Re: optimal sql

From: Peter van Rijn <p.vanrijnREMOVE_at_THISzhew.nl>
Date: Mon, 9 Dec 2002 13:12:35 +0100
Message-ID: <uv92dcl2781oc1@corp.supernews.com>

Christoph,

It partly depends on how big your tables are. It may also depend on Oracle version.
It is quite easy to test their effectiveness by just doing some testing yourself.

Just use the autotrace facility:

SQL> set autotrace traceonly explain

and it will supply the execution plan for your queries, without really executing it.

SQL> set autotrace off

disables the facility.

hth,
Peter

"Christoph Seidel" <chris666.seidel_at_gmx.de> schreef in bericht news:at1uho$vq9ai$1_at_ID-143718.news.dfncis.de...
> Billy Verreynne wrote:
>
> >> how can i find all records by pk in table 1 which are not in table 2
> >
> > I usually use something like:
> >
> > SELECT
> > t1.key1,
> > t1.key2
> > FROM table1 t1
> > MINUS
> > SELECT
> > t2.key1,
> > t2.key2
> > FROM table1 t2
>
> is this more efficient than not in or not exists?
>
> and: this assumes that t1 and t2 have the same fields to select?
>
>
Received on Mon Dec 09 2002 - 06:12:35 CST