Study: Hard Drive MTBF Ratings Highly Exaggerated

[Kab]

http://www.dailytech.com/article.aspx?newsid=6404

Well whadya know its marketing... Ooo and money in there too

Judging by all those HDD failures, I think this was fairly obvious but anyhoo

I love my Maxtor's, have been of excellent QoS to me since day one, and that day is 5-6 years back! The rest of them never stood the test of time to tell the tale...

[Gas Pipe Jimmy]

That was a great read!
Thanks.

[ianm2]

quote

"According to the study, the number one cause of drive failures was simply age. The longer the drive has been in operation, the more likely it will fail."

well you don't say.

occasionally, there is the odd funded study that says things like " it has been proven that girls prefer pink and boys like blue" and that sort of thing has been funded with public money.

that sounds a little similar in an obvious kind of way.

I haven't had any probs with my few hard drives, I had a maxtor, which I always thought were slightly poorer relations, and a samsung spinpoint, whilst at times I had the impression its a bit slow, expecting miracles from sata II and all that, again, may be either marketing or my unreal expectations, its worked flawlessly.

I think they are again, much of a muchness, and you pays your money, read your reviews, listen to jonnyguru.com and other people, and takes your chances. sometimes I think you can't go wrong with a big brand, sometimes

edited to add:

I suppose with most pc parts, hard on drives esp. you continue to buy what you have found reliable and work well, one slightly good rule of thumb is to buy a product which makers specialise exclusively in, ie. say a seagate or WD hard drive instead of a samsung, who are jacks of all trades, and its only the slightly 'unknown' unestablished brand that would make me think twice about buying another, say instead of a hitachi which I would probably go for next time, but I have few problems with it, as its been really great. Maxtors fluid bearing too is attractive.

Its nice to hear someone happy with my first ever hard drive, maxtor, which was reliable, if a bit noisy, I always do get the impression maxtor are the poor relations to wd and seagate, is this true? justified?

[Kab]

No one of the main h/w sites really reviews failure rates in connection to HDDs on a technical level, and there have been EE studies focusing on HDD failure rates finding that the design is the main aspect of vulnerability that render them prone to failure in the first place. This isn't the first time, hardly. Engineering level analysis is what really focuses on facts and physics, and there was one carried out by IBM not long ago with the same or very similar findings. A few corporations also studied this trend due to massive costs with lost h/w and data in many cases.
On a side, I found zero problem with my Maxtors, all of them (5) but my Raptor in RADI0 was the first to go out of all my drives which includes a wide array of them. Heavier than normal use is one candidate for destruction

[ianm2]

It reminds me a little of a car magazine 10 yrs ago, who purely reviewed cars on the COMPARITIVE basis that 1. if they were pretty fast 2. they handled well, 3/ they had a snob value,

they would get a top review.

whereas, if you broke down at night in the middle of nowhere lashing down with wind and rain, they didn't seem to care, unlike a well know consumer focused magazine which rates living with the car and reliability and owner satisfaction as the most important.

I know which I would prefer after the odd breakdown, I ain't a fan of japanese things for nothing

trouble is, mag. reviews I don't think can test componets for the hours a consumer will likely keep them running for.

yes I suppose a 'scientific' exhaustive study to establish all failure mechanisms is useful, even if it did bring up the obvious in one instance.

radio? don't you mean raid 10? or raid zero its a bit hard to tell

it depends what is defined by 'normal' use, that varies between just about everyone who uses it, is there a 'normal' for hard drive use?

[Bbq]

Like just about every review of a computer system on the internet consists of running benchmarks on an expensive computer, or "I plugged it in and it works". In that case, [H] consumer reviews the whole aspect (especially customer service) and i wish more people would do that.

[Kab]

Originally Posted by ianm2

radio? don't you mean raid 10? or raid zero its a bit hard to tell

LOL! Yes, that was supposed to be RAID0.
Maxtors for me did perfect on desktop systems but complete failure in servers (don't ask me why). Seagate were the ones that did best overall in servers for moi Obviously it differs from one user to another, but there are pretty interesting known findings.
Like what Google found here and here's the study paper itself dated Feb 2007.

[doublejack]

I have a retro Dos gaming box that is made from uber cool old skool parts. For storage it's got an original Adaptec 2940, and two SCSI drives - a Micropolis 2gb and a Quantum Fireball 1gb. Those drives date to circa 1994, have been in service since new and are still kicking.

But I've had many drives fail on me over the years. It's happened to me so often that now I can usually tell a failure is imminent just from a change in the noises they make.

I think the manufacturers have taken a beating on RMA's and have noticed that warranties are getting shorter and shorter as a result. Used to be every drive sold had a 3 year warranty minimum, but go to a B&M store and a majority of drives they stock now only have 1 year of coverage. That alone tells you something about drive failures and MTBF.