Poll - dead hard drives - My: 500GB - ST3500320AS - 75 days - Dead

[jeffy]

My drive bought in February this year, used for recording analog TV videos, started approx. 2 hours ago giving weird sounds, then Windows XP gave me some notices like cannot save $Bitmap, $Mft, the data have been lost...

Seagate Sea Tools won't run a long test, a short test fails,
the drive capacity is shown 0GB (last LBA 0). It just happened all of a sudden. The tests from a bootable SeaTools CD or WD DLG (Windows only) will not test it either.

Let me ask you in the poll:

[kurt]

had also a seagate (250 gb I believe) which made some click noises after 2-3 years or so... replaced it before I could lost data...

[Avenger007]

I had a 40 GB Quantum Fireball from my Dell PC back in 2001. It lasted 14 months; just 2 months after warranty on parts expired. I think it was due to heat.
Inferior parts is just one reason why I will never buy any PC from Dell or any other PC manufacturer (except for Alienware, Falcon Northwest, etc..., but these have their own issues like poor price/performance ratios).

[setarip_old]

@jeffy

Do you have an alternate connection available?

Do you have a replacement cable that you can swap in?

[jeffy]

setarip_old: Yes, I have a replacement cable. Yes, I have several SATA ports available where any other drive I have is fully functional and passes SMART test, short test, long test.

But not this one. It will show in BIOS its type, but it will not run any SeaTools for DOS (or Windows) test, not even when run outside of the OS. Short test fails immediately, long test shows 99 errors and terminates in a few seconds. It failed out of a sudden, no SMART warning, no SMART attribute bad, but now it won't even read any SMART attribute. It spins, that's all. HDTune won't show any SMART data and BIOS will warn about the Drive Failure.

Unfortunately, it's not the first drive. In last twelve months, I lost this 500GB Seagate, 250GB Seagate (system drive bought December 2007, lost January 2008), WD 320GB (lasted one week), WD 320GB (lasted one month) - is it all? - (I don't want to check the invoices now) - it's weird, it's frustrating and no one knows why. All electric parameters are okay, no drive shown any signs of electrical damage or of a shock, each and every one was cooled. The PSU is appropriate, Fortron, measured with the wattmeter, sufficient and working without any problems. They just failed, all of a sudden, but this Seagate is the only one that failed completely (0GB and max LBA 0 in SeaTools says all).

I've changed the sellers, I've changed two manufacturers (WD & Seagate), and to be honest, out of more than 20 drives I had in 5 years, more than a half failed (usually just getting bad sectors, some had trouble spinning).

Thank you to all for your effort, I appreciate it.

[setarip_old]

Mysterious.

If I were confronted with your situation and had already made all of the tests and determinations as you have, I'd try placing my system in a different environment - either in the same building, or better still, a different building...

[*.mp4 guy]

That sort of drive turnover heavily indicates something pathologically wrong with the system. I have never had a hard drive just give up and stop working entirely, the worst I've ever had was a poorly ventilated seagate drive in a dell dimension desktop become unreliable after ~2 years (surprisingly, it never got all that bad, its actually still semi usable, and I still use it from time to time in "scrap systems" to test out the hardware, without investing any extra money, though it inevitably becomes unstable after a few months on the same os install)

Off the top of my head, here are factors that could be causing problems.

- Failing motherboard components, if youve blown a few capacitors, some of your controller chips could be unreliable, this is technically possible without blown caps, but unlikely. similar problems could be caused by, a short, a corrupt bios (perhaps indicating a failing battery), extremely excessive heat, or voltage damaged chips. It is also possible, but unlikely that there is some form of design defect with your motheboard, or perhaps and add on card that could be responsible for the problem.

- failing/unreliable powersuply, voltage ripple, or "dirty power" is not always emediately obvious in monitoring programs, and even if it is technically withing the acceptable voltage range, dirty enough power will cause premature component death, and stability issues in general. Conversely, your power supply could be of acceptable quality (a really good unit like a pc power and cooling turbo cool 510 would take it upon itself to fix the problem, but plenty of workable units woudn't) and the problem may be that the power in the building that you are feading to your computer is of exceptionally bad quality; in this case, you can buy a unit to clean the power before you feed it to your pc.

- A strong source of electro magnetic fields, or interference is near your computer (or sometimes neer it). Common sources of emf are televisions, electric motors, microwaves, equipment that utilizes large amounts of power, or magnets/electromagnets in general; such as defibrilators, loadspeakers, and MRI machines. Strong EMF can cause bit errors on nearby magnetic media, such as floppies, tapes, credit card magnetic strips, and hard disks. occasionally improperly functioning crt computer monitors may cause significant levels of EMF, though this is highly unlikely.

