Backblaze reports Q3 2017 HDD Failure results - includes 12 TB Drives as well

Can someone please explain what "Drive days" means? 'cause if that's the time the drive has been operating... some of them have been working since over 8600 years ago... )

 

If you have 100 drives and run them for a week, you have 700 days of usage.

 

Take home message is that you don't want a 4TB Seagate drive!

 

So: "Drive days" / "Drive count" = single drive operating time?
For the Seagate in question: 3.152.542 / 33.510 = 94 days ???

So every 3 Seagate drives out of 100, are dead after just 3 months of non-stop work???

 

Not all drives were installed on the same day, so at best thats an average, but its unclear if there is a bunch of really old ones and a bunch of new ones, for example.

 

Interesting stats... gotta love Seagate..

 

As you can see, Seagate sells way more drives than any company.

 

Yeah. Seagate must have built in their 4TB drives some ingenious timed breaking functionality, to sell more drives as replacements.

 

Guess you aren't too smart, then

 

Holy cock those Seagate 4TB drives. So if I understand it correctly, 1/3 of them fail within a year of usage?

 

You, sir, look like a really honest and intelligent person. People like you make the world a better place.

 

Sell a lot of drives, get a lot of dead drives... makes sense I guess.

After all 4 of my seagates had the click of death I won't be purchasing that brand again ( ◉ - ◉ )

 

I have all my 14 WD drives dead, from seagate only 4 and those were from same problematic 3TB batch and all of them were showing symptoms in SMART prior of being unusable unlike WD which just died instantly or with very little warning from SMART.
In my main rig currently running 6 Seagate drives (3/6/10TB) with average 20k-40k hours of operations running (24x7) operations and none showing any issues so far.
Though none of those are cheapest desktop drives variants.

My future disks most likely will be again 10TB+ Helium based Baracuda pro.

 

That seems acurate, i have changed hundreds in CCTV machines

 

Every Seagate I've ever owned has failed.

 

wat WD you been using ? Green or Blue i guess?
if yes, wonder if any relation to hdd park that wd claims for saving power

i use WD black, i know price abit high, but i am prefer paying abit more for reliability
most HDD depends on workload, should at least working without any issue for 16k hours (2years)
on light workload/use then we can expect it last 4~5years which is within hdd warranty (black series)

green and blue only have 2/3years warranty, so we can simply the company also confidence their product working within the warranty

 

Makes me feel a little better about my recent Toshiba purchase. It was Between them and HSGT.

 

To my understanding, this data is based on whenever the drives report a failure, regardless of whether or not the failure was just a fluke or if drive is still usable. When I worked at IBM, a lot of companies didn't feel like taking their chances with any errors; they just toss the drives as soon as they report any semblance of inadequacy.
In my experience, Seagate has a lot more SMART stats, and they're more picky about what they deem a failure vs competitors. If you're someone who doesn't like to take chances with data, Seagate is good because they let you know of a potential problem long before it actually fails, but, most of the time the problems they report can be ignored.

As for whether or not Seagate drives actually fail more often than other drives (to the point where they are incapable of being used or trusted) I'm not really sure. I personally have used HDDs from all brands where they had long-term or major SMART errors and they operated just fine for many years, or have yet to actually fail. And no, I don't use these drives for anything important.

 

If they mix together new and old drives, I don't see how this tables can be of any use ....

Example:
Brand A has 50 disks 5 years old and 50 1 year old. Then 20 disks fail so fail rate is 20% on average 250 disk years.
Brand B has 100 disks 3 year old. Then 20 disks fail, so fail rate is 20% on average 250 disk years.

Brand A and B would look identical in those tables but A is actually more reliable than B in the case only old disks fail, or terrible worse than B if the failing ones are the newer......