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: Question about using a pl/sql table to improve performance of a query

Re: Question about using a pl/sql table to improve performance of a query

From: Daniel Morgan <dmorgan_at_exesolutions.com>
Date: Tue, 04 Jun 2002 15:59:29 GMT
Message-ID: <3CFCE3DA.67984D41@exesolutions.com>


Kenny Yu wrote:

> Set operations are slow, not optimized.
>
> "Daniel Morgan" <dmorgan_at_exesolutions.com> wrote in message
> news:3CFB8D8E.B1B1C4FA_at_exesolutions.com...
> > Ryan Gaffuri wrote:
> >
> > > Also, are there times when loading a table into a pl/sql table in
> > > memory might speed up performance of a select statement that is based
> > > on many join conditions? For example, I need to indentify duplicate
> > > data in two tables based on about 30 of the 40 columns in the two
> > > tables. Im currently doing a join on each of these 30 columns to look
> > > for the duplicates. This is really slow. Would using a pl/sql table
> > > improve performance?
> >
> > Why a join?
> >
> > Why not set operators such as INTERSECT or MINUS?
> >
> > Daniel Morgan
> >

They can also be the fastest solution to a problem. And I have metrics and explain plans to prove it.

I would not be so fast to dismiss them without a comparison.

Daniel Morgan Received on Tue Jun 04 2002 - 10:59:29 CDT

Original text of this message

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