Good point! I should have also stated that since this is a migration
database, there is only about 5-10% of the total number of expected
concurrent production users on it right now. I'm trying to head off
what could very well be a major showstopper come go-live.
Thanks for the feedback!
From: Allen, Brandon [mailto:Brandon.Allen@(protected)]
Sent: Wednesday, December 20, 2006 11:08 AM
To: Jesse, Rich; oracle-l@(protected)
Subject: RE: How to reproduce an ORA-7445
Probably not much point at all because they're always going to ask for
it to be reproduced with tracing on and as a DBA that provides tech
support myself, I usually find myself in the same position. It's often
very difficult if not impossible to troubleshoot a problem that can't be
reproduced and on top of that, there is the issue of triage - how much
time are you going to spend working on a problem that can't even be
reproduced when you've got 10 other fires to fight. If it can't be
reproduced, then chances are it won't be a problem again anyway so your
time would be better spent elsewhere. Of course, if it does end up
recurring, hopefully the next time you will have a better idea of what
to look for and compare with the first occurrence to see what the exact
conditions were that caused the problem - and each time you will
hopefully get closer to being able to reproduce it.
So what's the point of filing an SR unless the problem is 100%
reproducible? I'm not trying to be sarcastic, just pragmatic.