Is there a problem with head parks on WD Green HDDs?

Silencing hard drives, optical drives and other storage devices

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
rseiler
Posts: 66
Joined: Sun Dec 28, 2003 1:14 pm

Post by rseiler » Tue Mar 24, 2009 12:33 pm

That excuse sounds like, well, an excuse.

"It was not designed for this and so the outcome can be unpredictable."

I'm thinking it was designed for exactly what it was obviously designed for and that the "outcome" is perfectly apparent (disabled, 25.5 seconds, whatever).

And wouldn't it be a rare drive that fails in such a way that even a utility like this wouldn't be able to reach it? If it powers on, you should still be able to use the utility to set it back to what it was--before RMA'g. Your data could be lost to all but the most advanced disk recovery outfit, and this utility should still be able to access the drive's flash memory (or whatever this setting is stored in).

lunadesign
Posts: 120
Joined: Sat Jun 10, 2006 11:49 pm

Post by lunadesign » Tue Mar 24, 2009 12:43 pm

For what its worth, I was on the phone with WD Level 2 tech support in California the other day for another issue and briefly mentioned the firmware update issue. Apparently a consumer drive like the WD10EADS can have in the neighborhood of 30 variants. Apparently they are constantly varying things (sometimes at the request of manufacturers like Dell and HP) so its hard to have one set of firmware work for all the variants.

At least that's what I was told. To me, it seems a bit spendy manufacturing-wise to have all those variants for a product that doesn't have huge margins.

I'm guessing the enterprise drives are more rigidly controlled (much fewer variants) and that's why its easier to release firmware updates for those.

WilliamG
Posts: 1
Joined: Tue Mar 24, 2009 4:46 pm
Location: Minnesota

Post by WilliamG » Tue Mar 24, 2009 4:47 pm

I've been driven mad for ages with these WD drives parking heads all the time. I've got two sets of WD Studio II drives each with two of the 1TB Green drives in them. Is there a way to update the drives so that the head parking no longer occurs?

FYI, I'm running Mac OS X 10.5.6. If someone has a method to fix this, I'd be eternally grateful!

silentbobbo
Posts: 108
Joined: Sat Mar 31, 2007 1:45 am

Post by silentbobbo » Thu Mar 26, 2009 12:23 pm

Here is a response to my question as to why 500GB models aren't addressed on their KB article 3263:
I apologize for the inconvenience; that feature cannot be modified on those drives.
I basically wrote them that Load_Cycle_Count is increasing by 1 per minute on my WD5000AACS-00ZUB0 drive and this:
I would like to request
* a solution similar to the one provided in KB article 3263;
* or a utility, which would make the drive to wait longer than 8 seconds, before entering the "idle 3" mode;
* or reassurance that my worries are invalid.
As already noted by others, there seems to be neither solutions, nor confirmation, nor denial at WD support...

casebuyer
Posts: 119
Joined: Mon Jan 29, 2007 10:30 pm

Post by casebuyer » Fri Mar 27, 2009 6:37 am

if this is the case simply I won't even consider buying WD Green Series at all in my next purchase
Either samsung f1 1 TB or seagate 1.5 TB

So did seagate firmware update solved the issue on 1.5 TB or is it still buggy drive with freezing issues

lunadesign
Posts: 120
Joined: Sat Jun 10, 2006 11:49 pm

Post by lunadesign » Fri Mar 27, 2009 11:25 pm

For those curious about the 2TB WD20EADS models, I just picked up a pair and tested one. I can confirm that the WD20EADS operates just like WD10EADS - the 8 second idle timer is present *and* Load Cycle Count is not being incremented when it happens. The clicking noise is about the same, perhaps a teeny bit quieter than the WD10EADS units.

trxman
Posts: 145
Joined: Wed Dec 15, 2004 5:45 pm
Location: system

Post by trxman » Sun Mar 29, 2009 1:37 pm

I'm realy disappointed.

