Repeated drive corruption

[cranioscopical]

In my current system I have 4 x WD 500 GB SATA 7200 drives.

Here's my problem:

• File corruption on 2nd drive. 
• System freeze
• Chkdsk fixes on restart
• Repeat above
• WD tools say 2nd drive is A-OK
• MFT went bad on 2nd drive
• Chkdsk bails out
• Reformat drive
• Repopulate with fresh files
• WD tools say drive is A-OK
• File corruption on 2nd drive. 
• Gently whisper naughty words.

So, since data is far more important to me than even the hint of a dodgy drive, I'll replace the 2nd drive.

Can anyone suggest what the heck might be happening, though?
(In the past I've had WD drives fail immediately after WD tools say they're fine, so I haven't much faith in the A-OK reports.)

Thanks in advance!



Any useful experiences to share?
Advice, other than the obvious step of junking the drive?
I really wouldn't like to see this happen again when I put in the new drive.

[40hz]

Most likely the drive is screwed up. WD is not IMHO as reliable as they used to be.

If you want to eliminate other possibilities, you could also try the following:

Start simple. Try doing a partition boot sector repair on the drive followed by a reformat. Here's instructions if you don't already know how to do it:

http://pcsupport.about.com/od/fixtheproblem/ht/newbootsector.htm

Other possibilities:

- Firmware issues. Check to see if your BIOS or chipset firmware needs updating? I've found a lot of problems (80%) get fixed by installing the most recent BIOS and chipset drivers.

- Bad/marginal data cable. Swap with known good.

- Power. Try switching wire from PS to drive.

- Bad SATA port/controller on mobo. Try switching drive to an alternate port and retry.
  (Also check SATA settings in BIOS.)

Good luck, and let us know how you make out.

[cranioscopical]

Thanks for the feedback, I'll follow up on your ideas.

[cmpm]

You could try these two options after saving your data to a safe place.

http://www.dban.org/

http://www.lehigh.edu/security/gwscan/gwscan.htm

gwscan will make your drive raw, and need formated
but if it's something software related it will kill it

[40hz]

Oh yeah, one other thought. Check the connector "pins" on the drive. Give the two connectors a little wiggle and see if it feels loose, or there are any cracks in the board they're part of. I did slightly crack a connector one time when I was plugging in the cable too forcefully (probably because I wasn't pushing it on straight) and the drive would format for a while before it failed.

(The manufacturer denied a warranty claim on that drive BTW. And sad to say - rightly so! )

[cranioscopical]

Thanks again.
I'm hurrying at the moment but from a VERY quick look it looks as if
these (http://www.dban.org/) and (http://www.lehigh.edu/security/gwscan/gwscan.htm)
Will wipe anything they find.
Will I have an option ONLY to wipe the second of 4 drives, do you know?

TIA

[Shades]

If these tools come in iso format, burn them to cd. Turn the PC off (switch at the back!), then make sure only the bad drive is connected to the mainboard. Without a physical connection it is impossible for this software to mess up your other drives and their content. Set the BIOS to boot from CD and (re)boot your system.

Follow the instructions from the tool, wait until it is finished and turn machine off (switch at the back!). Connect all drives again and boot your system (without the boot cd of course) and Windows should recognize a clean/pristine drive.
Use drive management console to partition and format the drive.

And tadaaa, cranioscopical is a happy camper again 

[mouser]

if it were me, i might buy another drive and swap it for the one that's acting up.. if no more errors, then the one you removed is evil and needs to be removed from your house.  if the new one exhibits same behavior then its something else like mobo, cable, etc. but most important thing for me would be to stop trusting that drive immediately.  view anything on it as potentially loseable.

[cmpm]

Thanks again.
I'm hurrying at the moment but from a VERY quick look it looks as if
these (http://www.dban.org/) and (http://www.lehigh.edu/security/gwscan/gwscan.htm)
Will wipe anything they find.
Will I have an option ONLY to wipe the second of 4 drives, do you know?

TIA

If you have a drive you do not want affected by these two programs, disconnect them from the computer.

Test each drive separately on the machine.
If you want to go to all that wirk.
Otherwise mouser's suggestion is workablr as well.

Again if you run nuke and boot or gwscan,
remove any drives you do not want wiped.
Though there are options I don't take those chances.

[cranioscopical]

Thanks everyone!

As I've been told, I don't have the brains I was born with.
Not thinking at all, of course I should pull the other drives 

Just to be clear, the troublesome drive is being replaced... no question. 
Data (and reliability) are worth far more to me than the cost of a replacement drive.

While replacing, I will check out all the other points, though (connections/cable etc).
I'll put the troublesome drive into a different computer when I run the tests as I need the machine that it's now in to be functioning.

My original query was aimed at avoiding a recurrence of this with a replacement drive (this one's only 6 months old).

[4wd]

Start simple. Try doing a partition boot sector repair on the drive followed by a reformat. Here's instructions if you don't already know how to do it:

http://pcsupport.about.com/od/fixtheproblem/ht/newbootsector.htm

The only problem with that is it will only work if the drive has a boot partition, ie. an Active Primary partition.
Because of the exceedingly stupid way both MS OS's and BIOS's default to "oh look, an IDE drive that must be the boot drive" mode, the only drive I have in my system marked for Active is also the only one with a Primary partition - all others use logical partitions.  Saves me from Windows stupidly marking pure data drives as Active which then results in when you take the drive out the computer won't boot.

Other possibilities:
- Firmware issues. Check to see if your BIOS or chipset firmware needs updating? I've found a lot of problems (80%) get fixed by installing the most recent BIOS and chipset drivers.
- Bad/marginal data cable. Swap with known good.
- Power. Try switching wire from PS to drive.
- Bad SATA port/controller on mobo. Try switching drive to an alternate port and retry.
  (Also check SATA settings in BIOS.)

My money's on a cable problem, (which means I'll be wrong ), backup the data on both the corrupting drive and another then just swap the connectors, (power and data), over on them and see how it goes.

If you can, spend the money on SATA data leads that have a locking clip, as supplied with all Gigabyte motherboards.  SATA connectors would have to be the most poorly designed connectors ever wrt computers.  I've had more problems with SATA connectors in 2 years than IDE in 10 years.

Does the eventlog give any extra info ?

Just to be clear, the troublesome drive is being replaced... no question. 
Data (and reliability) are worth far more to me than the cost of a replacement drive.

The only problem with that thought is, it might not be the drive - so you will have solved nothing while spending more money.  The simple act of swapping the cables will prove it into the system, (motherboard, cables or OS), or the drive.

While replacing, I will check out all the other points, though (connections/cable etc).
I'll put the troublesome drive into a different computer when I run the tests as I need the machine that it's now in to be functioning.

My original query was aimed at avoiding a recurrence of this with a replacement drive (this one's only 6 months old).

If it's only 6 months old it's still under warranty, just get it replaced.

[f0dder]

Yeah, SATA connectors are really stupidly designed - they wiggle loose too easily, and it's also too easy to break the connector (both the cable-end as well as the connector on your disk drive). A pity really, since it's otherwise pretty nice, slim cables and all.