Re: 57 minute SELECT

Поиск
Список
Период
Сортировка
От David Johnston
Тема Re: 57 minute SELECT
Дата
Msg-id 1380764921339-5773187.post@n5.nabble.com
обсуждение исходный текст
Ответ на Re: 57 minute SELECT  (Samuel Stearns <sstearns@staff.iinet.net.au>)
Ответы Re: 57 minute SELECT  (Samuel Stearns <sstearns@staff.iinet.net.au>)
Список pgsql-performance
Samuel Stearns-2 wrote
> EXPLAIN:
>
> QUERY PLAN
>
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>  Hash Join  (cost=408.53..1962721.39 rows=98068 width=126) (actual
> time=30121.265..3419306.752 rows=1929714 loops=1)

You are selecting and returning 2 million records...how fast do you want
this to run?  For some reason I read 57 seconds initially - I guess 57
minutes is a bit much...but the most obvious solution is RAM.

Might want to include buffers output in the explain as well but:

I'm doubting the contents of your result fit into the server memory so your
disk is involved which will severely slow down processing.

Hopefully someone more knowledgeable and experience will chime in to help
you.

David J.




--
View this message in context: http://postgresql.1045698.n5.nabble.com/57-minute-SELECT-tp5773169p5773187.html
Sent from the PostgreSQL - performance mailing list archive at Nabble.com.


В списке pgsql-performance по дате отправления:

Предыдущее
От: Samuel Stearns
Дата:
Сообщение: Re: 57 minute SELECT
Следующее
От: Samuel Stearns
Дата:
Сообщение: Re: 57 minute SELECT