Are desktop class hard drive failures vs enterprise class hard drive failures as expected with the desktop drives being higher? Based on our own subjective and objective experience in the drive industry, we have long suspected that desktop hard drives and enterprise drives fail at roughly the same rate. This goes against the conventional wisdom that says enterprise drives last longer. The conventional wisdom is that all you have to do is look at the two types of drives to see the better construction of the latter – heavier metal enclosures and the like. We’ve read claims and specifications that say many enterprise drives have bearings at both the top and bottom of the spindle, which reduces vibration and leads to longer life.
A series of data released by online backup provider Backblaze says that 78% of the drives they purchase and run 24×7 are living longer than 4 years. But the bigger surprise is that their experience mirrors our own – that enterprise drives actually fail slightly more than the consumer drives! To be fair the sample of enterprise drives have only been run for 2 years (they may be more reliable as time goes on) and the sample size is much smaller than for the consumer drives. Since backblaze uses consumer drives for most of their online backup, they have more statistically relevant failure data for those. Still, enough drives are involved to find the data compelling… and somewhat damning for those expensive SAS drives.
|Enterprise Drives||Consumer Drives|
|Drive-Years of Service||368||14719|
|Number of Failures||17||613|
|Annual Failure Rate||4.6%||4.2%|
When viewed in this manner and from a reliability perspective, it appears the extra cost of enterprise drives may not be worth it.
A more recent review of hard drive life can be found over at Comparitech’s website