Backblaze publishes data on specific hard drive reliability

vbimport

#1

We’ve just posted the following news: Backblaze publishes data on specific hard drive reliability[newsimage]http://static.myce.com//images_posts/2014/01/blog-fail-drives-manufacture-95x75.jpg[/newsimage]
[quote][/quote]
Read the full article here: http://www.myce.com/news/backblaze-publishes-data-on-specific-hard-drive-reliability-70246/

            Please note that the reactions from the complete site will be synched below.

#2

looks like sticking to western digital drives would be a good idea for me even though they are slightly more expensive. bit worried about all the 3tb seagates I have.


#3

Interesting numbers, but clearly skewed by the nature of the drives being used. Seagate and WD both produce a number of cheaper consumer level drives that fail more often. If these folks are buying drives based solely on price, they will be getting more of the cheaper drives and fewer of the more reliable models. So when grouped by drive maker, the brand that sells the most cheap drives will appear to be less reliable.

This type data is far more useful when tabulated by drive model than by maker.


#4

@CDan, agreed. There’s a lot of pertinent info missing.


#5

I’d rather choose to imagine the publisher of that statistics’s unreliable. Somebody should publish data on the reliability of websites and statistics posted on the web and dedicated (and highly costly) analysis and consultation in PDF format on the reliability of HDD and SSD.


#6

I don’t see [B]any[/B] reason to doubt the source’s reliability. They are simply sharing real world experience while using many thousands of these drives. And the most problems they have had were with Seagate drives, and yet they will continue to purchase from Seagate for certain models, particularly the 4TB ones. If they were trying to denigrate Seagate, I don’t see why they would continue buying their drives.

Like them, I wonder about the performance of the Hitachi drives in the future, now that WD has bought them. Does HGST produce consumer drives now? All I’ve seen recently have been enterprise grade. Ah, answered my own question…they do have the Deskstar and Coolspin lines.


#7

I don’t see any reason to trust any source’s realiability on the reliability of HDD and SSD. Have you ever found a reliable source comparing the reliability of statistics on the reliability of HDD and/or SSD?

Perhaps StorageReview’s reliability statistics of HDD’s rather well qualified, much more reliable than a single-source reliability statistics published randomly.

Most products Seagate has sold during the recent few years were 500GB-2TB HDDs. It’s reasonable to estimate those products are responsbile for most of Seagate’s phenomenal amount and ratio of sales and profits. NO company can survive a single quarter with 15% FAILURE rate (as opposed to RETURN rate) for mainstream products. I’m not implying a certain user - even if that user’s been using 10,000 units which’s still less than 0.001% of the total circulated in the recent few years - lying with the intention to make Seagate look bad (and no one could or can within foreseeable years.)

The exact opposite’s more likely to be true. Low failure rate’s the key to profitability especially when a hardware manufacturer that designs and makes the major parts and assembles them into final products intends to maintain an exceptionally large market share by selling at very competitive prices to both OEM and retail.


#8

I see no reason not to trust their data. They are simply reporting what they encountered after buying and using thousands of drives of all three manufacturers.

We can make all the asumptions we want about their motives, but they are no more than that. Assumptions. If anyone has any evidence that they might be related to some HDD manufacturer in any way I’m sure he won’t keep it for himself.


#9

That sounds like a joke. That title and interpretations are what count. Whether that data is true or not is hardly relevant. To interprete a few hundreds or dozens of specific models in a certain IDC as a reliable data to determine the reliability of the specific model or manufacturer is more than unreliable.


#10

[QUOTE=CDan;2716979]Interesting numbers, but clearly skewed by the nature of the drives being used. Seagate and WD both produce a number of cheaper consumer level drives that fail more often. If these folks are buying drives based solely on price, they will be getting more of the cheaper drives and fewer of the more reliable models. So when grouped by drive maker, the brand that sells the most cheap drives will appear to be less reliable.

This type data is far more useful when tabulated by drive model than by maker.[/QUOTE]

Cheap drive doesn’t a unreliable drive make.

And as the text said - all of the drives are “consumer” grade drives, which offer the best capacity/price -ratio, so grouping by manufacturer is fair.

When buying a new drive, I - as one - will go to the one with the cheapest capacity/price -ratio (if not special usage - portable or SSD speed needed)

The most puzzling thing to me is what is said about the WD Green 3TB and Seagate LP 2TB drives. “…two specific drives that would not work well in their particular hardware configuration”

What kind of hardware configuration you will need that it destroys drives? At first I thought that in such a data storage company those “Green/Low power” drives would spin up and down repeatedly so they would fail immaturely. But wouldn’t the drives spin up and down at a elevated pace even at home usage? So those two drives would be not recommended (except for certain uses).


#11

well they have pods with 25 drives in each of them, from what ive read the older revisions don’t have much vibration dampening whilst consumer drives need a low vibration environment afaik.