Jump to content


Photo

WD Green 3TB Observations

AskSR

  • You cannot start a new topic
  • Please log in to reply
17 replies to this topic

#1 Brian

Brian

    SR Admin

  • Admin
  • 5,298 posts

Posted 21 October 2013 - 08:47 AM

Via email from Marcin:

 

I just wanted to share my observations regarding the above HDD, and your testing process. I have 8 of these HDDs, but over the 2y course I had 16 of them - they fail so often. And not only that, they also silently corrupt data - returning rubbish and forgetting to mention that...
I've seen a few reviews of these drives, but none of them, including yours, mention this issue. I had been able to stumble upon it after a fortnight of use, and I have not been really stressing them. This concerns me - how thorough these reviews are? All of them consider performance, none - reliability. It is very nice to know how fast the drive can chew your data, but at the point of chewing the speed is rather irrelevant.

I am curious to hear your opinion about this - are the review just a gimmick, a flashy advert, a bit of slideware, or are they meaningful?

Looking forward to hearing from you,
M.W.

 

 

When we review hard drives we generally only have one sample, and even if we have more, long term endurance testing won't be relevant for us because we can't stress the drives long enough to fail. We do on occasion have a drive fail during testing, but there are so many reasons for failure it's hard to say if one out of 10 throws an error that the drives fail 10% of the time. It's something we'd like to do more of but there's not a great methodology to work from. 


Brian

Publisher- StorageReview.com
Twitter - @StorageReview

 

#2 [ETA]MrSpadge

[ETA]MrSpadge

    Member

  • Member
  • 742 posts

Posted 21 October 2013 - 01:34 PM

Thanks for posting, Brian.

 

I can understand your point here, and even support it: to see meaningful failure rates with 1 - 2 years you'd at least have to stress-test 1000 units. For every model.. which should make it pretty clear that this is just impossible.

 

But Marcin's point is also very important: if an HDD has a significantly higher failure rate than others, all other characteristics pale in comparison. Especially now with the slow evolution of HDDs, which keeps the older models useful for far longer - if they haven't failed yet. In the past SR had the reliability database to get around the problem of "too few samples". Personally I haven't used it in years.. not ure in which state it is.

 

MrS


#3 Brian

Brian

    SR Admin

  • Admin
  • 5,298 posts

Posted 21 October 2013 - 02:19 PM

It's state is unwell, lol. 

 

We've been wanting to redevelop that - probably need to start from scratch and maybe poll the community here as to what it should include to get a working document. 


Brian

Publisher- StorageReview.com
Twitter - @StorageReview

 

#4 dietrc70

dietrc70

    Member

  • Member
  • 106 posts

Posted 22 October 2013 - 10:12 PM

Simply reading user posts and Newegg reviews has given me a "bad vibe" about the reliability of the 3TB WD Greens.

 

It would be great to have real hard data, but detailed small-sample reports like this one are a lot better than nothing.


Edited by dietrc70, 22 October 2013 - 10:14 PM.

#5 FastMHz

FastMHz

    Member

  • Member
  • 403 posts

Posted 23 October 2013 - 10:52 AM

I have personal experience seeing a large proportion of WD Green failures vs. any other drive.  In fact, there are a few dead greens on my shelf right now.


Production: Vishera 8350/32gb RAM/Dual SSD/VelociRaptor/Radeon 7750
Gaming: Phenom II 955/16gb RAM/SSD/VelociRaptor/Radeon 7950
Retro: K6-2 550/256mb RAM/160gb HDD/CompactFlash/3DFX/ATI AIW Pro/SB16/DB50XG
http://www.fastmhz.com

#6 continuum

continuum

    Mod

  • Mod
  • 3,568 posts

Posted 24 October 2013 - 09:49 PM

Problem with user posts and Newegg/Amazon/whoever user reviews is that usually only the people who are having issues will start posting, those who have no issues don't say a peep, so your data is by its very nature quite skewed.

 

I wouldn't trust Newegg reviews any further than I could throw them.

