From owner-freebsd-questions@FreeBSD.ORG Wed Aug 31 13:19:25 2005 Return-Path: X-Original-To: freebsd-questions@freebsd.org Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 879BE16A43B for ; Wed, 31 Aug 2005 13:19:25 +0000 (GMT) (envelope-from freebsd@meijome.net) Received: from sigma.octantis.com.au (ns2.octantis.com.au [207.44.189.124]) by mx1.FreeBSD.org (Postfix) with ESMTP id 163B443D45 for ; Wed, 31 Aug 2005 13:19:24 +0000 (GMT) (envelope-from freebsd@meijome.net) Received: (qmail 7186 invoked from network); 31 Aug 2005 23:19:24 +1000 Received: from 203-217-79-78.dyn.iinet.net.au (HELO ?192.168.13.3?) (203.217.79.78) by sigma.octantis.com.au with (DHE-RSA-AES256-SHA encrypted) SMTP; 31 Aug 2005 23:19:24 +1000 Message-ID: <4315AE5A.8000902@meijome.net> Date: Wed, 31 Aug 2005 23:19:22 +1000 From: Norberto Meijome User-Agent: Mozilla Thunderbird 1.0.6 (Windows/20050716) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Emanuel Strobl References: <200508310046.20808@harrymail> <200508311251.06884@harrymail> In-Reply-To: <200508311251.06884@harrymail> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: freebsd-questions@freebsd.org Subject: Re: GbE NICs besides em (recommendation wanted) X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 31 Aug 2005 13:19:25 -0000 Emanuel Strobl wrote: > > Thank you for the offer, but I thought people had some simple test results > in mind. If you next time use rdump or large NFS transfers to another GbE > connected (and fast enough) box just watch the system load (I use systat > -vm 1) and see what card causes what interrupt load. em cards can't > transfer (real files over FTP/NFS) more than 200mbit/s on a > Coppermine@866, at this level the system load is 100% of which ~80% is > interrupt systemload :( > Would device polling help in this case?