On May 12, 3:31 pm, Niall Litchfield <niall.litchfi..._at_dial.pipex.com>
wrote:
> DA Morgan wrote:
> > Robert Klemme wrote:
>
> >>> What does cost have to do with response time?
> >>> Are you looking to speed something up or is this just like "tuning the
> >>> cache hit ratio".
>
> >> assumed IO cost => Plan => Performance
>
> >> where "=>" means influences.
>
> >> robert
>
> > A very weak assumption. I would recommend the books written by Tom Kyte,
> > Cary Milsap, and Jonathan Lewis as therapy.
>
> I'm not sure that I would call it a very weak assumption, in fact I'd
> call it really quite a good assumption. Up until very recently Oracle's
> costing model was based pretty much entirely on attempting to cost
> things in units of IO. Now these days there are also attempts to cost
> the CPU resources being used if system stats have been gathered, but
> never the less IO cost is a very large contributor to most calculated
> costs, the cost drives the plan which is adopted and plan choice can
> hugely influence performance.
>
> --
> Niall Litchfield
> Oracle DBAhttp://www.orawin.info/services
Does lower cost mean faster response time every time? I would say the
assumption lies therein. I've seen many cases where developers work
solely from cost and find that a much higher cost plan run magnitudes
faster.
Received on Sat May 12 2007 - 16:54:01 CDT