[BilboFett]

Hey Jeffy, Bilbofett here from the crazy rainbow lines virtualdub forum.
I installed a 500 watt power supply, Antec, earthwatts, top of the line. As a precaution, I ran chkdsk on all 3 of my drives: 1 Seagate 400 gig 7200 IDE, 1 500 gig IDE, and 1 500 gig SATA, all Seagate.
Guess what... EVERY drive had errors on them. Deleting orphan file... recovering free space in the volume bitmap, etc. etc.
It finally finished, I went into Windows, and was curious... so I ran it again at reboot time.
More errors! WTF.
I have 3 theories:
a) seagate drives are being made like crap now with that 5 year warrranty
b) seagate drives get corrupted in vista if you have the right hardware combination
c) When I originally partitioned out all my drives, using an underpowered power supply, with not enough wattage to support all my devices, every partition was corrupted.
I'm not sure about theory C, but it would at least explain how chkdsk finds errors every single time I run it on a reboot. I find it very hard to believe somehow I got 3 bad hard drives from Seagate, including the brand new replacement drive I just got.
Btw, since I got the new power supply in, BIOS detects all my drives instantly, and the IDE port always lists DMA mode 5 for the drives as it should now.

[jeffy]

setarip_old & *.mp4 guy: Thank you again for your suggestions and comments. Adding something more:

Last year when the drive failed (after several weeks) I took it to the seller and they were kind enough to copy all the possible data to a brand new drive I bought from them that day. One day after: they couldn't believe, but the drive had a first bad sector and started playing up - guess what - another RMA. I have changed in 5 years: five motherboards (four ASUS ones, one MSI), tried several power supplies (last three: Fortron 350W, Fortron Epsilon 700W, Seasonic S12-II 500W) and nothing solved it yet. Except of the location, (which sounds really odd), I've changed everything. It is a single one desktop system.

None of the RMA stated what could be the cause, no electric damage or shock damage was found. None of the motherboards had a single blown capacitor. What I found for sure: newer and larger hard drives are failing more often (at least for me). An 8GB drive was good for 4 years, but 250, 320, 500GB ones - some of them have already failed.

But I know I am not alone. Two of the notebooks the technician I know had on his service desk (one cost about 4000, the other about 2000 USD) failed in a single month.
Both of their hard drives were bad in a single month. Different location, several tens of miles away. Two server hard drives (73GB, 500 USD each) bought for another location, one DOA, the other gone bad in a few weeks. Again, in a different location.
I find it hard to belive, but is it just luck?

Maybe it is, just a single Google search (NewEgg is for example only, try whatever site):
http://www.google.com/search?q=+site%3Awww.newegg.com+newegg+doa+%22hard+drive%22&hl=en&safe=off
"Bought 4, had 1 DOA "
"Pros: none, was DOA (Dead on arrival) "
"Cons: 4th drive was DOA"

http://forum.doom9.org/showthread.php?p=955412
DailyTech: "Study says failure rates 15 times that of what manufacturers indicate"

TGDaily: "In the end, Google's research does not solve the problem of predicting when hard drives are likely to fail. However, it shows that temperature and high usage alone are not responsible for failures by default. Also, the researcher pointed towards a trend they call "infant mortality phase" - a time frame early in a hard drive's life that shows increased probabilities of failure under certain circumstances. "

Direct link:
http://labs.google.com/papers/disk_failures.pdf

[Shinigami-Sama]

I had my seagate laptop drive a few months ago, its sucks being out a laptop...

it has a total of 3% bad sectors
from 192 bad sectors to a full 3%(of an 80gb drive) in a day and a night
thankfully I saved all my data off it before it finally died completely, only to have one of the discs I burned die the next day when I went get a file off it (toc unreadable...)
drive was 2.6 years old...

[communist]

Drives that died on me so far (over the last 7 years):
2x (or 3?) IBM DTLA (aka 'DeathStar'-Series), lifetime 4-5 months
2x Samung SpinPoint (40 & 80GB), lifetime 2 & 4 years
1x Seagate 10GB, lifetime 7+ years (got it 2nd hand)

In different computers and environments. You'll either get over it or start doing backups :)

[Inventive Software]

I posted no yesterday, but today I dropped my laptop without shutting it down completely, and the hard drive seems very much dead, it sounds like it's trying to be read, but nothing's actually being read. The drive in question is a Hitachi TravelStar 80GB 2.5" SATA drive. I'm gutted cause I had lots of data on it, especially uni stuff, and at the moment, no way of recovering it before deadlines. :(

[Nematocyst]

I think the last drive I actually lost data from was a 1.0 GB Western Digital. I had trouble also with a 1.6 GB WD, but was able to back it up before it went dead. Now that I'm doing regular backups, it's not surprising that I haven't lost any drives or had any act up with important data. (Murphy's Law)

I have 3 WD drives now (2 80's and 1 160) that are causing controller errors to appear in event viewer on one machine. But they don't do this on other machines. I don't trust them, but it may just be that the IDE controller on that mobo is picky.

