From owner-freebsd-scsi@FreeBSD.ORG Fri Feb 18 20:30:52 2005 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 775C016A4CE for ; Fri, 18 Feb 2005 20:30:52 +0000 (GMT) Received: from svm.csie.ntu.edu.tw (svm.csie.ntu.edu.tw [140.112.90.75]) by mx1.FreeBSD.org (Postfix) with ESMTP id 721F743D2F for ; Fri, 18 Feb 2005 20:30:51 +0000 (GMT) (envelope-from rafan@svm.csie.ntu.edu.tw) Received: from svm.csie.ntu.edu.tw (localhost [127.0.0.1]) by svm.csie.ntu.edu.tw (8.13.1/8.13.1) with ESMTP id j1IKTFpC071888; Sat, 19 Feb 2005 04:29:15 +0800 (CST) (envelope-from rafan@svm.csie.ntu.edu.tw) Received: (from rafan@localhost) by svm.csie.ntu.edu.tw (8.13.1/8.13.1/Submit) id j1IKT1be071887; Sat, 19 Feb 2005 04:29:01 +0800 (CST) (envelope-from rafan) Date: Sat, 19 Feb 2005 04:29:01 +0800 From: Rong-En Fan To: Rong-En Fan Message-ID: <20050218202901.GA71842@svm.csie.ntu.edu.tw> References: <20050204102558.GA2001@svm.csie.ntu.edu.tw> <1108667040.656.20.camel@spirit> <20050218071746.GA41419@svm.csie.ntu.edu.tw> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20050218071746.GA41419@svm.csie.ntu.edu.tw> User-Agent: Mutt/1.5.7i cc: delphij@delphij.net cc: ob@e-Gitt.NET cc: scsi@freebsd.org Subject: Re: Problem with mpt(4) and Infortrend RAID 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: Fri, 18 Feb 2005 20:30:52 -0000 On Fri, Feb 18, 2005 at 03:17:46PM +0800, Rong-En Fan wrote: > On Fri, Feb 18, 2005 at 03:04:00AM +0800, Xin LI wrote: > > Have you got a chance to test this, I have gotten some similar problem > > and got rid of them with the following patch applied (credit goes to > > gibbs@ who fixed this last November): > > > > http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/cam/cam_xpt.c.diff?r1=1.142&r2=1.142.2.2 > > > > This might be important since I think this has affected many users who > > wants to use multiple LUN's, and thus warrants a new errata candidate > > for 5.3-RELEASE. > > > > Thanks in advance! > > I have tested this with 5.3-p5: > > 1. 2 LD -> 1 channel, 320MB/s ok (with patch) > 2. 2 LD -> 1 channel, 320MB/s failed (without patch) > > I simulate io by benchmark/blogbench This box now serves as NFS server with 20 clients. and has some heavy io (about 30MB/s, lasting for 5 hours or so) with this patch works fine. (2 channels, each channel is 1 LD running at 320MB/s). Though you said this patch solves multiple LUN problem, I think this also solve my problem here. Please consider merge to RELENG_5_3. Thanks. Regards, Rong-En Fan