Re: turn WAL off.

From: Chris Gamache <cgg007(at)yahoo(dot)com>
To: pgsql-general(at)postgresql(dot)org
Subject: Re: turn WAL off.
Date: 2004-05-30 14:02:28
Message-ID: 20040530140228.9520.qmail@web13811.mail.yahoo.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Slightly off topic for this thread... I figured I give it a whirl...

I've often wondered what sort of performance increase one would get by placing
the WAL on a solid-state drive like a 2 or 4GB TiGi. Has anyone tested this
type of setup for a performance gain? For a 2GB drive it runs ~$3000. It would
really have to make a difference... I'm strongly cosidering testing this out.

CG

--- Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Alexander Cohen <alex(at)toomuchspace(dot)com> writes:
> >>> Is it possible to turn WAL completely off. For good reasons, i dont
> >>> ever want to use it. How can i turn it off?
> >>
> >> You can't. The postgresql team is not in the business of producing
> >> databases that can't handle a power failure.
> >>
> >> Any hints as to what your "good reason" is?
>
> > I need a small cluster. Thats the main reason. 30 Mb with no data in it .
> > is pretty large, to me at least. And im not using it in a manner that a
> > power failure will matter.
>
> To be blunt, you don't want Postgres. Consider Berkeley DB or tinysql
> or (holds nose) MySQL. What you're after isn't within the design goals
> for this project, either as to disk footprint or disinterest in power
> failure behavior.
>
> regards, tom lane
>
> ---------------------------(end of broadcast)---------------------------
> TIP 1: subscribe and unsubscribe commands go to
> majordomo(at)postgresql(dot)orgt)---------------------------
> TIP 1: subscribe and unsubscribe commands go to majordomo(at)postgresql(dot)org



__________________________________
Do you Yahoo!?
Friends. Fun. Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/

In response to

Browse pgsql-general by date

  From Date Subject
Next Message Bastian Blank 2004-05-30 14:16:38 COPY FROM fails to trigger rules
Previous Message Jurgen Defurne 2004-05-30 10:18:05 This is another testmail