Re: Stored procedure speed

Hi Jerry,
Thanks for replying. Actually the non stored proc querry runs faster(true, only milliseconds) sometimes! And that's something I don't understand. Maybe since one has to use ref cursors in order to get resultsets explains this. And sorry for my email address, but I'm so fed up with spam that I hesitate to put even my name in messages.

DM
Remove spam_spam. from my email address.

Jerry Gitomer wrote:

> A copy of this message was emailed to DM at
> <379F2FFA.3C81272F_at_spam_spam.idirect.com>...
>
> Nice theory ... and if all of your queries run against very small
> tables and return only one row it is valid.
>
> The catch is the longer your query runs the smaller percentage of
> total time is spend parsing and as you discovered in your own
> case you quickly reach a point where the parse speed improvement
> becomes insignificant.
>
> regards
> Jerry Gitomer
>
> DM wrote in message <379F2FFA.3C81272F_at_spam_spam.idirect.com>...
> >Hello everybody,
> >I wanted to speed up a sql statement execution so I embedded it
> in a
> >stored procedure and the result set is passed through a ref
> cursor to
> >the client. I got no speed improvement. Does anyone know more
> about
> >this? I mean, the general wisdom is that a stored procedure will
> give
> >you a 20% increase in speed upfront because the sql is already
> parsed,
> >the datapath set up. It seems that it's not the case. The sql is
> a join
> >and outer join of 4 tables with a parameter and it retrieves 899
> rows.
> >Thanks,
> >Max Mera
> >
> >
Received on Wed Jul 28 1999 - 13:27:06 CDT