Getting Oracle Trace files in DB

[KPS]

We just recently installed an Oracle 9.2.0.3 DB on an HP-UX 11.11 server. Since starting the DB, we notice we're getting trace files in the udump area. We're also getting the following messages:

Dump file /u01/oracle/autosys2db/9.2.0/admin/udump/autosys2_ora_2595.trc
Oracle9i Enterprise Edition Release 9.2.0.3.0 - 64bit Production
With the Partitioning option
JServer Release 9.2.0.3.0 - Production
ORACLE_HOME = /u01/oracle/autosys2db/9.2.0
System name: HP-UX
Node name: imhqu27
Release: B.11.11
Version: U
Machine: 9000/800
Instance name: AUTOSYS2
Redo thread mounted by this instance: 1
Oracle process number: 13
Unix process pid: 2595, image: oracle@imhqu27 (TNS V1-V3)

*** SESSION ID:(12.40146) 2003-07-24 14:21:16.282
*** 2003-07-24 14:21:16.282
ksedmp: internal or fatal error
ORA-00600: internal error code, arguments: [12333], [14], [0], [97], [], [], [],
[]
Current SQL information unavailable - no session.


Would anyone know if this would have anything to do with OS or Kernel Params??

Thx,
-Ken

[Steven E. Protter]

This is a very generic errror.

Oracle picked up an error but had no idea from where it came.

We last got it off 9ias and were forced to rip out and reinstall the whole product.

If you take that route, never use the Ignore button and be meticulous about following the minutae of the install instructions.

SEP

[twang · Accepted solution]

The ora-600[12333] is reported when a request is being received from a network packet and the request code in the packet is not recognized. The three additional values report the invalid request values received.

Unfortunately, the error may have a number of different root causes so its not so easy to tell you what happened. For example, a network error may have caused bad data to be received, or the client application may have sent wrong data, or the data in the network
buffer may have been overwritten. Since there are many potential causes of this error, it is ESSENTIAL TO HAVE A REPRODUCIBLE testcase
to correctly diagnose the underlying cause. If operating system network logs are available, it is advisable to check them for
evidence of network failures which may indicate network transmission problems.

[Alexander M. Ermes]

Hi there.
If possible, report this error to Oracle.
Open up a TAR in Oracle Metalink and upload the appropriate alert log and trace files.
The arguments for ORA-00600 will be kept a secret by Oracle, so make notes avout the arguments and the solution.
Rgds
Alexander M. Ermes

[Yogeeraj_1]

hi,

As Alexander M. Ermes said, it is preferable that you log a tar with Metalink. They will be asking you to upload the trace file as well as other RDA output files.

Am sure they will be able to help you fix it.

best regards
Yogeeraj

[R. Allan Hicks]

It could very well have something to do with the kernel. Oracle's install doc has a boat load of kernel parameter changes. Take them very seriously.

As the other respondents have told you, it's a generic oracle problem. One of those "I don't have a clue how I got here so I'm gonna quit" messages.

If you have read the fine manual throughly and followed all the steps, then you only option is to open a tar. I'm running 9.2.0.2 on 11.00 without problems.

[Volker Borowski]

Hi Ken,

I found a SAP Note (505246) which is relating this error (600 [12333]) to lack of shared_pool_size.

When parsing complex (large) statements (i.e. lots of inlist operands), it get some sort of memory error, sometimes accompanied by
ORA-4031 / ORA-3113.

The SAP note suggests to increase the shared_pool_size to 400M+.

Since the statement is not part of your trace, it might be difficult to decide it's complexness :-)

No idea if this helps
Volker