Which 4 TB drive should I buy?


Recommended Posts

Does anyone have any opinions to share about the new Seagate 4TB NAS HDDs?  Anyone using them in their server?  I really like the WD reds and WD in general as I've had bad luck with many Seagate drives in the past, but seeing as WD Reds top out at 3TB, these new Seagates look pretty tempting...

 

This review suggests that the Seagate 4TB NAS HDD bests the 3TB WD Red in almost every category:

 

http://www.storagereview.com/seagate_nas_hdd_review

 

 

Anyone know when WD will release a 4TB Red drive to compete?

Link to comment
  • Replies 131
  • Created
  • Last Reply

Top Posters In This Topic

Does anyone have any opinions to share about the new Seagate 4TB NAS HDDs?  Anyone using them in their server?  I really like the WD reds and WD in general as I've had bad luck with many Seagate drives in the past, but seeing as WD Reds top out at 3TB, these new Seagates look pretty tempting...

 

This review suggests that the Seagate 4TB NAS HDD bests the 3TB WD Red in almost every category:

 

http://www.storagereview.com/seagate_nas_hdd_review

 

 

Anyone know when WD will release a 4TB Red drive to compete?

 

I just put two in one of my Microservers and also one of the SG 4TB DMs. All passed the preclear and are humming along nicely. They do run cool (as reported, not physically measured). I also have two Hitachi 7200rpm drives in the same box and they run much hotter (reported).

 

I would have filled the box with 5900/5400rpm 4TB Hitachis, but I can't seem to buy them in the UK. So, to avoid power and heat issues, I decided to use the SGs:

 

iqIGKkj.png

Link to comment
  • 2 weeks later...
  • 2 weeks later...

I've taken a different tack and opted for the WD4000F9YZ. These are the WD's low-end enterprise series (SE).

Up to now I'm very pleased with what I got.

 

You pay a little premium, but you get: enterprise-scale reliability and MTBF, 5 years warranty, super cool operation, and (surprise) extreme silence. They have a stabilizing mechanism that makes them really quiet with almost no noticeable wobble. Me happy.

Link to comment

I've taken a different tack and opted for the WD4000F9YZ. These are the WD's low-end enterprise series (SE).

Up to now I'm very pleased with what I got.

 

You pay a little premium, but you get: enterprise-scale reliability and MTBF, 5 years warranty, super cool operation, and (surprise) extreme silence. They have a stabilizing mechanism that makes them really quiet with almost no noticeable wobble. Me happy.

 

The WD4000F9YZ is indeed a very nice drive.    I almost used these in a system I recently built, but opted for the 4GB Seagate NAS units instead, as I wanted a 1TB/platter drive.    Although the seek times for this drive are better (due to its 7200rpm speed); the sustained transfer rate on the Seagate NAS units is actually better because of the higher areal density.    The Seagate also runs a good bit cooler, thanks to its much lower power requirements [4.8w vs. 9.5w].    Note that the new 4TB WD Red is also a 1TB/platter drive, and only draws 4.5w, although it also spins a bit slower than the Seagate unit, so won't have quite as good a sustained transfer rate.

 

Link to comment
  • 1 month later...

I've taken a different tack and opted for the WD4000F9YZ. These are the WD's low-end enterprise series (SE).

...

You pay a little premium,...

 

Nice drive but 54% more expensive than a green this side of the pond and that with them having a 10% sale.

Link to comment

 

Nice drive but 54% more expensive than a green this side of the pond and that with them having a 10% sale.

 

Indeed. They do come at a premium.

 

My personal experience with the WD green drives was nothing short of horrific. One NAS build I did had just 4 drives (2TB - EARS). I had to go through 6 drives replacements (six!!!) to get four drives that would just pass the SMART extended test. And then they started dying.

I would not recommend them.

Link to comment

 

Nice drive but 54% more expensive than a green this side of the pond and that with them having a 10% sale.

 

Indeed. They do come at a premium.

 

My personal experience with the WD green drives was nothing short of horrific. One NAS build I did had just 4 drives (2TB - EARS). I had to go through 6 drives replacements (six!!!) to get four drives that would just pass the SMART extended test. And then they started dying.

I would not recommend them.

I agree.  I have 28 greens that are only suitable for backups.  12 I killed when a fan stopped running on a parity check and the drives got to 60+c - they still work but have a wide assortment of smart errors/ata errors and slower access (by half).  But 12 others just died with smart errors and a general slow down in access.  Only have 4 greens left that are worth anything.
Link to comment
  • 2 weeks later...
  • 2 weeks later...
  • 2 months later...

I use my WD-Greens (3x 2TB) since April 2010 and they are running without any issues.

The Quality Range seems wide - some of my friends have 50% dropout - strange.  :-\

 

 

I've had five 2TB green drives since 2010 and all but one of those are still going. One got fried when a power supply died but the other 4 and the replacement for the fried one are all still going.

Link to comment

Any specific comments on the 4TB WD Red vs 4TB Seagate NAS drive?  Both are similarly priced, and I am looking at slowly moving my arrays from desktop drives to NAS drives just to hedge my bets for long(er) term reliability.

