News - Major cloud service provider publishes HDD reliability chart

[HEXUS]

Results garnered from its use of over 27,000 HDDs in the past three years.

Read more.

[Dooms]

Very interesting. I've never had much faith in Hitachi before.

Previously all my drivers where the Samsung F series and after being on 24/7 for nearly 3 years they are still going strong *touch wood* shame you can't get them any more - Experience before that seems to agree with their findings of WD drives, long as it didn't die quickly then it was generally good for a long time

[this_is_gav]

Very interesting. I've never had much faith in Hitachi before.

I still don't have any faith in Hitachi. I seem to have a laptop with a failed/failing hard drive land on my desk a few times a month, and it's scary how many of them have Hitachi hard drives.

That said, up until last year I'd seen one Western Digital drive fail (and that was about 7 years old and absolutely coated in dust), but in the last year I've had 3 myself (I use a lot of hard drives in my respective machines at home and work) and a few from laptops too.

The shocking level of reliability in hard drives is what irks me most about the computer industry at the moment. You can stress to people the importance of backups, but hard drives are getting less reliable as technology (in theory) improves, and for me that is unacceptable.

[zdn1042]

Hmm.. Seagate failing around the 20 month period. That's one to watch out for.
The only HDDs I've used were WD and Samsung and both have served me quite well. Honestly have not considered getting Hitachi before in buying HDDs. And I'm quite impressed with Hitachi's figures. Let's see what will happen with WD & Hitachi...

[MaverickWill]

Well, I guess that explains why Seagate dropped their standard warranty period down to just 12 months!

[kompukare]

I still don't have any faith in Hitachi. I seem to have a laptop with a failed/failing hard drive land on my desk a few times a month, and it's scary how many of them have Hitachi hard drives.

Ah but do you have a comparable bunch of other laptops? Because the way (some) people treat their laptops I am not surprised their HDDs fail. Whereas most desktop HDDs never get moved while on and that makes a big difference. Of course, Hitachi are very common because they are cheap. The caveat with the small sample size people are likely to come across is that there may be very big reliability differences between different models from the same manufacturer or even batches. Certainly those IBM 'DeathStars' skewed the results for DeskStar and similar that Seagate Barrucada a few years ago did too. Having said all that, Hitachi may very well be unreliable...

[this_is_gav]

I still don't have any faith in Hitachi. I seem to have a laptop with a failed/failing hard drive land on my desk a few times a month, and it's scary how many of them have Hitachi hard drives.

Ah but do you have a comparable bunch of other laptops?

We're a school. We've got all sorts!

[crossy]

Interesting figures - I've just got a Seagate SSHD drive to install (to replace a badly performing WD Black) but given those results I'm wondering if I should maybe downgrade it to a media or NAS drive instead. The start-strong-and-fail-increasingly is very worrying to me, although the drive I've got is a 1TB so maybe okay.

Not been that impressed with Hitachi laptop drives, they never seem to be as quick as the competing products from Seagate and especially WD. Although my own laptop is currently running with a Samsung SSD, (with secondary storage on a RAID0 pair of WD Blue's)

I tend to flip/flop between Seagate and WD drives, (although SSD's are now all Samsung), and apart from the WD Black, (see above - my one WD Black drive performs worse than a three-year old Samsung F series), I've got a slight preference for WD. Looks like it was a smart move, especially if WD drives tend to fail early but otherwise continue strong - which I'd regard as the best way to fail, (although "no failures" is obviously preferable).

Smart money though surely is to avoid drives with short warranties. Interestingly enough, the warranty period seems to be clear on WD's site, but carefully hidden on Seagate's desktop site, (unless you have a SSHD where it's clearly stated as "three years limited warranty").

[Noli]

I generally preferred WD and avoided Hitachi. A spate of a few WD failures early doors (as the survey shows likely) caused me to switch to Seagate. Even though they're ok for now (after a year and a half or so) - it looks like I'll switch back to WD next time round. After all, they've got much better stats with 27,000 drives then I could ever muster!

73.5% vs 95% is nowhere near ok vs discounts you can get with Seagate vs WD prices....

[Mama Sumae]

Having recently bought two new HDDs I was surprised to see the oldest one I'm still using is Hitachi branded. 10 years going strong and got some rough treatment too, removed it from a Freecom external drive that used to follow me whenever I travelled.

WD reliability can't questioned but it does carry a premium due to its reputation. Seagate's 1 year warranty for a HDD is ludicrous, surely not fit for purpose and I wont go anywhere near them.

[davidcrofter]

Ouch - I can see a lot of Seagate drives hitting the classifieds in the coming weeks lol ... I think I have about 4 of the 3TB flavour populating my rig in one shape or another.

Be interesting to see if these were mainstream drives or you would imagine some sort of enterprise range (do Seagate even have that??).

[DanceswithUnix]

Be interesting to see if these were mainstream drives or you would imagine some sort of enterprise range (do Seagate even have that??).

These were mainstream drives either desktop or removed from USB carriers.

I think all the manufacturers do an enterprise range, but the aren't cheap: http://www.scan.co.uk/products/2tb-w...-hotswap-ready

The extra features are for use in huge raid arrays, so for home use you probably don't need them.

