From owner-freebsd-scsi@FreeBSD.ORG Mon Oct 18 18:18:55 2004 Return-Path: Delivered-To: freebsd-scsi@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 53F9616A4CE for ; Mon, 18 Oct 2004 18:18:55 +0000 (GMT) Received: from email.eurowings.com (email.eurowings.com [193.96.182.7]) by mx1.FreeBSD.org (Postfix) with ESMTP id C06BC43D3F for ; Mon, 18 Oct 2004 18:18:54 +0000 (GMT) (envelope-from h.kipp@eurowings.com) Received: from localhost (localhost [127.0.0.1]) by email.eurowings.com (Postfix) with ESMTP id 19A0814E0C6 for ; Mon, 18 Oct 2004 20:18:54 +0200 (CEST) Received: from EXCH5.eurowings.com (unknown [10.100.1.37]) by email.eurowings.com (Postfix) with ESMTP id BF9EA14E0C9 for ; Mon, 18 Oct 2004 20:18:51 +0200 (CEST) X-MimeOLE: Produced By Microsoft Exchange V6.0.6556.0 content-class: urn:content-classes:message MIME-Version: 1.0 Date: Mon, 18 Oct 2004 20:18:29 +0200 Message-ID: X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: Interesting SCSI-Problem with Quantum Atlas10K3 (from freebsd-stable) Thread-Index: AcSmYLEp6F+UUQTlTDCbz5uRfwS37gO3XMIN From: "Kipp Holger" To: "Matthew Jacob" , X-Virus-Scanned: by amavisd-new at eurowings.com Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable X-Content-Filtered-By: Mailman/MimeDel 2.1.1 Subject: [SOLVED] RE: Interesting SCSI-Problem with Quantum Atlas10K3 (from freebsd-stable) X-BeenThere: freebsd-scsi@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: SCSI subsystem List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 18 Oct 2004 18:18:55 -0000 It was after all a defect disk drive (one of those three drives had an 'oscillating' writing performance (ie between 0 bytes/sec and maximum throughput) which gave all sorts of interesting behaviour within vinum as raid 5. This was very hard to debug under FreeBSD 4.10 and might explain why changing the available command queue length first looked like solving the problem... (because it changed the timing issues of the bad drive). Regards, Holger Kipp -----Original Message----- From: Matthew Jacob [mailto:mjacob44@yahoo.com] Sent: Wed 29.09.2004 22:12 To: Kipp Holger; freebsd-scsi@freebsd.org Cc:=09 Subject: Re: Interesting SCSI-Problem with Quantum Atlas10K3 (from = freebsd-stable) It's not a command queue length issue. The message indicates that we had a disk command timeout that when we turned around and called the interrupt service routine that the command did get serviced after all. You didn't say what kind of MPT h/w you're running on. =09 =09 __________________________________ Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! http://promotions.yahoo.com/new_mail=20