Hardware Detectability of HDD failure?

[gifi4]

One of my HDD's tends to be grinding when in certain directories. I immidiately considered this to be a sign of an incoming failure and am about to head off to back up all my data off that drive. Before I got around to that, I ran both Seatools (I have a WD HDD but Seatools is compatible) and Data Lifeguard Diagnostic which is for WD drives (And supposedly others) and all test were sucessfull with no signs of failure.

The Data Lifeguard Diagnostic had to do a 2 and a half hours long scan of all sectors of the drive and the result was perfect.

I'll continue on to backup my drive but I'm just curious if I really should be worried?
As I said it's only in certain directories: My One Piece folder is neatly organised into each arc and only two of the arcs cause a crazy amount of LOUD grinding. (There are also other directories)

When the drive grinds, it tends to take a VERY long time and basically any of the files in that directory aren't available for use.

 

[FAST6191]

It could be a case of extreme fragmentation* but most consider it a warning of impending drive failure. On drive failure though it is not necessarily all or nothing but corruption can sneak in as well which is not fun to recover from in any case let alone as a single normal user type machine.

*if you grabbed from torrents and/or extracted rar files when you have little space remaining that is often a good recipe for fragmentation and on top of that if you have thumbnails and metadata parsing turned on then that gets more likely- what happens if you navigate there by command line and do a DIR command?

2 and half hours for (assuming your signature is accurate) a terabyte drive is very much on the low side of things (for my money that is probably 100 gigs range for time on a proper scan). Moreover drives will seek to remap sectors (I have actually seen drives hose themselves by overflowing their reallocation section but that is a different matter) so depending upon what you did you might have forced it to remap (this would also account for the time taken to parse directories).

Your scanning tools probably just use SMART readings and a few choice vendor readings which do much the same thing - generally if a SMART reading pops up I will pay attention to it and the raw data is provides is useful as well but I do not expect it to tell me a thing.
http://research.google.com/archive/disk_failures.pdf if you are curious (long story short google, which notably runs a lot of stuff on consumer grade hardware, more or less dissed SMART).

 

[gifi4]

It has nothing to do with fragmentation. Simply because it was only 5% fragmented a few weeks back when I ran the defragger...(Plus it's 0% Fragmented now)
The HDD is usually low on space. 10-20GB remaining is normal for that drive (Movies,TV,Anime etc are all stored there) and is why the drive was even fragmented.
The scanning tools used SMART but those were the basic tests.

Edit: The results of the test I ran:

Test Option: EXTENDED TEST
Model Number: WDC WD1002FAEX-00Y9A0
Unit Serial Number: WD-WCAW32179115
Firmware Number: 05.01D05
Capacity: 1000.20 GB
SMART Status: PASS
Test Result: PASS
Test Time:
23:01:27, December 08, 2012

Edit again: The description of the test is:

EXTENDED TEST performs a Full Media Scan to detect bad sectors. Test may take hours for large drive.

More detailed version:

EXTENDED TEST: For most computers the DLGDIAG Extended Test takes 30 minutes to an hour to test one drive. Larger drives take longer to test. The performance of the computer also affects the test time. If a bad sector is detected under Win9x/Me environments, the test may stop responding for a few minutes and then continue. This is not a destructive test unless users select repair function when bad sectors are detected.

 

[gifi4]

The grinding is crazy in another area of the HDD. It took atleast 3 mins for it to stop and allow me to access the files. I've got a 1TB external that I'm about to backup the entire contents of the drive on to. I'll be looking at investing in a 2TB internal to replace the 1TB but the price is crazy at the moment. Nearly $200... I got my 1TB for $60 before the floods in Thailand and the prices haven't come down. As long as my data is backed up I don't mind using the drive 'til it's death.

 

[Coto]

Run HDD Regenerator from Hiren Boot CD (10.x and up), beware, since you have a 1TB+ drive it would take hours if not days. But it'll tell you how many bad sectors you could have. Also, this may be me, but when I grab a disk and shake it gently, and I hear its needle tiltling, it's a bad sign. It may not be dead, but it'll soon, that's fpr sure.