[ik9000]

Having recently bought two new HDDs I was surprised to see the oldest one I'm still using is Hitachi branded. 10 years going strong and got some rough treatment too, removed it from a Freecom external drive that used to follow me whenever I travelled.

WD reliability can't questioned but it does carry a premium due to its reputation. Seagate's 1 year warranty for a HDD is ludicrous, surely not fit for purpose and I wont go anywhere near them.

Why they're not using ES2 drives puzzles me. Ok they're double the price, but reliability is supposed to be way better than 7200 series drives. Also the stats are skewed by inconsistent sample sizes for the respective manufacturers. Plus they don't differentiate between failures of different models. Going to the source article here shows most of the seagate failures are in one particular drive - the barracuda green which they rate at 120% failure rate? Something not right there. And they only had 51 of those drives. Compared to over 4000 of the ST3000DM001 which have a failure rate of 9%. The 7200.11 series don't look so good but note this is the 7200.11 series and therefore we can't compare that to the 7200.12, .13 and .14 series which have been released since. Unless I'm mistaken all the Seagate drives they are comparing are 7200 rpm vs the 5400rpm of the Hitachi drives. This also could skew the results. Then we get the recent mergers and acquisitions in the HDD manufacturer sphere with WD buying Hitachi, Seagate buying Samsung, and rumours that Toshiba are now making old Hitachi designs... new drives may not be anything like the 3-4 year old designs these guys are using. NB the 7200.11 reported in the article are listed as being run for 3.8 years. This is waayyy older than the Hitachi drives listed at 1.9 years. Of course their failure rate is going to be increased. The only drive with a usage time equivalent are the 80 WD Green WD10EACS but they have N/A data on failure rate. Not zero, just Not available.

What this actually tells me is that the company uses mainly Seagate drives, and rags them into the ground because they consider them best bang for buck, and are willing to factor in some down time if/when some of them do break.

In the article they state [of seagate] "Their overall failure rate isn’t great, but it’s not terrible either"

and in the forum comments they state

Double the reliability is only worth 1/10th of 1 percent cost increase. I posted this in a different forum:

Replacing one drive takes about 15 minutes of work. If we have 30,000 drives and 2 percent fail, it takes 150 hours to replace those. In other words, one employee for one month of 8 hour days. Getting the failure rate down to 1 percent means you save 2 weeks of employee salary - maybe $5,000 total? The 30,000 drives costs you $4 million.

The $5k/$4million means the Hitachis are worth 1/10th of 1 percent higher cost to us. ACTUALLY we pay even more than that for them, but not more than a few dollars per drive (maybe 2 or 3 percent more).

Moral of the story: design for failure and buy the cheapest components you can.

So they buy cheap. They need to add charts of failure rate normalised against unit cost, and average lifetime.

[davidcrofter]

These were mainstream drives either desktop or removed from USB carriers.

I think all the manufacturers do an enterprise range, but the aren't cheap: http://www.scan.co.uk/products/2tb-w...-hotswap-ready

The extra features are for use in huge raid arrays, so for home use you probably don't need them.

Yeah guess I should have read the article. Why would they use consumer drives in an enterprise environment (apart from cost)??

[davidcrofter]

Why they're not using ES2 drives puzzles me. Ok they're double the price, but reliability is supposed to be way better than 7200 series drives. Also the stats are skewed by inconsistent sample sizes for the respective manufacturers. Plus they don't differentiate between failures of different models. Going to the source article shows most of the seagate failures are in one particular drive - the barracuda green which they rate at 120% failure rate? Something not right there. And they only had 51 of those drives. Compared to over 4000 of the ST3000DM001 which have a failure rate of 9%. The 7200.11 series don't look so good but note this is the 7200.11 series and therefore we can't compare that to the 7200.12, .13 and .14 series which have been released since. Unless I'm mistaken all the Seagate drives they are comparing are 7200 rpm vs the 5400rpm of the Hitachi drives. This also could skew the results. Then we get the recent mergers and acquisitions in the HDD manufacturer sphere with WD buying Hitachi, Seagate buying Samsung, and rumours that Toshiba are now making old Hitachi designs... new drives may not be anything like the 3-4 year old designs these guys are using. NB the 7200.11 reported in the article are listed as being run for 3.8 years. This is waayyy older than the Hitachi drives listed at 1.9 years. Of course their failure rate is going to be increased. The only drive with a usage time equivalent are the 80 WD Green WD10EACS but they have N/A data on failure rate. Not zero, just Not available.

What this actually tells me is that the company uses mainly Seagate drives, and rags them into the ground because they consider them best bang for buck, and are willing to factor in some down time if/when some of them do break.

In the article they state [of seagate] "Their overall failure rate isn’t great, but it’s not terrible either"

and in the forum comments they state

So they buy cheap. They need to add charts of failure rate normalised against unit cost, and average lifetime.

That struck me as strange as well. With the Seagate unreliability why do they seem to still be using them in vast numbers - you would think a professional business like this would pay the premium and have mainly Hitachi or WD ... but no it seems.

[ik9000]

That struck me as strange as well. With the Seagate unreliability why do they seem to still be using them in vast numbers - you would think a professional business like this would pay the premium and have mainly Hitachi or WD ... but no it seems.

but yet the Seagate drives listed are average 3.8 years old vs average 1.9 years on the slower Hitachi drives. It just can't be a comparable sample.