Backblaze report shows larger hard drives are more reliable than small ones

Oct 23, 2018

John Aldred

John Aldred is a photographer with over 20 years of experience in the portrait and commercial worlds. He is based in Scotland and has been an early adopter – and occasional beta tester – of almost every digital imaging technology in that time. As well as his creative visual work, John uses 3D printing, electronics and programming to create his own photography and filmmaking tools and consults for a number of brands across the industry.

Backblaze report shows larger hard drives are more reliable than small ones

Oct 23, 2018

John Aldred

John Aldred is a photographer with over 20 years of experience in the portrait and commercial worlds. He is based in Scotland and has been an early adopter – and occasional beta tester – of almost every digital imaging technology in that time. As well as his creative visual work, John uses 3D printing, electronics and programming to create his own photography and filmmaking tools and consults for a number of brands across the industry.

Join the Discussion

Share on:

Hard drives are something that all of us have. Some of us just have the one our computer came with, while others will have a whole mountain of them. I’m in the latter camp, although I do try to keep things to a minimum. Every few years I replace all my storage drives as faster, cheaper, larger capacity drives are released.

So, every few years I’m in the market for new drives. So, I look at large consumers of hard drives. Consumers like Backblaze, who just released their Q3 2018 report, showing their hard drive failure rates and they’re quite surprising. At least, they surprised me. It seems that those huge 8TB, 10TB and 12TB drives are actually far more reliable than smaller drives.

Blackblaze says that as of September 30th, 2018 they had 99,636 spinning hard drives. Of those, 1,866 were boot drives and the remaining 97,770 were data drives. Their review looks at the quarterly and lifetime statistics of the data drives in operation in their data centres.

For the purposes of the test, Backblaze didn’t count drives that had been used simply for testing and those models for which they didn’t own at least 45 samples. That leaves 97,600 drives.

You’ll often hear people say that larger drives (and larger memory cards) are more prone to failure than smaller ones. Of course, this is nonsense. I’ve never really considered that the capacity of a storage device had anything to do with its failure rate. Simply the tech and build quality. And, obviously as larger drives are generally newer tech, they’re going to be more reliable – at least, that’s my theory.

It just makes sense to me to replace all my drives with larger ones. My storage needs are always growing, and fewer larger drives are less expensive to run than many smaller ones. “Don’t put all your eggs in one huge basket!” people say, but just buy several, backup and don’t sweat it is what I say.

That seems to be the approach Backblaze has taken, too. They replaced their 3TB, 4TB and a handful of 6TB drives with 3,600 new 1121TB drives. They ended up with 584 fewer drives but with over 40 petabytes more storage space.

The lifetime failure rates of the 8TB, 10TB and 12TB models show an extremely good reliability. And the overall annualised failure rate of 1.71% is the lowest Backblaze says they’ve ever achieved.

Even more interesting, though, is how well Seagate seems to be doing with these larger capacity drives. Especially considering the rather high 4TB drive failure rates from Q3 last year. It looks like they might have finally gotten their act together. It looks like Backblaze has almost entirely eliminated Western Digital from their lineup now.

So, if you’re looking for large capacity backup drives in the near future, Seagate may be the way to go.

Check out Backblaze’s complete Q3 2018 report on their website.

[via Techspot]

Filed Under:

Tagged With:

Find this interesting? Share it with your friends!

John Aldred

John Aldred

John Aldred is a photographer with over 20 years of experience in the portrait and commercial worlds. He is based in Scotland and has been an early adopter – and occasional beta tester – of almost every digital imaging technology in that time. As well as his creative visual work, John uses 3D printing, electronics and programming to create his own photography and filmmaking tools and consults for a number of brands across the industry.

Join the Discussion

DIYP Comment Policy
Be nice, be on-topic, no personal information or flames.

Leave a Reply

Your email address will not be published. Required fields are marked *

3 responses to “Backblaze report shows larger hard drives are more reliable than small ones”

  1. ForSquirel Avatar
    ForSquirel

    I mean, is this is even surprising? When you have a larger surface area to paint you’re less likely to repaint the same area twice.

  2. Ahmet Avatar
    Ahmet

    Two examples from the above:

    HGST 12 TB drive average run days: 80 (total of 0% died)

    Seagate 4 TB drive average run days: 1772 (total of 13% died)
    What do the data say about reliability? Nothing. OK, it says HGST 12 TB most probably will run for 80 days without failure.
    To know how reliable these drives are, we should run them still they all die. Then we can say something. The more you run a mechanical drive, the higher the chances is will fail, and it is not linear.

    I think if you run 97 thousand HDDs you better go for bigger ones, since the power consumption will be lower, the operation will be less complex, you need fewer people to run the whole thing, and since the prices of the drives are dropping, you will be better off buying one 12 TB than three 4 TB drives ($/TB).

  3. ext237 Avatar
    ext237

    Because larger drives were installed this quarter, the 3 and 4TB drives were installed in 2013/14.

    A hard drive that runs continuously in a heavy access environment for 4 to 5 years are going to fail much sooner than drives installed a month ago. Your headline is misleading and your data analysis is based on correlated causal relationships, not control normalized comparison data.

    It’s like a headline that reads “People who die in car crashes are more often riding in a car than those who do not.” It is factually correct, but a useless (and obvious) correlation statistic.