Why is that newer hard drives are so prone to failures?

[baghu]

I've had 6 HDD failures in the past year, all from various brands like Seagate, WD, Hitachi, but my old 40 GB, 20 GB HDDs still work perfectly if I power them up, even after being in service for >30,000 hrs.

Even the RMA drive WD sent me goes in just a few hours!

[fzabkar]

Your replacement drive was probably refurbished. It could have had a prior history of head/media problems.

[data-medics]

Personally I think it's related to data density. The new drives are trying to write data so small that it's becoming unreliable. Even the slightest bump will destroy these new drives.

I personally don't trust anything over 500Gb but that's just me.

[fzabkar]

ISTM that data-medics is probably right. The flying height of today's drives must be a lot lower than in those older models, and today's higher track density would be placing much greater demands on the servo system and on the motor bearing.

To get an idea of the progression in technology, it might be worth following the changes in the Hardware ECC Recovered SMART attribute in Seagate's models. AFAICT, the normalised value for this attribute has steadily declined with newer models, to the point that Seagate no longer appears to report it. ISTM that HDDs are nowadays digging data out of noise.

[pcimage]

I also suspect that modern drives are of generally poorer quality, built to a price rather than quality, to meet the "demand" for cheap drives.

Especially drives like some WD Blue drives, which feel very flimsily built, and all Seagate slim desktop drives which are cheap and nasty too

[guru]

Agree, Seagate desktop "seem" to be the latest and greatest "CRAPWARE"

[HaQue]

in contrast, a few of my collegues that run some ridiculous storage setups are raving about Seagate Flash based storage

[labtech]

Seagate Flash based storage

Some people called it "immortal".

[fzabkar]

in contrast, a few of my collegues that run some ridiculous storage setups are raving about Seagate Flash based storage

That's SandForce IP, isn't it? (SandForce -> LSI -> Avago -> Seagate).

[HaQue]

..........

sf.jpg (41.55 KiB) Viewed 14915 times

Western Digital also acquired Virident that does enterprise flash storage, so will be interesting to see where it goes from here.

From a DR standpoint, being enterprise means that (hopefully) when the Sandforce controller goes belly-up, the DR is from backups.

If they start selling cheap consumer ones, then hopefully some bright spark can get some SF DR happening.
This is interesting, wonder how?:
"Along with the PCIe interface, enhanced, newer versions of SandForce’s excellent wear-leveling, maintenance and error correction engines are on board here as well. In fact, the controller can now handle a full die-level failure in the Flash array and still maintain data integrity. "

I personally think a lot of people are underestimating the speed in which Flash based storage become the standard choice, SSD's or PCIe.

statements like "There’s obviously a long way to go before HDDs will be completely put out to pasture" I think are flippant.

Seagate have also used other controllers, one I had not heard of very much at all, and haven't seen any cases for...Link-a-Media (LAMD), use in Seagate 600 SSD :

http://www.storagereview.com/seagate_600_ssd_review

getting back on topic, I don't think it is possible yet to judge which SSDs are more trouble-prone, or if they are indeed more prone to failure or not.

[rehaandrew]

You can't really say anything, but yes i agree today HD's are more subjected to failure. My 40gb seagate gone just like that. I couldn't even recover the data inspite of using some softwares.

[baghu]

Thank you, everyone, for the great, informative answers. I went through all of them and now it all makes sense to me. Thanks a bunch for a goldmine of information here! Have a great day!

[fzabkar]

It's an old thread, but I recently saw the SMART report for Seagate's recent ST6000DM004-2EH11C 6TB drive:

http://www.techsupportforum.com/forums/ ... 1501285606

The drive is error-free and 7 hours old.

Code:

Attribute ID   C3 / 195
Attribute name   Hardware ECC Recovered (aka ECC On the Fly Count)
Current      2
Worst      2
Threshold   0
Raw Value   0006A3FE6C

[scratchy]

I've had an hitachi 4 tb running more or less continuously for about 4 years. Gets a lot of data written, deleted etc and is still running fine. I guess it's luck of the draw.