ORA-00600[kglhdunp2_2]错误一例

一套 AIX上的10.2.0.3 数据库出现了ORA-00600: internal error code, arguments: [kglhdunp2_2]错误,详细日志如下:

 

ORA-00600: internal error code, arguments: [kglhdunp2_2], [0x7000007A061F8A0], [
3], [0x7000007F4AEC160], [0x7000007A061F990], [0x7000007A06639A8], [1000], [18]
Sat Aug 20 05:11:26 2011
Fatal internal error happened while SMON was doing Unpin KGL handles with depend
ency.
Sat Aug 20 05:11:26 2011
Errors in file /u01/app/oracle/product/10.2.0/admin/bdump/prod_smon_4915426.trc:
ORA-00600: internal error code, arguments: [kglhdunp2_2], [0x7000007A061F8A0], [
3], [0x7000007F4AEC160], [0x7000007A061F990], [0x7000007A06639A8], [1000], [18]
SMON: terminating instance due to error 474
Instance terminated by SMON, pid = 4915426
Sat Aug 20 05:26:15 2011

ksedmp: internal or fatal error
ORA-00600: internal error code, arguments: [kglhdunp2_2], [0x7000007A061F8A0], [3],
[0x7000007F4AEC160], [0x7000007A061F990], [0x7000007A06639A8], [1000], [18]
----- Call Stack Trace -----

ksedst <- ksedmp <- ksfdmp <- kgerinv <- kgesinv
<- kgesin <- kglhdunp2 <- kglScanDependencyHa <- ndles4Unpin <- ktmmon
<- ktmSmonMain <- ksbrdp <- opirip <- opidrv <- sou2o
<- opimai_real <- main <- start

Stack trace matches Bug 7254367----
 kgesinv <- kgesin <- kglhdunp2 <- kglScanDependencyHa <- ndles4Unpin
 <- ktmmon <- ndles4Unpin <- ksbrdp <- opirip <- opidrv
 <- sou2o <- opimai_real <- main <- start

 

对比stack call 确认为Bug 7254367 – SMON crashes instance with OERI[kglhdunp2_2] (Doc ID 7254367.8)

 

Bug 7254367  SMON crashes instance with OERI[kglhdunp2_2]

Affects:

    Product (Component)    Oracle Server (Rdbms)
    Range of versions believed to be affected    Versions < 11
    Versions confirmed as being affected    

        10.2.0.4

    Platforms affected    Generic (all / most platforms affected)

Fixed:

    This issue is fixed in    

        10.2.0.3 Patch 30 on Windows Platforms
        10.2.0.4 Patch 14 on Windows Platforms
        10.2.0.5 (Server Patch Set)

Symptoms:

Related To:

    Internal Error May Occur (ORA-600)
    Instance May Crash
    ORA-600 [kglhdunp2_2]

    (None Specified)

Description

    The SMON process may fail due to an ORA-600 [kglhdunp2_2]
    leading to an instance crash.

ORA-600 [kglhdunp2_2]

PURPOSE:
  This article represents a partially published OERI note.

  It has been published because the ORA-600 error has been
  reported in at least one confirmed bug.

  Therefore, the SUGGESTIONS section of this article may help
  in terms of identifying the cause of the error.

  This specific ORA-600 error may be considered for full publication
  at a later date. If/when fully published, additional information
  will be available here on the nature of this error.

SUGGESTIONS:

  If the Known Issues section below does not help in terms of identifying
  a solution, please submit the trace files and alert.log to Oracle
  Support Services for further analysis.

  Known Issues:

  Bug# 7254367   See Note:7254367.8
      SMON crashes instance with OERI[kglhdunp2_2]
      Fixed: 10.2.0.5, 10.2.0.4.P14, 10.2.0.3.P30

  Bug# 3749490   See Note:3749490.8
      KGL linked lists may become corrupt if process killed at specific time
      Fixed: 9.2.0.6

  Bug# 3724548   See Note:3724548.8
      OERI[kglhdunp2_2] / OERI[1100] under high load
      Fixed: 9.2.0.6, 10.1.0.4, 10.2.0.1:

 

解决方案是升级到10.2.0.5 或者 在原AIX 10.2.0.3.0的基础上安装7254367补丁:

 

We are hitting the bug 7254367

It is fixed in 10.2.0.5

There is a patch available for 10.2.0.3.0 on AIX

The patch 7254367 is available via My Oracle Support

To implement the solution, please execute the following steps:
1) Please download and review the read me for Patch 7254367
2) Please apply Patch. 7254367 in a test environment.
3) Please retest the issue.
4) If the issue is resolved, please migrate the solution as appropriate to other environments.

If you are going to apply the latest patchset in the same release ( 10.2.0.5 patchset or any patch on 10.2.0.3 ) , the other things like SGA recommendations / settings do not change.
So please go ahead and apply the patch.

Applying latest patchset is strongly recommended .

Why latest Patch set?

a)Latest Patch set has resolved many Bugs
b)Interim patches are provided on latest Patch set on Supported releases
c)Interim Patches are “not” rigorously tested but Patch Sets are tested rigorously.
d)We always recommend latest Patch set.
e) Refer following note.

Article-ID: Note 742060.1
Title: Release Schedule of Current Database Patch Sets

To obtain latest patchset from metalink
1.1 Go to metalink
1.2 Click on patches and updates
1.3 Click “Quick Links to the Latest Patchsets, Mini Packs, and Maintenance Packs”
1.4 Choose Oracle Database
1.5 Choose Your operating System
1.6 Choose & click 10.2.0.5

We also recommend you to apply latest PSU ( patchset update ) on top of latest patchset to make the DB stable .


Posted

in

by

Tags:

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *