Backblaze: Hard Drive Stats for Q2 2017

Megalith

Megalith

24-bit/48kHz
Staff member
Joined
Aug 20, 2006
Messages
13,000
Backblaze has published their statistics from the period of April 1, 2017 through June 30, 2017. The company has now added a trend indicator to their failure rate table: a few hard drive models managed to show a decrease in failures, but a 6TB Seagate (ST6000X000) did just the opposite. The company has also compared enterprise drives with consumer drives and found that both had nearly the same annualized failure rate based on runtime/hours.

When looking at the quarterly numbers, remember to look for those drives with at least 50,000 drive hours for the quarter. That works out to about 550 drives running the entire quarter. That’s a good sample size. If the sample size is below that, the failure rates can be skewed based on a small change in the number of drive failures. As noted previously, we use the quarterly numbers to look for trends. So this time we’ve included a trend indicator in the table. The “Q2Q Trend” column is short for quarter-to-quarter trend, i.e. last quarter to this quarter. We can add, change, or delete trend columns depending on community interest.
 
D-Money said:
Seagate drive managing to show an *increase* in failure rate? I thoroughly believe that.
Click to expand...

Sadly, They're not what they used to be. I had a couple newer seagates failed on me with less workload than older ones that were used for video editing, transferring large video files and gaming.

Or atleast I'm just unlucky with seagate picks.
 
The company has also compared enterprise drives with consumer drives and found that both had nearly the same annualized failure rate based on runtime/hours.
Click to expand...

As expected. The difference between the "good, better, best" (or consumer, prosumer, enterprise) models is often not one of innovation, but rather one of warranty.

They offset the increased likelihood of drive failures above each tier with the higher prices.
 
Oh Seagate.

facepalm.jpg
 
Seagate 3.5" HDDs are questionable, especially on those with crappy SMR.
Their 2.5" HDDs are solid though. Can't speak as to their latest 12.5mm 2.5" line (up to 5TB) that uses SMR and not PMR though.
 
I have about 12-15 dead 3.5" hard drives in my basement (when I get bored I will eventually remove the magnets so I can play with them...). One of them is just 320 GB (I think). 9 of these are Seagates. I have a few Maxtor (ohh - remember them) and only 1 Western Digital. Granted, I was using some of the "green" drives in a NAS which is a no-no. I've switch to using WD Red's. Knock on wood - haven't had any of these fail yet.
 
This makes me sad. I worked at Seagate and can say that there are a lot of good people who work their ass off trying to reduce failure rates.

What you're seeing is probably a result of the consolidation of the 2 Chinese factories. I can't be sure, but drives that had been previously manufactured at the now closed site are probably having some issues at the new site.

If that's not applicable due to the drives being manufactured in Thailand, I would attribute it to seagate's head sensitivity to particulate contamination relative to WD and Hitachi. Seagate heads are a lot more sensitive to particulate contamination than their competitors. We had a lot better contam reduction than everyone, but that was because it had to be done or yields would be even worse.

Honestly, when I first started working there, it seemed like we were positioned well. But over time, most of the new hires started looking elsewhere due to upper management not being able to incorporate the new generation of engineers very well. I know of at least 10 other engineers who were all hired within a year or two of me who have left for greener pastures. And that's just at one R&D site.

Helium blindsided them and their acquisition of Xyratex immediately blew up in their face due to a large drop in demand. When you buy a company for their test rigs and end up not needing them, that's rough.
 
Bigballingpr24 said:
This makes me sad. I worked at Seagate and can say that there are a lot of good people who work their ass off trying to reduce failure rates.

What you're seeing is probably a result of the consolidation of the 2 Chinese factories. I can't be sure, but drives that had been previously manufactured at the now closed site are probably having some issues at the new site.

If that's not applicable due to the drives being manufactured in Thailand, I would attribute it to seagate's head sensitivity to particulate contamination relative to WD and Hitachi. Seagate heads are a lot more sensitive to particulate contamination than their competitors. We had a lot better contam reduction than everyone, but that was because it had to be done or yields would be even worse.

Honestly, when I first started working there, it seemed like we were positioned well. But over time, most of the new hires started looking elsewhere due to upper management not being able to incorporate the new generation of engineers very well. I know of at least 10 other engineers who were all hired within a year or two of me who have left for greener pastures. And that's just at one R&D site.

Helium blindsided them and their acquisition of Xyratex immediately blew up in their face due to a large drop in demand. When you buy a company for their test rigs and end up not needing them, that's rough.
Click to expand...

had 3 seagate drives in the last 10 years 2 failed with in 18 months the that still "works" has single digit transfer speeds after it got to ~80% capacity and never recovered looks like its a fimeware issue but i can't fix it because dispite it being the same mode its was an oem drive with a slightly different firmware version so the updater won't work ....
 
WhoBeDaPlaya said:
BackBlaze blocked at work :|
Thank God for SSH tunnelling >.>
Click to expand...
That's why you gotta be the playa controlling the block :cool:

Gavian said:
Blackblaze is the reason why I tend to stick with HGST.
Click to expand...
BlackBlaze was confirmation bias for me regarding Seagate and HGST. One always failed and one, no matter how old, always worked. But I've been personally buying Toshiba's which still are usually at the top of reliability.
 
Last edited:
Spidey329 said:
As expected. The difference between the "good, better, best" (or consumer, prosumer, enterprise) models is often not one of innovation, but rather one of warranty.

