Hi,

I Read on Many sites about Count (*) and Count(1). But there are
conflicts in the documents.
Some where it has been specified like count (*) gives better
performance and somewhere it is like Count (1) gives better
performance.

Can anybody please help me to understand how it works and which one is
better for performance?

Any help would be appreciated.

Thanks

On May 3, 9:21 am, Priya <supriyarsh...@gmail.com> wrote:
> Hi,
>
> I Read on Many sites about Count (*) and Count(1). But there are
> conflicts in the documents.
> Some where it has been specified like count (*) gives better
> performance and somewhere it is like Count (1) gives better
> performance.
>
> Can anybody please help me to understand how it works and which one is
> better for performance?
>
> Any help would be appreciated.
>
> Thanks

http://www.oracledba.co.uk/tips/count_speed.htm, Google archives,
numerous other places. To sum up: there are no significant differences
in performance between count(*) and count(1) beyond usual experiment
error, and Oracle does exactly the same amount of work for both.

Regards,
Vladimir M. Zakharychev
N-Networks, makers of Dynamic PSP(tm)
http://www.dynamicpsp.com

On May 3, 7:02 am, "Vladimir M. Zakharychev"
<vladimir.zakharyc...@gmail.com> wrote:
> On May 3, 9:21 am, Priya <supriyarsh...@gmail.com> wrote:
>
> > Hi,
>
> > I Read on Many sites about Count (*) and Count(1). But there are
> > conflicts in the documents.
> > Some where it has been specified like count (*) gives better
> > performance and somewhere it is like Count (1) gives better
> > performance.
>
> > Can anybody please help me to understand how it works and which one is
> > better for performance?
>
> > Any help would be appreciated.
>
> > Thanks
>
> http://www.oracledba.co.uk/tips/count_speed.htm, Google archives,
> numerous other places. To sum up: there are no significant differences
> in performance between count(*) and count(1) beyond usual experiment
> error, and Oracle does exactly the same amount of work for both.
>
> Regards,
> Vladimir M. Zakharychev
> N-Networks, makers of Dynamic PSP(tm)
> http://www.dynamicpsp.com

Actually slightly more work for COUNT(1), since the optimizer has to
substitute COUNT(*) internally.

If any document recommends something as idiotic as using COUNT(1)
instead of COUNT(*) I think you can safely disregard any other advice
in that document.