I might not use Backblaze’s cloud storage service, but I sure do love the company’s transparency when it comes to hard drive reliability. I’ve reported on the company a couple of times before, and always look forward to its latest reports to see which vendor’s doing well, and which vendor’s not.
This time around, the company’s report includes results for 8TB drives; in particular, 2,720 Seagate ST8000DM002, and 45 HGST HUH728080ALE600. To date, 0 of the HGST drives have failed, but 3 of the Seagate ones have. Given the massive delta between the number of 8TB drives deployed between the two vendors, it’s impossible to make conclusions at this point, but an annualized failure rate of 3.06% for the Seagate drives is still pretty high.
Compare that annual failure rate to the 4TB Seagate ST400DM000 drives that Backblaze uses: 34,693 of them. Despite having more than ten times the number of drives, these 4TB models have an annual failure rate of 2.66% – and those drives have been in use for much longer. It’ll be very interesting to see Backblaze’s next report, to see how both of the 8TB models fare.
The aggregate annual failure rate for all of the hard drives Backblaze has in use is 2.02%, which isn’t too bad: it means that one in every fifty hard drives have failed, which is a good average for those home users who deploy no more than a handful of them. It’d be great to see the failure rates even lower, but hard drives are mechanically complex, and Backblaze’s drives are going to be used much more than the average drive in our homes.
In case you’re curious about the amount of storage Backblaze has in use right now, I’d first warn you to set your cup of coffee down and refrain from taking a sip until you look over the graph below:
That’s right – Backblaze has a staggering 250 petabytes of storage in use. That’s equal to 250,000 terabytes, 250,000,000 gigabytes, or just enough to sustain Kanye West’s ego.