Re: Slow performance

Поиск
Список
Период
Сортировка
От Kai Otto
Тема Re: Slow performance
Дата
Msg-id A2F2DADB84065C41B4948C619B5E482E021A9E27@md-mail01.medis.local
обсуждение исходный текст
Ответ на Re: Slow performance  ("Kevin Grittner" <Kevin.Grittner@wicourts.gov>)
Список pgsql-performance
Hi all,

Thanks for the replies and sorry for the late response, I have been away
for a few days.

Concerning the performance: 1 ms per row seems slow knowing that the
entire database is less then 64MB and therefore should easily fit into
memory and the client (pgAdmin III) runs on the server.

I am going to test another database to check the performance of the
hardware.

-Kai

> -----Original Message-----
> From: Kevin Grittner [mailto:Kevin.Grittner@wicourts.gov]
> Sent: Wednesday, August 31, 2011 8:59 PM
> To: Kai Otto; pgsql-performance@postgresql.org
> Subject: Re: [PERFORM] Slow performance
>
> "Kai Otto" <kotto@medis.nl> wrote:
>
> > Time taken:
> >
> > 35.833 ms (i.e. roughly 35 seconds)
>
> Which is it?  35 ms or 35 seconds?
>
> > Number of rows:
> >
> > 121830
> >
> > Number of columns:
> >
> > 38
>
> > This is extremely slow for a database server.
> >
> > Can anyone help me in finding the problem?
>
> > "Seq Scan on "Frame"  (cost=0.00..9537.30 rows=121830 width=541)
> > (actual time=0.047..93.318 rows=121830 loops=1)"
> >
> > "Total runtime: 100.686 ms"
>
> Assuming 35 seconds for the 121 K rows, it would seem that you're
> taking less than 1 ms per row on the database server, which may not
> be too bad, depending on how many of them are read from disk.  The
> rest of the time would seem to be in the network and the client.
> That's where you need to fix something if you want it to be faster.
>
> With only a fraction of 1% of the run time being on the database
> server, any attempt to tune things there can't improve performance
> by more than that fraction of a percent.
>
> -Kevin

В списке pgsql-performance по дате отправления:

Предыдущее
От: "Tomas Vondra"
Дата:
Сообщение: Re: Sudden drop in DBb performance
Следующее
От: Craig Ringer
Дата:
Сообщение: Re: Embedded VACUUM