Subject: Re: [performance] fast reads on a busy
 server
From: Hannu Krosing <hannu@2ndQuadrant.com>
To: Willy-Bas Loos <willybas@gmail.com>
Cc: pgsql-cluster-hackers@postgresql.org, pgsql-performance@postgresql.org
In-Reply-To: 
 <CAHnozTgLOqcAZWJu482Nh9K0f2BobXh3LPGYvycvr+Tb8prozw@mail.gmail.com>
References: 
 <CAHnozTgLOqcAZWJu482Nh9K0f2BobXh3LPGYvycvr+Tb8prozw@mail.gmail.com>
Content-Type: text/plain; charset="UTF-8"
Date: Wed, 27 Jun 2012 09:34:16 +0200
Message-ID: <1340782456.12666.45.camel@hvost>
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit

On Wed, 2012-06-27 at 00:16 +0200, Willy-Bas Loos wrote:
> Hi,
> 
> I've read this:
> http://wiki.postgresql.org/wiki/Prioritizing_databases_by_separating_into_multiple_clusters
> 
> But it doesn't really say anything about memory.
> If i can fit an extra cluster into it's shared buffer, it should have
> fast reads, right?
> Even if i don't have seperate spindles and the disks are busy.

Check if you are CPU-bound. On a database which fits fully you may
already be.

> This is on a Debain server, postgres 8.4

And if possible, upgrade to latest pg (9.1). On some operations this
already may give you a considerable performance boost

> Cheers,
> 
> WBL
> -- 
> "Quality comes from focus and clarity of purpose" -- Mark Shuttleworth
> 

-- 
-------
Hannu Krosing
PostgreSQL Unlimited Scalability and Performance Consultant
2ndQuadrant Nordic
PG Admin Book: http://www.2ndQuadrant.com/books/