After years of using WD, I have to advise everyone to use some other drives. :(

Go for Samsung F1, folks. I hope there will be some 2TB versions soon.

Benjiro
Posts: 14
Joined: Wed Jul 09, 2008 6:11 pm
Location: Belgium

Post by Benjiro » Thu Apr 02, 2009 11:36 am

lunadesign wrote:For those curious about the 2TB WD20EADS models, I just picked up a pair and tested one. I can confirm that the WD20EADS operates just like WD10EADS - the 8 second idle timer is present *and* Load Cycle Count is not being incremented when it happens. The clicking noise is about the same, perhaps a teeny bit quieter than the WD10EADS units.
*uch* No WD 2TB for me then.

casebuyer:

I'm considering the Seagate or Samsungs. Problem is, Seagate there track record is also not very good with the lasted serie. And the Samsung is relative new.

darthpingoo
Posts: 3
Joined: Wed Feb 06, 2008 2:31 pm
Location: France

Post by darthpingoo » Sun Apr 19, 2009 1:54 pm

Benjiro wrote:
lunadesign wrote:For those curious about the 2TB WD20EADS models, I just picked up a pair and tested one. I can confirm that the WD20EADS operates just like WD10EADS - the 8 second idle timer is present *and* Load Cycle Count is not being incremented when it happens. The clicking noise is about the same, perhaps a teeny bit quieter than the WD10EADS units.
*uch* No WD 2TB for me then.

casebuyer:

I'm considering the Seagate or Samsungs. Problem is, Seagate there track record is also not very good with the lasted serie. And the Samsung is relative new.
It's stupid !

I've an Hitachi T7K250 that have 3 years old, ~12000 hours running, ~2000 power on/off cycle but ~20 000 load/unload cycle... And it's still running OK.

But one things to now is that Hitachi have a similar system than the WD's NoTouch(tm) since several years.

This system may have dramaticaly increased reliability of the unloading head feature. And it's why WD let idle3 at 8 seconds on newer hardware revision but stopped to count it as an unload cycle.

If it was really a problem, WD had added the fix on newer model. It makes sense.

As far as I know, Samsung is the manufacturer with the lower reliability so...

shunx
Posts: 341
Joined: Sun Oct 06, 2002 1:20 pm
Location: Vancouver

Post by shunx » Sun Apr 26, 2009 1:46 am

I just received this email response from WD's support after bringing up the discussion here:

This information is not accurate. The unit will not reach its designed park lifespan due to the IntelliPark feature. No such DOS program [re: wdidle3.exe] has been created or provided by WD in order to enable/disable a firmware feature such as IntelliPark. The reason why this feature was created and placed on the drives, was in order to lower power consumption in order to be more eco-friendly. The unit will not encounter a sudden death due to this feature. Rest assured.

Monkeh16
Posts: 507
Joined: Sun May 04, 2008 2:57 pm
Location: England

Post by Monkeh16 » Sun Apr 26, 2009 12:06 pm

darthpingoo wrote: I've an Hitachi T7K250 that have 3 years old, ~12000 hours running, ~2000 power on/off cycle but ~20 000 load/unload cycle... And it's still running OK.
12k? That all?
shunx wrote:I just received this email response from WD's support after bringing up the discussion here:

This information is not accurate. The unit will not reach its designed park lifespan due to the IntelliPark feature. No such DOS program [re: wdidle3.exe] has been created or provided by WD in order to enable/disable a firmware feature such as IntelliPark. The reason why this feature was created and placed on the drives, was in order to lower power consumption in order to be more eco-friendly. The unit will not encounter a sudden death due to this feature. Rest assured.
And thus, why I will not be purchasing a WD Green series. Ever.

whiic
Posts: 575
Joined: Wed Sep 06, 2006 11:48 pm
Location: Finland

Post by whiic » Mon Apr 27, 2009 6:37 am

My three GP's (all 1TB, two first generation 00ZJB0 (4-platter) and one second generation 00D6B0 (3-platter)) are still working fine. Unload cycle counts are as follows:

GP #1: 0x873F7 = 553975 cycles, 0x2716 = 10006 hours
GP #2: 0x54189 = 344457 cycles, 0x2325 = 8997 hours
GP #3: hidden ("0x4C"), 0x1516 = 5398 hours

Bad sectors on all HDDs = zero. (No reallocated, no events, no pending, no uncorrectable, no other errors in SMART. All pass offline test and extended self-test. No Windows Event Viewer errors.)