My Q6600 machine hasn't exhibited any drive issues since I built it in Aug 07. It has a 320 Seagate SATA, a 200 maxtor IDE, and a 160 WD IDE.

BTW, the Google study (http://www.google.com/search?hl=en&q=google+hard+disk+OR+drive+failure+heat&btnG=Google+Search) has shown that heat isn't a hard drive killer.

[Shinigami-Sama]

actually that reminds me I had a maxtor external drive that died on me this time last year I prolly should've saved the enclosure but I threw it all out I think



I have 3 WD drives now (2 80's and 1 160) that are causing controller errors to appear in event viewer on one machine. But they don't do this on other machines. I don't trust them, but it may just be that the IDE controller on that mobo is picky.


my current (only) harddrive is doing that too
I've tracked it down so far to the drive wanting to turn itself off then getting a read/write request being slow to respond

at least most of them

[RickB]

I always back up my files to a seperate harddrive and to Boxnet. For photos its Flickr. Touch wood I've never yet had a problem with my hard drives.

[leeperry]

do you o/c like crazy ?
is your PCI really 33 MHz ?

I've got a a 160GB SATA Seagate that was supposedly dead from what windows said(it even took it off the device manager at some point "error saving MFT$, FO" :eek: )

DOS Seatools was able to do a short test and everything was fine, but a long test would freeze immediately.

I put the FSB back to 200, and I was able to back it up completely...but I got a low budget mobo....weird thing it worked perfectly for like one year ?!

[jeffy]

do you o/c like crazy ?
is your PCI really 33 MHz ?

I've got a a 160GB SATA Seagate that was supposedly dead from what windows said(it even took it off the device manager at some point "error saving MFT$, FO" :eek: )

DOS Seatools was able to do a short test and everything was fine, but a long test would freeze immediately.

I put the FSB back to 200, and I was able to back it up completely...but I got a low budget mobo....weird thing it worked perfectly for like one year ?!

Yes, my PCI is 33.33 MHz, PCIe locked at 100 MHz.
Of course, I have tried the non-overclocked settings and a big NO, SeaTools won't test this 500GB drive, won't report its capacity, won't show its S.M.A.R.T. data, it will just show its type, that's all.

The problems described here appeared on different systems, some of them were never overclocked:
http://forum.doom9.org/showthread.php?t=129829&highlight=autochk

[anti]

How old is your house? Bad power or bad luck. :(

8-10 500GB Seagates
5-6 250GB Seagates
3 80GB Seagates
2 1TB Seagates
1 160GB Seagate (IDE) -- Failed
1 80GB WD
2 30GB WD -- Both Failed
Various Fujitsu 6GB,8GB,10.5GB,15GB and 30GB maybe about 7-8 of them all failed.
2 30GB Maxtors -- Failed
1 2GB WD -- Failed
5 1GB Quantum Fireballs
3 20GB Quantum Fireballs
1 20GB IBM Deskstar

Everything I can think of off memory for failure rates of hard drives.

[burfadel]

The problem at the moment is that the drives aren't all made in one country! I bought 3 supposedly identical Seagate 320gb drives a couple of years ago, one was made in Thailand, one in Singapore, and one in China!

A friend recently has a 1tb WD drive fail, which of course wasn't a nice thing!

What happens is the drive manufacturers naturally want to make the most money from the most competitive component, and what suffers is the quality of the drive. Much of this is related to the factory in which its made, or seems to be.

When I next buy hard drives which will be soon, I'm only getting the new Samsung 32mb cache drives, they seem to be more consistently reliable at the moment (and they perform well).

[jeffy]

@burfadel:
My failed 500GB Seagate was made in China.
It was the model with 32MB cache, check the thread title.
My failed 250GB Seagate was made in China.
Other mentioned failed drives (Seagate & WD) were not made in China (Taiwan, Malaysia).

[jeffy]

@jeffy

Do you have an alternate connection available?

Do you have a replacement cable that you can swap in?

Today I had a chance to try it in the company on a completely different motherboard; it would not even identify itself in the BIOS, yet on my board it would... Well, it spins, but it's gone anyway.

[CWR03]

I've only had one hard drive ever fail on me, which was a new Maxtor. It started clicking, then failed to show in Windows, after only a few hours of use. The only other drive trouble was one that was improperly ventilated and got so hot it seized up, but after relocating it with good airflow it's still spinning and working with no bad sectors two years later.

[anti]

Speedfan is a decent way to keep track of HD temps.

I don't think any of my Seagates were made in China, most of them are OEM though.