Should I get a 2TB WD Caviar Green?

[bononos]

LMGTFY:
http://www.anandtech.com/show/3517
http://www.hardwaresecrets.com/article/27

Old 2008 and 2009 articles explains this fairly non-technically. You understanding of point-to-point is flawed or you wouldn't consider bus saturation inconsequential. The PHY layer signaling is entirely different between Rev 2 and Rev 3.

There are cases where Rev 2 is quite sufficient. There are cases where it is not. Misunderstanding how multipliers function, present on the vast majority of consumer SATA implementations, is why you think saturation isn't occurring on either 3 or 6Gb/s bus with 3Gb/s drives. There are perfectly valid reasons why someone should get a drive with a 6Gb/s interface......

I think my understanding of point to point and sata is quite correct and your 2nd link (HWS article) also uses the same term to describe sata.

The articles do mention cache bursts hitting the limits of sata2 for 64Mb cache drives and the use of sata port multipliers for storage racks putting the sata architecture under heavy strain. I'm not sure if the cache burst issue qualify as 'bus saturation', the port multiplier seems like it does but its not in normal usage and I would have overlooked it.

Edit-
And just for clarification- why did you say that theres no such thing as HW raid? Did you mean that for windows systems only or..?

 

[reallyscrued]

LMGTFY:
http://www.anandtech.com/show/3517
http://www.hardwaresecrets.com/article/27
There are cases where Rev 2 is quite sufficient. There are cases where it is not. Misunderstanding how multipliers function, present on the vast majority of consumer SATA implementations, is why you think saturation isn't occurring on either 3 or 6Gb/s bus with 3Gb/s drives. There are perfectly valid reasons why someone should get a drive with a 6Gb/s interface.

I read your links and I'm a little lost.

I see that multiplying SATA channels into multiple physical ports can create a bottleneck if the onboard interface runs at 150 mbps, so the new standard moves this up to 300mb/s and allows more devices to send data simultaneously. Great.

How does that affect a single drive where SATA 3gb/s vs SATA 6gb/s only see performance increases (maybe) in burst mode? Wouldn't it be fine to have a bunch of a SATA 3gb/s drives connected to a SATA 6gb/s controller and not really lose any performance?

From the links you provided, it sounds to me more like choosing a motherboard with SATA 6gb/s is a big deal. Would OP really gain anything if he chose the one SATA interface over the other in his 2 TB Green desktop environment?

 

[murphyc]

I think my understanding of point to point and sata is quite correct and your 2nd link (HWS article) also uses the same term to describe sata.

The articles do mention cache bursts hitting the limits of sata2 for 64Mb cache drives and the use of sata port multipliers for storage racks putting the sata architecture under heavy strain. I'm not sure if the cache burst issue qualify as 'bus saturation', the port multiplier seems like it does but its not in normal usage and I would have overlooked it.

Let me better qualify by saying link saturation, meaning the link between a particular disk and port. That article was written almost three years ago and said drive cache reads were on the cusp of saturating the link.

So maybe this is a better article explaining the benefit of rev 3.

http://www.zdnet.com/blog/storage/what-does-6-gbit-sata-mean-to-you/631

And then this FAQ, page 2, on the two new NCQ commands.
http://www.sata-io.org/documents/SATA-Revision-3.0-FAQ-FINAL.pdf

As for port multiplier, there's still a reason why you'd want SATA Rev 3 disk if you have a Rev 3 controller and multiplier, even though bandwidth is split up due to the use of the multiplier.

Anyway, not everyone is going to be in a situation where a SATA Rev 2 vs 3 interface on the disk will matter. But the idea it's just marketing fluff is ridiculous.

And just for clarification- why did you say that theres no such thing as HW raid? Did you mean that for windows systems only or..?

