Slashdot Mirror
More on IBM 75GXP Drive Fiasco
FolkImplosion writes "Internal documents have been released suggesting that IBM was well-aware that its click-of-death 75GXP hard drives had a failure rate of as much as 10 times that of its competitors. IBM apparently sold drives it knew were faulty into distribution, and reportedly planned to deal with any issues with marketing spin rather than a fixing the problem. This new information should help bolster a class action suit that accuses IBM knowingly shipped defective 75GXP drives with abnormally high failure rates." The lawfirm pursuing the class action suit has a page of information, including the latest news report (pdf) on information coming out in the suits. See also our original story about the drive failures.
This can't do IBM's reputation much good - fortunately for them the damage should be minimised since they no longer sell hard disks directly (so little loss of business)
I've had one of these babies in my Dell 4100 for
years and never a problem. There was a firmware
update released ages ago.
It's not rarely corporate policy to release faulty products. (Microsoft freaks, step aside, please.)
What happens is that internal politics turn problems into cover ups. Someone, somewhere decides that it's more logical to ignore the issue than to address it. The falacy continues up the line, since decisions are often based on information from 'down the line'.
The best thing IBM can do is to issue a general recall, offer generous replacement policies ("bring it in, we'll fix it on the spot") and try to recover their image as a reliable drive manufacturer. Otherwise their HD business is down the drain.
Oh wait! They sold it to Fujitsu! OK, sue their asses!
Ceci n'est pas une signature
I had one. After about a year it failed, in summer so I suspect it had been running too hot. More and more bad sectors with the familiar scraping sound.
But after I got tired of running scandisk for hours to mark bad sectors daily, I erased it with IBM's DFT (drive fitness test).
And it has been fine ever since.
It looked like the heat made it lose its calibration, unable to find the exact position on the disk for some sectors.
Don't we build cheaper things that are less reliable so that you have to buy new ones more often. Many industries already do that. They can make a light bulb that goes for 10 or 20 years no problem. And I can buy one that is garunteed to go for 5 years. But there are still ones that have 2500 hours. This concept isn't something new.
Evolution or ID?
that two of three HDs failed within 2 years.. :)
Atleast I want to give credits to IBM for an excellent replacement procedure, I have received two new drives without any hazzles what so ever. Impressive actually, considering the trouble I've had trying to get replacement ASUS Graphic cards etc...
Probable impossibilities are to be preferred to improbable possibilities.
Aristotele
Forget IBM drives, Seagate or Maxtor produce the best quality hard drives
I've noticed that everyone who is for abortion has already been born - Ronald Reagan
People, get real. This is all about the lawyers.
The lawyers are suing IBM. They are paying all the costs. The "class" is made up of losers who lend their names by affirming they bought a "defective drive."
In the end, the lawyers will get to keep 30% to 50% of the settlement or award (the cash component); the losers will get a coupon for discounts on the purchase of IBM stuff.
If you feel you have been wronged by because your 1,000,000 hour MTBF drive will only last 900,000 hours, simply tell 10 of your friends and don't buy any IBM stuff.
Believe me, that's a lot more painful to IBM and a lot less destructive to our society.
.. the same problem applied to the 60GXP and the earliest 120GXP drives (a friend just had his 120GXP click to death the other week) aswell.
The problem can be solved with a software upgrade in the drive.
This site has it all: http://www.pheuron.de/index.htm?deathstar.htm
it's in my head
Does anyone still have one of these? I purchased two 30 gig 75GXP's - first one died after about 2 months. The second after 6.
I RMA'd them - drive 1 was RMA'd a total of 6 times; drive 2, 7 times.
I got so disgusted with dealing with them that I replaced them outright with larger Maxtor drives and haven't had a problem since.
I sold one of the IBM drives on ebay to some poor sucker - the other one is sitting on a shelf waiting to be taken to the firing range.
If a manufacturer only guarantees an item for one year (thought I notice 90 days is often the case in the US - it would be illegal here), as what point does a failure rate occuring after that first year constitute a problem with the product. If 50% fail within two years can this actuallt be seen as a problem with the product. In our modern-age ,Just what does guarantee actually mean...???
- - Sha la la la . . .
I have 3 of the 3-year warranty 120GB/8MB models and 1 has failed within a year. Hitachi apparently doesn't do advance-replace so I bought a new drive to replace the failed one (they are in raid5) and shipped the failed one back. When I get a replacement I guess I'll have a 4-drive array. :) I think these drives are pretty good though, and a 3-year warranty makes me feel better.
