JAVA修改dmp版本号12g到11,又见bug，动态调整SGA参数导致数据库crash

最近遇到这样一个事情，在动态的调整sga的一个参数的时候，数据库就down了：

SQL> alter system set DB_KEEP_CACHE_SIZE=1g;

alter system set DB_KEEP_CACHE_SIZE=1g

ERROR at line 1:

ORA-03113: end-of-file on communication channel

SQL>

SQL>altersystemsetDB_KEEP_CACHE_SIZE=1g;

altersystemsetDB_KEEP_CACHE_SIZE=1g

ERRORatline1:

ORA-03113:end-of-fileoncommunicationchannel

SQL>

对应的，在alertlog中的报错为：

Sat Nov 13 03:56:00 2010

Errors in file /u01/app/oracle/product/10.2.0/admin/MICPRD/bdump/micprd_mman_3264732.trc:

ORA-00600: internal error code, arguments: [kmgs_update_target_size_1], [0], [], [], [], [], [], []

Sat Nov 13 03:56:01 2010

Errors in file /u01/app/oracle/product/10.2.0/admin/MICPRD/bdump/micprd_mman_3264732.trc:

ORA-00600: internal error code, arguments: [kmgs_update_target_size_1], [0], [], [], [], [], [], []

Sat Nov 13 03:56:01 2010

MMAN: terminating instance due to error 822

Sat Nov 13 03:56:01 2010

Errors in file /u01/app/oracle/product/10.2.0/admin/MICPRD/bdump/micprd_mmon_2969868.trc:

ORA-00600: internal error code, arguments: [kmgs_perform_parameter_updates_1], [0], [], [], [], [], [], []

Sat Nov 13 03:56:03 2010

Errors in file /u01/app/oracle/product/10.2.0/admin/MICPRD/bdump/micprd_mmon_2969868.trc:

ORA-00600: internal error code, arguments: [kmgs_parameter_update_timeout_1], [600], [], [], [], [], [], []

ORA-00600: internal error code, arguments: [kmgs_perform_parameter_updates_1], [0], [], [], [], [], [], []

Instance terminated by MMAN, pid = 3264732

SatNov1303:56:002010

Errorsinfile/u01/app/oracle/product/10.2.0/admin/MICPRD/bdump/micprd_mman_3264732.trc:

ORA-00600:internalerrorcode,arguments:[kmgs_update_target_size_1],[0],[],[],[],[],[],[]

SatNov1303:56:012010

Errorsinfile/u01/app/oracle/product/10.2.0/admin/MICPRD/bdump/micprd_mman_3264732.trc:

ORA-00600:internalerrorcode,arguments:[kmgs_update_target_size_1],[0],[],[],[],[],[],[]

SatNov1303:56:012010

MMAN:terminatinginstanceduetoerror822

SatNov1303:56:012010

Errorsinfile/u01/app/oracle/product/10.2.0/admin/MICPRD/bdump/micprd_mmon_2969868.trc:

ORA-00600:internalerrorcode,arguments:[kmgs_perform_parameter_updates_1],[0],[],[],[],[],[],[]

SatNov1303:56:032010

Errorsinfile/u01/app/oracle/product/10.2.0/admin/MICPRD/bdump/micprd_mmon_2969868.trc:

ORA-00600:internalerrorcode,arguments:[kmgs_parameter_update_timeout_1],[600],[],[],[],[],[],[]

ORA-00600:internalerrorcode,arguments:[kmgs_perform_parameter_updates_1],[0],[],[],[],[],[],[]

InstanceterminatedbyMMAN,pid=3264732

对应的tace文件中的call stack为：

----- Call Stack Trace -----

calling call entry argument values in hex

location type point (? means dubious value)

-------------------- -------- -------------------- ----------------------------

ksedst+001c bl ksedst1 000000000 ? FFFFFFFFFFFC7D4 ?

ksedmp+0290 bl ksedst 1048DFFC0 ?

ksfdmp+0018 bl 03F5B014

kgerinv+00dc bl _ptrgl

kgeasnmierr+004c bl kgerinv FFFFFFFFFFFCB00 ? 000000000 ?

101045830 ? 000000010 ?

