From owner-freebsd-questions  Fri May 18 12:37:31 2001
Delivered-To: freebsd-questions@freebsd.org
Received: from stuart.microshaft.org (ns1.microshaft.org [208.201.249.2])
	by hub.freebsd.org (Postfix) with ESMTP id 27B7337B424
	for <freebsd-questions@FreeBSD.ORG>; Fri, 18 May 2001 12:37:28 -0700 (PDT)
	(envelope-from jono@networkcommand.com)
Received: from localhost (jono@localhost)
	by stuart.microshaft.org (8.9.3/8.9.3) with ESMTP id MAA39631;
	Fri, 18 May 2001 12:37:23 -0700 (PDT)
	(envelope-from jono@networkcommand.com)
Date: Fri, 18 May 2001 12:37:22 -0700 (PDT)
From: "Jon O." <jono@microshaft.org>
X-Sender: jono@stuart.microshaft.org
To: Paul Herman <pherman@frenchfries.net>
Cc: Kevin Oberman <oberman@es.net>, "Jon O." <jono@microshaft.org>,
	freebsd-questions@FreeBSD.ORG
Subject: Re: FreeBSD benchmark question 
In-Reply-To: <Pine.BSF.4.33.0105181933060.29842-100000@husten.security.at12.de>
Message-ID: <Pine.BSF.4.10.10105181057280.39019-100000@stuart.microshaft.org>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-freebsd-questions@FreeBSD.ORG
Precedence: bulk
List-ID: <freebsd-questions.FreeBSD.ORG>
List-Archive: <http://docs.freebsd.org/mail/> (Web Archive)
List-Help: <mailto:majordomo?subject=help> (List Instructions)
List-Subscribe: <mailto:majordomo?subject=subscribe%20freebsd-questions>
List-Unsubscribe: <mailto:majordomo?subject=unsubscribe%20freebsd-questions>
X-Loop: FreeBSD.ORG

Thanks for all the suggestions. Here is what I've found. 

I mounted the filesystem async with no improvment.
Softupdates don't help.
vmiodirenable didn't help.
hw.ata.wc didn't help.

You are correct that the genetic searches aren't real disk intensive but
rather CPU intensive. 

Are there some shell scripts or some way to isolate what the bottleneck
is? 

The goal is to get these searches done in the fastest time possible. Data
loss from caching or otherwise is not an issue. 

So, if you have any more suggestions or scripts to run on both machines to
isolate the disk, cpu, etc. that would be great. 

Thanks,
Jon

 
On Fri, 18 May 2001, Paul Herman wrote:

> On Fri, 18 May 2001, Kevin Oberman wrote:
> 
> > Another significant issue if you are using IDE disk is write cache. It
> > is currently turned off by default to prevent possible data loss. You
> > might want to test with it on (or turn it off in Linux, if you can).
> 
> The way Jon O. described the run ("genetic searches"), it sounded
> like it wasn't doing much writing to the disk, so write cache,
> softupdates, etc. will do little help if it's just reading data
> from a disk.  It sounded like it was I/O bound and profited from a
> large disk cache (read "reading from disk".)  I'm pretty sure Linux
> caches directory entries but FreeBSD doesn't by default.  Be sure
> you have:
> 
>   sysctl -w vfs.vmiodirenable=1
> 
> enabled.  There was some debate about whether or not this should be
> turned on by default, and the general consensus was that it should.
> 
> With vmiodirenable your disk cache will work Just Like Linux.  ;-)
> 
> -Paul.
> 



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message