Sure, I'll monitor this, dont happen too often. Thanks.

Regards,
Saša Stamenković


On Thu, May 27, 2010 at 8:27 PM, Thomas D. <[email protected]> wrote:

> Hi,
>
> Bill Karwin wrote:
> > I wrote the following on the February thread with Ulf Wendel that you
> > linked to, but I think it bears repeating:
> >
> > [...]
> >
> > Peter seems to say that the "overhead" of using prepared statements is
> > that they are 14.5% *faster* than using a non-prepared query
> > execution, at least in this simple test.   I'd expect the relative
> > difference probably diminishes with a more complex query or a larger
> > result set.
>
> Peter was testing *real* prepared statements.
> Real prepared statements are prepared on the DBE server-side.
>
> Most people think, that when they are using PDO, which force them to use
> prepared statements in most cases, that they are using *real* prepared
> statements, but they aren't! PDO is just another abstraction PDO just
> emulate prepared statements. That means, preparation is done client-side.
> It's more a validation than preparation.
> This has *nothing* to do with server-side prepared statements. Most of the
> benefits a DBE is offering you when it is supporting prepared statements
> cannot be used by client-side prepared statements.
>
> That's why client-side prepared statements from a high performance
> viewpoint are real show stopper. They do nothing than wasting time ;-)
>
> For more:
> <http://blog.ulf-wendel.de/?p=187#pdo>
>
>
> > In any case, we should be careful about citing round-trips as a
> > significant performance factor, because it discourages people from
> > using prepared queries when they should.
>
> *FULL ACK*
>
> If anyone is not using some kind of abstraction like Zend_Db*, everyone
> should prefer PDO. Not because it is faster (it isn't, but it isn't really
> slower!), but because it is more secure by design, when you work with
> prepared statements (because it is enforcing some kind of security).
>
> But you should understand it. You should know why you are using PDO and why
> you should prefer it. You shouldn't use something, just because someone has
> told you "It's better" ;-)
>
>
> --
> Regards,
> Thomas
>
>

Reply via email to