FFFFFFFFFFFCBA0 ?

kmgs_update_target_ bl kgeasnmierr 110190110 ? 110440040 ?

size+006c 104A818A0 ? 100000001 ?

000000000 ? 000000000 ?

12E0BE826D694B2F ?

000000000 ?

kmgs_update_paramet bl kmgs_update_target_ 900000000083144 ?

er_ctx+0354 size FFFFFFFFFFFCDA0 ?

kmgs_update_param_m bl kmgs_update_paramet 000000002 ? FFFFFFFFFFFD468 ?

anual_helper+0078 er_ctx 010008000 ? 110011970 ?

kmgs_update_param_m bl kmgs_update_param_m 000000000 ? 000000001 ?

anual+0064 anual_helper 000000002 ?

kmgsdrv+16cc bl kmgs_update_param_m 9000000000F6F64 ?

anual

ksbabs+03a8 bl _ptrgl

ksbrdp+0408 bl _ptrgl

opirip+03fc bl 03F59B54

opidrv+0448 bl opirip 110296850 ? 410298190 ?

FFFFFFFFFFFF960 ?

sou2o+0090 bl opidrv 3202AB393C ? 4A0144924 ?

FFFFFFFFFFFF960 ?

opimai_real+0150 bl 01FC4EF4

main+0098 bl opimai_real 000000000 ? 000000000 ?

__start+0098 bl main 000000000 ? 000000000 ?

-----CallStackTrace-----

callingcallentryargumentvaluesinhex

locationtypepoint(?meansdubiousvalue)

----------------------------------------------------------------------------

ksedst+001cblksedst1000000000?FFFFFFFFFFFC7D4?

ksedmp+0290blksedst1048DFFC0?

ksfdmp+0018bl03F5B014

kgerinv+00dcbl_ptrgl

kgeasnmierr+004cblkgerinvFFFFFFFFFFFCB00?000000000?

101045830?000000010?

FFFFFFFFFFFCBA0?

kmgs_update_target_blkgeasnmierr110190110?110440040?

size+006c104A818A0?100000001?

000000000?000000000?

12E0BE826D694B2F?

000000000?

kmgs_update_parametblkmgs_update_target_900000000083144?

er_ctx+0354sizeFFFFFFFFFFFCDA0?

kmgs_update_param_mblkmgs_update_paramet000000002?FFFFFFFFFFFD468?

anual_helper+0078er_ctx010008000?110011970?

kmgs_update_param_mblkmgs_update_param_m000000000?000000001?

anual+0064anual_helper000000002?

kmgsdrv+16ccblkmgs_update_param_m9000000000F6F64?

anual

ksbabs+03a8bl_ptrgl

ksbrdp+0408bl_ptrgl

opirip+03fcbl03F59B54

opidrv+0448blopirip110296850?410298190?

FFFFFFFFFFFF960?

sou2o+0090blopidrv3202AB393C?4A0144924?

FFFFFFFFFFFF960?

opimai_real+0150bl01FC4EF4

main+0098blopimai_real000000000?000000000?

__start+0098blmain000000000?000000000?

经查metalink，应该是这以下的2个bug有关：

Bug 6737235: DATABASE WITH SGA_TARGET CRASH WHEN ALTER SGA COMPONENTS

Bug 6146397: ORA-00600 [KMGS_UPDATE_TARGET_SIZE_1], [0] INSTANCE TERMINATED BY MMAN

Bug6737235:DATABASEWITHSGA_TARGETCRASHWHENALTERSGACOMPONENTS

Bug6146397:ORA-00600[KMGS_UPDATE_TARGET_SIZE_1],[0]INSTANCETERMINATEDBYMMAN

我操作的数据库是aix上oracle 10.2.0.3数据库。目前这2个bug都还没有在10203上的fix，只是建议升级到10204或者用workaround的解决。workaround是用alter system scope=spfile的方式修改，再重启数据库生效。

这让我感觉10g上的自动内存管理不太靠谱，虽然自动内存管理可以动态调整buffer cache和其他几个pool如share pool之间的比例，但是如果一个系统是成熟的系统，很好的进行了绑定变量，那么就不太会发生各个池子之间比例的变动。如果连动态的调整一下sga中的一个组件，都会导致MMAN将数据库down掉的话，那oracle是不是太脆弱了呢？