Western Digital drives have been extremely reliable for us: No failures. I haven't needed drive technical support for a long time, but a while ago WD had the best technical support. 3 year warranty on Retail boxed drives.
I have a 10 drive array (75GXP series) plugging away at about 20% load for a full year now, no failures. I have always had luck with IBM drives when I keep them chilled. If you want to keep your 75gxp drives happy, or any drives for that matter, keep them cool.
I don't keep a lid on my coffee so when I walk around I look busy -me
IBM and Hitachi are two of the few manufacturers that still offer 3-year warranties on IDE hard drives. They are also two of the few mainframe manufacturers. I had thought that there was a correlation between these two facts -- perhaps that mainframe manufacturers regard storage as something more sacred or mission-critical than your average hard drive manufacturer. I am disappointed that IBM would knowingly ship drives with a too-high rate of failure. This is not consistent with their mainframe heritage.
Here at work use IBM through and through. Not many problems with any Think Centres but some batches of them come with Western Digital drives inside. We replace faulty drives with Maxtor drives and have had the best results with them, although there are some that are very noisy (but haven't and won't die). Loads vary due to segment: either workstation (client server and word, no big stress) or student workstation (they've seen the worst).
Personally, I have had horrid luck with the Western Digital JB (i.e. special edition) series. All have failed on me and I wouldn't trust them outside of an IDE RAID mirrored situation. About a dozen Maxtor drives have yet to die. I also have Fujistu, Seagate and the old IBM 8.4 original quit drives (best drive ever, IMO).
From what I see at deal sites like fatwallet and bens bargain's, overall drive quality has gone down in the past year or so. Some drives are over abused though, as I've seen people rip DVD's and re-encode to the same drive and wonder why they fail after about 8 months...
-Phil
Shoot questions, first ask later...
The biggest question is whether or not IBM knew the drives were failures and decided to make money on the drives selling them and then dump the business when lawsuits and RMAs were getting greater than the profit generated by their sells.
To those that RMAed more than 1 time, don't you feel like you are humping a pillow when that just can't give you satisfaction? Okay, bad part makes it to you. Okay, the replacement part is bad then we have a Quality Assurance issue. It is time to prevent the continue loss of time and break down to a new purchase.
In addition, those that RMAed the drives should not be part of the class action since IBM would not have profitted on a low margin being compounded with continuously replacing the drives. We make $2 each, but spend 70$ building them. We are 68$ in the hole for each one RMAed. See what I am saying?
At that point it is just Punative.
I think the problem was obvious when looking through forum posts of users who had 75GXP drives. It wasn't the number of failures as much as it was the number of failures per user. Sure, some people didn't have any problems, but many others had multiple drive failures, and the failure distribution was statistically abnormal.
As much as I'm happy that this is out in the open and that there's a class action suit, what will the users get out of this? In the tech world, two+ years is an eternity. Will they get the typical $20 voucher towards a new Hitachi drive while the class action lawyers get the millions? I had two fail on me in two months (on my VIA 686B south bridge while they worked fine on someone else's AMD south bridge). I had to fight IBM red tape as they kept trying to pawn me off on Acer but couldn't even give me the right point of contact at Acer (but I finally got through after six months).
You know what they say... in a lawsuit, the only winners are the lawyers.
We had a 20GB IBM Deskstar (probably 60GXP series) HDD fail at work recently; it was part of a mirrored (RAID 1) array, so no serious harm was done.
:-D
Being inquisitive folk, we cracked open the case to see what was inside. The cause of the failure was abundantly clear: the head assemblies had scraped the shiny, magnetic coating off about 2/3rds of the disk surface (on both sides) revealing the glass platter. I've never seen a glass platter before - they are so cool!
The extent of damage was equally impressive; our "museum" of salvaged bits includes various head-crashed platters of considerable vintages, but this disk will certainly take pride of place in the collection.
I'm in a similar situation - I have 4 of the 3-yr-warranty 180GB/8MB models in a RAID 5 array (with the addonics scsi to ide adapters). I've had them for about 8 months now, no problems. I also have an older IBM 120GB drive that still works fine. My parents and sister have had 40GB and 80GB IBM drives with no problems for years, except when I dug around in my parents system and when I put it back together, their 40GB seems to have some really bad sectors... I was able to recover most of the data off the drive though... I blame this incident on myself and not IBM (since it failed right after I dickered around with it).
Also, I should note that my RAID 5 of the 180GBs gets VERY heavy use.
