From: Kirt Thomas <kremovethisspamthingthomas@gfsiinc.com>
Newsgroups: comp.databases.oracle.server
Subject: On Indexes and Commits
Organization: Evil Order of the Dark Duck
Reply-To: kremovethisspamthingthomas@gfsiinc.com
Message-ID: <lvd2ntcneshf4nqvqf3vif6as71u5o5k05@4ax.com>
X-Newsreader: Forte Agent 1.8/32.548
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Lines: 14
Date: Wed, 08 Aug 2001 08:16:22 -0500
NNTP-Posting-Host: 205.162.49.134
X-Complaints-To: abuse@sky.net
X-Trace: alpha.sky.net 997276566 205.162.49.134 (Wed, 08 Aug 2001 08:16:06 CDT)
NNTP-Posting-Date: Wed, 08 Aug 2001 08:16:06 CDT


We're sitting here puzzling over this :)

Our developers use a table to store data during the execution of a
report, they may end up inserting several to tens of thousands of
records into this table to be reported on (of course this is a bad,
but that's neither here nor there right now).  The way the dev's are
handling the records in this 'temp' table is to do no commit, and
issue a rollback after the report is done - effectively cancelling the
inserts.  This allows multiple reports to run at the same time as each
report only see's it's own sessions data.  

My assertion is that doing things this effectively negates any indexes
on the table.  I don't see how they could be used w/out commiting the
data.  Is this true?  Thanks.

