RE: 10.2.0.2 Core dumps

From: Schultz, Charles <sac_at_uillinois.edu>
Date: Tue, 16 May 2006 08:33:32 -0500
Message-ID: <565F609E6D736D439837F1A1A797F34171D635@ADMINMAIL1.ui.uillinois.edu>

If you end up opening a case with Oracle and getting it solved, I would be interested in hearing about it. We have a Mutex problem with 10.2.0.2, which is a pain because when it bites, one cannot log in via sqlplus, not even as SYS (or internal).

From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Stefan Knecht
Sent: Tuesday, May 16, 2006 8:27 AM
To: oracle
Cc: oracle-l
Subject: Re: 10.2.0.2 Core dumps

Can't reproduce at will, seems to occur irregularly - though I had 7 the past few days:

Mon May 1 21:01:19 2006
SERVER COMPONENT id=PATCH_BGN: timestamp=2006-05-01 21:01:19 SERVER COMPONENT id=UPGRD_END: timestamp=2006-05-01 21:40:28

Thu May 11 15:17:16 2006
ORA-07445: exception encountered: core dump [kglhpd_internal()+1298]
[SIGSEGV] [Address not mapped to object] [0x1525438] [] []
Thu May 11 15:45:30 2006
ORA-07445: exception encountered: core dump
[qesmmIUnRegisterWorkArea()+628] [SIGILL] [Illegal operand] [0x8EA33E0]
[] []

Thu May 11 15:45:35 2006
Sat May 13 15:45:21 2006
ORA-07445: exception encountered: core dump [kcbget()+1662] [SIGSEGV]
[Address not mapped to object] [0xE] [] []
Sat May 13 15:47:04 2006
ORA-07445: exception encountered: core dump [kcbgcur()+2685] [SIGSEGV]
[Address not mapped to object] [0xE] [] []
Sat May 13 15:48:04 2006
ORA-07445: exception encountered: core dump [kcbgcur()+2685] [SIGSEGV]
[Address not mapped to object] [0xE] [] []
Mon May 15 11:13:44 2006
ORA-07445: exception encountered: core dump [kglhpd_internal()+1298]
[SIGSEGV] [Address not mapped to object] [0x14C63C8] [] []

The times it happened basically reflect the times of activity on the system. As i stated earlier, it's a test system, and it's not being used daily. And even in repeating the very same application procedures; sometimes it dumps, sometimes it don't. Beats me.

Though I would do some very extensive testing before going into production with 10.2.0.2, especially if you run Intermedia as well (like I do on this system).

Stefan

On 5/16/06, oracle <oracle_at_digistar.com> wrote:

	Stefan Knecht wrote:
	> Hello all
	>
	> Anyone else experience unusually often occurring coredumps and
segfaults
	> on 10.2.0.2 <http://10.2.0.2 > (linux) ?
	
	Feh - when is this occuring?  I'm getting ready to go to
production with
	a bunch of RAC environments on RHAS4 64bit Opterons...

(worried look).

Thanks

--
http://www.freelists.org/webpage/oracle-l

Received on Tue May 16 2006 - 08:33:32 CDT