Re: When should one rebuild an index?

Bob Jones wrote:

> Perhaps, instead of looking at stats, we just need an Ace Director to
> orchestrate DBAs what to use or not to use in all situations.

Cute Bob. Really cute. The advice you are hearing from Oakies, from Aces, from internal Oracle resources is the same. If you wish to ignore it that is certainly your choice.

None of us make a dollar if you take our advice. None of us lose a wink of sleep if you ignore it. The simple fact is that sooner or later your employer will contact one of us to bail the bilge. Whether you will still be there when that happens is irrelevant tous. But a cautionary note to you and others who think attitude trumps aptitude. There is a reason why Oracle is being so successful selling the "you need fewer DBAs" mantra. CIOs and CTOs see too many dollars wasted on people that don't read. Too many dollars wasted on people that go to classes and still keep doing things the old way. Too many dollars wasted on people that think they know better than Tom Kyte (still not using bind variables), Richard Foote (still rebuilding indexes), Jonathan Lewis (still tuning based on BHCR), etc.

What I am telling you in simple terms is that you can run ANALYZE INDEX from now until you are a member of the gray-hair-no-hair crowd like me and you will not once see anything that will clearly indicate that your database will benefit from an index rebuild. It is the wrong tool. You can start figuring that out for yourself here if you wish: http://download.oracle.com/docs/cd/B28359_01/server.111/b28286/statements_4005.htm#SQLRF01105.

If you think I'm wrong write up a White Paper and I will be happy to provide you an opportunity to present it at a conference. If you just want to stand on the sidelines and throw stones the line forms to the right.

-- 
Daniel A. Morgan
Oracle Ace Director & Instructor
University of Washington
damorgan_at_x.washington.edu (replace x with u to respond)
Puget Sound Oracle Users Group
www.psoug.org