Seagate 3TB Drives Are Problematic

[Nashemon]

I bought three drives with the ST3000DM001 model number from Best Buy over the past 2 or 3 years. I found out about this study a week after buying the 3rd one earlier this year, so it was still within return policy, so I returned it and got the 4TB one which fared better in the article's testing. I've not had any issues with the two other two 3TBs drives I still have, but don't expect to as I trust Backblaze's analysis in their test's faults. I test them monthly though to get a jump on them if they start to fail.

 

[Muse]

smart tests typically take a long time on drives this big, I think mine does an extended offline test in about 12hrs.
I've never used seatools to do the test however.

What do you use? I really need some ideas here. I think I should use something else. I have WD Data Lifeguard, I could use that.

I bought three drives with the ST3000DM001 model number from Best Buy over the past 2 or 3 years. I found out about this study a week after buying the 3rd one earlier this year, so it was still within return policy, so I returned it and got the 4TB one which fared better in the article's testing. I've not had any issues with the two other two 3TBs drives I still have, but don't expect to as I trust Backblaze's analysis in their test's faults. I test them monthly though to get a jump on them if they start to fail.

What do you use to test them? How do you do it?

 

[Soulkeeper]

What do you use? I really need some ideas here. I think I should use something else. I have WD Data Lifeguard, I could use that.What do you use to test them? How do you do it?

I use the smartctl command on my linux system
part of the smartmon tools package on most distros http://www.smartmontools.org/
"It should run on any modern Darwin (Mac OS X), Linux, FreeBSD, NetBSD, OpenBSD, Solaris, OS/2, Cygwin, QNX, eComStation or Windows system."
You could use a livedvd to run this as well.

 

[Muse]

I use the smartctl command on my linux system
part of the smartmon tools package on most distros http://www.smartmontools.org/
"It should run on any modern Darwin (Mac OS X), Linux, FreeBSD, NetBSD, OpenBSD, Solaris, OS/2, Cygwin, QNX, eComStation or Windows system."
You could use a livedvd to run this as well.

Thank you, I'll look into that.

 

[Muse]

smart tests typically take a long time on drives this big, I think mine does an extended offline test in about 12hrs.
I've never used seatools to do the test however.

Well, this time I hit [OK] on the Time is taking too long message, and allowed the test to continue. the message said I could abort, but this time I didn't abort (this is on the new drive I received yesterday). Right now, after running for over 16 hours the Seatools Long Generic Test is showing as 43% done. Maybe it will PASS afterall.

I moved the Backup Plus 3TB USB HD that FAILed the Short Generic Test and the Long Generic Test yesterday on my XP desktop to my Win7 32bit laptop and reran those Seatools tests. It FAILed the Short Generic Test but didn't promptly FAIL the Long Generic Test running on the Win7 system. It's been almost 2 hours now, so maybe that will PASS, don't know, but it's a bad sign that it failed the Short test, certainly. So, now I don't know which of these drives are actually are worse or even certainly bad. I think I should run more tests and try to keep track of the results in the days ahead, maybe call Seagate again and determine what to do. I wonder if I should bother using WD Data Lifeguard Diagnostics. I can run it because I have other WD USB HDs to keep the app runnable. Seems it will run on any drive, even SSD, so long as it detects one WD drive.

 

[B-Riz]

The 3TB Seagate's are rated for 8 hours of use for X number of days; these are not server drives, they are meant to be used in a mid to light duty desktops.

I ran 2 in RAID 0 for 1 year + and now have 4 in RAID 0; no issues.

Have good cooling and let Windows power down the drives when they are not being used.

 

[bradley]

The 3TB Seagate's are rated for 8 hours of use for X number of days; these are not server drives, they are meant to be used in a mid to light duty desktops.

I ran 2 in RAID 0 for 1 year + and now have 4 in RAID 0; no issues.

Have good cooling and let Windows power down the drives when they are not being used.

OK, but all you are confirming is that cost-cutting measures made Seagate an inferior choice. I personally believe the Seagate 3TB failures are attributable to substandard parts combined with aggressive power management. It's a deadly combination if you value your HDD data.

 

[B-Riz]

OK, but all you are confirming is that cost-cutting measures made Seagate an inferior choice. I personally believe the Seagate 3TB failures are attributable to substandard parts combined with aggressive power management. It's a deadly combination if you value your HDD data.

On the flip side, the spec sheets outline what you are buying.

http://www.seagate.com/www-content/...am/en-us/docs/desktop-hdd-ds1770-5-1409us.pdf

2400 power on hours and 300,000 load / unload cycles.

The BackBlaze results are worst case usage scenarios.

If you need NAS or server HDD's, buy them.

Don't buy a light duty low rated HDD and then be sad when it fails from running way out of spec.

