Tile Floors And Hard Disks

[Michael W]

For almost the last six months my Windows PC has been resting beside my desk directly on the tile floor of my apartment. A couple of times now I've had to run scandisk (the computer has two hard disks) to mark off bad sectors. The computer is now making funny scratching noises again usually indicative of trouble it's having doing a disk write and I suspect the sofa I pushed quickly across the floor an hour ago may not've helped. Anyway, it looks like I'll be running scandisk again tonight before bed to fix these latest apparent problems. An obvious low-tech solution to having this come up again would seem simply to add some padding for the computer to rest on and act as a sort of shock absorber -- maybe a few old magazines would do the trick. Does all this add up or should I be taking some other action?

Edit: On second thought these are five year old IBM 75 gig disks. Am I maybe just seeing the beginnings of a total crash and should prepare accordingly.

Edited April 28, 2006 by Michael W

[Guest endure]

If you have IDE drives and they're openly reporting bad sectors you need to back them up and change them TODAY. Once an IDE drive has got to the stage where it can't remap bad sectors internally it's on the way out.

[Michael W]

If you have IDE drives and they're openly reporting bad sectors you need to back them up and change them TODAY. Once an IDE drive has got to the stage where it can't remap bad sectors internally it's on the way out.

Hmmm, not what I wanted to hear but also my suspicion in the post edit I just made. Yes, they are IDE drives (IBM, supposedly among the better makers). Any recommendations for a replacement(s)?

[Guest endure]

I can't really recommend any particular make. I tend to buy what's cheapest at the time. If you're looking for quiet drives the Samsung Spinpoints are quite good.

[Michael W]

If you have IDE drives and they're openly reporting bad sectors...

Well if by openly you mean a big white Windows error message or a Blue Screen of Death , no I'm not getting that, only the scratching noises which stop on their own as I presume the OS stops trying to write to that sector and chooses another. It's only after this and as a precaution that I run scandisk and at that point the bad sectors are actually reported and marked off. Does that allow me to breath a little easier or am I grasping for straws and still need to get a new hard disk Right Away?

[cobra]

If you suspect excessive vibration from the local environment you'd probably be seeing write errors to optically drives as well, I don't think that's the issue.

Drives are cheap, I replaced both of mine (I run a stripped RAID 2 config) with Seagate's because I couldn't stand the high pitched drive whine of the cheap Western Digital drives, the Seagates use liquid bearing technology and are absolutely noiseless. Cost about $45 a piece for the 60g,

Change the drives out or just wait until they fail outright, because you back up your data religiously, right ?

[Guest endure]

If you have IDE drives and they're openly reporting bad sectors...

Well if by openly you mean a big white Windows error message or a Blue Screen of Death , no I'm not getting that, only the scratching noises which stop on their own as I presume the OS stops trying to write to that sector and chooses another. It's only after this and as a precaution that I run scandisk and at that point the bad sectors are actually reported and marked off. Does that allow me to breath a little easier or am I grasping for straws and still need to get a new hard disk Right Away?

IDE drives have got a little hidden area stashed away that they use to remap bad sectors. If it's got to the stage that they're reporting bad sectors it means that they've used up that hidden area and things are getting desperate. It's best to replace them asap - it's also good to make sure all your data is backed up now.

[Firefoxx]

5 years is a fairly long time for hard drives to function. Not only that, but IBM is notorious for a large batch of defective drives... the Deskstar series. People have gone so far as to call them the "deathstar" series, since they have a very high rate of failure. This is said to be one of the reasons IBM stopped producing hard drives.

Of course, you should backup your data ASAP and replace the drive. It's doubtful that vibration caused the failure... heat is a more common cause these days. Today's drives run *extremely* hot... to the point that it's uncomfortable to touch them when they're running. I've seen several drives fail from inadequate cooling, and inadequate cooling is the norm in today's cases. A good case will have a fan directly cooling the hard drives with air sucked from outside the case.

Any of today's main brands will do (WD, Seagate, Maxtor), as their performance is similar. More important is the warranty (where, who, how, how long). If you want the most capacity for the buck, you should get 200-250GB (do the math).

Edited April 29, 2006 by Firefoxx

[autonomous_unit]

Having your computer on the floor shouldn't really cause vibrations unless the computer is literally bouncing around... the drive has internal rubber shock absorbers for small vibrations. Have you kept it clean inside? Being on the floor, it might fill up with dust and then get hotter than it would normally...

As for bad blocks, you might try to get a S.M.A.R.T. utility program to query the drive condition. I don't know anything about windows, but under Linux you would use the "smartctl" tool. There are two important measures for bad blocks: the reallocated sector count and the pending sector count.

The reallocated sector count is what people are referring to above about the "stash" for the drive to remap errors on its own. The pending sector count are bad sectors that the drive is still trying to remap. There is actually a period of time between when the drive detects an error and when it remaps which is the time when it desparately tries to re-read the bad block (It cannot write it to a new location until it can get lucky and read the old block data). During this time, the drive may make terrible noises as it rapidly seeks and seeks to keep trying to read the bad block. It will also seem relatively unresponsive during this time.