"Hardware RAID" simply means dedicated hardware, which runs software RAID. It's all software RAID ultimately. I think distinguishing RAID implementations as software RAID vs hardware RAID, rather than about the specific implementation, confuses/ignores the real issues.

e.g. hardware RAID is not inherently superior to software RAID. Many products described as being hardware RAID (by techs, users, reviewers, sometimes even the vendor) in fact using linux md RAID. More interesting and relevant is the feature/cost/complexity/openness breakdown.

 

[murphyc]

Wouldn't it be fine to have a bunch of a SATA 3gb/s drives connected to a SATA 6gb/s controller and not really lose any performance?

In the original poster's context as a backup drive, this is probably true. But read the SATA rev 3 FAQ, and you'll see there are other features of rev 3 other than its transfer rate going to 6Gb/s.

Would OP really gain anything if he chose the one SATA interface over the other in his 2 TB Green desktop environment?

Depends on what's being backed up, but probably not, and almost doesn't matter anyway because the OP said performance wasn't the consideration. All I was really pushing back on was the "purely marketing" comment.

 

[reallyscrued]

you'll see there are other features of rev 3 other than its transfer rate going to 6Gb/s.

All I was really pushing back on was the "purely marketing" comment.

Ah, I see. Point taken then.

 

[lsv]

I have the following WD Green drives, no issues for 2 years. Very slow, maybe 70-80megs/sec from drive to drive but they're quiet and that's important in the studio. The SSD makes up for any slowness anyway. These are pure storage.

(Drive 2) WDC WD10EARS-00Y5B1 ATA Device, 931.51GB size, 1 partition(s), IDE interface
(Drive 3) WDC WD10EARS-00Y5B1 ATA Device, 931.51GB size, 1 partition(s), IDE interface
(Drive 4) WDC WD20EARS-00MVWB0 ATA Device, 1.82TB size, 1 partition(s), IDE interface
(Drive 5) WDC WD20EARX-00PASB0 ATA Device, 1.82TB size, 1 partition(s), IDE interface

 

[thelastjuju]

:thumbsdown::thumbsdown::thumbsdown:

god damn.. those drives are absolute garbage.. I can't believe people actually buy drives that were specifically DESIGNED to fail early. I love WD's black and blues, but they should be ashamed for releasing the green series.

 

[lsv]

:thumbsdown::thumbsdown::thumbsdown:

god damn.. those drives are absolute garbage.. I can't believe people actually buy drives that were specifically DESIGNED to fail early. I love WD's black and blues, but they should be ashamed for releasing the green series.

How exactly are these drives designed to fail early?

 

[murphyc]

:thumbsdown::thumbsdown::thumbsdown:

god damn.. those drives are absolute garbage.. I can't believe people actually buy drives that were specifically DESIGNED to fail early. I love WD's black and blues, but they should be ashamed for releasing the green series.

Haha. Troll.

 

[thelastjuju]

How exactly are these drives designed to fail early?

The "Intellispin"feature, that's how.

WD decided to take advantage of the "green" marketing trend, where they claimed the drive was designed to save you energy costs.. even though this amounts to single digit wattage savings, saving you maybe ten cents per month.

Basically, the drive's spindle shuts itself down every few minutes to save you the 3 watts that hard drive would otherwise idle at. The initial spinning up is THE MOST stressful thing for a hard drive to do, and so that explains why the green drives are aging at a significantly faster rate than any other drive, and why they have absurdly high failure rates.

If you are at the 2 year mark, expect those drives to fail any day now. Why do you think they only offer a 2 year warranty while they offer a 5 year warranty with the black drives? Just a coincidence?

 

[thelastjuju]

Haha. Troll.

Yes, and the 45% of people in the last 6 months giving the drives either a 1 or 2 egg rating claiming it died abruptly must also be trolling.

http://www.newegg.com/Product/Produc...82E16822136514

 

[murphyc]

The "Intellispin"feature, that's how.

Intellispin? What are you talking about?

WD decided to take advantage of the "green" marketing trend, where they claimed the drive was designed to save you energy costs.. even though this amounts to single digit wattage savings, saving you maybe ten cents per month.

Show your math.

