From owner-freebsd-hackers  Tue Mar 11 09:09:48 1997
Return-Path: <owner-hackers>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id JAA19704
          for hackers-outgoing; Tue, 11 Mar 1997 09:09:48 -0800 (PST)
Received: from nyx.pr.mcs.net (nyx.pr.mcs.net [204.95.55.81])
          by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id JAA19691;
          Tue, 11 Mar 1997 09:09:39 -0800 (PST)
Received: from nyx.pr.mcs.net (localhost [127.0.0.1])
          by nyx.pr.mcs.net (8.8.5/8.8.4) with ESMTP
	  id LAA04536; Tue, 11 Mar 1997 11:09:15 -0600 (CST)
Message-Id: <199703111709.LAA04536@nyx.pr.mcs.net>
X-Mailer: exmh version 1.6.9 8/22/96
To: Tom Samplonius <tom@sdf.com>
cc: Michael Hancock <michaelh@cet.co.jp>, Brian Tao <taob@vex.net>,
        "matthew c. mead" <mmead@goof.com>, isp@freebsd.org,
        hackers@freebsd.org
Subject: Re: freebsd as a news server? 
In-reply-to: Your message of Mon, 10 Mar 1997 19:46:37 -0800.
             <Pine.NEB.3.94.970310194240.27200A-100000@misery.sdf.com> 
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Tue, 11 Mar 1997 11:09:15 -0600
From: Chris Csanady <ccsanady@nyx.pr.mcs.net>
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk


>
>On Tue, 11 Mar 1997, Michael Hancock wrote:
>
>> On Mon, 10 Mar 1997, Tom Samplonius wrote:
>> 
>> > On Tue, 11 Mar 1997, Michael Hancock wrote:
>> > 
>> > ...
>> > > I've been using -i 3072 -b 4096 -f 1024 -a 8 for non-binary newsgroups a
nd
>> > > the default for binary newsgroups.
>> > > 
>> > > This way news articles on average will fit in a block.
>> > 
>> >   You always want maxcontig to beigger than 8.  It has nothing to do with
>> > file allocation, only have many sectors are transfered at once to the
>> > driver.
>> 
>> And the magic number is ____.
>
>  As big as possible.  All the disk drivers should be able to handle at
>least 128.  I don't really know how big it can be.  256?  1024?
>
>  A too small maxcontig will hamper scatter/gather.  I noticed that by
>looking at the sps and tps values from iostat that it was never
>transfering more than 16 sectors per transfer.  I bumped up maxcontig, and
>noticed a nice performance increase (at least for my application).

I thought that there is currently a 64K io limit imposed due to some
bogosity.  Any know if this ever got fixed?

If this is so, I assume 8 is the maximum if you use 8K blocks.  Or 128, if
they are talking about sectors in tunefs(8)..  I don't, but would like to
know. :)

Chris

>
>  The default maxcontig setting for 2.1.x is just way too small.
>
>Tom
>