From owner-freebsd-hardware Sat Sep 20 11:26:26 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id LAA20410 for hardware-outgoing; Sat, 20 Sep 1997 11:26:26 -0700 (PDT) Received: from po1.glue.umd.edu (root@po1.glue.umd.edu [128.8.10.97]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id LAA20404 for ; Sat, 20 Sep 1997 11:26:23 -0700 (PDT) Received: from aliasing.eng.umd.edu (crb@aliasing.eng.umd.edu [129.2.98.137]) by po1.glue.umd.edu (8.8.7/8.8.7) with ESMTP id OAA08888; Sat, 20 Sep 1997 14:26:12 -0400 (EDT) Received: from localhost (crb@localhost) by aliasing.eng.umd.edu (8.8.7/8.8.7) with SMTP id OAA01386; Sat, 20 Sep 1997 14:26:07 -0400 (EDT) X-Authentication-Warning: aliasing.eng.umd.edu: crb owned process doing -bs Date: Sat, 20 Sep 1997 14:26:05 -0400 (EDT) From: "Christopher R. Bowman" X-Sender: crb@aliasing.eng.umd.edu To: Mark Mayo cc: freebsd-hardware@FreeBSD.ORG Subject: Re: HD failure; possible causes?? In-Reply-To: <19970919230016.40089@vinyl.quickweb.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-hardware@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Fri, 19 Sep 1997, Mark Mayo wrote: All that air conditioning, and power conditioning and UPS junk all sits on one side of the disk drives, not between them, I mean if you follow the current it comes out of the wall into the ups then into the redundant power supply, then to the drives, in all likely hood the drive power is tapped off them same point, their is no isolation between them, so what if the head crash of one drive cause a transient dip or spike in the regulated output of the power supply inside the case, air conditioning and UPS wont really help this. I am an electrical engineer (least that's what my degree sais) and this means that I am absolutlely posatively qualified to speak on this subject with NO credability whatsoever, so please take what I say with a large pile of salt. I am just speculating since it is unlikely you will ever really know what happened. > Hi all. Well, this weekend I had a couple of CCD disk arrays go belly up, > and I'm curious if anyone has had any experience with multiple disks > crashing at once. > > Originally I thought one of the disks in the pair had crashed, but when I > sent the drives to a data recovery place the guy informed me that both > disks were pooched - both of them had the heads physically touch the > media, destroying the platters and all data of course. > > One of the disks was an old Seagate Baracuda (the 2XL I think) and one > was a relatively new Quantum (Fireball). I've had several of the older > Baracudas blow up in the past, which is what I assumed happened here. > Needless to say I was quite surprised to learn that the Quantum suffered > the same death - at the same time presumably... According to the > data recovery guy, when 2 disks fail like this simultaneously in about > 80% of the cases he sees the cause of the problem was not the disks > themselves, but the computer or disk tower. My disks were held in a > pretty decent tower from Open Storage Solutions, with redundant power > supplies and redundant cooling. Typically, very reliable disk towers and > I've never had a problem with them in the past. Both te computer and the > disk tower were plugged into a nice 1400VA APC UPS - so I really doubt there > was a power surge or anything that got to the drives... > > Naturally I'm quite a bit worried that this might happen again, or to other > disks in the tower... So this leaves me with a possible problem with the > computer. The guy said "the motherboard"... I'm having a really hard time > imagining how the computer could have cause the destruction of 2 disks. Yet > the odds seems low enough for 2 disks to die in unison (plus the advice of > the "expert") that I'm trying to imagine where the problem could lie. > Maybe the old Adaptec 2940 SCSI controller? The only reasonable event I > can think of would be some sort of static discharge along the SCSI cable. > > Anyhow, if any body has any experience with drive failures I'd love to hear > from you!! This whole event has me very nervous! :-) > > TIA, > -Mark > > > -- > ------------------------------------------------------------------------ > Mark Mayo mark@quickweb.com > RingZero Comp. http://vinyl.quickweb.com/mark > > finger mark@quickweb.com for my PGP key and GCS code > ------------------------------------------------------------------------ > "The newest book, The Dilbert Future, took a broader view, describing how > idiots will threaten every aspect of business, technology and society in > the future." --Scott Adams > --------- Christopher R. Bowman crb@Glue.umd.edu My home page