Samsungs... well, they don't have unload count issues.

F1 7200rpm #1:
OS drive, just recently one uncorrectable sector appeared. Some odd crashes (but might be unrelated as well). Offline datacollection routine fails with fatal error. But now the uncorrectable sector disappeared... odd. And it's NOT displayed as reallocated either.

Very low vibration sample. :)

F1 7200rpm #2:
High/medium vibration sample. "Retired" to Topfield PVR because it made my computer case resonate.

F2 5400rpm:
Flaky sequantial transfer rates from the first verify pass. No problems with offline datacollection or SMART selftests. It's just read or verify passes that are jagged. Disabling indexing helped a lot but there's still downward spikes on the same areas, just not as big as indexing turned on.

Speed (outside these downward spikes) is equal to that of F1 7200rpm. 105 MB/s from a 5400rpm drive. 500GB platters are wheeee~. And with only 2 of them at low rpm, it's quieter than my 3 and 4 platter (250GB or 333GB per platter) GreenPowers. Noise character is also smoother.

But, I still feel a LOT more confident with my WDs. Samsungs tend to die in VERY odd ways. This uncorrectable error stuff is odd. Why don't they ever reallocate? Always the friggin' uncorrectables and bad sectors always visible to OS level. Or maybe the reallocated sectors are just hidden from SMART (despite the attribute being still present even though unused). Samsungs scare the shit out of me. If I have to decide whether I like to have hidden bad sectors or hidden unload counts in my SMART reports, I pick the latter any day.

I think the GP issue is way blown out of proportion. Trying to reduce cycles is wise (i.e re-configure SpeedFan to disable tempeture monitoring on GreenPower HDDs) but to boycott the drive and favour other manufacturers (with their own, bigger reliability issues) is just plain stupid. If you pick Samsung because it's cheaper, you're just curious, because you're unsure what density will the 1TB WD be (as we may never know when they go from 3 to 2 platters), they are all valid reasons (and they ALL played a role why I bought F2 EcoGreen) but I would definitely not favour Samsung (let alone a noisy and potentially firmware suicidal Seagate) over WD due to reliability concerns.

ult
Posts: 2
Joined: Sat Apr 25, 2009 7:40 am
Location: US

WD5000BEVT

Post by ult » Tue Apr 28, 2009 1:17 pm

Since we're all talking about similar tweaking issues, I thought I'd share my experience:

I've had two of the WD5000BEVT. Yes, they were quiet while running (other than the whoosh sound of alternating waves) but that was a minor standard drive activity noise to me compared to other manufacturers' seek noise.

Out of the two WD5000BEVT I've had, one vibrated substantially and the other didn't vibrate at all during use. Both were same models. Both manufactured in Jan 09 and at the same plant. (characters after the dash in model number dictated the site I'm told, mine were 11ZAT) This sounds great and I wanted to love these very fast performing but cooler temperature WD drives...

However, what caused me to return *both* of them was the head park noise (audible tick clunk) that occurred roughly every 4-8 seconds of inactivity (aka idle3 mode accounts for the heads to follow their park/unload routine).

Couple that with the Mac's aggressive drive power saving characteristics (google it and you'll undoubtedly read more about it elsewhere) and you have a symphony of idle, tick, activity, idle, tick,
activity, idle, tick.... you get the picture. Consequently, they were all experiencing very high load/unload counts from the SMART data fields *everytime* it parked. I mean, very high.

Ok, off to try to resolve this...