 

Both are excellent choices.  I've built system using both of these, and they've all been rock-solid.  Based on my Kill-a-Watt measurements, the WD Reds are marginally more power-efficient, but not by enough that it matters.    The Seagates are a bit faster -- 5900rpm vs. 5400rpm, but with their 1TB platters both drives have sustained transfer speeds well above what a Gb network can achieve, so there's no effective difference when used as UnRAID data drives (although an all-Seagate array will have a slightly better parity check time than an all-WD array).

 

I tend to favor the WD Reds, but for no specific reason other than personal preference.

 

Link to comment

I've noticed that Seagate drives tend to have a performance drop in the first 20-25 GB range, at least in the desktop class regardless of model.

 

Measured how?  There's no mechanical reason that should be the case ... in fact the very outermost cylinders are the fastest area of the disk.  Some benchmarks (e.g. HD Tune) take a second or two to stabilize their statistics display, which can APPEAR to show a slow initial value; but this is a display anomaly and not an actual performance issue.    I've certainly never seen what you described ... on Seagate or any other drives.  Just curious how you've measured this?

 

 

Link to comment

I've noticed that Seagate drives tend to have a performance drop in the first 20-25 GB range, at least in the desktop class regardless of model.

 

Measured how?  There's no mechanical reason that should be the case ... in fact the very outermost cylinders are the fastest area of the disk.  Some benchmarks (e.g. HD Tune) take a second or two to stabilize their statistics display, which can APPEAR to show a slow initial value; but this is a display anomaly and not an actual performance issue.    I've certainly never seen what you described ... on Seagate or any other drives.  Just curious how you've measured this?

 

I first noticed it using my diskspeed script which performs a read test across the disk - the first check at the start was always lower than subsequent checks. Then I wrote a script to track the parity speeds every 15 seconds, there is a steady drop until after the 20 GB part over multiple runs utilizing both 32bit and 64bit UNRAID OS.

 

http://strangejourney.net/UNRAID/32vs64.html (zoomable, click-n-drag to zoom)

 

root@NAS:/boot/scripts# hdparm -t --offset 0 /dev/sdj

Timing buffered disk reads (offset 0 GB): 484 MB in  3.07 seconds = 157.66 MB/sec

 

root@NAS:/boot/scripts# hdparm -t --offset 5 /dev/sdj

Timing buffered disk reads (offset 5 GB): 490 MB in  3.00 seconds = 163.08 MB/sec

 

root@NAS:/boot/scripts# hdparm -t --offset 10 /dev/sdj

Timing buffered disk reads (offset 10 GB): 488 MB in  3.01 seconds = 162.05 MB/sec

 

root@NAS:/boot/scripts# hdparm -t --offset 15 /dev/sdj

Timing buffered disk reads (offset 15 GB): 492 MB in  3.00 seconds = 163.73 MB/sec

 

root@NAS:/boot/scripts# hdparm -t --offset 20 /dev/sdj

Timing buffered disk reads (offset 20 GB): 492 MB in  3.01 seconds = 163.52 MB/sec

 

root@NAS:/boot/scripts# hdparm -t --offset 25 /dev/sdj

Timing buffered disk reads (offset 25 GB): 530 MB in  3.00 seconds = 176.38 MB/sec

 

root@NAS:/boot/scripts# hdparm -t --offset 30 /dev/sdj

Timing buffered disk reads (offset 30 GB): 540 MB in  3.01 seconds = 179.41 MB/sec

 

root@NAS:/boot/scripts# hdparm -t --offset 35 /dev/sdj

Timing buffered disk reads (offset 35 GB): 538 MB in  3.01 seconds = 178.78 MB/sec

 

 

 

 

Link to comment

I wonder if this has to do with the drive's cache "settling" ... or even the internal UnRAID drivers.

 

What happens if you do this in reverse?  ... i.e. use offsets of 35, 30, 25, 20, ... 0    If the same areas have the same speed differences, then it's indeed the location on the disk that's somehow causing this -- but otherwise it's likely got something to do with the initial caching.

 

Link to comment

I wonder if this has to do with the drive's cache "settling" ... or even the internal UnRAID drivers.

 

What happens if you do this in reverse?  ... i.e. use offsets of 35, 30, 25, 20, ... 0    If the same areas have the same speed differences, then it's indeed the location on the disk that's somehow causing this -- but otherwise it's likely got something to do with the initial caching.

 

Running in reverse, the speed was within 0.5MB/sec for each reported values when running in ascending order with the exception of the 0 GB offset which reported 163.44 MB/sec

Link to comment

Very interesting results.  Of course none of these are poor speeds ... but it's nevertheless interesting that the outermost cylinders are showing these unanticipated results.

 

Have you confirmed that this only happens with Seagate drives??

 

It does not happen with my Hitachi drive. Another user has confirmed they see the drop in their Seagate drives but not WD.

Link to comment

Really strange results.  I can't think of any technical reason for it ... unless they've reserved a few spare sectors on the outermost cylinders so relocations in that area can be closer to the failed sectors and thus higher performance.    That would explain what's happening, as there'd be fewer active cylinders on the outermost cylinders, thus a slightly reduced data transfer speed.

 

Link to comment
  • 1 month later...

Anyone using the WD Red 4tb drives, model WD40EFRX?  These are made for NAS systems so I'm thinking of buying a couple of them.

 

Thanks!

I just ordered 5 more on the newegg deal that came out today.  I have 8 of them in a 19 drive array.  The rest of the drives are 3TB Red and 1 4TB Black.
Link to comment
  • JorgeB unpinned this topic

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.