From owner-freebsd-scsi Tue Feb 1 14:20:56 2000 Delivered-To: freebsd-scsi@freebsd.org Received: from caspian.plutotech.com (caspian.plutotech.com [206.168.67.80]) by builder.freebsd.org (Postfix) with ESMTP id D19903F75; Tue, 1 Feb 2000 14:20:51 -0800 (PST) Received: from caspian.plutotech.com (localhost [127.0.0.1]) by caspian.plutotech.com (8.9.3/8.9.1) with ESMTP id PAA62239; Tue, 1 Feb 2000 15:21:04 -0700 (MST) (envelope-from gibbs@caspian.plutotech.com) Message-Id: <200002012221.PAA62239@caspian.plutotech.com> X-Mailer: exmh version 2.1.0 09/18/1999 To: Greg Lehey Cc: "Justin T. Gibbs" , "Justin T. Gibbs" , Gary Palmer , scsi@FreeBSD.org, up@3.am, Wilko Bulte Subject: Re: hardware vs software stripping In-reply-to: Your message of "Wed, 02 Feb 2000 08:22:14 +1030." <20000202082214.S76348@freebie.lemis.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 01 Feb 2000 15:21:04 -0700 From: "Justin T. Gibbs" Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >That doesn't correspond to any of the definitions I have seen. Where >did you get it from? This is from the RAID Advisory board's RAID book. Take a look at www.raid-advisory.com for ordering details. >> In RAID4, it is supposed to be a multiple of your transaction size > >Where do you get the term "transaction" from? I haven't seen it in From the dictionary? 8-) The point is that your system is such that you may be able to satisfy a request by only reading one component of the stripe. >any RAID documentation. In ufs, there is no fixed size. Sure there is, the block size (i.e. 8k.) But then again, you wouldn't usually use RAID 3 or 4 for a filesystem. >> so you can perform partial read (assuming you don't need parity >> verification) > >When would you need parity verification for reads? When you are paranoid about the disks or some other portion of your system screwing up the data without telling you. >> and RMW operations to update the parity for updating contiguous >> transactions that are not as large as the stripe. > >I'd call both of these RAID-4, considering that RAID doesn't use the >term "transaction". Sure it does. In RAID-3, your transaction size *is* the stripe size. In RAID-4, it may be less than the stripe size. >>>> Pluto uses a RAID-3 system in its video server products and it is >>>> certainly not striped on a byte level. >>> >>> So how exactly is it striped? >> >> Our stripe size is 1-2MB with the per-drive stripe component size >> varying depend on the number of drives in the system. > >So what's the difference from RAID-4? I can accept the fact that this >is the way you use the term "RAID-3", but it conflicts with all >documentation I have seen, and you haven't presented any other >evidence. Here's what the RAID Advisory's RAID book has to say: RAID Level 3 Raid Level 3 adds redundant information in the form of parity to a parallel access striped array, permitting regeneration and rebuilding in the event of a disk failure. One "strip" of parity protects corresponding strips of data on the remaining disks. Raid Level 3 provides high data transfer rate and high data availability, at an inherently lower cost than mirroring. Its transaction performance is poor, however, because all RAID Level 3 array member disks operate in lockstep. RAID Level 4 Like RAID Level 3, RAID Level 4 uses parity concentrated on a single disk to protect data. Unlike RAID Level 3, however, a RAID Level 4 array's member disks are independently accessible. Its performance is therefore more suited to transaction I/O than large file transfers. Raid Level 4 is seldom implemented without accompanying technology, such as a write-back cache, because the dedicated parity disk represents an inherent write bottleneck. -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message