另外，在查询关于这个ora-600的相关报错信息的时候，还查到了另一个有意思的bug：

MMON Terminates Instance With ORA-600 [kmgs_pre_process_request_6] or ORA-600 [kmgs_update_target_size_1] When Resizing Caches [ID 373802.1]

MMONTerminatesInstanceWithORA-600[kmgs_pre_process_request_6]orORA-600[kmgs_update_target_size_1]WhenResizingCaches[ID373802.1]

这个是Bug 4433838，影响10201至10203的所有平台，这个bug虽然和我的情况有点出入，文档说是MMON进程终止实例，我的情况是MMAN进程终止实例。但文档说触发这个bug的cause是sga的大小为4G的整数倍。呵呵，碰巧，我的sga_target和sga_max_size为12G，正是4G的倍数。而且ora-600之后的参数和trace文件中的信息也和我类似。

因此，综上，在10204以下的版本设置调整SGA的时候，需要注意以下几点：

1、SGA大小不要设置成4G的整数倍。

2、调整SGA内的组件的时候，不要动态的调整，利用scope=spfile再重启数据库进行修改。

附：

Bug 6737235: DATABASE WITH SGA_TARGET CRASH WHEN ALTER SGA COMPONENTS

显示 Bug 属性 Bug 属性

类型 B - Defect 已在产品版本中修复 -

严重性 2 - Severe Loss of Service 产品版本 10.2.0.3

状态 36 - Duplicate Bug. To Filer 平台 23 - Oracle Solaris on SPARC (64-bit)

创建时间 08-Jan-2008 平台版本 -

更新时间 11-Aug-2009 基本 Bug 6192679

数据库版本 10.2.0.3

影响平台 Generic

产品源 Oracle

显示相关产品相关产品

产品线 Oracle Database Products 系列 Oracle Database

区域 Oracle Database 产品 5 - Oracle Server - Enterprise Edition

Hdr: 6737235 10.2.0.3 RDBMS 10.2.0.3 MEMORY MGMT PRODID-5 PORTID-23 6192679

Abstract: DATABASE WITH SGA_TARGET CRASH WHEN ALTER SGA COMPONENTS

*** 01/08/08 10:47 pm ***

TAR:

----

PROBLEM:

--------

The parameter setting are as follows:

NAME TYPE VALUE

------------------------------------ ----------- ---------

shared_pool_size big integer 608M

sga_max_size big integer 12G

sga_target big integer 12G

db_cache_size big integer 9G

*.java_pool_size=209715200

*.large_pool_size=209715200

*.streams_pool_size=50331648

DEFAULT buffer cache in the v$sga_dynamic_components shows the below value :

CURRENT_SIZE MIN_SIZE USER_SPECIFIED_SIZE

------------ -------- -------------------

1.1744E+10 1.1744E+10 9663676416

When try to increase the shared_pool_size to 700M :

SQL> alter system set shared_pool_size=700M;

alter system set shared_pool_size=700M

ERROR at line 1:

ORA-2097: parameter cannot be modified because specified value is

invalid

ORA-4033: Insufficient memory to grow pool

SQL> select name from v$database;

select name from v$database

ERROR at line 1:

ORA-3135: connection lost contact

After the above error message the database crashes.

below is the message in the alert log :

Errors in file

/u01/app/oracle/admin/LSTDDB1/bdump/lstddb1_mman_12427.trc:

ORA-600: internal error code, arguments: [kmgs_update_target_size_1],

[0], [], [], [], [], [], []

When try to decrease the size of the default buffer cache getting:

SQL> alter system set db_cache_size=10G;

alter system set db_cache_size=10G

ERROR at line 1:

ORA-3113: end-of-file on communication channel

and the database crashes.

the alert shows the following:

Errors in file /u01/app/oracle/admin/LSTDDB1/bdump/lstddb1_mman_11825.trc:

ORA-600: internal error code, arguments: [kmgs_pre_process_request_6], [6],

[700], [576], [3], [0x67C7B9398], [], []

DIAGNOSTIC ANALYSIS:

--------------------

WORKAROUND:

-----------

None

RELATED BUGS:

-------------

Bug 6192679

REPRODUCIBILITY:

----------------

Everytime at customer site

TEST CASE:

