It's all I use for server storage drives. Haven't had one fail. All the models I've ever deployed are 1TB and 2TB drives. It's amazing how these modern hard disks are somewhat competitive with low-end SSD's (aside from access time)
funny as i did not realize i had a backblaze HDD it is i accuity have the ST31500341AS drive in my system right now (1300 days power on time 7500 stop start count due to HDD power save spin down) and its failing slowly bad sectors are racking up and "reported uncorrectable errors" are now starting to happen (not gone up from when i started copying the data to a WD RED 4TB same 1607) don't think i lost any data itself as "read error rate" and "hardware ECC recovered" are still the same (not that the stuff on it is that important to lose any way) just 70 extra relocations in last day meant it was time to copy data to another drive
... NCQ streaming feature which enables isochronous data transfers for multimedia streams while also improving performance of lower priority transfers. This feature could be very useful for media server and video editing use-cases....
I've got a pair of WD 1TB Re drives in my NAS... full, so been wanting to put a pair of 4TB or 6TB drives in.... the WD Reds would be great, except I'm also seeing a lot of issues.
I really don't recommend you follow backblaze statistics. They are collected under one use-case: cold storage. It's also important to note that all of their drives are consumer drives that lack firmware to change the harmonic vibration in cases with that many (72!!) drives. Most consumer drives are rated for installation in cases for 2-8 drives, with enterprise drives rated for up to 16 drives per chassis. Real data centers don't use pods like backblaze (who make their own) because no matter how you dampen the vibrations, these drives are not engineered to work in a large chassis together.
All their data shows is Seagate drives suck as vibration resistance, with WD being slightly better and Hitachi cleaning the show. You know why? WD Red models and ALL Hitachi drives have the platter shaft locked at the TOP of the drive case; Seagate drives just have the shaft locked at the motor (bottom axis) which is for the most part adequate as long as vibration isn't an issue. This makes the drives cheaper, quieter, cooler and more efficient, and entirely adequate for consumer applications.
I'm not too worried though since my NAS is running Raid-1 Drive fails, RMA it. In 3-5years I'll likely either upgrade/replace the drives and/or get a new NAS+drives
My findings are that Backblaze statistics are RIGHT on target. In fact they may be too kind to Seagate. I threw 12 Seagate drives in the trash this spring that were 1.5, 2 and 3TB capacities with 3tb drives being the worst drives I have ever come in contact with. I would say that my 3TB Seagate drive failures are higher than 75% in 24 months. I own a service and repair company and observe Seagate drives fail in droves, particularly the 3TB ones. Seagate 4tb drives are vastly more reliable so far but I stopped buying Seagate drives PERIOD because of the thousands of dollars I lost dealing with them. My primary drive is HGST now. I have a couple dozen of their 4, 5 and 6TB NAS drives in a pair of Win 2012R2 servers and a X99 Workstation. I also use them in client servers and workstations for backup images. In 3 years I have yet to have a HGST drive fail or even receive a D.O.A. Great drives.
Formerly know as 'deathstar' when still owned by IBM due to consistent early failure, I don't get that they didn't change the name. Personally I don't trust 'consumer reviews' blindly, because there is massive cheating with paid fake reviews on the internet, and because each person has its own usage pattern and perhaps a good experience with a brand is based on a use quite dissimilar to mine.
I had 6 of those 60GB "Deathstar" drives and each and every one of them eventually failed with the "click of death". I was able to RMA them under warranty and every one failed except for one which is sitting in a box somewhere.
^ this. The first-generation giant magneto resistant (GMR) heads were not reliable. This hasn't been an issue for a decade. IBM/Hitachi/HGST-Toshiba has changed hands so many times (not owned by WD) that very little IBM technology, if any at all, remains in these new drives.
I posted my deahtstar remark as a reply to a comment about the deskstar, and am aware it's not a WD company. I have no idea why it did not show up as a reply, probably the poorly codedantiquated comment system to blame?
And my primary point is that it's odd that the company changed name twice but they kept that name with which people had bad memories, it's simply a weird thing to do.
It's a pretty good name. The same reason Ford brought back the Taurus nameplate. Even though it was associated with a crappy product at the end of its run, it still has good history behind it.
