From owner-freebsd-current  Thu Apr 11 23:31:21 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id XAA28742
          for current-outgoing; Thu, 11 Apr 1996 23:31:21 -0700 (PDT)
Received: from trane.uninett.no (trane.uninett.no [129.241.1.16])
          by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id XAA28735
          Thu, 11 Apr 1996 23:31:18 -0700 (PDT)
From: sthaug@nethelp.no
Received: from localhost (localhost [127.0.0.1]) by trane.uninett.no (8.7.3/8.7.3) with SMTP id IAA21633; Fri, 12 Apr 1996 08:29:54 +0200 (METDST)
Message-Id: <199604120629.IAA21633@trane.uninett.no>
X-Authentication-Warning: trane.uninett.no: Host localhost [127.0.0.1] didn't use HELO protocol
To: nathan@netrail.net
Cc: terry@lambert.org, freebsd-question@FreeBSD.org,
        freebsd-current@FreeBSD.org
Subject: Re: What is the max network IO on a FreeBSD box?
In-Reply-To: Your message of "Thu, 11 Apr 1996 19:52:17 -0700 (MST)"
References: <199604120252.TAA00634@phaeton.artisoft.com>
X-Mailer: Mew version 1.03 on Emacs 19.28.1
Mime-Version: 1.0
Content-Type: Text/Plain; charset=us-ascii
Date: Fri, 12 Apr 1996 08:29:54 +0200
Sender: owner-current@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

> T1 is slow; it shouldn't be a problem.  The best number's I've seen
> for FDDI is 60-something  Mbit/S... 2/3's of one wire bandwidth.
> 
> You should go over the list archives for the -current list to get the
> actual performance figures on the 100 Mbit/S tests (I didn't run them,
> I'm just quoting from memory).

One datapoint which may be of interest: I have (repeatedly, consistently)
been able to get 75-80 Mbit/s between two P133s with SMC Etherpower 10/100
cards (100BaseTX) connected back to back with a TP crossover cable. Tests
performed with both netperf and ttcp give essentially the same results.
You can view my netperf results in the netperf database at

	http://www.cup.hp.com/netperf/numbers/NetperfBrowse.html

look for FastEthernet network type.

It should (at least in principle) be possible to do better with FDDI, due
to higher MTU.

Steinar Haug, Nethelp consulting, sthaug@nethelp.no