When is it time to replace a hard drive?

In October last year some of the startup files on a Windows XP PC were corrupted and the computer was caught in a continual reboot loop. I ran the Seagate disk utility (Seatools) and this revealed 56 cable test errors, so I replaced the IDE cable. I booted the PC up using the XP CD and performed a System Restore from within the Recovery Console. This got the PC up and running again.  
The same PC got caught up in a continual reboot loop again. At first I couldn't even communicate with the hard drive, even after removing it and attaching it to another PC using an IDE to USB adaptor and PSU. Ended up returning it to the PC and amazingly (to me) running a diskchk from within the Recovery Console regained contact with the hard drive. A system restore from Safe Mode had the PC up and running again.
The Seagate HDD passed a complete test with the Seatools utility (apart from file system errors, but these  always seem to be reported on XP, or maybe NTFS, PCs) . However, due to it's troublesome history, I had a look at it's SMART attribute values and am concerned. I would like some help interpreting the results:
Raw read error rate: Type = Prefailure,  Threshold = 25, Value = 62, Worst = 59, Raw = 94320242, Status = OK
Seek error rate: Type = Prefailure,  Threshold = 30, Value = 78, Worst = 60, Raw = 68414776, Status = OK.
Also note that the PC's 5 year hardware warranty is about to expire.
What do you reckon?
Cheers.
TorDurBarAsked:
Who is Participating?
 
TelnetServicesCommented:
Hi,
SMART attrubute values are read in decreasing order - ie 253 is best and 1 is worst.
Manufacturers often use 100 (or sometimes 200) as the nominal factory value.  With the values you quote, I would suggest that both read and seek errors look like they are worse than a new disk - but certainly within the 'OK' values for the drive.

HOWEVER - more important than the SMART attributes at any one time, is the rate at which they are getting worse.... if thi snumber is dropping steadily with time, then replace the hard disk.

A number lower than the threshold number on any 'prefailure' attribute, would indicate a drive that is about to fail.

If you are worried - swap it out - disk storage is cheap these days - you should be able to buy 320G or more for US$100 or less....  You might however have a challenge getting it replaced under warranty, as it's not actually failed yet.
0
 
PUNKYCommented:
When you acknowledge such issue, then it might be to late for replace the hard drive. Always backup your important data into CD/DVD and keep in safe place that I would recommend. If you have an external drive, that is even better, you could clone whole system to it.

The electronics device can be failed at anytime, so backup is must. Your system is still under warranty, then use that benefit now for replacement.
0
 
TorDurBarAuthor Commented:
Thanks for getting back to me on this.
Punky, already backed up the data, did it the first chance I got.
Telnet Services, I appreciate your feedback. Here's a bit more info, which suggests the drive is still OK.
Oct 2006 results:
Raw read error rate: Type = Prefailure,  Threshold = 25, Value = 62, Worst = 59, Raw = 115990145, Status = OK
Seek error rate: Type = Prefailure,  Threshold = 30, Value = 77, Worst = 60, Raw = 57180915, Status = OK.
Aug 2007 - Just finished a disk defrag to work the drive a little (the previous results were from before the disk defrag):
Raw read error rate: Type = Prefailure,  Threshold = 25, Value = 64, Worst = 59, Raw = 135815601, Status = OK
Seek error rate: Type = Prefailure,  Threshold = 30, Value = 78, Worst = 60, Raw = 68578407, Status = OK.
0
Live Q & A: Securing Your Wi-Fi for Summer Travel

Traveling this summer? Join us on June 18, 2018 for a live stream to learn about the importance of Wi-Fi security and 3 easy measures you can start taking immediately to protect your private data while using public Wi-Fi. Follow us today to learn more!

 
nobusCommented:
maybe this can help :
http://www.urltoy.com/asc.htm
0
 
mastooCommented:
Although I rely on Smart as being better than nothing, a good smart status is not the same as saying the drive is in good shape.  That google study recently was interesting because it showed smart catching fewer drive failures than expected:

http://www.pcworld.com/article/id,129238/article.html

So, ditto TelnetServices.  In in doubt, replace it.
0
 
TorDurBarAuthor Commented:
Thanks to everyone for their input. I awarded the points to TelnetServices since the comparison of SMART attributes almost a year apart did indeed indicate that the hard drive in question was probably  still OK. To double-check this I purchased the SpinRite utility (since I also wanted something for testing and repairing future hard drives) and after over 11 hours of testing the hard drive, no errors were reported. Cheers.
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.