SMART Questions (HDD's)

[Livewire]

So earlier today I heard one of my hdd's cycle down then spin back up; I've since ran some varying apps that do some SMART checking, but I have some questions cause I'm all of a sudden concerned.

I originally thought it was my 120gb dying (it apparently says 60% health in one app, however the hell it's gettingthat), but it actually seems it's the 320gb - Disk Checkup is showing a predicted TEC (Threshold Exceeded Condition) for both Temperature Difference from 100, and for Raw Read Error Rate.

For the former, it's thinking august 30th, the latter, 7 september.



I guess my question is how bad is that, and just how quickly should I be replacing the drive?



And for an overall question, if it's not actually "that bad," what are the SMART attributes I should be watching for problems on?

[lhyman]

any day now your drive will die, buy a new one

forget about what any program says, when this happens

So earlier today I heard one of my hdd's cycle down then spin back up

you only have a few days left...

[Livewire]

any day now your drive will die, buy a new one

forget about what any program says, when this happens



you only have a few days left...

Main reason I was asking is cause the drive hadn't been accessed within 20 minutes and I had forgotten to disable the "Shutdown harddrives after X minutes" option by accident. Not entirely sure it was actually fully spun up the first time when I heard it spin down >_<

[lhyman]

doesn't matter, when ever you hear strange sound coming from a drive or loud clicking noises, the drive is dying.... let me ask you how old is the drive? if it's more than two years old it could die any day (that is why you can not get a guarantee on a drive for longer than 2 years) however, I have seen drives with smart errors last for more than 5 years... some people with smart drive errors just turn smart off in the bios, but, you should still be prepared with a replacement drive and a backup of your data, what are you going to do if it dies tomorrow?

[Livewire]

If it would happen to die tomorrow I'll lose some data I can recover in about 12 hours (all the important stuff's been on DVD's since last year); I'll end up using the leftover space on my 640 until a replacement drive gets here.

That drive is easily 9 years old, I can't remember what PC I gutted it from; been waiting for one of them to start showing wear.


The question's less about whether it needs replacing and more about how quickly I should be looking; I gather I should probably get this order in by next weekend (as in placed and received).

But back on the original question as well, for the other two drives in this system, what SMART attributes are the best indicators for when a drive is approaching failure? I know SMART's not perfect, but what's a good indication the drive is close to dying?

[lhyman]

I don't like smart, I find it lies a lot.. Most of the time you boot up the machine and it says smart status OK, then you run a program that analyzes the smart and it say drive failure imminent, then you (I) ignore it and the drive works perfectly for a year or two with no errors..

however... I never let my drive goto sleep (turnoff after x min in the windows power settings)
while, that being set, it has been my experience that if the drive powers off and powers on by itself, or makes a clicking or cracking noise that you can hear while you are working... then you only have days left in the drives life... you should have enough time to do your backup....

The question's less about whether it needs replacing and more about how quickly I should be looking; I gather I should probably get this order in by next weekend (as in placed and received).

you could turn your maching off for a few days.. but you might run the risk of the drive not booting up after it cools down...

YOUR PRIORITY right now should be your back up !!!!!

[Smith6612]

If SMART started showing heat related shutdowns, I would check to see what temperature your drive is running at with HWMonitor. The drive should not be over 42C! Otherwise, anticipate that drive to die any day now if there is a mechanical or controller based problem with the drive if it starts to spit up any more SMART errors. For now, back up your stuff and keep the disk drive cool. A warm drive on a failing disk will cause the disk to fail faster. Some side information, the error you saw about data reallocation, many hard drives hide some parts of the disk which are never used until your drive needs them. This usually takes place when a sector of the disk is corrupt or unusable. When this happens, the BIOS on the drive will use one of those "reserved" clusters, making it seem like nothing is wrong with the drive. Only SMART data will let you know about that failure until the disk starts to run out of reserved sectors, which is when your data starts to corrupt. I had to take a look at a PC which had the entire middle of the disk unusable, with all of the reallocated sectors filled (half of the were unusable as well). Their drive didn't have any heat related issues but instead was most likely defective. The person winded up buying a new PC anyways, and they gave me the hard drive to incinerate with a sledge hammer, after of course the drive got DBAN'd.

Otherwise, SMART is a reliable way to look for failed drives or drives on their last hours (pretty much, the only way to find out unless mechanical issues are going on or your data is corrupting slowly). The BIOS itself if it has SMART ability can warn you of possible failures, however I've seen drives die where the BIOS didn't even warn the user. I wound up booting into a Linux disk and installing an application inside of the Live CD from the Internet (they had DSL, fortunately) and having the program read off the SMART data from the drive. The drive had high temp instances, disk head smashing against the drive's sides frequently before death, and had many reallocated clusters of disk space. I wound up buying them a SATA II drive in replacement for their PATA drive, and also added some better cooling to their case over the crappy OEM builder's cooling design and since then their drive has been rock solid.