Years ago WD released a DOS executable (wdidle3) for older generation drives that increases the idle3 timer or threshold to accommodate these setups. Using a vendor specific feature set command, it modifies the drive's parameters to wait longer before invoking Idle 3. You'll also hear about the RE2IdleModeUpd utility that performed this for the WD1000FYPS, 7500 and 7501 drives.

This is *not* to be confused with WDSpinup which is designed to lower the spinup current so as not to overload USB power specs at once. (has nothing to do head park noise)

This is *also not* to be confused with the jumper setting that enables spread spectrum clocking (SSC). (which decreases EMI and addresses crosstalk concerns in multi-drive installations. Think server.)

This is *also not* to be confused with modifying the AAM (Automatic Acoustic Management) to change the drive's acoustic behavior. As my drives were set to enabled and maximum performance (FEh), changing this setting to enabled and minimum noise (80h) didn't affect the time/noise before the next head park tick clunk. The ATA spec issues this change for vendors to tweak activity noise/performance not the noise related to movement of the heads to park or idle position.

Okay, off to try hdapm, which is a Mac utility that prevents the drive from spinning down by issuing an ATA subcommand to change the drive's APM (Advanced power management). Great! No head park noise! But then it effectively disables the power management and the head doesn't park but at the cost of heat and power usage. Another but is that it needs to be run at each OS boot as the setting isn't static. There are very easy ways to do this on a mac using a tool called lingon but I won't get into that here. Ultimately the hdapm tool does work to eliminate the noise but, again, at a cost of heat/power usage.

Okay, so either disable the power management (hdapm worked just fine) or use the WDIDLE3 utility. Well, I was not able to get the DOS mode utility to run on a freedos boot disk on a dell laptop as it was an older machine that didn't offer a native SATA port on the motherboard thus DOS wasn't able to have complete low level access to the SATA/USB connected drive. Yes, I've spent many hours and many different versions of boot disks and drivers to no avail.

Okay, with FREEDOS on a macbookpro, I was able to run the WDIDLE3 utility and it saw the drive.

The commands are:

/S[<Timer>] Set timer, units in 100 milliseconds (1 to 255). Default=80.
/D Disable timer.
/R Report current timer.


As many other Macbookpro owners may find (and other PC owners with certain sata chipsets), the utility only works to allow you set the timer from 1 to 255. (aka 25 seconds) But the DISABLE timer switch
doesn't work via this setup. If I knew someone who had a newer PC desktop or PC laptop that natively had SATA onboard, then I'm confident the /D switch would work, but not via the Macbookpro's
implementation. If you live near DC and would like to help, let me know.


Alternatives that people have tried but to no avail so far:

* WD to acknowledge issues with these drives in certain setups (mac) and issue a Mac firmware utility to allow customers to change the idle3 setting on the newer Scorpio 2.5 drives. (it works in other setups, but WD officially doesn't advise using this utility on newer drives as they've since removed it from their site and tech support is tough about releasing it as well )

* Apple would maybe step in and resolve the aggressive drive issue (linux suffers from this as well from what I've heard)

* Switch to XP

* Switch drive vendors. Sigh.

So, either live with it and try to convince yourself that the tick clunk noise and load/unload counters aren't important until you get near 600,000 or switch manufacturers. After countless, countless, countless hours trying to get the wdidle3 /D switch utility to work on these unique setups, I've resorted to saving my frustration and what's left of my sanity by buying the seagate momentus 500g 5400.6 . I hope the 5400.6's head park tick clunk noise isn't as bad as WD.

I'm sure in your setup your WD drive's head park clunk noise isn't an issue. Your mileage may vary. Mine certainly did.

