dhpnbm2p Has Possible CPU Hardware Problem

[Jeff Clauss]

Can someone shed some insight on this problem.
We are told that a problem we are having with Orcal is caused by a possible CPU problem. The Oracal issue is: ORA-07445: exception encountered: core dump [pdz7M82_Validate()+432] [SIGSEGV] [Address not mapped to object] [0x97FFFFFFFFFEEE80]
The solution is: This is caused by an hardware issue, specifically with the CPU. There is a known problem with a set of HP Itanium processors and the solution is to identify and change these processors. My question is there any information detailing the identification of the CPUs, serial numbers, engineering date codes, ect?

[Dennis Handly]

This looks like a software bug to me. Do you have a URL that claims otherwise?

[Avinash20]

Did you got the core analysed by the Oracle support !!

[Jeff Clauss]

Yes the core analysis was done Oracle support. They report an HP incident number 1213378136 (HP-UX 11.23 - 10GR2 - fails to create DB catalog: script fails.) WTEC 4000130242.

[Gordon Crone]

We had a similar issue, bug was in something called the pxnmove. It was in an rx4640 4-way, and HP ended up changing all 4 cpus.

This was what we had gotten from oracle:
The ora-7445 pxnmove points to a faulty CPU on HP. From NOTE 398526.1:

Cause
This is caused by an hardware issue, specifically with the CPU. There is a known
problem with a set of HP Itanium processors and the solution is to identify and change these processors.

The stack trace points to the problem described in Bug 5304741 CATPROC.SQL CAUSE
S ORA-7445[PXNMOVE()+33].

This is reported as HP incident number 1213378136 (HP-UX 11.23 - 10GR2 - fails t
o create DB catalog: script fails.)
WTEC 4000130242.