Basically, the drive's spindle shuts itself down every few minutes to save you the 3 watts that hard drive would otherwise idle at.

5.5 watts vs 0.8 watts = 4.7 watts. But I've not ever had these drives go idle every few minutes. This is a setting up to the system, not the disk. It takes 10+W to spin up, so your math makes zero sense that the drive would sleep every few minutes. It would consume more power in the long run.

The initial spinning up is THE MOST stressful thing for a hard drive to do, and so that explains why the green drives are aging at a significantly faster rate than any other drive, and why they have absurdly high failure rates.

Conjecture. No proof.

If you are at the 2 year mark, expect those drives to fail any day now. Why do you think they only offer a 2 year warranty while they offer a 5 year warranty with the black drives? Just a coincidence?

2 year warranty on blues. What's your logic now? That blues are also crap?

Amateur.

 

[murphyc]

Yes, and the 45% of people in the last 6 months giving the drives either a 1 or 2 egg rating claiming it died abruptly must also be trolling.

Those would be anecdotes. Not the vast useless conjecture you're supplying.

 

[thelastjuju]

Oh wow, so I was wrong.. it saves you a whopping 4 watts instead of the 3 watts I initially though.

I don't even think its minutes by the way.. its every few SECONDS, the green drives are designed to "park" themselves to conserve power. intellipower, intellepark, intelleseek, whatever they are calling it.. there's nothing intelligent about it.

All this approach leads to is a perpetual load/unload cycle, which is why I said they age faster than normal drives. Hard drives are limited to a certain amount of these cycles, it is not infinite.. so it makes perfect sense why such an inordinate amount of Green drives fail prematurely.. because they reach that limit at at such an accelerated rate.

 

[murphyc]

Oh wow, so I was wrong.. it saves you a whopping 4 watts instead of the 3 watts I initially though.

4.7 != 4.

I don't even think its minutes by the way.. its every few SECONDS, the green drives are designed to "park" themselves to conserve power

You're daft. This is head parking, it does not spin down the drive. It does not put it into idle. It is designed to keep the head off the platter when not in use. This isn't even a power saving feature.

intellipower, intellepark, intelleseek, whatever they are calling it.. there's nothing intelligent about it.

IntelliPower, is WDC's way of not specifying the RPM of the disk, so they can releast 5400-7200RPM disks without telling you what you're getting. That's it. Has nothing to do with what you're talking about. Nothing to do with head parking or spin down.

All this approach leads to is a perpetual load/unload cycle, which is why I said they age faster than normal drives.

Conjecture. You don't actually know this. You're speculating. You don't have any of the terms, behaviors, mechanisms correct, let alone causation.

Hard drives are limited to a certain amount of these cycles, it is not infinite.. so it makes perfect sense why such an inordinate amount of Green drives fail prematurely.. because they reach that limit at at such an accelerated rate.

You do not have access to a scientific sample in order to draw such conclusions. This is why what you're doing is called trolling. You don't know what you're talking about, Mr. "Designed To Fail Early". But I'm gonna guess you're too dumb to realize that's libel.

 

[slayernine]

I've seen higher than average failures on green series drives, I think the 7200RPM is just better. No more cheap drives for me.

 

[murphyc]

Two years old.
http://www.macobserver.com/tmo/article/french_retailer_data_offers_ssd_failure_rates/

 

[thelastjuju]

I've seen higher than average failures on green series drives, I think the 7200RPM is just better. No more cheap drives for me.

Anyone who has done the research would have seen this. ()

If you want to save money and take the RISK on what are clearly inferior drives, be my guest..

Its funny how this WD Green apologist accuses me of "trolling" and "speculating" based off conjecture.. and then he cherry picks a random data sample, from a French Mac site that admits the study was "a muddy comparison in a number of ways" to try and prove his point. :whiste:

.. as if there is such thing as an official failure rate database.. Obviously you have to rely on other people's experiences and testimonies. I've seen enough horror stories and reported failures to know to stay away from WD Green drives, and that the "energy saving" nonsense is doing nothing but shortening their lifespans.