I wonder if drive temp has anything to do with these problems? My 180GBs and 120GB are right next to two case fans, they stay really cool...
Cheers,
Dave
moo
I don't know about the desktop drives, but my employer uses the Travelstar 20 and 40 Gb laptop drives. Our manufacturer in Korea used them because they were manufactured there and there was no import duty on them.
I should say "used to use" them. About 30% of those drives died by the time our customers received the product they were installed into (voicemail systems - 24/7/365 uptime required). Now our customers are pissed and threatening to sue us.
We ordered our manufacturer to switch back to Toshiba (which is what we used previously). They work.
Microsoft's VP of Customer Service is Helen Waite. If you are having problems with their products go to Helen Waite.
I've purchased at least 6 of the 60GXP line. I've had a single failed drive in the past three years. I still need to RMA that thing, but I've just been lazy about it.
I'm not sure that the 60GXP fall victim to the same failured that the 75GXP did. It seems to be that when IBM went to the 20GB platter sizes, and redisigned their drives, things got better.
I have in the past had data corruptions on one or two occasions. I'm pretty sure that this was from my Highpoint RAID controller though. After I stopped using it, I never had another problem with the drive.
I still can't help but believe that this is why IBM sold off the mechanical storage division to Hitachi. IBM claimed that they were going to work on some newer types of storage, but we've seen very little so far.
I got on IRC right after bringing home my Hitachi Deskstar 180GB drive (I wrote it as 160 elsewhere, oops), to brag about Fry's price of $70 after rebate.
The response? "You bought a DeathStar?"
Needless to say, that didn't please me much. Nor did Fry's lying to me about how they'd handle the rebate when the store didn't have the forms on hand, but that's an issue with them...
Get off my launchpad!
I bought one, and admit it was fast, loved it.
Read the article on the drives being defective, and didn't want to believe it.
Then my drive made a few clicks, but didn't crash imediately.
Turned my computer off, and ordered a new Seagate Barracuda IV. Copied data over. Never used that 75GXP to this day. Still sitting in a box.
IBM owes me. I had to get 2 HD's in a year, rather than one.
I don't have any mod points today, somebody make sure this AC gets noticed. Tort litigation is out of control. It should be the last contingency when a company will not be reasonable, not the first line of defense. Say no to greed.
Auto-reply to ACs: "Truly, you have a dizzying intellect."
". . . A plus B plus C equals X. If X is less than the cost of a recall, we don't do one."
Mikey-San
Karma: +Eleventy billion (mostly affected by watching Celebrity Jeopardy)
The last time I sent them back, I had high hopes that Hitachi (who bought the hd business from IBM) would finally read my letter and send me some new replacements but no luck.
They are now on their way to the nearest landfill.
If you can read this sig - the bitch fell off.
Actually, temperature was an issue. If my memory is working right, the issue with the 75GXP disks was in the thermal compensation... that is, they incorrectly compensate for thermal expansion of the platters through r/w head positioning. It seems like this is actually one of the most common problems with 2.5" disks too... I've had many of those fail, and I can read from them for a few minutes until they get warm... Nothing like trying to recover the CEO's laptop using only 'dd' :)
"except when I dug around in my parents system and when I put it back together, their 40GB seems to have some really bad sectors"
did you send the sectors to time out?
Tis better to be silent and thought a fool, than to open your mouth and remove all doubt --Abraham Lincoln
Your point is not without merit, especially with regards to coupon settlements. If coupons aren't good enough for the lawyers, they aren't good enough for the plaintiffs.
That being said, the class action lawsuit does benefit society with respect to one thing. It strongly influences how often a recall is done on shoddy or unsafe merchandise when it would otherwise not be done. By making not recalling known defective products more expensive than recalling known defective products, the public (which paid for those products in the first place) benefits. Without these class action lawsuits companies would shaft their customers on a far more routine basis than they already do.
- You can use it as a fridge magnet. Keep a pry bar around to detach it
- Stick one on a toolbelt, it's strong enough to hold a magnet!
simply tell 10 of your friends
;-)
I dont have 10 friends, you insensitive clod !
Probable impossibilities are to be preferred to improbable possibilities.
Aristotele
There are multiple issues:
1) Temperature is a big one. At higher temperatures, the lubricant between the platter and the heads thins out so much that the hydrostatic wedge is no longer thick enough to protect against surface irregularities - causing oxide to flake off and pile up on the head. The new firmware upgrade causes the head to wiggle slightly from track to track while idling, thus preventing the loss of lubricant wedge.
2) Compensation for thermal expansion - this is another firmware issue.
