From owner-freebsd-isp Mon Mar 5 4:37: 5 2001 Delivered-To: freebsd-isp@freebsd.org Received: from slarti.muc.de (slarti.muc.de [193.149.48.10]) by hub.freebsd.org (Postfix) with SMTP id 3A05C37B71A for ; Mon, 5 Mar 2001 04:37:01 -0800 (PST) (envelope-from jhs@jhs.muc.de) Received: (qmail 23929 invoked from network); 5 Mar 2001 12:36:56 -0000 Received: from jhs.muc.de (193.149.49.84) by slarti.muc.de with SMTP; 5 Mar 2001 12:36:56 -0000 Received: from park.jhs.private (localhost [127.0.0.1]) by jhs.muc.de (8.11.0/8.11.0) with ESMTP id f24JYAu08057; Sun, 4 Mar 2001 19:34:10 GMT (envelope-from jhs@park.jhs.private) Message-Id: <200103041934.f24JYAu08057@jhs.muc.de> To: nino@inode.at Cc: freebsd-isp@FreeBSD.ORG Subject: Re: Dead link problems (with crossover cable) From: "Julian Stacey" Organization: Vector Systems Ltd - Munich Unix & Internet consultancy X-Web: http://www.jhs.muc.de http://bim.bsn.com/~jhs/ In-reply-to: Your message of "Fri, 02 Mar 2001 21:07:14 +0100." <20010302210714.A31469@TK147108.geizhals.at> Date: Sun, 04 Mar 2001 20:34:10 +0100 Sender: owner-freebsd-isp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi, "Marinos J . Yannikos" wrote: > I seem to have a non-reproducible but reoccurring problem: > > Host A: single CPU, Intel 443GX+ board with D-Link DFE-530TX ('vr') > Runs FreeBSD 4.1.1-STABLE > Has a Mylex ExtremeRAID 1100, no other PCI cards. > > Host B: SMP system, dual Intel 82557 on-board ethernet > SuSE Linux 7.0 > > They are connected with a crossover cable, both ends set to 100mbps > full-duplex. Mode Byte maybe ? I had back-to-back direct wired FBSD-3.4 systems (with 3com cards) last summer that sometimes started in the wrong mode, (admittedly, not in mid use like you). I guessed mine was a 10/100 thing, & by asserting the media byte (man xl to see one example of a `media' byte, but better see manual for your card), the problem went away. Presumably either end forces a bandwidth re-evaluation at least at every reboot/reset if not more often, so it's probably worth forcing the media asertion byte if available, rather than trusting to any auto-negotiation of 10/100/full/half etc. I recall for top speed, all 8 wires need to exist, but for slower modes less wires will do ? If you want to investigate what different 100M modes either end might be attempting, here's clues from my notes in: http://bim.bsn.com/~jhs/standards/rs232.v24 Halsall Book: 100Base4T=UTP=Unshielded Twisted pair Cat3 33.3Mbps/pair 100BaseX= STP= Shielded Twisted pair Cat5 http://margo.student.utwente.nl/stefan/hwb/hwb.html http://www.obbit.se/~mike/wordlist.htm#patchcord It's been a while since I looked at 100 modes, & I've got a cold, so no more energy to check things, but hope some of the above might help, Good Luck. Julian - Julian Stacey Unix Consultant - Munich Germany http://bim.bsn.com/~jhs/ Considering Linux ? Try FreeBSD with 4500 packages ! Ihr Rauchen => mein allergischer Kopfschmerz ! Kau/Schnupftabak probieren ! Vintage Computer Fest Europa 28&9 Apr 2001 http://www.vintage.org/vcfe To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-isp" in the body of the message