 

 

That said WD Green 3TB's, especially the older SATA 3Gbps units, according to some forums, seem to have a significantly higher failure rate than the newer WD Green 3TB SATA 6Gbps units. I have zero data to back this up and this is not my opinion, but that's the noise I am starting to see.

 

And it's utterly impractial for any review site to do real reliability testing on harddisks. Maintaining a large enough random sample size and stressing them hard enough would completely tie up the resources of every major review site on the internet to just test a small fraction of modern harddisks currently on the market...


#7 FastMHz

FastMHz

    Member

  • Member
  • 403 posts

Posted 25 October 2013 - 09:22 AM

I agree with that, review sites aren't set up for such a massive long term operation.  Now if Google would release the brands and fail rates from their datacenters full of consumer drives, that would be a real win!!  But I doubt they ever would, as I'm sure HDD manufacturer lawyers would have a field day :-(


Production: Vishera 8350/32gb RAM/Dual SSD/VelociRaptor/Radeon 7750
Gaming: Phenom II 955/16gb RAM/SSD/VelociRaptor/Radeon 7950
Retro: K6-2 550/256mb RAM/160gb HDD/CompactFlash/3DFX/ATI AIW Pro/SB16/DB50XG
http://www.fastmhz.com

#8 Kevin OBrien

Kevin OBrien

    StorageReview Editor

  • Admin
  • 1,441 posts

Posted 25 October 2013 - 09:46 AM

I have personal experience seeing a large proportion of WD Green failures vs. any other drive.  In fact, there are a few dead greens on my shelf right now.

 

 

One additional observation, we normally have a fairly average distrobution of drive failures (we consider any drive with even a single bad sector to fall into the defective category) across standard 5400 and 7200RPM HDDs. Now for WD Greens and Seagate LPs, we had a higher # of the drives fail, although we also had more samples in general and put them into NAS use which got many into the 10-15k hour range. Same can't be said about the WD Red or Seagate NAS category... still rocking all original review samples.


#9 continuum

continuum

    Mod

  • Mod
  • 3,568 posts

Posted 25 October 2013 - 03:52 PM

Google did do that one datacenter report a few years ago, but you have to keep in mind that even then it is still very specific to conditions within Google's user case, hardware environment, and datacenter specifics. Translating those numbers over to typical consumer use is very easily misleading in many scenarios-- best case it definitely is useful, but yeah... not a pancea.


#10 [ETA]MrSpadge

[ETA]MrSpadge

    Member

  • Member
  • 742 posts

Posted 26 October 2013 - 04:37 AM

A few or even a few 10 bad sectors aren't bad for an older drive (in a non-mission critical environment). I'd only consider replacing a drive if their number starts to increase. Having said that.. the only 2 two drives I'd experienced doing this are a WD Green 640 GB with just a few 100 power-on-hours (but >3 years since buying it) at work and my sister's external WD elements 1.5 TB - also a Green inside, drive was infrequently used with mostly static content and just out of warrenty. Both went over 1000 defective sectors when trying to format them and became really slow - so not even useful for invaluable data any more.

 

Of course other drives have failed too, but these were usually really old, with the appropriate usage. And non showed this failure mode. Well.. it's just a limited personal experience, but fits "WD Green Observations".

 

MrS


#11 baruch

baruch

    Member

  • Member
  • 3 posts

Posted 03 November 2013 - 03:23 PM

I've been interested in this too and started to write a tool to monitor the disks, it can both help the user to monitor his disks health and report back to a central database from which it may be possible to extract insights on large disk populations and their behavior over time. The client side of that project is working and I've yet to muster enough interest to bring forth enough motivation to complete this.

 

The client source lives in https://github.com/baruch/disksurvey


#12 antymat

antymat

    Member

  • Member
  • 5 posts

Posted 05 November 2013 - 01:17 AM

Problem with user posts and Newegg/Amazon/whoever user reviews is that usually only the people who are having issues will start posting, those who have no issues don't say a peep, so your data is by its very nature quite skewed.