My 75GB 75GXP failed slowly, I dunno if it was related to the main source of failure everyone was seeing... It made for some interesting surprises, like waking up to an empty Desktop because the directory had been mysteriously renamed Desotop (amazingly everything within was intact).
I prefer HGST drives, but really just the older ones like the 2TB and 3TB 5400RPM Deskstar Coolspins. They were nearly as fast as most 7200RPM drives of the day.
The He6 is just too expensive, although they seem to be very reliable and that was the real concern at the beginning for that technology.
And until recently, Hitachi/HGST didn't have a 6TB non-Helium model available. They are clearly behind Seagate and WD on density, but ahead of them on reliability. Sometimes keeping it old-school is the best path. Look at Mazda with Skyactive. No turbo charging, no fancy dual-clutch or CVT transmissions, just a modern lightweight engine with as much friction removed as possible mated to a tweaked slushbox and it still beats every single competitor in fuel economy, aside from Hybrids, while also being among the few vehicles manufactures that have not had transmission or engine related recalls.
At the current rate, it'll be only a few years before NAND hits parity with magnetic storage in capacity and price.
3D (verticle) NAND and Intel's 3D X-Point are going to revolutionize storage as we know it.
X-Point will slowly make NAND a cheap commodity for the consumer sector (abandoned by the enterprise sector for X-Point) and NAND will be so cheap to produce by that point (it's 30 year old technology) that it'll be the common medium. It has already killed virtually every other form of portable storage (floppy disks, tapes, CD/DVD/Bluray, etc) with the only worthy exception being large capacity (2-4TB) 2.5" portable hard disks. It's only a matter of time before 4TB SSD's cost nearly the same as 4TB hard disks, and hard disk platter density is already hit some physical barriers, hence the need for shingled recording and other reliability/performance sacrifices.
Reliability would not be a factor. Large financial institutions are using them for transaction processing, and few things require more write reliability than that.
The advantage of SSD is, according to legend at least, a more consistent life time: just retire them at x% of warranty, less infant mortality and random death (getting run over by a bus, in human terms).
Why is the article listing the WD red pro with the addition 'star NAS'? Since when is WD using the 'star' term? I can't imagine them doing so, and especially see a company market something with 'red star' these days :) A quick web search shows only anadtech adds 'star' the the name.
According to the benchmarks in this article – as a DAS drive the WD Red Pro seems nice – but it for some reason seems not as good when used as a NAS drive, agreed on that.
Colour me disappointed that there is no analysis of acoustic behaviour. My NAS is by far the loudest device in my room due to the 4 WD Red (non pro) drives in it, and they're supposed to be one of the quietest drives available.
The problem might also lie with your NAS: insufficient decoupling will lead to very nasty vibration, as is insufficient dampening or the use of not stiff enough components.
Important details are missing from the article. It should be the first thing covered for drives of such capacities - make it clear if they are using a shingled write method requiring rewrites of large blocks for small random writes.
As far as I know Apple was the first mass consumer company to embrace the practice. Or at least the best known for it. Usually gadgets of other companyes were expected to be superceeded by advancing technology, not designed to fail after x time... One of the most famous examples was the Ipod Nano case.
The practice existed before but apple put it in the spotlight.
Why not moving to a 10gbe for the nas test? I think that for a soho scenario could make sense, with a direct 10gb connection between workstation and nas and classic 1gb link among remaining clients. Just my 2 cents
Except for upper end enterprise it's still probably too early to add 10gbe to a nas test suite.
A few years back I remember reading that there was a process threshold that a lot of people were expecting to dramatically drive down the high power cost of sending a 10gb signal over twisted pair connections. Unfortunately I don't remember if it was 14nm; in which case we should start to see lower power hardware within a year or 10 nm; in which case we'll probably end up waiting until closer to 2020 before getting in in our boxes.
Yes, that's absolutely true, my nas waste same power on the 10gbe adapter and on cpu and that's crazy. Still, with HD and then 4k videos, raw pics near or over 50MB each, old good hdd disks which continue to improve, 1gbe is going to be too limiting... 2020 is really far away! ☺
We could well see copper Ethernet being replaced by then. IBM just announced in May a 100 Gb/s transceiver using the first fully integrated wavelength multiplexed chip. The optical components are on-chip along side the electronic components. This integration is what will bring the price of optical Ethernet down to parity with copper. It will be a huge bump in bandwidth and power reduction. It is telling that the 100G consortium is already increasing the max distance to 1 km.
