From owner-freebsd-questions@FreeBSD.ORG Mon Mar 31 08:35:05 2003 Return-Path: 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 9D8CA37B405 for ; Mon, 31 Mar 2003 08:35:05 -0800 (PST) Received: from ntli.com (pc1-glfd2-4-cust59.glfd.cable.ntl.com [81.99.187.59]) by mx1.FreeBSD.org (Postfix) with ESMTP id ED18A43FBF for ; Mon, 31 Mar 2003 08:35:03 -0800 (PST) (envelope-from william@palfreman.com) Received: from aqua.lan.palfreman.com (localhost [127.0.0.1]) by ntli.com (8.12.3p2/8.12.3) with ESMTP id h2VGftuG010610; Mon, 31 Mar 2003 17:41:56 +0100 (BST) (envelope-from william@palfreman.com) Received: from localhost (william@localhost)h2VGfta4010607; Mon, 31 Mar 2003 17:41:55 +0100 (BST) X-Authentication-Warning: aqua.lan.palfreman.com: william owned process doing -bs Date: Mon, 31 Mar 2003 17:41:55 +0100 (BST) From: William Palfreman To: "Jack L. Stone" In-Reply-To: <3.0.5.32.20030331082034.01414bf8@sage-one.net> Message-ID: <20030331171052.W249@ndhn.yna.cnyserzna.pbz> References: <3.0.5.32.20030331082034.01414bf8@sage-one.net> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII cc: freebsd-questions@freebsd.org Subject: Re: Interface collisions X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 31 Mar 2003 16:35:08 -0000 On Mon, 31 Mar 2003, Jack L. Stone wrote: > For the first time within the past few days, I've noticed collisions being > reported on the public NIC for one of the servers. I'm not sure if it means > the switch or the NIC is the culprit, so not sure which component may need > to be replaced. > > Name Mtu Network Address Ipkts Ierrs Opkts Oerrs Coll > rl1 1500 00:40:33:5b:bb:5f 6816063 0 7494432 0 > 66977 That is normal for half duplex. All it means is that you were using 100Mbit full duplex and now you are using either 100Mbit/half duplex or 10Mbit half duplex. I did think for a minute that you might have a mismatch between a hard-set full duplex port on the switch, and the nic not being able to - possibly for cable quality reasons - and then going into 10Mb/half mode. But don't think that is the case because you would get a terrible performance hit (down to 500Mbts-ish IME), and many more collisions than you have, plus lots of error packets. So, you are either using an old Cisco Cat 5000, in which case you will need to relay on black magic and illogical combinations of hard setting and auto-neg on both the switch and nic end, or you have a kinked cable causing a reflection back, or some other kind of badness - in which case the switch and nic should both auto-neg down to 10Mb/half, and you should be fine until you change the cable. I've had cables like that for the last three years without problem. If your rl nic was bad I'd expect it just to fail, rather than dropping down to half duplex. Do a big scp transfer over the cable and try to see what speed you get. If you are not too far from 100Mb do you care? -- W. Palfreman. I'm looking for a job: Tel: 0771 355 0354 http://www.palfreman.com/william/ for my CV.