Having many pending sectors is bad, since it indicates something has caused sudden damage. Having one is annoying, but not critical. Running out of reallocated sectors is the big problem, as once that happens the future errors will just accumulate and cause problems for the OS which isn't really prepared to handle that these days.

[astral]

I doubt that moving the sofa had much effect.

A tile floor should be fairly solid.

Get that disk backed up and chaged as soon as possible.

[Crossy]

I doubt that having your PC on a solid floor is causing problems.

Vibration does kill drives, we were losing drives after 2-3 months when we first started installing computers in railway equipment rooms. Heavy vibration from the trains was the cause, but it was readily felt and pretty continuous. Isolated rubber mounts cured the problem.

As others have said, I suspect your drive is on its way out, time to bite the bullet and replace it.

[Firefoxx]

Yep, a railway equipment room is a whole different ball game from your apartment room... you'd better have industrial-strength equipment to fit the task. As for the apartment room, unless you have something akin to a perpetual contruction site using jackhammers next door, you don't need to worry about vibration causing a failure.

[Michael W]

IBM is notorious for a large batch of defective drives... the Deskstar series. People have gone so far as to call them the "deathstar" series, since they have a very high rate of failure.

Yep, that's what it turns out I have (the things you learn...).

I found a Windows version of smartmontools here and when I ran smartctl -i /dev/hda it reported:

Model Family: IBM Deskstar 40GV & 75GXP series (all other firmware)

Device Model: IBM-DTLA-307075

It also said:

==> WARNING: IBM Deskstar 40GV and 75GXP drives may need upgraded SMART firmware

and provided some URLs for updates. IBM has even been sued because of the quality of this drive. A bunch more links here for those interested.

Anyway, from one of the IBM links smartctl gave I downloaded an identification utility which said I needed new firmware. I ended up Googling my firmware version (TXADA50C) and found this discussion thread and from a link there went to the Dell drivers update page and found this.

Now the only problem here is in order to run it, the update has to be installed on a floppy disk from which you boot as part of the update process. Well, I haven't used a floppy in years and didn't think I had any lying around, having dumped them all during The Big Move. I was about to resign myself to going to Foodland and buying a box of 10 when I remembered that in my collection of pre-digital photos I always chose the option of a set of JPG scans included on a floppy. So I pulled out a set of photos from a New Years 2002 trip to Pattaya, found the floppy, backed it up, installed the update on it, and ran it by booting from the floppy. It seemed to run fine, smartctl said I now had firmware version TXADA5AA and the IBM identification utility also said I was up to date.

Supposedly all this should fix the scratching noise by moving the drive heads around from time to time while idling. But I imagine at best all I've done is buy a little time and I'll still be replacing these drives. I should actually consider myself lucky in comparison to this guy's experience.

[meadish_sweetball]

I recently started a similar thread as you, and my hard drive is now history. I was lucky to have backed up all essential information from it before it really kicked the bucket. You should try to do the same.

Incidentally, do you know what hard disk it was? A Hitachi Deskstar (IBM/Hitachi apparently collaborate when it comes to hard disks).

[Guest endure]

Supposedly all this should fix the scratching noise by moving the drive heads around from time to time while idling. But I imagine at best all I've done is buy a little time and I'll still be replacing these drives. I should actually consider myself lucky in comparison to this guy's experience.

Without wishing to excuse IBM at all the guy made a bit of a rod for his own back. He installed a RAID 0 array which is inherently more unreliable than a single drive. Lose one drive and you've lost both. He also wasn't doing any backups. I've had a 60GXP 40GB running continuously in my machine for 4 years now.

[Michael W]

Incidentally, do you know what hard disk it was? A Hitachi Deskstar (IBM/Hitachi apparently collaborate when it comes to hard disks).

I think Hitachi took over IBM's PC hard disk division some time (6 months, a year?) after I got my pair of matching 75 gig drives. Here's the entire output from smartctl -i /dev/hda:

smartctl version 5.36 [i686-mingw32-98se] Copyright © 2002-6 Bruce Allen

Home page is http://smartmontools.sourceforge.net/

=== START OF INFORMATION SECTION ===

Model Family: IBM Deskstar 40GV & 75GXP series (all other firmware)

Device Model: IBM-DTLA-307075

Serial Number: YSDYSF54205

Firmware Version: TXADA5AA

User Capacity: 75,329,372,160 bytes

Device is: In smartctl database [for details use: -P show]

ATA Version is: 5

ATA Standard is: ATA/ATAPI-5 T13 1321D revision 1

Local Time is: Sun Apr 30 09:23:14 2006 PST

==> WARNING: IBM Deskstar 40GV and 75GXP drives may need upgraded SMART firmware.

Please see http://www.geocities.com/dtla_update/ and

http://www-3.ibm.com/pc/support/site.wss/d...ocid=MIGR-42215 or

http://www-1.ibm.com/support/docview.wss?uid=psg1MIGR-42215

SMART support is: Available - device has SMART capability.

SMART support is: Disabled

SMART Disabled. Use option -s with argument 'on' to enable it.