 

Oh, I don't know. There are >300 positive reviews of these HDDs on Amazon. True, they look like written by some PR agency - still they are there. And more and more of them.

 

I wouldn't trust Newegg reviews any further than I could throw them.

 

This applies to positive reviews anywhere. Unless it is a proper test, it is just bragging about how lucky the owner is...


#13 antymat

antymat

    Member

  • Member
  • 5 posts

Posted 05 November 2013 - 01:41 AM

I will reply to myself (after I've located the password ;) ).

 

The funny thing about these 3TB Green disks is - they do not develop bad sectors - maybe one or two did. No - they just return rubbish data.  With no read errors whatsoever. The bad sector reported as bad is OK - that's what the RAID is for. But bad sector reported as correct will bypass RAID, and wreak havoc. And this is what happens with WD 3TB Green disks. So - however I like the baruch's idea of gathering data from many users, this will not work. Anything relying on correct disk error reporting will not work in this case. 
 
If I may suggest a test for 3TB Reds - if you have them in a RAID on a NAS, try writing 3-5TB of 4-8GB files, compute SHA1 or even MD5, and just start checking continuously (for a week or maybe two...). This is how I've found the issue with 3TB Greens. Maybe keep the atime on, this will add some small writes. 
 
The latest change to my WD 3TB green saga is - after 2 disks failed a few weeks ago I got the replacements, one of them was dead. 3 replacements in a month with 8 drives.
This has never happened with any other disks. 

#14 continuum

continuum

    Mod

  • Mod
  • 3,568 posts

Posted 06 November 2013 - 02:11 PM

As many drives as have been sold, I would expect 30,000+ positive reviews, not a piddling 300. ;)


#15 jtsn

jtsn

    Member

  • Member
  • 94 posts

Posted 07 November 2013 - 02:12 AM

The funny thing about these 3TB Green disks is - they do not develop bad sectors - maybe one or two did. No - they just return rubbish data.  With no read errors whatsoever.

 

The bit error rate on consumer drives is only 10e-14. So rubbish data can just happen, if your drives become big enough and you put enough data through them. 3 TB is already 24e12 bits, so if your read or write the complete drive once, there is a 1 in 2.4 (42 %) chance, that you get an undetected error just flying by...


#16 antymat

antymat

    Member

  • Member
  • 5 posts

Posted 07 November 2013 - 02:30 AM

 

The bit error rate on consumer drives is only 10e-14. So rubbish data can just happen, if your drives become big enough and you put enough data through them. 3 TB is already 24e12 bits, so if your read or write the complete drive once, there is a 1 in 2.4 (42 %) chance, that you get an undetected error just flying by...

 

Er... 10e-14 uncorrectable bit error chance. Not undetectable. This is crucial difference. 


#17 antymat

antymat

    Member

  • Member
  • 5 posts

Posted 08 November 2013 - 03:59 PM

Oh, I observed yet another drive failing. It is 4'th in a month; 2 worked for <200h. Half of my drives died in a month.  :huh: 

 

The real problem with these drives is that I do not need 8 paper weights, and they are too light for a door stop.


#18 cbrworm

cbrworm

    Member

  • Member
  • 131 posts

Posted 15 November 2013 - 11:26 AM

I had similar luck with the original WD15EARS that I put out in the field - I only used a few, but they all came back within a couple years.  They didn't corrupt data, they just died.  I have been having higher failure rates with mainstream drives from all manufacturers over the last few years than I have seen since the mid-late 90's.  In fact, I still have old 5 platter 250gb Hitachi's all over the place that just keep going (and meowing).  Many of my customers have older, pre FDB IDE drives that just run 24/7 and never fail.  I build them a new machine with good cooling and a year later the new drive craps out.

 

BTW, I have a WD15EARS that works perfectly and only has about 40 hours of use on it that I want to sell if anyone is interested...






Also tagged with one or more of these keywords: AskSR

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users