From owner-freebsd-net@FreeBSD.ORG Mon Jan 24 18:02:07 2005 Return-Path: Delivered-To: freebsd-net@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 8587916A4CE for ; Mon, 24 Jan 2005 18:02:07 +0000 (GMT) Received: from lennier.cc.vt.edu (lennier.cc.vt.edu [198.82.162.213]) by mx1.FreeBSD.org (Postfix) with ESMTP id 19CE943D46 for ; Mon, 24 Jan 2005 18:02:07 +0000 (GMT) (envelope-from gaylord@dirtcheapemail.com) Received: from vivi.cc.vt.edu (IDENT:mirapoint@evil-vivi.cc.vt.edu [10.1.1.12]) by lennier.cc.vt.edu (8.12.11/8.12.11) with ESMTP id j0OI26Pd010298 for ; Mon, 24 Jan 2005 13:02:06 -0500 Received: from [127.0.0.1] (locust.cns.vt.edu [198.82.169.14]) by vivi.cc.vt.edu (MOS 3.5.6-GR) with ESMTP id CLW47928; Mon, 24 Jan 2005 13:02:00 -0500 (EST) Message-ID: <41F537E1.40500@dirtcheapemail.com> Date: Mon, 24 Jan 2005 13:01:05 -0500 From: Clark Gaylord User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) Gecko/20041217 X-Accept-Language: en-us, en MIME-Version: 1.0 To: freebsd-net@freebsd.org Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Subject: [Fwd: Re: Making ICMP the default traceroute protocol?] X-BeenThere: freebsd-net@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Networking and TCP/IP with FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 24 Jan 2005 18:02:07 -0000 Marian Durkovic wrote: > seems that in today's networking environment the original traceroute >concept utilising high UDP ports no longer works - since those ports >are now typically blocked by firewalls. > > However, when traceroute is performed using ICMP protocol, the results >are much better. > > Therefore, I'd like to propose to patch > >src/contrib/traceroute/traceroute.c > > so the ICMP protocol is the first one in I disagree. Firstly, IWFs tend to also block ICMP. Secondly, routers sometimes queue ICMP differently than UDP (not just in their own processing, which they almost always do, but also in their forwarding), giving even more distortion to these data than they naturally possess otherwise. In particular, if filtering happens, this becomes obvious; if differential queueing happens, it is difficult to notice that is likely what is happening as it doesn't break the trace, it just distorts the data. Finally, knowing that there is some IWF between me and the destination is usually a good indication of where a performance problem resides. ;-) This is most likely to make a difference at the end hop itself, though of course filtering can happen anywhere along the path. If you are finding that your destinations tend to need ICMP, I'd recommend aliasing traceroute to "traceroute -I". --ckg