RE: How to reproduce an ORA-7445?

How about sennding signal 4 (illegal instruction), 8 (floating point exception) or 11 (segmentation violation) to an oracle process of your choice? The ORA-7445 error is generated from within a signal handler. If oracle process receives an external signal that it cannot handle, it will generate ORA-7445. That way of generating signals will keep Oracle developers occupied for a while, as they will frantically trie to reproduce the phantom menace causing the signal.

-----Original Message-----

From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Jesse, Rich
Sent: Wednesday, December 20, 2006 11:51 AM To: oracle-l_at_freelists.org
Subject: How to reproduce an ORA-7445?

Hey all,

In our migration of a new ERP system in 10.2.0.2.0 on AIX, a user hit an
ORA-7445 (with no parameters!). I file an SR, upload the resulting
trace file, troubleshooting happens, and it's determined we've hit a BUG
due to issues with NLS translation of dates in binds. I'm then told
that unless we can reproduce the problem, there will be no patch.

Setting aside my knee jerk of steam shooting out my ears, I started
thinking about this situation. Applying the logic that it's only a bug
if it can be reproduced, I believe the only way I can attempt to
accomplish that is to trace every session, either at the client or the
server, just in case it errors out in the future. And even then, that's
hardly guaranteed, as the issue could depend on factors only existing at
the time of the original error (e.g. SGA status, OS memory, etc).

So what's the point of filing an SR unless the problem is 100%
reproducible? I'm not trying to be sarcastic, just pragmatic.

TIA,
Rich

"It's a Christmas Miracle!" -- Jenny Meyer, Better Off Dead
--

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

--

http://www.freelists.org/webpage/oracle-l Received on Wed Dec 20 2006 - 11:45:11 CST