----------

STACK TRACE:

------------

SUPPORTING INFORMATION:

-----------------------

24 HOUR CONTACT INFORMATION FOR P1 BUGS:

----------------------------------------

DIAL-IN INFORMATION:

--------------------

IMPACT DATE:

------------

*** 01/08/08 10:49 pm ***

*** 01/08/08 10:53 pm ***

*** 01/08/08 10:53 pm *** (CHG: Sta->16)

*** 01/09/08 07:51 am *** (CHG: Asg->NEW OWNER OWNER)

*** 01/09/08 08:53 am *** (CHG: Sta->10 SubComp->MEMORY MGMT)

*** 01/09/08 08:53 am ***

*** 01/10/08 08:20 pm ***

*** 01/10/08 08:20 pm *** (CHG: Sta->16)

*** 01/11/08 08:33 am *** (CHG: Sta->10)

*** 01/11/08 08:33 am ***

*** 01/28/08 10:43 pm ***

*** 01/28/08 10:43 pm *** (CHG: Sta->16)

*** 01/29/08 06:17 am *** (CHG: Sta->10)

*** 01/29/08 06:17 am ***

*** 01/29/08 06:42 am *** (CHG: Sta->16)

*** 01/29/08 06:42 am ***

*** 01/29/08 06:44 am *** (CHG: Sta->10)

*** 01/29/08 06:44 am *** (CHG: Sta->16)

*** 01/29/08 10:00 am *** (CHG: Sta->11)

*** 01/29/08 10:00 am ***

*** 01/29/08 10:00 am *** (ADD: Impact/Symptom->INTERNAL ERROR )

*** 01/29/08 10:01 am *** (CHG: Asg->NEW OWNER OWNER)

*** 01/29/08 10:01 am ***

*** 02/13/08 06:07 pm ***

*** 05/06/08 06:21 pm ***

*** 05/06/08 06:29 pm ***

*** 05/06/08 06:31 pm ***

*** 08/07/08 10:57 am *** (CHG: Sta->36)

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

Bug6737235:DATABASEWITHSGA_TARGETCRASHWHENALTERSGACOMPONENTS

显示Bug属性Bug属性

类型B-Defect已在产品版本中修复-

严重性2-SevereLossofService产品版本10.2.0.3

状态36-DuplicateBug.ToFiler平台23-OracleSolarisonSPARC(64-bit)

创建时间08-Jan-2008平台版本-

更新时间11-Aug-2009基本Bug6192679

数据库版本10.2.0.3

影响平台Generic

产品源Oracle

显示相关产品相关产品

产品线OracleDatabaseProducts系列OracleDatabase

区域OracleDatabase产品5-OracleServer-EnterpriseEdition

Hdr:673723510.2.0.3RDBMS10.2.0.3MEMORYMGMTPRODID-5PORTID-236192679

Abstract:DATABASEWITHSGA_TARGETCRASHWHENALTERSGACOMPONENTS

***01/08/0810:47pm ***

TAR:

----

PROBLEM:

--------

Theparametersettingareasfollows:

NAMETYPEVALUE

--------------------------------------------------------

shared_pool_sizebiginteger608M

sga_max_sizebiginteger12G

sga_targetbiginteger12G

db_cache_sizebiginteger9G

*.java_pool_size=209715200

*.large_pool_size=209715200

*.streams_pool_size=50331648

DEFAULTbuffercacheinthev$sga_dynamic_componentsshowsthebelowvalue:

CURRENT_SIZEMIN_SIZEUSER_SPECIFIED_SIZE

---------------------------------------

1.1744E+101.1744E+109663676416

Whentrytoincreasetheshared_pool_sizeto700M:

SQL>altersystemsetshared_pool_size=700M;

altersystemsetshared_pool_size=700M

ERRORatline1:

ORA-2097:parametercannotbemodifiedbecausespecifiedvalueis

invalid

ORA-4033:Insufficientmemorytogrowpool

SQL>selectnamefromv$database;

selectnamefromv$database

ERRORatline1:

ORA-3135:connectionlostcontact

Aftertheaboveerrormessagethedatabasecrashes.

belowisthemessageinthealertlog:

Errorsinfile

/u01/app/oracle/admin/LSTDDB1/bdump/lstddb1_mman_12427.trc:

