Sport & Auto
- About Future
- Digital Future
- Cookies Policy
- Terms & Conditions
- Investor Relations
- Contact Future
Whenever talk turns to the topic of hard drives, inevitably you'll find that some people swear by brand X while others will only buy HDDs from brand Y. Their reasons are often anecdotal, and usually influenced by a bad experience with a particular model or brand. A person may say something like, "Each time I've plopped a brand Y HDD in my home server, it's crapped out after a year, but my brand X drive is still going strong after 10 years!" Unfortunately, two separate accounts of HDD longevity can (and often do) contradict each other, but there are other problems that prevent a logical conclusion, such as the lack of scientific data. What's much more meaningful is an ongoing study of 25,000 hard drives.
Backblaze, a cloud backup company, keeps over 25,000 HDDs spinning at all times, but even so, the topic of reliability is a complicated one. In Backblaze's experience, 26 percent of HDDs fail within the first four years. Beyond that, data is limited at this point because Backblaze has only been operating for five years.
Another thing that Backblaze has observed is what's called the bath curve, which plots out expected failure rates based on three different kinds of defects or failures: Factory defects resulting in "infant mortality," random failures, and parts that wear out resulting in failures after much use.
Since Backblaze began backing up data, it's noticed that the failure rate hovers around 5 percent for the first 18 months, then drops for a period, and then goes up "substantially" around the 3-year mark. The "infant mortality" rate isn't something Backblaze has seen much of in the real world, "but it does look like 3 years is the point where drives start wearing out."
There's a lot more to digest -- a LOT more -- in Backblaze's study. If you have some free time and feel like leveling up your HDD geek skillset, we highly suggest giving the study a read.