Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

Yesterday I came home from work to notice that not only had my PC reboot, it was stuck in the BIOS detecting hard drives. I eventually got the machine to boot but once I got into Windows it locked up. When I was able to get back into Windows (Win XP SP3) I noticed an error in the system event log: "The driver detected a controller error on \Device\Harddisk1\D" with Event ID: 11. I installed the SpeedFan utility and was able to capture the SMART information for both of my drives:

Hard Drive 0:

SpeedFan SMART

Hard Drive 1:

SpeedFan SMART

I don't know exactly how to interpret this data, but despite all the "OK"s I'm not entirely convinced that everything is ok. It seems like some of the values are beyond the thresholds which would indicate not ok.

Oddly enough, at this point the machine seems to be pretty stable. Any insight would be appreciated.

Edit: I appreciate everyone's concern, but I already have all my data backed up. At this point I'm just trying to repair the machine, data loss isn't a concern.

share|improve this question
6  
rather than worrying about SMART data capture you should immediately backup your data while you can, only then proceed with diagnosis. "pretty stable" can be a pretty treacherous assumption that may cost you dearly. –  Molly7244 Oct 1 '09 at 12:28
    
That is a good point. I actually keep regular backups and already backed up the few things that changed since my previous backup now that I am seeing some issues. At this point I'm just concerned with getting the machine back to a healthy state. I don't want to just replace the drives if that isn't the issue (is it possible that both drives are going bad at the same time?). –  Jason Oct 1 '09 at 12:52
    
possible? yes, likely? not really :) –  Molly7244 Oct 1 '09 at 12:59
1  
@Jason - I edited your post to put the images in there for you :) –  Jared Harley Oct 1 '09 at 14:24
add comment

6 Answers

I agree with mouviciel, save data first, then diagnose.

I would take that one step further- pull the hard disk, and attach it with a USB converter to a known good machine, -or- boot a Linux LiveCD or BartPE. Then, as you recover your data, your hard disk will not be burdened with O/S overhead, nor your O/S burdened with boot errors.

EDIT: I forgot to answer your question. Yes, your hard drive is dying. So is mine and everyone else's. The real question is, how fast is it dying?

share|improve this answer
1  
+1 the EDIT is a sound answer, though it doesn't really help ^^ –  Oskar Duveborn Oct 2 '09 at 12:33
add comment

As with others, the first thing is to backup your data (which you should do anyway). After that, feel free to try the built in windows disk/file system repair tools, "chkdsk" is one.

One other option is GRC's SpinRite. It's expensive, but it's an amazing tool and they offer a absolute satisfaction guarantee.

It can't fix actual hardware damage, but it can recover data from most drives, and gives you very thorough details about a drives condition.

share|improve this answer
    
I've used SpinRite to fix a notepad's boot drive 3 years ago, and it's still going strong. –  harrymc Oct 1 '09 at 13:05
    
You think $89 is expensive? I consider it very cheap compared to the cost of replacing what you've lost on your hard drive. –  GregD Oct 1 '09 at 14:55
1  
I consider $89 to be very expensive in comparison to free tools that may or may not work. Just because it's "expensive" does not mean it's a bad tool, and I think it's actually a low price for what it does. –  chills42 Oct 5 '09 at 14:50
    
Please DON'T use or recommend SpinRite except as a last resort! Try ddrescue or dd_rescue or a similar reputable data recovery tool first--preferably, something with a lot fewer exclamation points in its documentation. GRC's satisfaction guarantee will not cover any further damage or data corruption that SpinRite may cause by running your drive for a month or more as it tries to read, seek, and reread data on the damaged parts of your disk hundreds or thousands of times. –  rob Oct 6 '12 at 0:52
add comment

Looking at the SMART info for the drives, I would say that you could try to repair Windows, but it appears that the numbers are very bad. Especially if the drives are more than 2 years old, you could crash and never come back any time now.

Since you say your data is backed up, I'd cut your losses and but new drives. At the very least, buy a new boot drive. Some data errors can be fixed, sometimes.

share|improve this answer
add comment

here's a quite comprehensive guide: Diagnosing Bad Hard Drives

share|improve this answer
add comment

The OP has indicated that his data is already backed up. Usually, when stuff like this happens I have a canned lecture on backing up data but no need for that here.

It could be a hard drive failure. Hard drives just don't last forever. They have an MTBF (mean time between failure) but any individual hard drive may fail far short of that statistical expectation. If the hard drive is under warranty, and their diagnostic indicates problems, then you can get a warranty replacement. I've done this a bunch of times and never had a problem.

It just sucks to re-install the OS and your apps, but I can't see any way around it [unless you maintain images ready to go on identical drives]. Much better to get it over with than to wait for the day when the computer completely fails to boot.

Other than that, I can't offer any other advice except to make sure that you have a UPS for your machine. Power outages can cause all kinds of problems ranging from corruption of data to power supply and peripheral failure.

share|improve this answer
add comment

My experience would be to backup right now, and replace the drive. I've found that by the time SMART actually detects something, the drive is within moments of going belly up and being completely unrecoverable.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.