Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Apparently anomalous behaviour with a subquery - has anyone seen this?

Re: Apparently anomalous behaviour with a subquery - has anyone seen this?

From: Galen Boyer <galenboyer_at_hotpop.com>
Date: 11 Jul 2002 09:57:05 -0500
Message-ID: <ud6tuuxdo.fsf@grossprofit.com>


On Thu, 11 Jul 2002, richard.foote_at_bigpond.com wrote:
> Hey Galen, don't go and doubt me now, we're almost there buddy
> !!

Throw me a rope!

> This is not a good example as it makes no sense to write it as
> it has been.
>
> But if you agree that the var column in the subquery represents
> the var in T1 in the outer query,

Yes.

> which occurrence of var are we talking about. The answer is
> *the current one* as being processed in the outer query.
>
> Therefore we must execute the subquery checking for the current
> value of var for as many rows as we process in the outer
> query.

This is what doesn't make any sense. As you do the T1 subquery, you CAN'T look for values of var, cause var doesn't exist in T2.

> A correlated subquery !! If there were 10 rows in T1, the
> subquery would be executed 10 times. The execution plans and
> stats would confirm this.

I understand a correlated subquery, I just don't understand this particular one.

> I noticed a posting by Tom Kyte in this tread. He points to a
> very similar example in his link.

Haven't read his particular example.

Guess I'll go do that soon.

-- 
Galen Boyer
Received on Thu Jul 11 2002 - 09:57:05 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US