ORA-600:internalerrorcode,arguments:[kmgs_update_target_size_1],

[0],[],[],[],[],[],[]

Whentrytodecreasethesizeofthedefaultbuffercachegetting:

SQL>altersystemsetdb_cache_size=10G;

altersystemsetdb_cache_size=10G

ERRORatline1:

ORA-3113:end-of-fileoncommunicationchannel

andthedatabasecrashes.

thealertshowsthefollowing:

Errorsinfile/u01/app/oracle/admin/LSTDDB1/bdump/lstddb1_mman_11825.trc:

ORA-600:internalerrorcode,arguments:[kmgs_pre_process_request_6],[6],

[700],[576],[3],[0x67C7B9398],[],[]

DIAGNOSTICANALYSIS:

--------------------

WORKAROUND:

-----------

None

RELATEDBUGS:

-------------

Bug6192679

REPRODUCIBILITY:

----------------

Everytimeatcustomersite

TESTCASE:

----------

STACKTRACE:

------------

SUPPORTINGINFORMATION:

-----------------------

24HOURCONTACTINFORMATIONFORP1BUGS:

----------------------------------------

DIAL-ININFORMATION:

--------------------

IMPACTDATE:

------------

***01/08/0810:49pm ***

***01/08/0810:53pm ***

***01/08/0810:53pm ***(CHG:Sta->16)

***01/09/0807:51am ***(CHG:Asg->NEWOWNEROWNER)

***01/09/0808:53am ***(CHG:Sta->10SubComp->MEMORYMGMT)

***01/09/0808:53am ***

***01/10/0808:20pm ***

***01/10/0808:20pm ***(CHG:Sta->16)

***01/11/0808:33am ***(CHG:Sta->10)

***01/11/0808:33am ***

***01/28/0810:43pm ***

***01/28/0810:43pm ***(CHG:Sta->16)

***01/29/0806:17am ***(CHG:Sta->10)

***01/29/0806:17am ***

***01/29/0806:42am ***(CHG:Sta->16)

***01/29/0806:42am ***

***01/29/0806:44am ***(CHG:Sta->10)

***01/29/0806:44am ***(CHG:Sta->16)

***01/29/0810:00am ***(CHG:Sta->11)

***01/29/0810:00am ***

***01/29/0810:00am ***(ADD:Impact/Symptom->INTERNALERROR)

***01/29/0810:01am ***(CHG:Asg->NEWOWNEROWNER)

***01/29/0810:01am ***

***02/13/0806:07pm ***

***05/06/0806:21pm ***

***05/06/0806:29pm ***

***05/06/0806:31pm ***

***08/07/0810:57am ***(CHG:Sta->36)

Bug 6146397: ORA-00600 [KMGS_UPDATE_TARGET_SIZE_1], [0] INSTANCE TERMINATED BY MMAN

显示 Bug 属性 Bug 属性

类型 B - Defect 已在产品版本中修复 -

严重性 3 - Minimal Loss of Service 产品版本 10.2.0.3.0

状态 33 - Suspended, Req'd Info not Avail 平台 212 - IBM AIX on POWER Systems (64-bit)

创建时间 21-Jun-2007 平台版本 -

更新时间 17-Sep-2007 基本 Bug -

数据库版本 10.2.0.3

影响平台 Generic

产品源 Oracle

显示相关产品相关产品

产品线 Oracle Database Products 系列 Oracle Database

区域 Oracle Database 产品 5 - Oracle Server - Enterprise Edition

Hdr: 6146397 10.2.0.3 RDBMS 10.2.0.3.0 MEMORY MGMT PRODID-5 PORTID-212

Abstract: ORA-600 [KMGS_UPDATE_TARGET_SIZE_1], [0] INSTANCE TERMINATED BY MMAN

*** 06/21/07 05:05 pm ***

TAR:

----

PROBLEM:

--------

When dynamically allocating memory for running multiple sized buffer pools

statement fails and instance terminates

DIAGNOSTIC ANALYSIS:

--------------------

ALTER SYSTEM SET db_16k_cache_size='256M' SCOPE=BOTH;

ORA-600: internal error code, arguments: [kmgs_update_target_size_1], [0],

[],

MMAN: terminating instance due to error 822

WORKAROUND:

