Re: [PERFORM] Memory question on win32 systems - Mailing list pgsql-hackers

From Merlin Moncure
Subject Re: [PERFORM] Memory question on win32 systems
Date
Msg-id b42b73150805290516jd764995n4da110ade0c05dba@mail.gmail.com
Whole thread Raw
In response to [PERFORM] Memory question on win32 systems  (Sabbiolina <sabbiolina@gmail.com>)
List pgsql-hackers
On Thu, May 29, 2008 at 4:26 AM, Sabbiolina <sabbiolina@gmail.com> wrote:
> On Thu, May 29, 2008 at 2:11 AM, Douglas McNaught <doug@mcnaught.org> wrote:
>> 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

[this should really be on -perform]

This question gets asked a lot...by tweaking memory settings in
postgresql.conf you can only reserve memory for one thing at the
expense of the other.  Shared buffers, for example will reserve space
for the postgresql built in buffer cache instead of letting the
operating system use it's own cache.

Even if your shared buffers were set to the minimum setting, you would
still not be going to disk very much if your active working set of
files fit in RAM.

work_mem and other settings _can_ have a drastic impact on
performance, because they can force the server to use query plans
depending on how much memory it has for things like sorting.

You didn't give much detail...maybe if you gave examples of queries
you don't think are going fast enough (with explain analyze)?

merlin


pgsql-hackers by date:

Previous
From: "Guillaume Smet"
Date:
Subject: Re: Upcoming back-branch update releases
Next
From: Jorgen Austvik - Sun Norway
Date:
Subject: Re: pg_regress: referencing shared objects from tests