From owner-freebsd-hardware Tue Jan 4 10:37:13 2000 Delivered-To: freebsd-hardware@freebsd.org Received: from feral.com (feral.com [192.67.166.1]) by hub.freebsd.org (Postfix) with ESMTP id 7CC5E15265 for ; Tue, 4 Jan 2000 10:37:11 -0800 (PST) (envelope-from mjacob@feral.com) Received: from semuta.feral.com (semuta [192.67.166.70]) by feral.com (8.9.3/8.9.3) with ESMTP id KAA09977; Tue, 4 Jan 2000 10:34:57 -0800 Date: Tue, 4 Jan 2000 10:37:06 -0800 (PST) From: Matthew Jacob Reply-To: mjacob@feral.com To: Jim Doherty Cc: Mitch Collinsworth , hardware@FreeBSD.ORG Subject: Re: differences between SCSI and EIDE [was: wanna buy an EIDE harddisk ... 5400 or 7200 for home use (noise)] In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-hardware@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > On Tue, 4 Jan 2000, Matthew Jacob wrote: > > > > > On the other hand, the newer bigger drives are getting able to basically > > consume most available bus bandwidth. If the numbers I've seen recently > > for drives being able to do ~24MB/s off the platter are indicative of > > There is a big difference between raw bandwidth and actual usage > patterns. Accessing a lot of small files drives a lot of overhead > and you are lucky to drive more than a couple hundred KB. I once > copied a large news spool off of a 9 gig drive and had the copy > operation take over 8 hours on an AIX system. By unmounting the > filesystem and performing a copy of the logical volume it took > under an hour. The filesystem overhead in opening/closing/seeking > files has a big impact on being able to drive i/o. For everything > but writing/reading very large files the SCSI bus is not going to be > the limit. My guess is that the overhead in the standard i/o > libraries might even be able to get in the way, but I have not > measured it. Yes, but the head actuator mechanisms are the same for EIDE and SCSI, so seeking times are the same. The only other issue might be how much onboard buffering EIDE has compared to SCSI drives. A lot of the speed of operations you mention above are solved, btw, with both softupdates (for making directory operations B_ASYNC) and clustering. The usage patterns for multiple threads will be, of course, much different. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hardware" in the body of the message