I've been using the Western Digital's NAS with these drives, WD EX2100, since April. Very solid unit. Really easy to set up, and includes advanced features when needed such as link aggregation.
No, they are right - 10 in 10^15 is exactly the same as 1 in 10^14. It's written/marketed that way to confuse and look much better than it really is.
Using a less than sign doesn't really change the base order of magnitude - eg. 9 in 10^15 would be consistent with their PDF table claim and it is still almost 10 times worse than any enterprise drive at 1 in 10^15 URE, which is why non-enterprise are not worth buying at these fail-likely sizes if you value your data.
but if you getting unrecoverable read errors at that point you should replace the HDD as there is a problem with the drive, like i need to replace my segate {as the theme goes with seagate drives} as its slowly developing reallocated sectors, surprisingly its still working fine
They are talking about the expected number of unrecoverable read errors on a perfectly functional drive. Also, reallocated sectors occur on a write, not a read.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
62 Comments
Back to Article
jragonsoul - Monday, September 7, 2015 - link
I have used a few Red as storage drives before. Liking the 6TB bump.Samus - Monday, September 7, 2015 - link
It's all I use for server storage drives. Haven't had one fail. All the models I've ever deployed are 1TB and 2TB drives. It's amazing how these modern hard disks are somewhat competitive with low-end SSD's (aside from access time)leexgx - Wednesday, October 21, 2015 - link
funny as i did not realize i had a backblaze HDD it is i accuity have the ST31500341AS drive in my system right now (1300 days power on time 7500 stop start count due to HDD power save spin down)and its failing slowly bad sectors are racking up and "reported uncorrectable errors" are now starting to happen (not gone up from when i started copying the data to a WD RED 4TB same 1607) don't think i lost any data itself as "read error rate" and "hardware ECC recovered" are still the same (not that the stuff on it is that important to lose any way) just 70 extra relocations in last day meant it was time to copy data to another drive
imaheadcase - Monday, September 7, 2015 - link
Streaming Not Supported Not SupportedWhat does that even mean?
ganeshts - Monday, September 7, 2015 - link
We covered the meanings of those table entries in a previous review: http://www.anandtech.com/show/7258/battle-of-the-4...A55A551N 11B2P - Thursday, September 10, 2015 - link
so you're saying that the WD Red PRO's wouldn't be good for a central media server with up to 6 clients?ddriver - Monday, September 7, 2015 - link
Judging by consumer reviews, HGST Deskstar seems to have the upper hand when it comes to reliability.Souka - Monday, September 7, 2015 - link
I've got a pair of WD 1TB Re drives in my NAS... full, so been wanting to put a pair of 4TB or 6TB drives in.... the WD Reds would be great, except I'm also seeing a lot of issues.Granted, not bad as the Seagate drives!
Source: https://www.backblaze.com/blog/best-hard-drive/
Samus - Monday, September 7, 2015 - link
I really don't recommend you follow backblaze statistics. They are collected under one use-case: cold storage. It's also important to note that all of their drives are consumer drives that lack firmware to change the harmonic vibration in cases with that many (72!!) drives. Most consumer drives are rated for installation in cases for 2-8 drives, with enterprise drives rated for up to 16 drives per chassis. Real data centers don't use pods like backblaze (who make their own) because no matter how you dampen the vibrations, these drives are not engineered to work in a large chassis together.All their data shows is Seagate drives suck as vibration resistance, with WD being slightly better and Hitachi cleaning the show. You know why? WD Red models and ALL Hitachi drives have the platter shaft locked at the TOP of the drive case; Seagate drives just have the shaft locked at the motor (bottom axis) which is for the most part adequate as long as vibration isn't an issue. This makes the drives cheaper, quieter, cooler and more efficient, and entirely adequate for consumer applications.