Windows power management does not mean bad power management.

 

[stockwiz]

Why would anyone make a drive only capable of supporting 2400 power on hours? There are incandescent light bulbs that can last longer then that. I'll have to read the spec sheets closely when choosing drives I guess. Waste of landfill space just to save $10 in manufacturing cost that could just be passed onto the consumer anyways.

 

[B-Riz]

Why would anyone make a drive only capable of supporting 2400 power on hours? There are incandescent light bulbs that can last longer then that. I'll have to read the spec sheets closely when choosing drives I guess. Waste of landfill space just to save $10 in manufacturing cost that could just be passed onto the consumer anyways.

Like I pointed out earlier, it is for 8 hrs a day, I believe over 2 years, so, that is pretty much inline with a corporate desktop doing e-mail and Excel work...

Companies build for price points now, and these HDD mfgs now have distinct product lines to sell features, NAS, Server, Video, low power, etc.

Just like Intel and i3, i5, i7, something a little bit better than the previous one, for a little more $$$.

 

[AlienTech]

Hmm. I bought three Seagate Backup Plus 3TB USB hard drives one day at Costco a little over a year ago. The warranty is 2 years.

I've gotten several Delayed Write Failed errors using one of the drives. The drive has 827 hours of use on it, only 131GB of data, currently. After the failure a couple of days ago I ran Seatools tests on it (the Basic Tests, which Seatools says will not alter data, they are read only). The Generic Short Test passed, but the Generic Long Test was extremely slow and put up this message twice (I ran the test again having rebooted the computer, and had no other apps running):

------------------------------------------------------------------------------------------------------------
Long Test Time
------------------------------------------------------------------------------------------------------------

! Model ST3000DM001-1ER166 Serial: Z5000QEY

The test is taking a long time. This could be caused by large
drive capacity, excessive retries, multiple drives under test,
system overhead, or other applications running at the same time.
You may abort this test at any time.


[OK]

------------------------------------------------------------------------------------------------------------

As you can see in the above quoted Seatools' message, the drive model is ST3000DM001. This is the same as the ~5000 HDs that are reported as highly problematic in the link in the OP!

I'm wondering what to do now. I think the important data is backed up elsewhere, have to make sure about that. I am thinking about calling Seagate, think maybe I should run more tests including the Fix tests, which attempt to fix bad sectors. I got the idea of running the Fix test(s) from Seagate's How to use Seatools page.

One reason I bought the Seagate drives is because I've had 3-4 failures using Western Digital 2TB ELEMENTS USB hard drives. In short, I've found them highly unreliable. As well, the WD drives were warranted for only 1 year. I have a handful of Seagate IDE HDs that haven't failed me, had them for around 10 years.

I'm thinking that the message above could well assure that there are too many bad sectors, period. Now, I bought my ST3000DM001 HDs in May 2014, much later than the HDs with a high rate of failure discussed in the link at the OP, which are from ~2 years earlier. Hopefully, mine are better. Thoughts?

Just like the 2TB ST2000DM001, There are multiple versions of the 3TB ST3000DM001 drives. Firmwares and differences makes them problematic in some cases. Verified that both have different versions and these are not just the 666GB 3 platter 2TB drives and 1TB 2 platter 2TB drives.. You have to check the firmware to find the differences, they do not work in a raid system and drops to 25% of performance if done so. So do not just depend on model numbers or even what ever it is after the model number ST2000DM001-1CH164..

Yea thats what makes it so hard to figure out what is going on.

 

[MountainKing]

Never really trusted Seagate. Have a WD 1TB green at 39,000 hours currently and a 3TB WD Green at 19,000 hours power on. What is more funny is that the Green label is their most 'unreliable' HDDs
WD/Hitachi/Toshiba for me.

 

[bigboxes]

Never really trusted Seagate. Have a WD 1TB green at 39,000 hours currently and a 3TB WD Green at 19,000 hours power on. What is more funny is that the Green label is their most 'unreliable' HDDs
WD/Hitachi/Toshiba for me.

Never trust any of them. Backup all your data or risk losing your data. I've had many Seagate HDDs fail. I've also had failures of Toshiba, Samsung, Hitachi, Maxtor and, yes, even Western Digital. I remember when the most expensive drive in my possession was a WD drive and when it died it took a crapload of irreplaceable data with it. Sad boxes was I. I've had three WD Green drives go bad so it's no longer anecdotal at this point. Backup your data and get a third set and store them offsite if you really value your data.

I have an original WD Raptor 74GB that is still roaring along at 39k power on hours. I don't use it any longer, but keep it hooked up just to see how long it can go.

 

[tweakboy]

Seagate and maxtor are the hard drive Devil

 

[tweakboy]

Play a game, the heck with benchmarks,,, and worrying about FPS.

