From owner-freebsd-isp Fri Nov 28 13:55:46 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id NAA11433 for isp-outgoing; Fri, 28 Nov 1997 13:55:46 -0800 (PST) (envelope-from owner-freebsd-isp) Received: from iworks.InterWorks.org (deischen@iworks.interworks.org [128.255.18.10]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id NAA11420 for ; Fri, 28 Nov 1997 13:55:41 -0800 (PST) (envelope-from deischen@iworks.interworks.org) Received: (from deischen@localhost) by iworks.InterWorks.org (8.7.5/) id PAA27779; Fri, 28 Nov 1997 15:58:13 -0600 (CST) Message-Id: <199711282158.PAA27779@iworks.InterWorks.org> Date: Fri, 28 Nov 1997 15:58:13 -0600 (CST) From: "Daniel M. Eischen" To: dg@root.com Subject: Re: Adaptec 2940 and FreeBSD 2.2.2-RELEASE Cc: freebsd-isp@freebsd.org, jeff@mercury.jorsm.com Sender: owner-freebsd-isp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > >The servers we have that keep blowing up have only a single internal DAT > >on the narrow bus and a single internal UW HDD on the wide bus. It most > >common for a tape error to lock up the machine, but they also lock up > >with the same symptoms when no backups are being done. > > Hmmm...Do you have a narrow->wide converter than has termination for the > upper (wide) data bits? I had a lot of problems here with using a narrow > drive on a wide controller until Justin pointed out that the upper bits have > to be terminated. I fixed the problem by putting a wide device at the end > of the same cable that the narrow device was on (thus providing full wide > termination). Or perhaps using the DAT drive to terminate the bus? Especially when running at Ultra SCSI speeds, it is essential to have proper (active) termination. Using the passive termination of a DAT or CD-ROM drive just doesn't cut it. FWIW, I couldn't even install FreeBSD with this kind of set up; I had to use an inline active terminator. I've also been recommending that people having these problems try disabling auto-termination in Adaptec BIOS and setting it manually. Justin would know better, but I think there could be a bug in the cable-detection algorithm when auto-termination is enabled. At least with the Linux aic7xxx driver, I've seen the cables improperly detected. Dan Eischen deischen@iworks.InterWorks.org