Souka - Tuesday, September 8, 2015 - link
Cool info, thanks!I'm not too worried though since my NAS is running Raid-1 Drive fails, RMA it. In 3-5years I'll likely either upgrade/replace the drives and/or get a new NAS+drives
MHz Tweaker - Wednesday, September 9, 2015 - link
My findings are that Backblaze statistics are RIGHT on target. In fact they may be too kind to Seagate. I threw 12 Seagate drives in the trash this spring that were 1.5, 2 and 3TB capacities with 3tb drives being the worst drives I have ever come in contact with. I would say that my 3TB Seagate drive failures are higher than 75% in 24 months. I own a service and repair company and observe Seagate drives fail in droves, particularly the 3TB ones. Seagate 4tb drives are vastly more reliable so far but I stopped buying Seagate drives PERIOD because of the thousands of dollars I lost dealing with them. My primary drive is HGST now. I have a couple dozen of their 4, 5 and 6TB NAS drives in a pair of Win 2012R2 servers and a X99 Workstation. I also use them in client servers and workstations for backup images. In 3 years I have yet to have a HGST drive fail or even receive a D.O.A. Great drives.Wwhat - Monday, September 7, 2015 - link
Formerly know as 'deathstar' when still owned by IBM due to consistent early failure, I don't get that they didn't change the name.Personally I don't trust 'consumer reviews' blindly, because there is massive cheating with paid fake reviews on the internet, and because each person has its own usage pattern and perhaps a good experience with a brand is based on a use quite dissimilar to mine.
melgross - Monday, September 7, 2015 - link
I only recall IBM having one particular drive that had a problem. Long time ago, and I don't remember the model, but it was a 75GB unit.bigboxes - Monday, September 7, 2015 - link
I had 6 of those 60GB "Deathstar" drives and each and every one of them eventually failed with the "click of death". I was able to RMA them under warranty and every one failed except for one which is sitting in a box somewhere.bigboxes - Monday, September 7, 2015 - link
75GXP and 60GXP were the bad models.Samus - Monday, September 7, 2015 - link
^ this. The first-generation giant magneto resistant (GMR) heads were not reliable. This hasn't been an issue for a decade. IBM/Hitachi/HGST-Toshiba has changed hands so many times (not owned by WD) that very little IBM technology, if any at all, remains in these new drives.Wwhat - Wednesday, September 9, 2015 - link
I posted my deahtstar remark as a reply to a comment about the deskstar, and am aware it's not a WD company.I have no idea why it did not show up as a reply, probably the poorly codedantiquated comment system to blame?
And my primary point is that it's odd that the company changed name twice but they kept that name with which people had bad memories, it's simply a weird thing to do.
Samus - Thursday, September 10, 2015 - link
It's a pretty good name. The same reason Ford brought back the Taurus nameplate. Even though it was associated with a crappy product at the end of its run, it still has good history behind it.Kjella - Tuesday, September 8, 2015 - link
The IBM Deskstar 75GXP series, but it had six models with 15/20/30/45/60/75GB capacity. Personally I had a 45GB drive die to the "click of death".Souka - Tuesday, September 8, 2015 - link
I think I had a number of these drives....they were the fastest around (or most performance per$$)Impulses - Tuesday, September 8, 2015 - link
My 75GB 75GXP failed slowly, I dunno if it was related to the main source of failure everyone was seeing... It made for some interesting surprises, like waking up to an empty Desktop because the directory had been mysteriously renamed Desotop (amazingly everything within was intact).MHz Tweaker - Wednesday, September 9, 2015 - link
Yes, I think I remember those GXP-Deathstars "click click clickety click"Samus - Monday, September 7, 2015 - link
I prefer HGST drives, but really just the older ones like the 2TB and 3TB 5400RPM Deskstar Coolspins. They were nearly as fast as most 7200RPM drives of the day.The He6 is just too expensive, although they seem to be very reliable and that was the real concern at the beginning for that technology.
And until recently, Hitachi/HGST didn't have a 6TB non-Helium model available. They are clearly behind Seagate and WD on density, but ahead of them on reliability. Sometimes keeping it old-school is the best path. Look at Mazda with Skyactive. No turbo charging, no fancy dual-clutch or CVT transmissions, just a modern lightweight engine with as much friction removed as possible mated to a tweaked slushbox and it still beats every single competitor in fuel economy, aside from Hybrids, while also being among the few vehicles manufactures that have not had transmission or engine related recalls.