Monkeh16
Posts: 507
Joined: Sun May 04, 2008 2:57 pm
Location: England

Re: WD5000BEVT

Post by Monkeh16 » Tue Apr 28, 2009 2:14 pm

ult wrote:But then it effectively disables the power management and the head doesn't park but at the cost of heat and power usage.
Both of which are frankly fairly insignificant.

ult
Posts: 2
Joined: Sat Apr 25, 2009 7:40 am
Location: US

Post by ult » Tue May 05, 2009 3:40 pm

as a followup to my previous scorpio blue 500g 2.5 drive posts, I'm starting to see evidence that WD's quality control may be a bit off. The store was out of stock on the seagate momentus 500g so i got another (third) WD 500g blue 2.5.

Turns out, the third WD drive didn't vibrate at all during use. (vibration was a problem for the first two I had). Which in turn I've noticed the drive is appears to be operating 2F degrees cooler and substantially quieter during use (seek and spin noise). Keep in mind it still had the clunk noise from the heads parking every 8 seconds. At least now I have a WD5000BEVT that appears to run "smoother" than the previous 2 I owned. All had the same model and batch numbers. All made in same factory. Just 1 out of 3 ran stable w/out excessive vibration.

As I previously mentioned, I wasn't able to get the WDIDLE3 utility to successfully issue the /D switch via DOS boot disk on a macbookpro or on a non-sata Dell laptop. I was able to get it to perform just fine when I connected it to a friend's desktop w/native sata onboard. I simply booted into DOS (via freedos bootcd) and issued the wdidle3 /D switch to disable the idle3 timer. Successful.

Now that the drive has been placed back in my macbook pro, no more clunk attributed to the heads parking thanks to the timer having been disabled.

Had wdidle3 not worked, I would have just used HDAPM until I could have been able to replace the drive with a different make. (seagate)

Way too much time, trouble & aggravation spent on fixed this WD scorpio's model idle timer head park noise in this setup. Way too much.

dt1000
Posts: 11
Joined: Wed Jun 18, 2003 6:12 am

Post by dt1000 » Tue May 05, 2009 11:15 pm

Oh... this doesn't sound good.

I just got a 1TB WD10EADS for my media centre PC.
Unfortunately, I can't find any utilities on the WD site to turn off the head parking... (Am I just being dumb?)

Does anyone know if this problem has been solved for this particular drive?

:(

Dan

Ralf Hutter
SPCR Reviewer
Posts: 8636
Joined: Sat Nov 23, 2002 6:33 am
Location: Sunny SoCal

Post by Ralf Hutter » Wed May 06, 2009 12:54 pm

dt1000 wrote:
Does anyone know if this problem has been solved for this particular drive?
Nobody actually knows if there even is a problem.

See this recent thread for a fairly succinct synopsis of the uselessness of this "Problems with Head Parks?" thread.

dt1000
Posts: 11
Joined: Wed Jun 18, 2003 6:12 am

Post by dt1000 » Wed May 06, 2009 8:33 pm

Thanks for the rumour control! :)

I'll run my drive for a month and let you know what I find...

Cheers,
Dan

NickBurns23
Posts: 4
Joined: Thu May 07, 2009 2:18 am
Location: NYC

Post by NickBurns23 » Thu May 07, 2009 2:24 am

Here's the deal. I want to buy a 1.5TB Western Digital Green drive. I don't care about whether or not the head parking issue actually harms the drives. I just don't want them to park every 8 seconds. I notice WD only offers this software to make adjustments to certain models, and I don't see the 1.5TB drives listed. Is there software from WD that will allow me to adjust the head parks from every 8 seconds to 25 on the 1.5TB drives?

dhanson865
Posts: 2198
Joined: Thu Feb 10, 2005 11:20 am
Location: TN, USA

Post by dhanson865 » Thu May 07, 2009 11:43 am

NickBurns23 wrote:Here's the deal. I want to buy a 1.5TB Western Digital Green drive. I don't care about whether or not the head parking issue actually harms the drives. I just don't want them to park every 8 seconds. I notice WD only offers this software to make adjustments to certain models, and I don't see the 1.5TB drives listed. Is there software from WD that will allow me to adjust the head parks from every 8 seconds to 25 on the 1.5TB drives?
Based on a message earlier in this thread the tool is not officially available for download because

1. It wasn't designed for modern drives
2. It may or may not have desired or undesired effects
3. It will modify settings in the firmware thus voiding the warranty on the drive.

