Date: Sun, 6 Jan 2002 14:23:06 +0000 From: Scott Mitchell <scott.mitchell@mail.com> To: Kevin Golding <kevin@caomhin.demon.co.uk> Cc: Cliff Sarginson <cliff@raggedclown.net>, FreeBSD Questions <freebsd-questions@FreeBSD.ORG> Subject: Re: Parts I recommend (formerly "Workstation and server-market") Message-ID: <20020106142306.C338@localhost> In-Reply-To: <4JD6R5A1lFO8EwPI@caomhin.demon.co.uk>; from kevin@caomhin.demon.co.uk on Sun, Jan 06, 2002 at 02:04:37PM %2B0000 References: <20020105195536.27073ca2.johann@broadpark.no> <20020105170230.636999bb.matthew@starbreaker.net> <20020106071439.GE1003@raggedclown.net> <0a5b01c19683$d1087880$0a00000a@atkielski.com> <20020106121009.A338@localhost> <4JD6R5A1lFO8EwPI@caomhin.demon.co.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, Jan 06, 2002 at 02:04:37PM +0000, Kevin Golding wrote: > In message <20020106121009.A338@localhost>, Scott Mitchell > <scott.mitchell@mail.com> writes > >I'd not be so sure that they're mechanically identical -- IME the failure > >rate for ATA drives is much higher than the SCSI ones, despite the fact > >that the SCSI server drives do a lot more work than the ATA units in the > >workstations. Of course most SCSI drives are going to be running at 10Krpm > >or above, while your typical ATA drive does 5400 or 7200rpm, so there's > >some extra noise for you right away... I suspect there might be some > >mechanical differences between a drive that has to run reliably for years > >at 15Krpm and one that's idle half the time at 5400rpm in someone's > >desktop. > > Isn't it exactly that reason? A constant velocity should result in less > wear than if it ran in a stop/start manner. If you make an ATA drive > spin all day long it should have a lot less mechanical wear and such, > SCSI would probably still have an edge since they're more likely to be > optimised for such things but the gap would be narrower. Maybe... I guess there's a balance between the wear caused simply by having the thing spinning, and the extra wear a start/stop cycle (and all the consequent thermal stress) causes... I wouldn't expect one daily power-cycle to dramatically shorten the *useful* life of the drive (it'll probably be replced by a bigger/faster one long before it actually dies), for example. All of our workstations are left on permanently, although the drives won't be doing much for half the day... they still die more often than the SCSI drives. Which is fine, really -- the SCSI drives cost a lot more so I expect them to last longer :-) > Also don't forget that servers tend to be looked after much more in > general, a lot of desktops take a fair bit of abuse, even if it's just a > bit more dust floating around inside the case it's one more thing a > desktop needs to deal with. Fair point -- I know my workstation tends to get kicked more often than it probably should. It gets pretty hot inside a lot of desktops, too. I'd still expect the ATA and SCSI drives to be mechanically different, however -- for example, IBM DeskStar (ATA) and Ultrastar (SCSI) drives look very different externally; it's a fair bet that some of the stuff inside the case is different too. If I ever have one die outside of warranty I'll open it up and take a look :-) Scott -- =========================================================================== Scott Mitchell | PGP Key ID | "Eagles may soar, but weasels Cambridge, England | 0x54B171B9 | don't get sucked into jet engines" scott.mitchell@mail.com | 0xAA775B8B | -- Anon To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020106142306.C338>