Re: ORA-00600: [25027], [27] in ALERT.LOG

From: bdbafh <bdbafh_at_gmail.com>
Date: Wed, 16 Apr 2008 05:53:03 -0700 (PDT)
Message-ID: <3ec669e1-9818-4c28-8ed0-00a77b2b6676@m73g2000hsh.googlegroups.com>


On Apr 16, 5:32 am, "isma.sanc..._at_gmail.com" <isma.sanc..._at_gmail.com> wrote:
> On Apr 15, 4:21 pm, bdbafh <bdb..._at_gmail.com> wrote:
>
>
>
> > On Apr 15, 8:01 am, "isma.sanc..._at_gmail.com" <isma.sanc..._at_gmail.com>
> > wrote:
>
> > > Hi,
> > > We have discovered in the alert.log some ora-00600 in our Oracle
> > > Database 10g (10.2.0.3.0) on Windows 2003 Server. The database is
> > > running, but the error is generating a big amount of dump files.
> > > In the dump files, we found the following information:
>
> > > *** SERVICE NAME:(SYS$USERS) 2008-04-10 13:31:19.527
> > > *** SESSION ID:(294.450) 2008-04-10 13:31:19.527
> > > *** 2008-04-10 13:31:19.527
> > > ksedmp: internal or fatal error
> > > ORA-00600: internal error code, arguments: [25027], [27], [0], [], [],
> > > [], [], []
> > > Current SQL statement for this session:
> > > SELECT STRING_VALUE FROM MGMT_CURRENT_METRICS WHERE TARGET_GUID = :B3
> > > AND METRIC_GUID = :B2 AND KEY_VALUE = :B1
>
> > > We have also found that Oracle defined automatic optimizer statistics
> > > collection job (GATHER_STATS_JOB) is not running properly.
> > > In the operation details, the job appears as STOPPED
> > > REASON="Job slave process was terminated"
>
> > > I don't know if there's any relation with the ORA-600 Error. But I
> > > don't know how to solve it.
>
> > > I've looked for information about this error in metalink but I
> > > couldn't find useful information.
>
> > > do you know how to solve it?
> > > Thank you for your help in advance.
> > > Ismael
>
> > There should have been a trace file generated (in bdump) by the
> > dbms_stats scheduled job failure.
> > There are several errors with the package dbms_stats in 10.2.0.3 on MS
> > Win 32, 2 of which I have seen directly.
>
> > If you manually gather statistics on this table as a foreground
> > process using dbms_stats.gather_table_stats, what errors occur?
>
> > Are you ready to deploy the 10.2.0.4 patchset?
> > There are fixes for dbms_stats in there (check the release notes) that
> > might or might not apply to your issue.
>
> > good luck.
>
> > -bdbafh
>
> When I manually gather statistics:
> dbms_stats.gather_table_stats(ownname => 'SYSMAN', tabname =>
> 'MGMT_CURRENT_METRICS');
> I get :
> ORA-03113: end-of-file on communication channel
> and a new ora-00600 in the alert.log. :-(

So its reproducible. That is in fact good news. Are you familiar with the ora-600 lookup tool? Its available on Metalink as Note: 153788.1

https://metalink.oracle.com/metalink/plsql/ml2_documents.showDocument?p_database_id=NOT&p_id=153788.1

But seeing as you already have a service request open ... upload the new files under the existing service request. There should be a trace file in udump.

You might decide to manually set and lock stats for the affected table as a work-around.

Are you planning on applying the 10.2.0.3 patch 20 anytime soon? Its provided as part of the Critical Patch Update for April 2008. You might find that the fix for this error has already been incorporated in a one-off patch.

-bdbafh Received on Wed Apr 16 2008 - 07:53:03 CDT

Original text of this message