" is not null group by "INSTANCE_NUMBER") order by val
----- PL/SQL Call Stack -----
object line object
handle number name
000000120DE28D48 12278 package body SYS.DBMS_STATS
000000120DE28D48 12868 package body SYS.DBMS_STATS
000000120DE28D48 14164 package body SYS.DBMS_STATS
000000120DE28D48 14463 package body SYS.DBMS_STATS
000000120DE28D48 14595 package body SYS.DBMS_STATS
000000120DE28D48 15121 package body SYS.DBMS_STATS
000000120DE28D48 21422 package body SYS.DBMS_STATS
check trace file e:\oracle\product\10.2.0\db_1\rdbms\trace\cdw_ora_0.trc for preloading .sym file messages
----- Call Stack Trace -----
calling call entry argument values in hex
location type point (? means dubious value)
-------------------- -------- -------------------- ----------------------------
kglophup+66 0000000000000000 000000000 0010385F2 000001FE8
000002000
kglHandleUnpin+159 CALL??? kglophup+66 000000001 01CFEF5D0 000000000
11D59D7B68
kglUnKeepHandle4Sta CALL??? kglHandleUnpin+159 D9EB95665868 000000000
ts+602 000000000 120DCA8040
kksFreeMarkedBktCur CALL??? kglUnKeepHandle4Sta 000000000 11C746C720
sorStat+960 ts+602 000000030 000000018
kksClearCursorStat+ CALL??? kksFreeMarkedBktCur 000000000 000000000 000000130
474 sorStat+960 01CFEF5D0
kksCheckToFreeCurso CALL??? kksClearCursorStat+ 00001F700 00001F5D8 02117D160
rStat+14 474 00001F5D8
kksLoadChild+15506 CALL??? kksCheckToFreeCurso 120435CD58 1100000000
rStat+14 11C7E3E3DC 000000024
kxsGetRuntimeLock+2 CALL??? kksLoadChild+15506 01CFF01D0 120BCE8908
016 02117E6F0 120BCE8908
kksfbc+18212 CALL??? kxsGetRuntimeLock+2 01CFF01D0 027A10058 02117E6F0
016 000000103
kkspsc0+2633 CALL??? kksfbc+18212 027A10058 000000003 000000108
0286F1E88
kksParseCursor+473 CALL??? kkspsc0+2633 01DD12870 0286F1E88 0000001AB
000000003
opiosq0+1867 CALL??? kksParseCursor+473 02117FC40 000000000 02117F900
----------------------------------------------------------------------------------------------------

another tracefile:
------------------------------------------------------------
Dump file e:\oracle\product\10.2.0\admin\cdw\bdump\cdw_pmon_3696.trc
Sat Dec 22 06:00:43 2012
ORACLE V10.2.0.5.0 - 64bit Production vsnsta=0
vsnsql=14 vsnxtr=3
Oracle Database 10g Enterprise Edition Release 10.2.0.5.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
Windows NT Version V6.1 Service Pack 1
CPU : 48 - type 8664, 24 Physical Cores
Process Affinity : 0x0000000000000000
Memory (Avail/Total): Ph:55515M/131058M, Ph+PgF:186489M/262115M
Instance name: cdw

Redo thread mounted by this instance: 1

Oracle process number: 2

Windows thread id: 3696, image: ORACLE.EXE (PMON)


*** 2012-12-22 06:00:43.107
*** SERVICE NAME:(SYS$BACKGROUND) 2012-12-22 06:00:43.107
*** SESSION ID:(1105.1) 2012-12-22 06:00:43.107
=================================================
KGL Child 20 Level 0 latch clean-up information:
=================================================
Operation = 18
kgllcpt1 = 000000120DCA7F10 kgllcpt2 = 00000011D7C1DDB8
kgllcpt3 = 000000120DCA8000 kgllcpt4 = 000000127F87312B
kgllcpt5 = 0000001210A7FB80 kgllcub4 = 35


Oracle Database 10g Enterprise Edition Release 10.2.0.5.0 - 64bit Production
Process Id: 0x0000068c Thread Id : 0x00000e70 Time : Sat Dec 22 06:00:43
Excp. Code: 0xc0000005 Excp. Type: ACCESS_VIO Flags: 0x00000000


------------------- Registers ----------------------------
ip=00000000058D4E34 sp=000000000F1BDF70 rp=0000000000000023
r1=0000000004BC8C80 r2=000000000519DC40 r3=0000000000000022
r4=0000000000000014 r5=000000000F1BDF70 r6=0000000000000023 r7=00000011D6B0A1E0
r8=000000000519DC40 r9=00000011D7C1DDB8 r10=000000120DCA7F10 r11=000000120DCA8000
r12=000000120DCA8000 r13=0000001210A7FB80 r14=000000127F87312B r15=000000000519DC40
------------------- End of Registers ---------------------


*** 2012-12-22 06:00:43.122
ksedmp: internal or fatal error
ORA-07445: 出现异常错误: 核心转储 [ACCESS_VIOLATION] [kglophup+66] [PC:0x58D4E34] [ADDR:0x127F873133] [UNABLE_TO_WRITE] []
check trace file e:\oracle\product\10.2.0\db_1\rdbms\trace\cdw_ora_0.trc for preloading .sym file messages
----- Call Stack Trace -----
calling call entry argument values in hex
location type point (? means dubious value)
-------------------- -------- -------------------- ----------------------------
kglophup+66 0000000000000000 000000000 000000000 000000000
000000000
kgllccl+731 CALL??? kglophup+66 00519DC40 11D6B0C8B0
000000014 000002580
kgllcu+408 CALL??? kgllccl+731 000000E70 0052E2037 000000540
0000005A0
kslilcr+926 CALL??? kgllcu+408 007BB08A0 1211387D68
003E74F14 11D59D7B68
ksl_cleanup+2010 CALL??? kslilcr+926 11D59D7B68 000000003
000000000 120524E700
ksuxfl+1021 CALL??? ksl_cleanup+2010 007BB08A0 000003CF0 0E9C2BFB4
000000000
ksuxda+66 CALL??? ksuxfl+1021 00000002B 1206250D30
000000001 0006D4A5A
ksucln+2423 CALL??? ksuxda+66 000000E70 0052E2037 000000540
000002000
ksbrdp+988 CALL??? ksucln+2423 004BCD730 004BCD74C 000000002
000000004
opirip+700 CALL??? ksbrdp+988 6F5C3A450000001E 003E481E8
00F1BFA10 000000000
opidrv+856 CALL??? opirip+700 000000032 000000004 00F1BFD50
000000000
sou2o+52 CALL??? opidrv+856 000000032 000000004 00F1BFD50
000000003
opimai_real+268 CALL??? sou2o+52 000000000 000000000 000000000
000000000
opimai+96 CALL??? opimai_real+268 000000000 000000000 000000000
000000000
-----------------------------------------------------------

if someone meet the same situation,do please let me know,great thanks!