{"id":87523,"date":"2019-01-22T08:00:51","date_gmt":"2019-01-22T16:00:51","guid":{"rendered":"https:\/\/www.backblaze.com\/blog\/?p=87523"},"modified":"2021-07-23T12:46:51","modified_gmt":"2021-07-23T19:46:51","slug":"hard-drive-stats-for-2018","status":"publish","type":"post","link":"https:\/\/www.backblaze.com\/blog\/hard-drive-stats-for-2018\/","title":{"rendered":"Backblaze Hard Drive Stats for 2018"},"content":{"rendered":"

\"Backblaze<\/p>\n

\n

We published our first “Hard Drive Stats” report<\/a> just over 5 years ago on January 21, 2014. We titled that report “What Hard Drive Should I Buy.” In hindsight, that might have been a bit of an overreach, but we were publishing data that was basically non-existent otherwise.Many people like our reports, some don’t, and some really don’t — and that’s fine. From the beginning, the idea was to share our experience and use our data to shine a light on the otherwise opaque world of hard disk drives. We hope you have enjoyed reading our reports and we look forward to publishing them for as long as people find them useful.<\/p>\n

Thank you.<\/p>\n<\/div>\n

As of December 31, 2018, we had 106,919 spinning hard drives. Of that number, there were 1,965 boot drives and 104,954 data drives. This review looks at the hard drive failure rates for the data drive models in operation in our data centers. In addition, we\u2019ll take a look at the new hard drive models we\u2019ve added in 2018 including our 12 TB HGST and 14 TB Toshiba drives. Along the way we\u2019ll share observations and insights on the data presented and we look forward to you doing the same in the comments.<\/p>\n

2018 Hard Drive Failure<\/a> Rates: What 100,000+ Hard Drives Tell Us<\/h2>\n

At the end of 2018 Backblaze was monitoring 104,954 hard drives used to store data. For our evaluation we remove from consideration those drives that were used for testing purposes and those drive models for which we did not have at least 45 drives (see why below). This leaves us with 104,778 hard drives. The table below covers what happened just in 2018.<\/p>\n

\"2018<\/a><\/p>\n

Notes and Observations<\/h3>\n

If a drive model has a failure rate of 0%, it means there were no drive failures of that model during 2018.<\/p>\n

For 2018, the Annualized Failure Rate<\/em> (AFR) stated is usually pretty solid. The exception is when a given drive model has a small number of drives (fewer than 500) and\/or a small number of drive days (fewer than 50,000). In these cases, the APR can be too wobbly to be used reliably for buying<\/a> or retirement decisions.<\/p>\n

There were 176 drives (104,954 minus 104,778) that were not included in the list above. These drives were either used for testing or we did not have at least 45 drives of a given model. We use 45 drives of the same model as the minimum number when we report quarterly, yearly, and lifetime drive statistics<\/a>. This is a historical number based on the number of drives needed to fill one Backblaze Storage Pod<\/a> (version 5 or earlier).<\/p>\n

The Annualized Failure Rate<\/em> (AFR) for 2018 for all drive models was just 1.25%, well below the rates from previous years as we\u2019ll discuss later on in this review.<\/p>\n

What\u2019s New in 2018<\/h3>\n

In 2018 the big trend was hard drive migration: replacing lower density 2, 3, and 4 TB drives, with 8, 10, 12, and in Q4, 14 TB drives. In 2018 we migrated 13,720 hard drives and we added another 13,389 hard drives as we increased our total storage from about 500 petabytes<\/a> to over 750 petabytes<\/a>. So in 2018, our data center techs migrated or added 75 drives a day on average, every day of the year.<\/p>\n

Here\u2019s a quick review of what\u2019s new in 2018.<\/p>\n