.. oh, and not to mention they are slow as turds!

 

[murphyc]

Anyone who has done the research would have seen this. ()

You've presented no research, no facts, no data.

Its funny how this WD Green apologist accuses me of "trolling" and "speculating" based off conjecture.. and then he cherry picks a random data sample, from a French Mac site that admits the study was "a muddy comparison in a number of ways" to try and prove his point. :whiste:

It was one Google search, there are others. You've presented none. Comments left by buyers at the point of sale by definition are not a scientific sample. You've presented no failure rate of your own. You've presented no failure rate for Caviar Blue or Black. You've suggested a 2 year warranty is indicative of low quality yet you don't subscribe the same criticism for the Caviar Blue which likewise has a 2 year warranty.

.. as if there is such thing as an official failure rate database.. Obviously you have to rely on other people's experiences and testimonies. I've seen enough horror stories and reported failures to know to stay away from WD Green drives, and that the "energy saving" nonsense is doing nothing but shortening their lifespans.

You should have a conversation with a lawyer about stating as fact that you know Western Digital intentionally designed, manufactures, and sells drives with the express purpose of shortened life. It is called libel unless you can prove with evidence what you're saying.

.. oh, and not to mention they are slow as turds!

The seek time and transfer rate is nearly identical to that of the Blue. Where's your data?

And if you'd done your research, you'd know any drive can die at any time. Green, Blue, Black, or any other brand, it doesn't matter. If you are not regularly running a SMART monitoring tool, you do not really care if the disk dies on you without notice. If you aren't doing backups, you don't care about the data. Not really.

 

[birthdaymonkey]

I did find the head park functionality on my WD Green annoying. Fortunately it can be easily switched off with the wdidle utility.

 

[murphyc]

My Caviar Black exhibits identical behavior. It's not just a green thing. In just a couple weeks over 12 months, SMART is reporting 586390 load cycle counts of the head parking. That's an average of 1606 events per day.

WD support told me they've not had a drive fail due to frequent actuator parking.

 

[bononos]

I don't know where the higher than average failures on Caviar Greens came from. Some searching brought up issues with the EADS green(possibly black/blue also?) model. The current model is EARX (earlier model is EARS). In fact Caviar greens are advertised as running cooler and more reliable (slower is a given). In any case it is not that slow to be considered bad, and its fast enough for normal builds. I think alot of the poor reputation it receives is because of the internet which tends to be very polarizing. Something which is average/subpar can turn out to be bad/horrible/shitty after rounds of chinese whispers.

 

[murphyc]

My best guess is that some of it has to do with RAID, because some people though they could build a cheap RAID with green drives. Some RAID implementations (dedicated control, a.k.a. hardware RAID) have the ability to correct for errors themselves, and expect the drive to quickly recover, in effect passing the error to the RAID hardware, rather than fixing it themselves - for WDC drives the features is TLER. If not enabled, the drive tries to do sometimes elaborate and lengthy correction that such RAID controllers then assume the drive is hosed and removes it from the array. Lotsa complaints about green drives and RAID. And then while TLER was disabled it was possible to re-enable. Newer green drives, the feature is permanently disabled. So some people might just be pissy. Even though WDC specifically says consumer RAID 0 and 1 are fine for green, blue and black drives. Business RAID is not. If you want to do business/enterprise class RAID you need to buy an RE drive.

 

[bononos]

Could be that the RAID issue got carried away until it became a general reliability issue. But from some of the threads around here, someone would be led to think that WD greens are not safe and fast enough for single drive budget pcs for surfing the net.

 

[Fjive]

i'm also looking for a data drive for my next build, 1.5TB to 2TB in size. coz 1TB is barely enough nowadays...was hoping to get a WD blue but the max size is 1TB...so what else is good & reliable these days? seagate? or the WD Green is sufficient enough?

btw, data drive will be storing movies, photos, mp3s, a few VMs