-----------

1) ALTER SYSTEM SET db_16k_cache_size='' scope=spfile

2) Restart instance

RELATED BUGS:

-------------

REPRODUCIBILITY:

----------------

TEST CASE:

----------

STACK TRACE:

------------

ksedst ksedmp ksfdmp kgerinv kgeasnmierr kmgs_update_target_size

kmgs_update_parameter_ctx kmgs_update_param_manual_helper

kmgs_update_param_manual kmgsdrv ksbabs ksbrdp opirip opidrv

sou2o opimai_real main start

SUPPORTING INFORMATION:

-----------------------

irp3_mman_1794226.trc

alert_IRP3.log

24 HOUR CONTACT INFORMATION FOR P1 BUGS:

----------------------------------------

DIAL-IN INFORMATION:

--------------------

IMPACT DATE:

------------

*** 06/21/07 05:13 pm *** (CHG: Sta->16)

*** 06/21/07 10:32 pm *** (CHG: Asg->NEW OWNER OWNER)

*** 06/22/07 06:26 am ***

*** 06/22/07 06:28 am *** (CHG: Sta->10)

*** 09/17/07 10:27 pm *** (CHG: Sta->33 SubComp->MEMORY MGMT)

*** 09/17/07 10:27 pm ***

Bug6146397:ORA-00600[KMGS_UPDATE_TARGET_SIZE_1],[0]INSTANCETERMINATEDBYMMAN

显示Bug属性Bug属性

类型B-Defect已在产品版本中修复-

严重性3-MinimalLossofService产品版本10.2.0.3.0

状态33-Suspended,Req'd Info not Avail 平台 212 - IBM AIX on POWER Systems (64-bit)

创建时间 21-Jun-2007 平台版本 -

更新时间 17-Sep-2007 基本 Bug -

数据库版本 10.2.0.3

影响平台 Generic

产品源 Oracle

显示相关产品相关产品

产品线 Oracle Database Products 系列 Oracle Database

区域 Oracle Database 产品 5 - Oracle Server - Enterprise Edition

Hdr: 6146397 10.2.0.3 RDBMS 10.2.0.3.0 MEMORY MGMT PRODID-5 PORTID-212

Abstract: ORA-600 [KMGS_UPDATE_TARGET_SIZE_1], [0] INSTANCE TERMINATED BY MMAN

*** 06/21/07 05:05 pm ***

TAR:

----

PROBLEM:

--------

When dynamically allocating memory for running multiple sized buffer pools

statement fails and instance terminates

DIAGNOSTIC ANALYSIS:

--------------------

ALTER SYSTEM SET db_16k_cache_size='256M' SCOPE=BOTH;

ORA-600: internal error code, arguments: [kmgs_update_target_size_1], [0],

[],

MMAN: terminating instance due to error 822

WORKAROUND:

-----------

1) ALTER SYSTEM SET db_16k_cache_size=''scope=spfile

2)Restartinstance

RELATEDBUGS:

-------------

REPRODUCIBILITY:

----------------

TESTCASE:

----------

STACKTRACE:

------------

ksedstksedmpksfdmpkgerinvkgeasnmierrkmgs_update_target_size

kmgs_update_parameter_ctxkmgs_update_param_manual_helper

kmgs_update_param_manualkmgsdrvksbabsksbrdpopiripopidrv

sou2oopimai_realmainstart

SUPPORTINGINFORMATION:

-----------------------

irp3_mman_1794226.trc

alert_IRP3.log

24HOURCONTACTINFORMATIONFORP1BUGS:

----------------------------------------

DIAL-ININFORMATION:

--------------------

IMPACTDATE:

------------

***06/21/0705:13pm ***(CHG:Sta->16)

***06/21/0710:32pm ***(CHG:Asg->NEWOWNEROWNER)

***06/22/0706:26am ***

***06/22/0706:28am ***(CHG:Sta->10)

***09/17/0710:27pm ***(CHG:Sta->33SubComp->MEMORYMGMT)

***09/17/0710:27pm ***

MMON Terminates Instance With ORA-600 [kmgs_pre_process_request_6] or ORA-600 [kmgs_update_target_size_1] When Resizing Caches [ID 373802.1]

修改时间 23-MAR-2010 类型 PROBLEM 状态 PUBLISHED

In this Document

Symptoms

Cause

Solution

References

Applies to:

Oracle Server - Enterprise Edition - Version: 10.2.0.1 to 10.2.0.3

Information in this document applies to any platform.

