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

Home -> Community -> Usenet -> c.d.o.tools -> Re: Very useful parameter - or did I dream it?

Re: Very useful parameter - or did I dream it?

From: Andrew Williamson <andrewweb_at_my-deja.com>
Date: Wed, 11 Oct 2000 10:14:21 +0100
Message-ID: <971255669.24396.0.nnrp-01.c30bdde2@news.demon.co.uk>

Thanks a lot guys...

It was the 8.1.5 docs I was looking through, though I do have some 8.1.6 (installation and administrators). The confusing part is where I saw it to begin with as even now I know what it is, I still can't find it... Oh well.

Thanks anyway.

AW

"Jonathan Lewis" <jonathan_at_jlcomp.demon.co.uk> wrote in message news:971206761.24505.1.nnrp-12.9e984b29_at_news.demon.co.uk...
>
> And it's only in 8.1.6 which is why you may be unable
> to find it right now.
>
> BTW
>
> >> eg.
> >> select xxx...yyy from zzz where columna = 'A' and columnb = :b1;
> >> select xxx...yyy from zzz where columna = 'A' and columnb = :b1;
> >> select xxx...yyy from zzz where columna = 'A' and columnb = :b1;
> >>
>
> In this example the second and third executions will reuse the path
> created on the first execution as the literals match.
>
> select xxx...yyy from zzz where columna = 'A' and columnb = :b1;
> select xxx...yyy from zzz where columna = 'B' and columnb = :b1;
> select xxx...yyy from zzz where columna = 'C' and columnb = :b1;
>
> It's only in the above that you could get the benefit of cursor_sharing;
> Oracle would rewrite this as (something like):
>
> select xxx...yyy from zzz where columna = :SYSBIND01 and columnb = :b1;
>
>
>
> --
>
> Jonathan Lewis
> Yet another Oracle-related web site: http://www.jlcomp.demon.co.uk
>
> James Lorenzen wrote in message <8rvmci$gf1$1_at_nnrp1.deja.com>...
> >It is "cursor_sharing"
> >HTH
> > James
> >
> >In article <971197492.7637.0.nnrp-01.c30bdde2_at_news.demon.co.uk>,
> > "Andrew Williamson" <andrewweb_at_my-deja.com> wrote:
> >> Hi
> >>
> >> I seem to remember while trawling through the Oracle8 init.ora
 parameters,
> >> that I came across something that helps SQL parse/execute ratios up by
> >> taking a slightly more 'relaxed' view as to whether a statement is in
 the
> >> cache or not..?
> >>
> >> I think its main area of help is where statements are repeatedly
 called with
> >> the same literal and thus we see the same statement in the cache
 multiple
> >> times for users.
> >>
> >> eg.
> >> select xxx...yyy from zzz where columna = 'A' and columnb = :b1;
> >> select xxx...yyy from zzz where columna = 'A' and columnb = :b1;
> >> select xxx...yyy from zzz where columna = 'A' and columnb = :b1;
> >>
> >> As it stands, because of the literal, this would be parsed each time,
 right?
> >> I thought I saw a parameter that helped ensure it wouldn't be?
> >>
> >> [We didn't write the code, it's supplied to us so and we can't change
 it so
> >> for us, bind variables are out in this instance.]
> >>
> >> Needless to say, I've had a *damn* good look through the
 documentation/net
> >> again, but I can't find hide-nor-hair of it.
> >>
> >> Did I dream it - or perhaps it wasn't in the init.ora?
> >>
> >> Anyone?
> >>
> >> Andrew
> >>
> >>
> >
> >--
> >Life is complex; it has real and imaginary parts.
> >
> >
> >Sent via Deja.com http://www.deja.com/
> >Before you buy.
>
>
Received on Wed Oct 11 2000 - 04:14:21 CDT

Original text of this message

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