Adul - Tuesday, September 8, 2015 - link
Not to mention taht they are great to drive cards :)Adul - Tuesday, September 8, 2015 - link
Cars! should not reply from phone.yeeeeman - Monday, September 7, 2015 - link
When do these ancient objects go extinct? And why didn't you include at least an ssd for comparison?damianrobertjones - Monday, September 7, 2015 - link
When you and I can afford to buy a 6TB SSD version. That probably means when we're dead.FunBunny2 - Monday, September 7, 2015 - link
I wonder if it's possible to stuff 6TB of NAND, at any node size, into 2.5" form factor?KateH - Monday, September 7, 2015 - link
Yes.http://arstechnica.com/gadgets/2015/08/samsung-unv...
FunBunny2 - Tuesday, September 8, 2015 - link
Zoweeeeee.Samus - Tuesday, September 8, 2015 - link
At the current rate, it'll be only a few years before NAND hits parity with magnetic storage in capacity and price.3D (verticle) NAND and Intel's 3D X-Point are going to revolutionize storage as we know it.
X-Point will slowly make NAND a cheap commodity for the consumer sector (abandoned by the enterprise sector for X-Point) and NAND will be so cheap to produce by that point (it's 30 year old technology) that it'll be the common medium. It has already killed virtually every other form of portable storage (floppy disks, tapes, CD/DVD/Bluray, etc) with the only worthy exception being large capacity (2-4TB) 2.5" portable hard disks. It's only a matter of time before 4TB SSD's cost nearly the same as 4TB hard disks, and hard disk platter density is already hit some physical barriers, hence the need for shingled recording and other reliability/performance sacrifices.
But even then, soon,
Souka - Monday, September 7, 2015 - link
SSDs aren't for NAS due to pricingIf you put a SSD in, many of the charts would be unreadable due to scaling issues.
Also, if the NAS is used for large number of writes, the SSD reliability would be a factor
melgross - Monday, September 7, 2015 - link
Reliability would not be a factor. Large financial institutions are using them for transaction processing, and few things require more write reliability than that.FunBunny2 - Tuesday, September 8, 2015 - link
The advantage of SSD is, according to legend at least, a more consistent life time: just retire them at x% of warranty, less infant mortality and random death (getting run over by a bus, in human terms).MrSpadge - Monday, September 7, 2015 - link
The strong DAS performance bodes well for the new 6 TB Black model, which is probably physically similar but with different firmware settings.Wwhat - Monday, September 7, 2015 - link
Why is the article listing the WD red pro with the addition 'star NAS'? Since when is WD using the 'star' term? I can't imagine them doing so, and especially see a company market something with 'red star' these days :)A quick web search shows only anadtech adds 'star' the the name.
Morawka - Monday, September 7, 2015 - link
disappointing performance. basically every other nas drive on the market is faster than these. HGST and Seagate dominating.star-affinity - Tuesday, September 8, 2015 - link
According to the benchmarks in this article – as a DAS drive the WD Red Pro seems nice – but it for some reason seems not as good when used as a NAS drive, agreed on that.Arbie - Monday, September 7, 2015 - link
We want eight and we won't wait.Gigaplex - Monday, September 7, 2015 - link
Colour me disappointed that there is no analysis of acoustic behaviour. My NAS is by far the loudest device in my room due to the 4 WD Red (non pro) drives in it, and they're supposed to be one of the quietest drives available.Souka - Tuesday, September 8, 2015 - link
It would be even louder with four of these WD Red Pro drives!nagi603 - Monday, October 5, 2015 - link
The problem might also lie with your NAS: insufficient decoupling will lead to very nasty vibration, as is insufficient dampening or the use of not stiff enough components.beginner99 - Tuesday, September 8, 2015 - link
Well with the Pro only $33 more it's a no brainer. The 5 years warranty alone will make that a profitable investment alone.Visual - Tuesday, September 8, 2015 - link
Important details are missing from the article. It should be the first thing covered for drives of such capacities - make it clear if they are using a shingled write method requiring rewrites of large blocks for small random writes.Morawka - Tuesday, September 8, 2015 - link
anyone remember when all seagate consumer drives had 5 year warranties.. it was great. now we are lucky to get a 3 year warranty.FunBunny2 - Tuesday, September 8, 2015 - link
planned obsolescence is a wonderful thing. just ask Apple.star-affinity - Tuesday, September 8, 2015 - link
In what way is Apple worse than others when it comes to ”planned obsolescence”?valinor89 - Tuesday, September 8, 2015 - link
As far as I know Apple was the first mass consumer company to embrace the practice. Or at least the best known for it. Usually gadgets of other companyes were expected to be superceeded by advancing technology, not designed to fail after x time... One of the most famous examples was the Ipod Nano case.The practice existed before but apple put it in the spotlight.