Say, while the disk is seeking data (those minutes you need to wait to retrieve files info), does the computer get stuck? (unresponsive), that could be some area of clusters severely damaged. And in such cases, it's a good idea to isolate those clusters and mark them as bad. (hdd regenerator does this automatically)

 

[gifi4]

Run HDD Regenerator from Hiren Boot CD (10.x and up), beware, since you have a 1TB+ drive it would take hours if not days. But it'll tell you how many bad sectors you could have. Also, this may be me, but when I grab a disk and shake it gently, and I hear its needle tiltling, it's a bad sign. It may not be dead, but it'll soon, that's fpr sure.

Say, while the disk is seeking data (those minutes you need to wait to retrieve files info), does the computer get stuck? (unresponsive), that could be some area of clusters severely damaged. And in such cases, it's a good idea to isolate those clusters and mark them as bad. (hdd regenerator does this automatically)

Hours?
Do you have an estimate?
(It'll take hours to backup the drive as is, so I'm trying to organise all the time necessary)

CPU: Intel - Core i7 2600k @ 3.4GHz
RAM: 8GB Dual Channel 1300MHz
GPU: Nvidia GeForce GTX 580 1536MB (Gigabyte)
MOBO: Intel DP67BA - LGA1155 (P67 chipset)
HDD's: 2 x 1TB Western Digital Caviar Black
SSD -Corsair Force GT 120GB (SATA3)
Optical Drive: LG - GH22NS70
OS: Windows 7 Enterprise 64bit
Case: CoolerMaster HAF X RC-942
Monitor: Benq G2450H 24"

Nope, the computer doesn't get stuck, I'd assume it's because my OS is on my SSD. That drive itself gets 'stuck'. I mean that it takes longer to do other thing like going to a different folder etc.

 

[Coto]

So any activity in between dirs/folders during those minutes-of-seeking data (ie: while opening file A, you try to get file B properties on a different folder/open file B) on the same drive will be unresponsive? (ssd excluded..!)

Uhm what if pagefile is causing you issues on such SSD? (fast access, but little remaining size). Since I have 8GB DDR3 too on my CQ-42, I suggest you to disable paging from disk and leave everything to RAM (i've been using this config since last year and believe me, at least 3-4GB are always FREE, not even used)

Also, try running that drive off an external SATA-USB like reader and see if the problem goes away.

The estimate will depend on total ammount of clusters=healthy clusters+dead clusters. If hdd regenerator finds a huge yet sequential ammount of dead clusters, like $1000E[BAD],$1000F[BAD],$10010[BAD] and those are shown as huge blobs of faulty clusters, leaving the hdd stuck, then you have a physical defect on the hdd plate.

 

[gifi4]

So any activity in between dirs/folders during those minutes-of-seeking data (ie: while opening file A, you try to get file B properties on a different folder/open file B) on the same drive will be unresponsive? (ssd excluded..!)

Uhm what if pagefile is causing you issues on such SSD? (fast access, but little remaining size). Since I have 8GB DDR3 too on my CQ-42, I suggest you to disable paging from disk and leave everything to RAM (i've been using this config since last year and believe me, at least 3-4GB are always FREE, not even used)

Also, try running that drive off an external SATA-USB like reader and see if the problem goes away.

The estimate will depend on total ammount of clusters=healthy clusters+dead clusters. If hdd regenerator finds a huge yet sequential ammount of dead clusters, like $1000E[BAD],$1000F[BAD],$10010[BAD] and those are shown as huge blobs of faulty clusters, leaving the hdd stuck, then you have a physical defect on the hdd plate.

Any activity completed on the same HDD whilst the hard drive is grinding is responsive however takes much longer and when I use my SSD or other HDD they work fine.
What do you mean by "issues on such SSD"? There are no issues on the SSD. It has a reasonable 30GB/120GB remaining simply because Steam is installed on it and the games take up a fair bit.
Basically, in order to run hdd regenerator, I need to find a day when the computer won't be needed by anyone in my household for several hours. Guess I could do it the week after Christmas because my father has a week off from work so he may want the computer during that week. (We have 1 computer however that's all we need as my dad just uses it for banking and whatnot)

 

[gifi4]

HDD is finally backed up. Took 12 hours to back up the entire 1TB -_-

I'll be running the regenerator ASAP but not just yet, it's going to be 41 degrees celcius tomorrow. Had the HDD backing up from 5AM today to 5PM -_- (5:03PM to be specific)