xfs
[Top] [All Lists]

Re: RAID Controller Caching and XFS

To: Seth Mos <knuffie@xxxxxxxxx>
Subject: Re: RAID Controller Caching and XFS
From: Utz Lehmann <ulehmann@xxxxxxxxxxxxxx>
Date: Thu, 6 Jun 2002 14:51:20 +0200
Cc: utz lehmann <xfs@xxxxxxxxxx>, Florin Andrei <florin@xxxxxxx>, linux-xfs <linux-xfs@xxxxxxxxxxx>
In-reply-to: <4.3.2.7.2.20020606071526.0431c8e0@xxxxxxxxxxxxx>; from knuffie@xxxxxxxxx on Thu, Jun 06, 2002 at 07:16:32AM +0200
References: <1023314217.32329.32.camel@xxxxxxxxxxxxxxxxxxx> <auto-000021147946@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <4.3.2.7.2.20020603111517.03b4e1b0@xxxxxxxxxxxxx> <20020605160422.A14886@xxxxxxxxxxxxxxxxxxxxx> <1023314217.32329.32.camel@xxxxxxxxxxxxxxxxxxx> <2
Sender: owner-linux-xfs@xxxxxxxxxxx
User-agent: Mutt/1.3.12i
Hi Seth

Seth Mos [knuffie@xxxxxxxxx] wrote:
> At 23:49 5-6-2002 +0200, utz lehmann wrote:
> >Hi
> >
> >Florin Andrei [florin@xxxxxxx] wrote:
> > > On Wed, 2002-06-05 at 13:04, Thor Lancelot Simon wrote:
> > > >
> > > > More importantly, you don't *want* to use readahead because it will 
> > > > waste
> > > > valuable cache space on the controller that could be used to cache small
> > > > writes.
> > >
> > > How safe is it to enable caching for writes on the controller (if you
> > > don't have a battery on it)?
> > > I mean, i configured my Mylex with "Write Through" for data-safety
> > > reasons, even though the other option ("Write Back") could offer
> > > improved performance.
> >
> >It's not (really) safe, see
> >http://oss.sgi.com/projects/xfs/mail_archive/200205/msg00510.html
> 
> He is talking about a add-in card with it's own battery backup. The mail 
> you referenced is about a externel raid device without a battery backup.

He sad, that the controller don't have a battery, look above. So it's the
same, a poweroutage cause lost of all unwritten data in the controller
cache.


utz


<Prev in Thread] Current Thread [Next in Thread>