3) Quality of the power supply. I've never had a problem with beefy name-brand ATX PSUs. Delta Electronics is good, Antec is OK.
4) The quality of the IDE connector and the molex power supply connector is poor. If you tweak the cables a bit, you risk a loose connection.
5) Overlooked, but may well be the main cause - the solder bumps on the PCB that make contact with the head electronics in the drive. The quality of the solder is critically dependent on its composition. It needs to be soft enough to maintain contact even after vibrations, thermal cycling, etc. If it's brittle (which happens in high-tin variants), the bumps no longer make contact after a while, causing sporadic failures.
You will often see one head drop out on multi-platter drives, while it happens less frequently on single-platter drives.
The fix for this is to unscrew and refasten the PCB, or even better, use a fine-tipped soldering iron to re-melt the bumps before refastening.
The best part about this post was going to the lawyers' website and seeing that they also sued Aiwa for their crappy mini systems. I owned 3 of them and every cd player on them broke. To this day when i see one at someone elses house, i ask them if their cd player works. And they say, "no..."... Glad to see that someone went after them for that.
The most problematic is the 75 GXP, which was available from 1 to 3 platters in capacities of 15, 30, 45, 60 and 75GB. The 75GB model has the most problems, the 15 GB the least. These have names like DJNA-30.., etc.
The 60GXP solves some problems, but still has issues in multi-platter configurations. These are available in 20, 40, 60, 80 and 100 GB capacities. The 20 has very few problems, the higher capacities have higher failure rates. These have names like DTLA-30.. or IC35..
The 120GXP also solves some problems, but introduces new ones. On the whole they're better than the 75GXP, but not necessarily better than the 60GXP.
So anyway, you can chalk me up as a Russian Roulett success story. No complaints here.
Having worked on many bad HDs, I keep this list of links to all the manufacturers HD testing programs:
a x.htm
l
Maxtor/Quantum
http://www.maxtor.com/en/support/downloads/powerm
IBM/Hitachi
http://www.hgst.com/hdd/support/download.htm
Seagate
http://www.seagate.com/support/seatools/index.htm
Western Digital
http://support.wdc.com/download/#dlgtools
Fujitsu
http://www.fcpa.fujitsu.com/download/hard-drives/
"TK-421, why aren't you at your post?"
They ALL suck. I've had within 90 day failures for all of the big name ATA drives. It's an inevitable byproduct of the ever-increasing data densities and the ever-tighter profit margins. They all have decent customer service (so far) when it happens as far as replacing the drive, but that doesn't help for the data.
As a result, I now require all new desktop computers for the college department I work for to come with three hard drives; a RAID-1 mirrored pair for the OS, and an external hard drive (ATA to USB/FW box + OEM drive) for doing daily backups on. (Weeklies go to the server, which has a more complicated backup arrangement). It's added about 15% to the cost of the build to the local white-box outfit... but now I no longer fear hard drive crashes.
I fear hard drive thieves. =)
//Information does not want to be free; it wants to breed.
It's funny. We have fora such as Slashdot and Groklaw, where information gained from hundreds of dedicated people are put to the common good. We have systems like Linux and GCC that are similarly the result of large common effort.
What is lacking is a good system for tracking hardware failures. All we have are anecdotes, until somebody gets sued and we can see their internal documentation.
Imagine, though, that there was a system for tracking hardware. Whenever you would get a drive, you'd put it into the common database. Serial number, model number, date and place of manufacture. Granted, this would take a few minutes. When (not if, they all die eventually) a drive fails, you would call up the record and mark it, along with (perhaps) the symptoms associated with the failure.
But, if tens of thousands of people did this, you would quickly build up a spectacular database of hardware reliability. You would be able to instantly see what drives were better than others, or see if quality was slipping or improving for popular drive makes as time went on.
Soon, much like the moderation system of Slashdot has truly raised the level of discussion here, you would find that the reliabilty ratings driven by this database would force manufacturers to make higher quality drives -- they'd know that they could never force crappy drives on the market as IBM is alleged to have done here.
Now, I'm not volunteering -- yet. But I could be interested. There would be insane pressures from the manufacturers to influence the results, and there would no doubt be some attempts made to stuff the ballot box. But, it would be a good thing.
Thad
I love Mondays. On a Monday, anything is possible.
IBM is a huge company. The people making hard drives probably have very little to do with the people making Linux other than the same style of standardized HR forms. IBM in particular seems to operate divisions with a good deal of modularity, from what I've heard from people working there.
May we never see th