Printed from http://www.electronista.com

BackBlaze analyzes desktop hard drives, notes manufacturer trends

updated 09:02 am EST, Tue January 21, 2014

Hitachi praised for reliability; Seagate cheaper, less reliable

Backup cloud services company BackBlaze have studied the platter-based hard desktop drives that have been in use since the launch of the service. At the end of 2013, the company had evaluated 27,134 drives from five different manufacturers across 15 different models spinning in its BackBlaze Storage Pods, and collated the data on drive reliability in a study, published today.

Referring to the testing method, BackBlaze says that it buys "the cheapest drives that will work. When a new drive comes on the market that looks like it would work, and the price is good, we test a pod full and see how they work. The new drives go through initial setup tests, a stress test, and then a couple weeks in production. If things still look good, that drive goes on the buy list. When the price is right, we buy it."

Of the vendors evaluated, Hitachi (now a subsidiary of Western Digital) had the highest survival rate, with 96.9 percent of purchased drives surviving after 36 months. After an initial batch of early failures in the first three months of life, Western Digital comes in second at 94.8 percent. Seagate trails in third, at 73.5 percent still alive after three years.

BackBlaze notes that it has not included Western Digital 3TB Green drives, and Seagate 2TB Low Power drives due to lack of suitability in the BackBlaze environment. The company notes that the drives "start accumulating errors as soon as they are put into production." Topping the failure rate is the Seagate Barracuda 7200 RPM 1.5TB capacity, at a 25.4 percent annual failure rate.



By Electronista Staff
Post tools:

TAGS :

toggle

Comments

  1. Makosuke

    Junior Member

    Joined: 08-06-01

    This study is exactly what I've been hoping some datacenter would do for years.

    Anyone can post anecdotal evidence about hard drive brands and models, but with something as statistically distributed as drives are, they're just that--anecdotes. Even clusters of failures are nearly meaningless in anecdotal terms--they might have all come off the same pallet that got dropped or frozen or something.

    But this--large numbers of drives run in real-world conditions for years--is the real deal. The manufacturer statistics are very interesting, but it's the model stats that are if anything even more useful, since they tried a number of consumer-grade drives from every manufacturer, and the results varied wildly depending on model and capacity.

    The only one I'm sad about is that they don't have any stats on Seagate's new(-ish) NAS drives, since I just bought four of them. The brand stats don't look very good compared to WD Reds (despite having one less platter at 4TB), but again, the huge variation between individual models and capacities might shake out differently, and it doesn't look like they have any hard numbers for 4TB Red drives, either, despite liking the 3TB model.

Login Here

Not a member of the MacNN forums? Register now for free.

toggle

Network Headlines

toggle

Most Popular

Sponsor

Recent Reviews

PenClic Bluetooth mouse

Windows 8 aside, computer users have been trained that a mouse is the proper way to navigate through the desktop for years. Trackballs ...

Cat B100

Cat is primarily known for its heavy-duty machinery used in the construction industry and farming, among other areas. What may not be ...

Linksys EA6900 AC Router

As 802.11ac networking begins to makes its way into more and more devices, you may find yourself considering an upgrade for your home ...

Sponsor

toggle

Most Commented

 
toggle

Popular News