From owner-freebsd-net Thu Sep 27 8:26:34 2001 Delivered-To: freebsd-net@freebsd.org Received: from panzer.kdm.org (panzer.kdm.org [216.160.178.169]) by hub.freebsd.org (Postfix) with ESMTP id CF63F37B422 for ; Thu, 27 Sep 2001 08:26:21 -0700 (PDT) Received: (from ken@localhost) by panzer.kdm.org (8.11.6/8.9.1) id f8RFPqg29115; Thu, 27 Sep 2001 09:25:52 -0600 (MDT) (envelope-from ken) Date: Thu, 27 Sep 2001 09:25:52 -0600 From: "Kenneth D. Merry" To: Matthew Rezny Cc: "net@freebsd.org" Subject: Re: Alteon gigE NIC (if_ti driver) problems Message-ID: <20010927092551.A29083@panzer.kdm.org> References: <20010926205541.10F5A37B417@hub.freebsd.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2i In-Reply-To: <20010926205541.10F5A37B417@hub.freebsd.org>; from mrezny@umr.edu on Wed, Sep 26, 2001 at 03:55:30PM -0500 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Wed, Sep 26, 2001 at 15:55:30 -0500, Matthew Rezny wrote: > I have some more information since my initial posting yesterday. I set > the NMBCLUSTERS back to default, which made no difference. Therefore, > moving from 4.3 to 4.4 is what drastically increased the frequency at > which the link goes down and back up. I also captured dmesg now I case > there is any useful information in it. The fluctuation of the link > occurs multiple times a second with the new kernel when attempting to > use the network with this machine. Any suggestions would be greatly > appreciated. Thanks. [ ... ] > ti0: mem > 0x82850000-0x82853fff irq 11 at device 6.0 on pci0 > ti0: interrupting at CIA irq 11 > ti0: Ethernet address: 00:60:cf:20:1e:99 [ ... ] > ti0: gigabit link up > ti0: link down > ti0: gigabit link up > ti0: link down > ti0: gigabit link up > ti0: link down > ti0: gigabit link up > ti0: link down > ti0: gigabit link up > ti0: link down > ti0: gigabit link up > ti0: link down > ti0: gigabit link up > ti0: link down > ti0: gigabit link up > ti0: link down > ti0: gigabit link up > ti0: link down > ti0: gigabit link up [ ... ] The most likely explanation for this is that you've got a bad/dirty piece of fiber. The number of mbuf clusters probably isn't going to affect this at all, unless you're dumping a lot of traffic over the NIC and increasing the number of mbufs allows you to dump more packets. So I'd say replace your fiber. Ken -- Kenneth Merry ken@kdm.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message