Hard drive failures: gradual or instant?

[symbol]

For those that have experienced hard drive failures, was it a gradual disintegration or did it happen instantly?

If it was gradual, how did you become aware of it and how much time did you have to take action (ie, data backup, drive replacement, etc) before it permanently gave out?

Finally, do SATA drives have better reliability than IDE's?

 

[Regs]

No difference in reliability because of the interface.

I had one drive that went dead instantly but this was caused from moving it from one case to another. It also had loose bearing noises before hand. So that gave me fair warning.

My 2nd hard drive that went bad just had a lot of fragmented sectors that were beyond repair.

 

[Texun]

It can go either way. Gradual failure could be a bearing whine that could last for months or years, or it could develop a few bad sectors and start crashing the OS or program. My backups are either already Ghosted from a fresh install or just bookmarks and the "My Documents" folder.

I haven't had any personal experience with SATA over standard IDE drives, but from what I have read there doesn't seem to be much difference with either one in terms of reliability.

Others will be along soon to comment on their experiences.

 

[Munky]

I've had 2 identical maxtor drives die on me, and it was gradual, so I knew the end was near. They'd either make weird clicking noises, and then crash, or become ridiculously slow. It lasted about a month or two, getting progressively worse, and then one day it just died for good.

 

[VirtualLarry]

There's really no set rule-of-thumb about this (although some mfg's drives may have a predisposition one way or the other about how they tend to go), but there are some warning signs that you should be aware of. The best way to do this is to run a SMART monitoring tool in the background, and check the stats every day. Another is to run HDTach's read test daily or weekly. Another is to be aware of any differences in the sounds that the HD makes. Lastly, be aware of strange pauses in the system, and anomolies with data-transfer rates, and any mysterious errors showing up in the System Event Log, referring to the disk or IDE controller.

 

[natto fire]

Originally posted by: Texun
It can go either way. Gradual failure could be a bearing whine that could last for months or years, or it could develop a few bad sectors and start crashing the OS or program. My backups are either already Ghosted from a fresh install or just bookmarks and the "My Documents" folder.

I haven't had any personal experience with SATA over standard IDE drives, but from what I have read there doesn't seem to be much difference with either one in terms of reliability.

Others will be along soon to comment on their experiences.

This is very true, I have had both types of failures myself. As for the SATA question, I'm pretty sure the mechanical parts of HDDs go first. Unless of course there is some kind of power supply problem that damages the PCB on the hard drive.

 

[DaveSimmons]

I've had a couple (one laptop one desktop) start to die with read-retry clicks gradually getting more frequent.

A couple of weeks ago I had a 2.5-year-old Hitachi die instantly, going into an endless loop of seek attempts.

 

[symbol]

Originally posted by: VirtualLarry
The best way to do this is to run a SMART monitoring tool in the background, and check the stats every day.

Which would you recommend?

 

[yezhou]

For me it was gradual. First I started experiencing weird things, such as when I try to open Outlook Express it'd take literally 5 minutes to start up. This happened with a few other programs as well. For the life of me I could not figure out what was wrong, I figured it must've been something software related but I was too lazy to reformat. Then one day I started up my computer and noticed one of the drives was not listed...it died.

 

[VirtualLarry]

Originally posted by: symbol

Originally posted by: VirtualLarry
The best way to do this is to run a SMART monitoring tool in the background, and check the stats every day.

Which would you recommend?

DTemp is small, lightweight, and unobtrusive. It hasn't been updated in a while though, so although it will show all of the drive's SMART attributes that are available, it won't show all of the names for newer drives' ones, just the numeric code.

Another one, that doesn't run continuously in the background, is Everest (used to be AIDA32).

One thing to note though - don't run MBM5 or ATITool with drive temp-checking enabled, alongside with another SMART tool - both of them trying to read the SMART data off of the drive at the same time might prove to be problematic.

You also normally can't read the SMART status of a RAID array directly, and in some cases, depending on the RAID drivers, it could hang the machine, just a fair warning.

 

[piasabird]

Many things can cause a hard drive to go bad. Just think IBM Death-Star. Then one company got a bunch of components that caused the electronics to go bad from their supplier. Hard Drives can also be severely damaged from both a lack of Air Flow and power fluctuation.

Many drives can be damaged in transit. Hard drives can easily be damaged when they are dropped or bumped or just sitting in a truck with bad suspension bouncing in a couple of pot holes. Some drives are more rugged and can absorb shock better.

When I think of a good quality drive I think Seagate and Maxtor. I dont know how good some of the other drive manufacturers are. Some newer compainies like Samsung are a bit questionable. Laptops are very proprietary and you never know what they put inside the things.

 

[phisrow]

It can go either way, as people have said. Weird noises are almost always bad. I've had some go gradually, start fine, become more and more coffee-grinder-like, eventually start throwing up CRC32 errors, then die horribly. I've also had them go suddenly. One moment fine, head crash, then the truly horrible sound of the read/write head actually flaying the data surface right off the platter. Even the professional recovery guys couldn't make anything of that one.

 

[airfoil]

Of the hard drive failures I have been afflicted with, one was an ouright failure while the other started letting out a SMART warning, giving me time to get data off. Overall, I'd say it's best to expect the worst - and have a backup policy that can take you through an abrupt failure.

DVD+RWs are probably the most cost effective way of backing up your data.

 

[3NF]

Gradual here. Both times I heard the dreaded "Click - Click - Click" sound. It was then I knew I needed to back up some stuff. Funny this was asked. The hard drive in my wife's laptop died for good last night .. I'm now in the process of trying to find a replacement for it.