 

[AlienTech]

True, I used to have this feeling that WDC drives were very reliable for some reason. But I have had just as many failures with them as the Seagates. Dont have any maxtors left, the last couple of them I junked them to get the magnets. They fail just after the warranty period so I got the feeling they made them that way. I would buy 2 identical ones in case one failed, I could swap parts to get them working but they changed that so it did not work. Very old maxtor drives were rock solid, seen them work for 5-10 years. Too bad samsung and hitachi went away.. Toshiba is the closest which uses premium parts but for how long. I learnt a trick to seagate, buy their drives when they come out since they use premium parts.. If you wait, you would get inferior ones as they reduce cost of production. The last 6 years or so has been this way. So the latest drive dont mean it is the best one. If there is a serious bug they would release a firmware update anyway. But they seem to have even stopped that now.

 

[MountainKing]

Never trust any of them. Backup all your data or risk losing your data.

Yes I bit my fingers once about 3 years ago when my main 250GB HDD crashed and I had no backup of 6 years worth of files which I can no longer obtain. What is more annoying is that I kept getting warnings about hard disk click sounds and 1-2 rare occasional freeze ups.

 

[Muse]

True, I used to have this feeling that WDC drives were very reliable for some reason. But I have had just as many failures with them as the Seagates. Dont have any maxtors left, the last couple of them I junked them to get the magnets. They fail just after the warranty period so I got the feeling they made them that way. I would buy 2 identical ones in case one failed, I could swap parts to get them working but they changed that so it did not work. Very old maxtor drives were rock solid, seen them work for 5-10 years. Too bad samsung and hitachi went away.. Toshiba is the closest which uses premium parts but for how long. I learnt a trick to seagate, buy their drives when they come out since they use premium parts.. If you wait, you would get inferior ones as they reduce cost of production. The last 6 years or so has been this way. So the latest drive dont mean it is the best one. If there is a serious bug they would release a firmware update anyway. But they seem to have even stopped that now.

My old Seagate 120, 180 and 200GB IDE internal HDs are still working. The 200GB throws a CAUTION in crystaldiskinfo but it keeps chugging along. I'm not sure what's going on with my Seagate Backup Plus 3TB USB HDs. One threw a FAIL in the short and long generic tests in Seatools a couple of times on my XP machine, FAILed the short test on my Win7 32bit laptop, but then passed it and passed the LONG test (talk about long, it was ~40 hours), which caused a "test is taking too long" message to pop up in Seatools, WTF! So, I'm unsure if any of my 3 Backup Plus drives is bad. I already have a replacement, so I'm sending in the drive that I complained about, otherwise they'll charge me for the replacement.

I've dismantled quite a few old HDs (failed or so small as to be not useful) in order to get the magnets. I have several uses for them.

I didn't know HDs had replaceable parts!

Samsung and Hitachi went away? They don't make HDs anymore?

Usually you'd think that production down the line eliminates original errors in production that became evident after the HDs were used over a protracted period by customers. So, you're saying they actually cheapen them as time goes on to save pennys here and there! Argh.

In the future (before buying) I think I'm going to just go out there and try to get anecdotal info, either customer reviews or site reviews such as the one linked in the OP.

 

[Muse]

I haven't yet sent back the HD after having received its replacement (Seagate Backup Plus 3TB USB). I'm preparing to do so. I bought 3 of them one day in May 2014 at Costco warehouse).

So, the story is I was getting "Windows - Delayed Write Failed" errors (a few, screwing up HDTV recordings), and the Long Generic test in Seatools was popping up a message saying the test was "taking too long." I called Seatools tech support and explained these things and the guy determined I should get an RMA, then transferred me to one of those folks and the whole thing started. However, it now appears that the "taking too long" message happens with the other drives too, after around 10-12 hours, and the test takes around 40 hours regardless. So, the drives might not be bad. Well, maybe the first of them is, based on the "Windows - Delayed Write Failed" errors.

Well, I have to send back one of the drives, so it will be the one with which I had those errors. One of the other drives FAILed the short and long test, but later PASSed them, so who knows what's up with that???

Well, I get a message a couple of days ago from Seagate:

Thank you for contacting Seagate Support.

I apologize for the problem you have encountered with your hard drive(s). *I understand that, due to the sensitive nature of your data, you would like to remove the platters before returning the drive(s) to us.*

You have been given authorization for a Certificate of Destruction (COD). *Please fill out the attached COD form and then reply directly to this email with the completed form. *You will also need to send a hard copy of the completed COD form with the top face plate (silver metal piece with a Seagate sticker) and the circuit board (the green board) of the drive(s). If this is an external drive, you will also need to send the external enclosure. *This COD gives you permission to remove and destroy the platters.

