Which hard disk is best? Here are some pointers

Backblaze has published another report on disk reliability

Backblaze uses a hell of a lot of hard drives for its cloud backup and storage services, and every year (since 2014) the firm produces a report on what it has found to be the most (and least) reliable drives, with the review of 2015 just having been published.

So what brand or disk size should you be looking for to give you the best odds of avoiding a drive failure that could cause a great deal of inconvenience, or even mess with your personal data (though surely not your business data – after all, SMBs out there have a cast-iron backup plan in place, right?).

According to Backblaze's stats, compiled across some 56,224 spinning HDDs, the most reliable brand is HGST with an overall cumulative failure rate (since the company started these surveys) of just 1%. Toshiba and Seagate both came in between 3% and 4%, whereas Western Digital was the clear loser in the reliability stakes with an overall failure rate of 7%.

But what disks were Backblaze buying the most of during last year? Seagate hard drives, not HGST.

Seagate supreme?
Backblaze noted: "Nearly all of the 16,000+ drives purchased in 2015 have been Seagate drives. Of the Seagate drives purchased in 2015, over 85% were 4TB Seagate drives."

4TB is the sweet spot, then, and Seagate drives were at a failure rate of just under 3% for this capacity, although HGST was under 1%.

However, HGST no longer produces these 4TB drives, having switched over to higher priced ones with better performance – and the Seagate 4TB drives are in plentiful stock and very competitively priced, with an "acceptable" failure rate, so these were the ones Backblaze bought in big numbers.}

Seagate's 6TB drives also performed well with a cumulative failure rate of 1.9%, compared to 5.8% for Western Digital.

It's also worth noting that with Seagate, Backblaze found that these hard disks were more likely than other vendors to signal an impending failure via SMART stats, which is good to know. It's obviously useful to have an early indication of disk trouble brewing, particularly for business-critical machines.

Tweet