Re: seqscan for 100 out of 3M rows, index present

Поиск
Список
Период
Сортировка
От ktm@rice.edu
Тема Re: seqscan for 100 out of 3M rows, index present
Дата
Msg-id 20130626204804.GC3264@aart.rice.edu
обсуждение исходный текст
Ответ на Re: seqscan for 100 out of 3M rows, index present  (Willy-Bas Loos <willybas@gmail.com>)
Список pgsql-performance
On Wed, Jun 26, 2013 at 10:36:10PM +0200, Willy-Bas Loos wrote:
> On Wed, Jun 26, 2013 at 10:31 PM, Jeff Janes <jeff.janes@gmail.com> wrote:
>
> >
> > Why is it retrieving 3.1 million, when it only needs 17?
> >
> >
> > that's because of the sequential scan, it reads all the data.
>
> cheers,
>
> willy-bas

Well, the two plans timings were pretty close together. Maybe your
cost model is off. Try adjusting the various cost parameters to
favor random I/O more.

Regards,
Ken


В списке pgsql-performance по дате отправления:

Предыдущее
От: Victor Yegorov
Дата:
Сообщение: Re: seqscan for 100 out of 3M rows, index present
Следующее
От: Sergey Konoplev
Дата:
Сообщение: Re: seqscan for 100 out of 3M rows, index present