Re: Re[2]: Hotsos Symposium, Day 1

  • From: Thomas Day <tomday2@xxxxxxxxx>
  • To: oracle-l@xxxxxxxxxxxxx
  • Date: Wed, 9 Mar 2005 10:36:47 -0500

I guess the reason that that jumped out at me was that I'm currently
trying to tune a database that uses a lot of 1=1 or 3=3 embedded in
it's packages and triggers.  Strangely, it never uses 2=2.

My assumption is that these expressions are used to force CBO not to
take a certain path.  My further assumption is that up-to-date
statistics will remove the need for these items.  Both assumptions are
unproven but that's the path I'm proceeding down.

I'm further assuming that Oracle does spend some resources evaluating
those expressions and that removing them, especially as they can get
executed thousands of times an hour, will (theoretically) improve
performance.

I guess what I'm looking for is --- is there a good, positive reason
to use an expression like 1=1?

Thanks.
--
//www.freelists.org/webpage/oracle-l

Other related posts:

  1. Home
  2. oracle-l
  3. Archive
  4. 03-2005