From owner-freebsd-questions Fri Apr 21 14:15:22 2000 Delivered-To: freebsd-questions@freebsd.org Received: from corinth.bossig.com (mail.dohboys.com [208.26.253.10]) by hub.freebsd.org (Postfix) with ESMTP id 80D5E37BDD4 for ; Fri, 21 Apr 2000 14:15:14 -0700 (PDT) (envelope-from kstewart@3-cities.com) Received: from 3-cities.com (unverified [208.26.241.228]) by corinth.bossig.com (Rockliffe SMTPRA 4.2.1) with ESMTP id ; Sat, 22 Apr 2000 14:19:22 -0700 Message-ID: <3900C4D0.10401817@3-cities.com> Date: Fri, 21 Apr 2000 14:14:56 -0700 From: Kent Stewart Organization: Columbia Basin Virtual Community Project X-Mailer: Mozilla 4.72 [en] (Windows NT 5.0; U) X-Accept-Language: en MIME-Version: 1.0 To: Erik Trulsson Cc: Artem Koutchine , questions@FreeBSD.ORG Subject: Re: SCSI vs UDMA IDE 32-bit References: <003601bfab80$82823c40$0c00a8c0@ipform.ru> <20000421204331.A1115@student.csd.uu.se> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Erik Trulsson wrote: > > On Fri, Apr 21, 2000 at 02:58:15PM +0400, Artem Koutchine wrote: > > I've made a little test. > > I took a bunch of files (~130 MB) and copied them from > > one filessystem to another. The HDD is IDE Quantum FB 10GB. > > While copying i run top and saw about 80% of cpu wasted on > > interupts. > > > > Then i anables 32bit access for the hdd and turned on DMA > > transfers. Repeated the test and saw only tiny 0.8% wasted on > > interupts, which is comparable to what SCSI takes. So, knowing > > that SCSI and IDE hdd are based on the same mechanical parts why > > should even use SCSI? Am I missing something? > > > > The advantages of SCSI are really noticable noticable when you have > several disks attached to the same controller. > With modern disks and controllers SCSI and IDE are just as good if you only > have a single disk attached. (And IDE disks are usually much cheaper.) You don't have to go that far before there is an advantage. I benchmarked my UDMA66 Maxtor against an older IBM-UW. The IBM was more than 10MB/s slower on sequential accesses of data. The IBM ran 8MB/s across all of "iozone's tests. I ran both tests with setiathome consuming all of the free cpu time. The Maxtor dropped below 750KB/s on random accesses. The IBM continued on at 8MB/s. I think the random test is consistent with a system build, which is the most time consuming I/O bound thing I am doing on this machine. You can imagine how the test would have looked with a 10K rpm IBM and a U2W controller :). Kent > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-questions" in the body of the message -- Kent Stewart Richland, WA mailto:kstewart@3-cities.com http://www.3-cities.com/~kstewart/index.html FreeBSD News http://daily.daemonnews.org/ SETI(Search for Extraterrestrial Intelligence) @ HOME http://setiathome.ssl.berkeley.edu/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message