From owner-freebsd-geom@FreeBSD.ORG Sun Dec 19 12:12:53 2010 Return-Path: Delivered-To: freebsd-geom@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 167B9106566B for ; Sun, 19 Dec 2010 12:12:53 +0000 (UTC) (envelope-from lulf@pvv.ntnu.no) Received: from decibel.pvv.ntnu.no (decibel.pvv.ntnu.no [IPv6:2001:700:300:1900::1:2]) by mx1.freebsd.org (Postfix) with ESMTP id 5DEEA8FC12 for ; Sun, 19 Dec 2010 12:12:52 +0000 (UTC) Received: from 38.247.251.212.customer.cdi.no ([212.251.247.38] helo=[192.168.1.2]) by decibel.pvv.ntnu.no with esmtpsa (TLS-1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.69) (envelope-from ) id 1PUI86-00083V-9w; Sun, 19 Dec 2010 13:12:49 +0100 Message-ID: <4D0DF6B7.4000904@pvv.ntnu.no> Date: Sun, 19 Dec 2010 13:12:39 +0100 From: Ulf Lilleengen User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.13) Gecko/20101208 Thunderbird/3.1.7 MIME-Version: 1.0 To: Lev Serebryakov References: <1701263071.20101219003441@serebryakov.spb.ru> In-Reply-To: <1701263071.20101219003441@serebryakov.spb.ru> Content-Type: text/plain; charset=windows-1251; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0.2 (/) X-Spam-Report: Status=No hits=0.2 required=5.0 tests=ALL_TRUSTED, TVD_RCVD_IP version=3.2.4 Cc: freebsd-geom@FreeBSD.org Subject: Re: Good configurable I/O benchmark? X-BeenThere: freebsd-geom@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: GEOM-specific discussions and implementations List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 19 Dec 2010 12:12:53 -0000 On 12/18/2010 10:34 PM, Lev Serebryakov wrote: > Hello, Freebsd-geom. > > > Is here any good I/O benchmark? What do I mean by `good': > > (1) Multiprocess workloads with different workloads per processe (thread) > (a) Sequential R or W. > (b) Random R and W with configurable R:W ratio. > (c) Configurable min/max/avg distance seek in random workloads. > (2) All above with configurable block size. > (3) All of above with configurable I/O queue depth. > I have used benchmarks/rawio in the past, which seems to work well, but not sure it fits all your requirements though. Ulf