Date: Tue, 26 Nov 2002 00:26:27 -0500
Reply-To: John <bentleyj@MINDSPRING.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: John <bentleyj@MINDSPRING.COM>
Organization: MindSpring Enterprises
Subject: Re: SAS/UNIX Work Area
It seems to me that you've identified the crux of problem--the bogus records
didn't force errors. Somehow, they became valid because the test criteria
didn't change. Was the Oracle database data changed between the 14th and
the 22d so that the previously bogus records' data could be valid? Are
there any business/transformation rules in the processing that are out of
your control that could have changed?
Can you rerun the tests again using the directories that worked properly,
that is, try to reproduce the first successful test?