Cloud backup provider is fast becoming one of my favorite companies in IT, if only because it shares so much information about its operations. On Tuesday, the company — famous for its open source storage-pod designs and infamous for getting blacklisted from stores including Best Buy and Costco during the 2010 hard drive shortage — released a new set of data regarding the lifespan of the hard drives on which its service runs.
Backblaze Distinguished Engineer Brian Beach provides a lot of detail in a blog post about the study, but here are the highlights:
- The company is storing 75 petabytes of customer data on more than 25,000 consumer-grade hard drives.
- It has been in business for 5 years and 74 percent of the hard drives it has ever deployed are still running.
- Annual failure rate for drives is 5.1 percent for the first 18 months, 1.4 percent for the next 18 months, and 11.8 percent during years three and four.
- By Backblaze’s estimates, more than half of its drives could still be running after 6 years — a finding most studies on hard drive lifespans to date would not have predicted.
The direct implications of the study beyond the borders of Backblaze’s business are unclear — its cloud backup service is certainly different than running a large web operation like Google, or even an application-serving storage service like Amazon S3 — but the numbers are interesting nonetheless. Taken as a whole, the efforts of companies like Backblaze and organizations like the Open Compute Project suggest a future where intrepid CIOs can legitimately question prior assumptions about how much hardware should cost, what it should look like and how long it should last.
And if you want to hear a little more about Backblaze’s history — its ups, downs, courtships from the CIA and blown marketing budgets — listen to our recent Structure Show podcast with Co-founder and CEO Gleb Budman.
[soundcloud url=”https://api.soundcloud.com/tracks/116861363″ width=”100%” height=”166″ iframe=”true” /]