Re: Configuring Shared Buffers

From: "Rafael Domiciano" <rafael(dot)domiciano(at)gmail(dot)com>
To: "Tino Schwarze" <postgresql(at)tisc(dot)de>
Cc: pgsql-admin(at)postgresql(dot)org
Subject: Re: Configuring Shared Buffers
Date: 2008-06-30 23:16:02
Message-ID: 3a0028490806301616v6882a40bi2b54dcf63b37e315@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-admin

Yes, the server just does S, U, I and D.
The queries is pretty simples, don't have huge joins across tables...
In this server I have around 500 tables, the largest are:
9 millions tuples | 5 millions tuples

This server is our "authenticator"; the response is need to be "imediatly".
Today, the monitoring of the server is done using the unix command:
loguptime
With the time we discovered that when loguptime is up than 1.0 the response
begin to late, and the applications begin to take "Time-out".

2008/6/30 Tino Schwarze <postgresql(at)tisc(dot)de>:

> On Mon, Jun 30, 2008 at 04:01:14PM -0300, Rafael Domiciano wrote:
> > The Postgres version is 8.3.3 and I am using Fedora Core 8.
> > I have in the actual server around 70 connections the same time. I am
> > assigning for this 100.
>
> And what does the server do? Mainly SELECT / UPDATE / INSERT / DELETE,
> how is workload distributed, are queries very complex or pretty simple
> are there huge joins across many tables?
>
> How many tables do you have and what are the largest ones? Are we talking
> about thousands of rows or millions or lots of millions?
>
> Tino.
>
> > > > Folks, I am configuring a new Postgres Server, that's gonna
> substitute
> > > the
> > > > critical server of the enterprise.
> > > > I have a good machine:
> > > > Quad-Core 2.5 Ghz
> > > > 4 Gb RAM
> > > > 1 Dedicated HD 300 Gb Sata for the PostgreSQL Directory
> > >
> > > If it's the "critical server", you'd want to use a RAID1 for the disk
> so
> > > your server and data survives a harddisk crash.
> > >
> > > > My question is:
> > > > How much I could assign for the "Shared Buffers" parameter?
> > > >
> > > > The default assigned now is 24 Mb.
> > >
> > > That's rather low. I suppose, you may safely say 200Mb here (check that
> > > you don't run into OS limits, they might need raising). You may also
> > > want to increase work_mem, but it depends on what the DB does, how many
> > > concurrent connections you have etc. pp. maintenance_work_mem may
> safely
> > > be set rather high - it speeds up vacuum etc. and is only used during
> > > such operations. Also look for effective_cache_size and max_fsm_pages.
> > >
> > > BTW: What version are you going to use?
> > >
> > > Tino.
> > >
> > > --
> > > "What we nourish flourishes." - "Was wir nähren erblüht."
> > >
> > > www.craniosacralzentrum.de
> > > www.forteego.de
> > >
> > > --
> > > Sent via pgsql-admin mailing list (pgsql-admin(at)postgresql(dot)org)
> > > To make changes to your subscription:
> > > http://www.postgresql.org/mailpref/pgsql-admin
> > >
>
> --
> "What we nourish flourishes." - "Was wir nähren erblüht."
>
> www.craniosacralzentrum.de
> www.forteego.de
>
> --
> Sent via pgsql-admin mailing list (pgsql-admin(at)postgresql(dot)org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-admin
>

In response to

Responses

Browse pgsql-admin by date

  From Date Subject
Next Message Alvaro Herrera 2008-06-30 23:19:41 Re: Configuring Shared Buffers
Previous Message Tom Lane 2008-06-30 20:05:36 Re: Proposal for restoring a dump into a database with a different owner