From owner-freebsd-stable Wed Jun 6 5:48:20 2001 Delivered-To: freebsd-stable@freebsd.org Received: from public.ndh.com (public.ndh.net [195.94.90.21]) by hub.freebsd.org (Postfix) with ESMTP id AD92537B403 for ; Wed, 6 Jun 2001 05:48:16 -0700 (PDT) (envelope-from t.seck@ndh.net) Received: from basildon (port1030.duesseldorf.ndh.net [62.40.8.30]) by public.ndh.com (8.9.3/8.8.0) with ESMTP id OAA24526 for ; Wed, 6 Jun 2001 14:48:11 +0200 (MET DST) Envelope-to: freebsd-stable@FreeBSD.ORG Received: from tmseck by basildon with local (Exim 3.22 #1) id 157cjE-0000Dp-00 for freebsd-stable@FreeBSD.ORG; Wed, 06 Jun 2001 14:48:04 +0200 Date: Wed, 6 Jun 2001 14:48:04 +0200 From: Thomas Seck To: freebsd-stable@freebsd.org Subject: Re: Strange SCSI error message Message-ID: <20010606144804.B724@basildon> Mail-Followup-To: freebsd-stable@FreeBSD.ORG References: <20010605123023.S4259-100000@merlin.th.physik.uni-bonn.de> <20010605125635.A1349@basildon> <20010606034157.B97958@mail.webmonster.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <20010606034157.B97958@mail.webmonster.de>; from karsten@rohrbach.de on Wed, Jun 06, 2001 at 03:41:57AM +0200 Organization: private site in Germany X-PGP-KeyID: DF46EE05 X-PGP-Fingerprint: A38F AE66 6B11 6EB9 5D1A B67D 2444 2FE1 DF46 EE05 Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Jun 06 2001, Karsten W. Rohrbach wrote: > there were some questions a while ago concerning the freebsd vs linux > scenario where a freebsd fails with SCB timeouts while a linux runs > along with apparently no problems at all. the quick answer for this > particular behaviour in freebsd is, that the io subsystem is designed to > put more load onto the scsi devices than the implementation of other > os'es io subsystems. so, basically, what happens is, that you only see > the timeouts under freebsd, of course, since it arbits the scsi bus more > often, and with more data in the same period of time compared to other > implementations. Yes, this might well be the problem. What bothers me is that these errors only appeared when the box had been _really_ idle for hours. > terminate correctly, symmetric terminators, active, for u160. use good > cables, no vobis crap ;-) Vobis? Hey! (Note: Vobis is a chain of computer shops in Germany notorious for selling *really* crappy HW for about 15 years now) > i use amphenol cables for external links and > gr-kabel (german distributor) for internal wiring. this solved most of > my scsi problems with u160. Well, I hope I never have to see these problems again, cross fingers. Thank you for your response and regards from the rainy rhinelands, Thomas Seck To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message