xfs
[Top] [All Lists]

Re: IDE write cache and journaling file systems

To: Jim Buzbee <James.Buzbee@xxxxxxxxxxxx>
Subject: Re: IDE write cache and journaling file systems
From: Simon Matter <simon.matter@xxxxxxxxxxxxxxxx>
Date: Fri, 19 Apr 2002 17:13:17 +0200
>received: from mobile.sauter-bc.com (unknown [10.1.6.21]) by basel1.sauter-bc.com (Postfix) with ESMTP id C895D57306; Fri, 19 Apr 2002 17:13:18 +0200 (CEST)
Cc: XFS List <linux-xfs@xxxxxxxxxxx>
Organization: Sauter AG, Basel
References: <3CBEE618.B220A393@echostar.com> <3CBFB9D3.831D2C98@ch.sauter-bc.com> <3CC02DFB.ACEA0E6A@echostar.com>
Sender: owner-linux-xfs@xxxxxxxxxxx
Jim Buzbee schrieb:
> 
> Simon Matter wrote:
> 
> ...
> 
> > Jim,
> >
> > It's obvious that lifetime is reduced with write caching off. Just
> > listen to the drive and compare between w/cache on / off. What helps
> > here is that Linux uses memory for caching effectively.
> >
> > Some weeks ago I brought an issue to this list. I got zero filled files
> > after _clean_ reboots and I thought it was the write cache of the IDE
> > drives not flushing correctly. In fact I got bitten by the 'remount
> > readonly bug' and it had nothing to do with the write chache being
> > turned on.
> 
> I think I missed some of the "remount readonly" discussion.  Can someone
> summarize what the problem is and the correct way we should be doing a
> controlled shutdown/reboot with XFS?

The problem was that I changed some xinetd services via ntsysv (RH
specific) and immediately rebooted. Something like this:

ntsysv ; shutdown -r now

Shutdown was okay and / got remounted ro. After reboot, all the files in
/etc/xinetd.d that were touched by ntsysv before shutdown were now
filled with zero. In the end it came out that it was a bug of this
kernel which is fixed now.

You may want to check this:
http://oss.sgi.com/projects/xfs/mail_archive/200203/msg00634.html

-Simon

> 
> ...
> 
> >
> > Cache flushing does only work if the drive honours the flushing request
> > correctly. What I recommend is trying to use drives which:
> > - really flush cache when requested to do so.
> > - flush cache on power failure.
> 
> The drive we are currently using does its best.  It has a "power
> failure" mode where the cache is flushed. I believe they re-order
> requests to get as much data on the platter as possible in this mode.
> We're trying to determine how how much time they have after the power is
> cut.
> 
> Jim Buzbee,
> Echostar Technologies
> 
> >
> > The latter is quite important because comsumers will just pull power to
> > turn it off. I don't know whether this feature is available with IDE
> > drives but I think it should be possible.
> >
> > -Simon
> >
> > >
> > > This has to be a common problem.  Does anyone have any strategies or
> > > words of wisdom?
> > >
> > > Jim Buzbee,
> > > Echostar Technologies



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