Rust keeps on spinin', spinin', spinin'; into the future

Subject: General Tech | January 29, 2019 - 12:39 PM |
Tagged: hdd, spinning rust, failure, mtbf

Backblaze have been one of the larger remote backup providers for quite some time now, and running their 104,778 HDDs gives them a lot of insight into drive failures.  They are kind enough to share their data with the public, to give a bit of insight into how many of their various drives have failed.  The data does not represent the findings of a proper scientific study but the sample size is large enough to give an idea how WD, Toshiba, Seagate and Hitachi drives fare in the wild. 

The Register is quite impressed with the overall durability of the drives, including the Toshiba 5TB drives, none of which have failed since the 45 of them were deployed in Q2 2016.

die.PNG

"Just 139 out of 10,000 12TB Seagate drives fail a year, and Western Digital's HGST brand has an even better rate of 51 in 10,000, according to cloud backup service provider Backblaze, which has 104,778 drives spinning in its data centre."

Here is some more Tech News from around the web:

Tech Talk

Source: The Register

Lies, damned lies and statistics

Subject: General Tech | January 26, 2015 - 01:15 PM |
Tagged: statistics, mtbf, hdd, backblaze

Backblaze is a moderately sized hosting company with about 40,000 disks set up in their own style of arrays called Storage Pods, which are open sourced so that you can build one yourself if you so desire.  Every once and a while they put out reliability numbers for the HDDs that they use in their arrays; the newest report just arrived for your perusal.  This is good as most reliability and market share studies are done by professional organizations which they tend to charge quite a bit for their findings as they do put a lot of effort into ensuring that their data is correct.  Unfortunately that also means that most people do not have access to the information and make judgments based on incomplete or incorrect data.  As The Register points out, 40,000 HDDs is a very small sample size compared to the market as a whole or even large hosting companies and so the data set you can see here may not be the best representation of the actual market failure rates projected from it may not be overly accurate.  On the other hand it is nice to have any data, especially when you are provided with the actual sample size and a definition of failure.  If you are really into the numbers game, spend some time researching the Mean Time Between Failure and Average Failure Rate and the ongoing debate on how to properly measure expected mortality rates among large drives.

quote-lies-damned-lies-and-statistics.png

"We're not entirely comfortable with cloud backup outfit Backblaze's data on disk drive reliability, but the company has just popped out another year's worth of analysis on which drives hang around longest. With due scepticism, let's have a look."

Here is some more Tech News from around the web:

Tech Talk

Source: The Register

Hard drives die too

Subject: General Tech | November 13, 2013 - 01:15 PM |
Tagged: mtbf, hdd

One thing you can do when you have 25,000 consumer level HDDs running is to amass accurate data on the failure rates of drives.  Backblaze has done exactly that and published their findings which match fairly closely to the predicted MTBF pattern of a spike in the beginning of the life cycle as flawed drives fail, a long period of reliability followed by another rise in failures as drives age beyond their expected lifespan.  They have told the Register that they intend to follow up with tests on enterprise grade disks to see if the premium you pay is a good investment.

images.jpg

"Cloud backup outfit Backblaze has cobbled together all the data it's gathered from the 25,000 or so disk drives it keeps spinning and drawn some conclusions about just how long you can expect disks to survive in an array."

Here is some more Tech News from around the web:

Tech Talk

Source: The Register