[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Gnumed-devel] postgresql considerations
From: |
Karsten Hilbert |
Subject: |
Re: [Gnumed-devel] postgresql considerations |
Date: |
Thu, 14 Jun 2012 16:23:05 +0200 |
-------- Original-Nachricht --------
> Datum: Thu, 14 Jun 2012 10:02:51 -0400
> Von: Dave Cramer <address@hidden>
> An: Karsten Hilbert <address@hidden>
> CC: address@hidden
> Betreff: Re: [Gnumed-devel] postgresql considerations
> On Thu, Jun 14, 2012 at 9:58 AM, Karsten Hilbert
> <address@hidden> wrote:
> >> I was wondering what influences GNUmed's performance with regards
> >> to database performance most.
> >
> > The proper answer to this question can really only be found by actual
> > investigation.
> >
> > Possible reasons are:
> >
> > - network latency
> > - many small queries in a use case which would benefit from fewer
> > larger queries
> > - one large queries in a use case which would benefit from many
> > smaller queries
> > - server capability (RAM, disk throughput)
> > - dark buffers
>
> What are dark buffers ?
http://www.bufferbloat.net/projects/bloat/wiki/Dark_buffers
Such might only apply when connecting to the public database
over the internet.
> > - configure PostgreSQL to not use fsync (but you'll lose ACID)
>
> This is your medical data, why would you ever not want to have ACID ?
You don't but technically it IS a way to improve performance on USB
thumb drives.
Karsten
--
Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir
belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de