Symptoms

The MMAN process terminates the instance with the the following error is reported in the database alert.log file:

ORA-00600: internal error code, arguments: [kmgs_pre_process_request_6]

ORA-00600: internal error code, arguments: [kmgs_update_target_size_1]

The call stack from the ORA-00600 trace file will be similar to:

ksddoa

kmgs_update_target_size

Cause

This is due to Bug 4433838

Similar issues have been filed as bugs which have also been determined to be duplicates of Bug 4433838 .

The error occurs when the parameter SGA_TARGET is set to an exact multiple of 4Gb.

For instance, if the SGA_TARGET value in HEXADECIMAL has 00000000 as the last 4 bytes then you can hit this problem. The bug indicates this happens on an exact multiple of 4G.

Example:

SGA_TARGET (20 Gb) = 21474836480 --> HEXADECIMAL -->( 0x500000000) --->(value in HEXADECIMAL has 00000000 as the last 4 bytes).

SGA_TARGET (4 Gb) = 4294967296 --> HEXADECIMAL -->( 0x100000000) --->(value in HEXADECIMAL has 00000000 as the last 4 bytes).

Solution

1. Upgrade to 10.2.0.4 or higher.

2 Apply the one-off Patch 4433838 depending on its availability for your patchset-level and platform

3. Use the workaround:

Set your SGA_TARGET parameter to any non-multiple values of 4 Gb.

In other words, ensure that your SGA_TARGET value in HEXADECIMAL does not have 00000000 as the last 4 bytes.

Example:

If your SGA_TARGET is set to 20 Gb (21474836480) = 0x500000000 (in Hexadecimal). The workaround consists of changing its value to 0x510000000 (in Hexadecimal) = 21743271936

If your SGA_TARGET is set to 4 Gb (4294967296) = 0x100000000 (in Hexadecimal). The workaround consists of changing its value to 0x100000001 (in Hexadecimal) = 4294967297

References

BUG:4433838 - ORA-600 [KMGS_PRE_PROCESS_REQUEST_6] TERMINATES INSTANCE WHEN RESIZING CACHES

BUG:6192679 - MMAN ORA-600[KMGS_UPDATE_TARGET_SIZE_1],[0] AND INSTANCE DOWN

显示相关信息相关的

产品

* Oracle Database Products > Oracle Database > Oracle Database > Oracle Server - Enterprise Edition

MMONTerminatesInstanceWithORA-600[kmgs_pre_process_request_6]orORA-600[kmgs_update_target_size_1]WhenResizingCaches[ID373802.1]

修改时间23-MAR-2010类型PROBLEM状态PUBLISHED

InthisDocument

Symptoms

Cause

Solution

References

Appliesto:

OracleServer-EnterpriseEdition-Version:10.2.0.1to10.2.0.3

Informationinthisdocumentappliestoanyplatform.

Symptoms

TheMMANprocessterminatestheinstancewiththethefollowingerrorisreportedinthedatabasealert.logfile:

ORA-00600:internalerrorcode,arguments:[kmgs_pre_process_request_6]

ORA-00600:internalerrorcode,arguments:[kmgs_update_target_size_1]

ThecallstackfromtheORA-00600tracefilewillbesimilarto:

ksddoa

kmgs_update_target_size

Cause

ThisisduetoBug4433838

SimilarissueshavebeenfiledasbugswhichhavealsobeendeterminedtobeduplicatesofBug4433838.

TheerroroccurswhentheparameterSGA_TARGETissettoanexactmultipleof4Gb.

Forinstance,iftheSGA_TARGETvalueinHEXADECIMALhas00000000asthelast4bytesthenyoucanhitthisproblem.Thebugindicatesthishappensonanexactmultipleof4G.

Example:

SGA_TARGET(20Gb)=21474836480-->HEXADECIMAL-->(0x500000000)--->(valueinHEXADECIMALhas00000000asthelast4bytes).

