From owner-freebsd-hackers@FreeBSD.ORG  Thu Sep 11 11:32:14 2003
Return-Path: <owner-freebsd-hackers@FreeBSD.ORG>
Delivered-To: freebsd-hackers@freebsd.org
Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125])
	by hub.freebsd.org (Postfix) with ESMTP id 0F62C16A52C
	for <freebsd-hackers@freebsd.org>;
	Thu, 11 Sep 2003 11:32:14 -0700 (PDT)
Received: from smtp5.andrew.cmu.edu (SMTP5.andrew.cmu.edu [128.2.10.85])
	by mx1.FreeBSD.org (Postfix) with ESMTP id 3A77E43F93
	for <freebsd-hackers@freebsd.org>;
	Thu, 11 Sep 2003 11:32:13 -0700 (PDT)
	(envelope-from eno@andrew.cmu.edu)
Received: from UNIX46.andrew.cmu.edu (UNIX46.andrew.cmu.edu [128.2.13.176])
	(user=eno mech=GSSAPI (0 bits))h8BIWBuO017664;
	Thu, 11 Sep 2003 14:32:11 -0400
Date: Thu, 11 Sep 2003 14:32:11 -0400 (EDT)
From: Eno Thereska <eno@andrew.cmu.edu>
To: Doug White <dwhite@gumbysoft.com>
In-Reply-To: <20030911110203.I61751@carver.gumbysoft.com>
Message-ID: <Pine.LNX.4.55L-032.0309111428050.22691@unix46.andrew.cmu.edu>
References: <3F5FCEB5.9010407@andrew.cmu.edu>
	<20030911110203.I61751@carver.gumbysoft.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
cc: freebsd-hackers@freebsd.org
Subject: Re: flush on close
X-BeenThere: freebsd-hackers@freebsd.org
X-Mailman-Version: 2.1.1
Precedence: list
List-Id: Technical Discussions relating to FreeBSD
	<freebsd-hackers.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-hackers>,
	<mailto:freebsd-hackers-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-hackers>
List-Post: <mailto:freebsd-hackers@freebsd.org>
List-Help: <mailto:freebsd-hackers-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-hackers>,
	<mailto:freebsd-hackers-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Thu, 11 Sep 2003 18:32:14 -0000

Doung,

Thanks for the quick reply. I am using the default options
(noasync, which according to the man pages means that
metadata I/O should be done synchronously, while
data I/O is asynchronous).

Unfortunately, with any of these options (nosync, async or using
soft updates), I think the flush-on-close semantics is an orthogonal
issue. For example, with the async option, with the flush-on-close
semantics, data and metadata are flushed.

Please let me know if you have any more hints on this.

Thanks
Eno

On Thu, 11 Sep 2003, Doug White wrote:

> Remove -fs. Don't crosspost, please.
>
> On Wed, 10 Sep 2003, Eno Thereska wrote:
>
> > In FreeBSD 4.4, I am noticing a huge number of calls
> > to ffs_fsync() (in /sys/ufs/ffs/ffs_vnops.c)
> > when running a benchmark like Postmark.
>
> Were you using softupdates, or the sync or async mount options?
>
> I believe this is correct (and safe) behavior for the default case.
>
> --
> Doug White                    |  FreeBSD: The Power to Serve
> dwhite@gumbysoft.com          |  www.FreeBSD.org
>
>