They offset the increased likelihood of drive failures above each tier with the higher prices.
Click to expand...
There's some truth to this, but it's worth noting that BackBlaze's usage pattern is not representative of many enterprise workloads. They mostly fill up a pod with data and then it sits there until there's an occasional recall of the backed up data. This is a lot like a small home NAS or other backup solutions. Since the disks are mostly sitting idle, it sort of makes sense that they'd all have similar failure rates. Than again as SSDs take more and more of the "real" workloads, HDD are increasingly become only used for longer term storage/backup.
 
Pardon my ignorance, but what formula are they using to calculate failure rate? I feel like the answer is staring me in the face but I'm at a loss.
 
steakman1971 said:
I have about 12-15 dead 3.5" hard drives in my basement (when I get bored I will eventually remove the magnets so I can play with them...). One of them is just 320 GB (I think). 9 of these are Seagates. I have a few Maxtor (ohh - remember them) and only 1 Western Digital. Granted, I was using some of the "green" drives in a NAS which is a no-no. I've switch to using WD Red's. Knock on wood - haven't had any of these fail yet.
Click to expand...
I've had no problems with WD Greens, with the caveat that ANY WD Green I deploy are wdidle /300 patched ;)
 
Yeah, Seagate isn't even an option for me. Apparently their IronWolf NAS drives are not bad, but I still wouldn't put them in my system.
 
Wasn't the issue before that their numbers were off because they bought considerably MORE Seagate drives than any other brand and thus failure rates appeared higher, but if you brought all the numbers to an equal level the rate was not as high, but still the worst brand of them all.
 
Elios said:
had 3 seagate drives in the last 10 years 2 failed with in 18 months the that still "works" has single digit transfer speeds after it got to ~80% capacity and never recovered looks like its a fimeware issue but i can't fix it because dispite it being the same mode its was an oem drive with a slightly different firmware version so the updater won't work ....
Click to expand...

Although I will probably go with HGST or WD today (toshiba), I had a Seagate drive last +12 years. It was a 40gig that during the last 4 years or so was basically on 24 x 7 because Win XP had gotten so slow to boot up, I just kept it on. Power savings profile was about 2 hrs, so not quite 24 x 7 but still powered on, but not always spinning. Finally failed one day when I was about to go online to play some Q3. Turned it off a couple of days before, it was the boot that finally killed it.

What happened to them?
 
I've never had particularly bad luck with seagate - one of their 8TB archive drives I had worked great, and still chugging along nicely after I gave it to my dad. *shrugs*

I slowly started moving back to WD Red series as thats what I've been stocking my Drobo with. If it's a numbers game and something fails I'm still covered since it's a 5 bay system.
 
I actually have pretty good luck with Seagate recently. I have 2x10TB Ironwolf drives and they work fine.
 
So I left Seagate for storage many years ago, however I've been using them for external storage and I'm doing so right now. I need to increase this storage and they seem to be the only game in town. I really hope I don't get burned.

Right now I have 1.5TB in a USB3 - connector powered unit. I will need to increase this and I can only find one other interface powered drive and all the 2TB or larger 2.5" drives I can find are Seagate.
 
kirbyrj said:
I actually have pretty good luck with Seagate recently. I have 2x10TB Ironwolf drives and they work fine.
Click to expand...

Got an Ironwolf 8Tb last month and its got about 11TB of writes on it so far and runs like a champ, pretty quick for a spinner too. Got it for a torrent download drive to replace the Samsung 1Tb that shit the bed, time will tell how good it is though.
 
WhoBeDaPlaya said:
I've had no problems with WD Greens, with the caveat that ANY WD Green I deploy are wdidle /300 patched ;)
Click to expand...

I just used CrystalDisk to set the power management to half way rather than far left. Seems to work just fine.
 
Bigballingpr24 said:
This makes me sad. I worked at Seagate and can say that there are a lot of good people who work their ass off trying to reduce failure rates.

What you're seeing is probably a result of the consolidation of the 2 Chinese factories. I can't be sure, but drives that had been previously manufactured at the now closed site are probably having some issues at the new site.

If that's not applicable due to the drives being manufactured in Thailand, I would attribute it to seagate's head sensitivity to particulate contamination relative to WD and Hitachi. Seagate heads are a lot more sensitive to particulate contamination than their competitors. We had a lot better contam reduction than everyone, but that was because it had to be done or yields would be even worse.

Honestly, when I first started working there, it seemed like we were positioned well. But over time, most of the new hires started looking elsewhere due to upper management not being able to incorporate the new generation of engineers very well. I know of at least 10 other engineers who were all hired within a year or two of me who have left for greener pastures. And that's just at one R&D site.

Helium blindsided them and their acquisition of Xyratex immediately blew up in their face due to a large drop in demand. When you buy a company for their test rigs and end up not needing them, that's rough.
Click to expand...
I've got a couple (or is it 3?) 4TB Seagate drives and I've had 0 issues with them, though I recently pulled the ones from the file server and replaced with 8TB WD drives, but if you look at the cumulative on 8TB drives, so far Seagate seems to be the most reliable. The older failure rates seem bad (though the worst of them is a model I used for 3 years without issue (it's one that is currently looking for a new home ;))
 
Gavian said:
Blackblaze is the reason why I tend to stick with HGST.
Click to expand...
Funny, because they're the reason I bought Seagates 3-4 years ago. In fact, that is the lesson I get from Back Blaze. Cheaper drives make more sense over the long haul.
 
These drives are in pods so very likey due to the way the seagate 6tb specific model are made it's more susceptible to Vibrations

but more likely it's faulty manufacturing defect like they had same problem with specific seagate 2tb model (and was also consistently unreliable in consumer use as well as I've had tone 2tb die as well witch was on blackblase list years ago)
 
You must log in or register to reply here.
Back
Top