SGA_TARGET(4Gb)=4294967296-->HEXADECIMAL-->(0x100000000)--->(valueinHEXADECIMALhas00000000asthelast4bytes).

Solution

1.Upgradeto10.2.0.4orhigher.

2Applytheone-offPatch4433838dependingonitsavailabilityforyourpatchset-levelandplatform

3.Usetheworkaround:

SetyourSGA_TARGETparametertoanynon-multiplevaluesof4Gb.

Inotherwords,ensurethatyourSGA_TARGETvalueinHEXADECIMALdoesnothave00000000asthelast4bytes.

Example:

IfyourSGA_TARGETissetto20Gb(21474836480)=0x500000000(inHexadecimal).Theworkaroundconsistsofchangingitsvalueto0x510000000(inHexadecimal)=21743271936

IfyourSGA_TARGETissetto4Gb(4294967296)=0x100000000(inHexadecimal).Theworkaroundconsistsofchangingitsvalueto0x100000001(inHexadecimal)=4294967297

References

BUG:4433838-ORA-600[KMGS_PRE_PROCESS_REQUEST_6]TERMINATESINSTANCEWHENRESIZINGCACHES

BUG:6192679-MMANORA-600[KMGS_UPDATE_TARGET_SIZE_1],[0]ANDINSTANCEDOWN

显示相关信息相关的

产品

*OracleDatabaseProducts>OracleDatabase>OracleDatabase>OracleServer-EnterpriseEdition

JAVA修改dmp版本号12g到11,又见bug，动态调整SGA参数导致数据库crash相关推荐

java修改另存excel_Word文档导入Excel总变形，调整格式你花了1小时，同事三步就搞定...
在进行文档编辑的时候我们都会用到Word,利用Word完成文档简历的编辑后,为方便文档的管理我们会把对应的简历文档复制粘贴到Excel当中.但是每次将Word内容直接复制到表格中后,我们都需要进行格式 ...
oracle数据库imp导入失败提示：“不是有效的导出文件, 标头验证失败”解决方法，修改dmp文件里oracle数据库版本号方法
报错信息如下,该报错主要是因为原来导出时的数据库版本和现有的数据库版本不匹配所致. C:\Users\Administrator>imp ncc2005zdh_0602/ufida file=D ...
Error:java: 无效的源发行版: 11
Error:java: 无效的源发行版: 11 1.问题描述 2.原因查找 3.解决办法 3.1 打开IDEA的File-Project Structure设置 3.2 修改Project SDK为自 ...
java修改字节码技术，Javassist修改class，ASM修改class
背景: 项目使用的Logback 1.1.11版本的类ch.qos.logback.core.rolling.helper.RollingCalendar的periodBarriersCrossed方 ...
Java编程思想学习笔记-第11章
<?xml version="1.0" encoding="utf-8"?> Java编程思想学习笔记-第11章 Java编程思想学习笔记-第11章 ...
CSDN实训 - 通过Java修改游戏存档
文章目录前言一.实现思路二.项目准备 1. 创建maven工程 2. 导入依赖 ①. JSON依赖 ②. Lombok依赖 ③. Junit4单元测试三.核心代码 1. 使用的对象 2. 读取 ...
安卓修改软件安装时间_安卓11正式发布
致力于成为您终身的苹果管家点击上方蓝字关注我们北京时间9月9日凌晨1点10分,Google低调放出了安卓11的正式版本,Pixel系列已经可以下载更新,支持Pixel 2系列及之后的所有机型. ...
安卓修改软件安装时间_安卓11正式发布：ColorOS秒级适配推送
北京时间9月9日凌晨1点10分,Google低调放出了安卓11的正式版本,Pixel系列已经可以下载更新,支持Pixel 2系列及之后的所有机型. 对于国内用户来说,更值得关注的当然是各家厂商自己的O ...
java 修改ip_如何用脚本快速修改IP地址(Netsh)
评论 # re: 如何用脚本快速修改IP地址(Netsh) 2006-03-30 15:36 david.turing (2006-03-30 15:12:48) 胡子鱼-SD(18684412) 另 ...

JAVA修改dmp版本号12g到11,又见bug，动态调整SGA参数导致数据库crash

JAVA修改dmp版本号12g到11,又见bug，动态调整SGA参数导致数据库crash相关推荐

最新文章

热门文章