Gigaplex - Sunday, September 13, 2015 - link
Apple is rarely the first to do anything. They certainly weren't the first to embrace planned obsolescence.Hannibal80 - Tuesday, September 8, 2015 - link
Why not moving to a 10gbe for the nas test? I think that for a soho scenario could make sense, with a direct 10gb connection between workstation and nas and classic 1gb link among remaining clients. Just my 2 centsDanNeely - Tuesday, September 8, 2015 - link
Except for upper end enterprise it's still probably too early to add 10gbe to a nas test suite.A few years back I remember reading that there was a process threshold that a lot of people were expecting to dramatically drive down the high power cost of sending a 10gb signal over twisted pair connections. Unfortunately I don't remember if it was 14nm; in which case we should start to see lower power hardware within a year or 10 nm; in which case we'll probably end up waiting until closer to 2020 before getting in in our boxes.
Hannibal80 - Wednesday, September 9, 2015 - link
Yes, that's absolutely true, my nas waste same power on the 10gbe adapter and on cpu and that's crazy. Still, with HD and then 4k videos, raw pics near or over 50MB each, old good hdd disks which continue to improve, 1gbe is going to be too limiting...2020 is really far away! ☺
Jaybus - Wednesday, September 9, 2015 - link
We could well see copper Ethernet being replaced by then. IBM just announced in May a 100 Gb/s transceiver using the first fully integrated wavelength multiplexed chip. The optical components are on-chip along side the electronic components. This integration is what will bring the price of optical Ethernet down to parity with copper. It will be a huge bump in bandwidth and power reduction. It is telling that the 100G consortium is already increasing the max distance to 1 km.Sivar - Tuesday, September 8, 2015 - link
I've been using the Western Digital's NAS with these drives, WD EX2100, since April. Very solid unit. Really easy to set up, and includes advanced features when needed such as link aggregation.[email protected] - Tuesday, September 8, 2015 - link
thats great can't wait to get my hands on thesewww.hardwarecomputerstore.co.uk
ex_User - Wednesday, September 9, 2015 - link
I believe you have an error in specs. "Non-Recoverable Read Errors / Bits Read" for Red Pro is "<10 in 10^15", not 10^14.http://www.wdc.com/wdproducts/library/SpecSheet/EN...
asmian - Thursday, September 10, 2015 - link
No, they are right - 10 in 10^15 is exactly the same as 1 in 10^14. It's written/marketed that way to confuse and look much better than it really is.Using a less than sign doesn't really change the base order of magnitude - eg. 9 in 10^15 would be consistent with their PDF table claim and it is still almost 10 times worse than any enterprise drive at 1 in 10^15 URE, which is why non-enterprise are not worth buying at these fail-likely sizes if you value your data.
ex_User - Thursday, September 10, 2015 - link
Well, that misleading marketing works very well on laymen like me -- only after your post have I noticed the <1/<10 trick.leexgx - Saturday, September 12, 2015 - link
i agree that <1/<10 is Very misleading, so they have reduced or not added more error correction on these drivesleexgx - Saturday, September 12, 2015 - link
but if you getting unrecoverable read errors at that point you should replace the HDD as there is a problem with the drive, like i need to replace my segate {as the theme goes with seagate drives} as its slowly developing reallocated sectors, surprisingly its still working fineGigaplex - Monday, September 14, 2015 - link
They are talking about the expected number of unrecoverable read errors on a perfectly functional drive. Also, reallocated sectors occur on a write, not a read.xicaque - Sunday, October 23, 2016 - link
I like my NAS Reds