Обсуждение: [PERFORM] Memory question on win32 systems

Поиск
Список
Период
Сортировка

[PERFORM] Memory question on win32 systems

От
Sabbiolina
Дата:
<p class="MsoNormal"><span lang="EN-US" style="">Hello, in my particular case I need to configure Postgres to handle
onlya few concurrent connections, but I need it to be blazingly fast, so I need it to cache everything possible. I've
changedthe config file and multiplied all memory-related values by 10, still Postgres uses only less than 50 Mb of my
RAM.</span><pclass="MsoNormal"><span lang="EN-US" style="">I have 4 Gigs of RAM, how do I force Postgres to use a
higherpart of such memory in order to cache more indexes, queries and so on?</span><p class="MsoNormal"><span
lang="EN-US"style="">Thanks!</span> 

Re: [PERFORM] Memory question on win32 systems

От
"Douglas McNaught"
Дата:
On Wed, May 28, 2008 at 7:05 PM, Sabbiolina <sabbiolina@gmail.com> wrote:
> Hello, in my particular case I need to configure Postgres to handle only a
> few concurrent connections, but I need it to be blazingly fast, so I need it
> to cache everything possible. I've changed the config file and multiplied
> all memory-related values by 10, still Postgres uses only less than 50 Mb of
> my RAM.

How are you measuring this?

> I have 4 Gigs of RAM, how do I force Postgres to use a higher part of such
> memory in order to cache more indexes, queries and so on?

Post the settings values you're using and people will be better able
to help you.

-Doug


Re: [PERFORM] Memory question on win32 systems

От
Sam Mason
Дата:
On Thu, May 29, 2008 at 01:05:22AM +0200, Sabbiolina wrote:
> I have 4 Gigs of RAM, how do I force Postgres to use a higher part of such
> memory in order to cache more indexes, queries and so on?

PG relies on the operating system to cache most disk accesses.  Looking
at the amount of memory a process directly uses isn't a good measure of
this, a better method would be to watch what the IO subsystem is doing.
If you've got lots of reads happening then the disk isn't being cached.

 Sam