Re: Optimizer question

From: Tyler D. Muth <tyler_muth_at_hotmail.com>
Date: 20 Sep 2002 05:36:48 -0700
Message-ID: <1afec35e.0209200436.76a6bb15_at_posting.google.com>


"dkrpata1" <dkrpata1_at_cox.net> wrote in message news:<awui9.72640$TX5.2224055_at_news1.east.cox.net>...
> Let's see how much dirt this brings up.
>
> Here's the situtation. We have a SQL statement that we force a hint on, but
> there are no statistics in the database. Let's even say there's another
> database that's the same way. From what I've read the docs say if you force
> a hint, then it will use the CBO, but since there are no statistics how does
> it figure the execution path to take. What assumptions, and are they
> documented anywhere (I have looked for about a day now) that the CBO will
> use. And why if it really does, would we see different execution paths for 2
> databases, both with no statistics that were collected on them.

When you force a hint on a table with no stats, the hint is not used. You will simply use the RBO.

Many factors will affect execution plan, but if you are looking to maintain the same query plan across multiple databases, pick up the following book, there's a whole chapter on it: http://www.amazon.com/exec/obidos/ASIN/1861004826/asktom-20/102-6049530-1609752

There's a wealth of other useful info in there too. Try http://asktom.oracle.com for this and other questions you may have.

Good luck. Received on Fri Sep 20 2002 - 14:36:48 CEST

Original text of this message