Re: test / live environment, major performance difference

Поиск
Список
Период
Сортировка
От Dave Dutcher
Тема Re: test / live environment, major performance difference
Дата
Msg-id 01d001c7ac43$ec686780$2e00a8c0@tridecap.com
обсуждение исходный текст
Ответ на test / live environment, major performance difference  (Christo Du Preez <christo@mecola.com>)
Ответы Re: test / live environment, major performance difference  (Christo Du Preez <christo@mecola.com>)
Список pgsql-performance
> -----Original Message-----
> From: Christo Du Preez
> Sent: Monday, June 11, 2007 10:10 AM
>
> I have narrowed down the problem (I think) and it's the query
> planner using different plans and I haven't got a clue why.
> Can anyone please shed some light on this?

Different plans can be caused by several different things like different
server versions, different planner settings in the config file, different
schemas, or different statistics.  You say the server versions are the same,
so that's not it.  Is the schema the same?  One isn't missing indexes that
the other has?  Do they both have the same data, or at least very close to
the same data?  Have you run analyze on both of them to update their
statistics?  Do they have the same planner settings in the config file?  I
would check that stuff out and see if it helps.

Dave


В списке pgsql-performance по дате отправления:

Предыдущее
От: Ansgar -59cobalt- Wiechers
Дата:
Сообщение: Re: test / live environment, major performance difference
Следующее
От: Vladimir Stankovic
Дата:
Сообщение: Variable (degrading) performance