For additional assistance, feel free to contact us at: www.seagate.com/about/contact-us/technical-support/

Regards,

Freda
Seagate Support

Now, my feeling about this was:

I don't remember asking them about data destruction or security concerns, although I have some. Maybe I did ask them, however I am prepared to just DBAN the drive and send it in. After all, I have a suspicion that the drive is actually OK and I don't feel right destroying the drive, they can presumably sell it as a refurb.

Then I get another email from them 2.5 hours later (...I had/have not contacted them back):

Thank you for contacting Seagate Support.

I apologize for the previous email that I sent to you.* I need you to ignore that information about the COD process.

Please reply to this email if you require further assistance or give us a call at 1-800-SEAGATE. We are available from 8:00 am to 6:00 pm*
Central time Monday - Friday.

Regards,

Freda
Seagate Support

WTF! OK, maybe they don't want me to dismantle the drive and confiscate the platters, don't know. Anyway, I figure to just DBAN it and send it back today.

 

[AlienTech]

In the future (before buying) I think I'm going to just go out there and try to get anecdotal info, either customer reviews or site reviews such as the one linked in the OP.

I just told you reviews and such are useless and here you are saying thats what I will use. There are a few Forums where people post real live results and experiences. But that is no guarantee either.. We look at posts but not on the dates and go by them, without thinking the messages from 6 months ago are not valid any longer.

And seagate used enterprise design originally to get their damaged rep a boost and people loved their drives. But that was too much.. So now they use a similar setup where new drives are far more reliable and lasts longer than the ones later on. It also gives them great reviews everywhere. But they are NOT APPLICABLE when you buy them. Its funny in a way.. someone who buys a drive 3 years ago and you buy a drive now, they will all fail 5 years from the date of original release.. Their drive lasts 5 years, yours only 2 years. Once you know this, keep it in the back of your mind and it all makes sense. I got burnt for the last 5 years on this and so fed up with it now. This has now made me shake and cause anxiety for purchases. I clearly am in distress with all the crap they pull. When you can not rely on a certain model numbers and when later models are junk, when they hide all these specs and when you get the stuff, even the specs you thought were valid are not in the drives you got. Clearly where there are such drastic usebality problems they should not use the same model numbers. Not only that but we used the serial numbers to figure out drive specs.. like how many heads, platters, speeds etc.. WELL guess what! they changed that too.. ie because people figured out their drives, they now change firmware where there is a huge performance hit in raid type setups by mixing new and older drives of the same models.

There must be a reason for all this, Mostly to fool non technical people and sell them vastly cheaper products.. I cant see why else they have such business practices. When they hide things, the only reason is to fool customers. Since they dont publish specs, you cant ask them about it. Even saw messages from reps saying those are trade secrets when people asking about them, its buyer beware now. Even making it far more difficult to use their forums now since that was where people posted this kind of stuff before.

 

[bononos]

........
Samsung and Hitachi went away? They don't make HDs anymore?

Usually you'd think that production down the line eliminates original errors in production that became evident after the HDs were used over a protracted period by customers. So, you're saying they actually cheapen them as time goes on to save pennys here and there! Argh.
......

Samsung's hdd division got sold to Seagate, Hitachi to WD.
I don't know if the previous poster was joking around about replaceable parts since its probably difficult to swap parts around without a clean room.

 

[bononos]

And here's a rebuttal to this very article....

Trust Backblaze's drive reliability data?
TweakTown's tacit defense of industry silence on drive reliability - and against the Backblaze report - fails on multiple levels. Here's why.
http://www.zdnet.com/article/trust-backblazes-drive-reliability-data/

Has backblaze got any raw data on which brand if drives got used in which version of enclosures? Or are all the drives mixed up and get placed everywhere?

 

[bononos]

I thought I came across a thread talking about the 3Tb Seagate issue, or maybe it was in general HW but I couldn't find it.

Anway According to this site link (you might have to feed it through google translate), the 3Tb which they opened up had a poorly fitted connector which let outside unfiltered air in which might have killed the drive. Maybe it is a general issue with the 3Tb.

 

[Topweasel]

Samsung's hdd division got sold to Seagate, Hitachi to WD.
I don't know if the previous poster was joking around about replaceable parts since its probably difficult to swap parts around without a clean room.

I don't know about Samsung. But Hitachi while under WD is still running as a separate Division making their own products.

 

[bigboxes]

Samsung's hdd division got sold to Seagate, Hitachi to WD.
I don't know if the previous poster was joking around about replaceable parts since its probably difficult to swap parts around without a clean room.

You can swap out circuit boards on matching drives, if that's the problem. No one can open up a drive and have it survive without a sterile environment.