You can search the web for a copy but you use it at your own risk as WD doesn't recommend you do so.

I'd suggest you get a WD green and not worry as it has a 3 year warranty or

get a WD blue with a 3 year warranty or

get a WD black with a 5 year warranty.

You'll save power with the green drive and it will be quieter.

Your initial purchase price will be lower on the blue but it will use more power in the long run.

The black will be about the same price as the green and have a longer warranty but it will use more power (about the same as the blue).

NickBurns23
Posts: 4
Joined: Thu May 07, 2009 2:18 am
Location: NYC

Post by NickBurns23 » Fri May 08, 2009 2:19 am

Thank you for the advice, however:
dhanson865 wrote:
I'd suggest you get a WD green and not worry as it has a 3 year warranty or
Well data loss is more of a worry to me than the cost of replacing a drive. I know "back up regularly" but it seems a bit ridiculous to have to by TWO hard drives for every one I want to use.
dhanson865 wrote: get a WD blue with a 3 year warranty or

get a WD black with a 5 year warranty.

You'll save power with the green drive and it will be quieter.

Your initial purchase price will be lower on the blue but it will use more power in the long run.

The black will be about the same price as the green and have a longer warranty but it will use more power (about the same as the blue).
I would be happy with getting the black or blue drives but WD doesn't make them in 1.5TB for some strange reason.
I'd also like the green drives because I want to use them for an HTPC/media server and the low PSU in the computer coupled with the heat and the power that will be running all day, I'd rather use green drives.

However, I am not planning on using the green drives for the operating system. I am using a solid state disk for that. Does that mean the head parking thing is a complete non-issue then? Is it any issue for drives being used just for high-def movie rips?

whiic
Posts: 575
Joined: Wed Sep 06, 2006 11:48 pm
Location: Finland

Post by whiic » Fri May 08, 2009 3:03 pm

"However, I am not planning on using the green drives for the operating system."

GreenPower might not be ideal for OS. Many think 5400rpm is inadequate for OS but I think that's just BS. The unload thingy might make it non-ideal for OS, though. Samsung F2 Ecogreen could be a good OS drive.

"I am using a solid state disk for that."

Or that. Or a laptop drive. All depends on demand for performance. Some cheal SSDs can be lousier than even low-rpm laptops in write-heavy applications.

"Does that mean the head parking thing is a complete non-issue then? Is it any issue for drives being used just for high-def movie rips?"

Let's go through an example:
[YuS-MaiWaiFu]_Jigoku_Shoujo_Mitsuganae_-_20_(H264)[1B7A67CB].mkv
It's a 720p, 23.967 Hz, 291 MiB, 24min 59sec.

So, average bitrate is 291/25 =11.64 MiB/min =0.194 MiB/s =199 KiB/s.

So, the video player needs around 200 kilobytes of data every second. This is on average, will vary quite a bit on what happens on screen. Good codecs with high compression with low loss will be quite peaky.

If your player requests data in 64 kilobyte chunks, it will request ~three chunks each second. When not much happens on the screen (outside high bitrate peaks), it still requests at least one packet per second. WD GreenPower unloads at 8 seconds. => Absolutely definitely no problem.

Constant idle = good.
Constant IO = good.
Regular or irregular access at 9+ second intervals = bad. (Though if it's 100+, the unload count doesn't increase alarmingly. The upper limit of "dangerous" interval is a lot more vague than the low limit which is anything above 8 seconds.)

Example #1 of bad access:
Linux. IO every 20 seconds. 3 cycles per minute. 1576800 per year.

Example #2:
SpeedFan. SMART polling every 60 seconds. 525600 per year.

Example #3:
Fileshare over slow connection. Assume 512 kbit/s upload bandwidth. Person named Jolly Roger seeds torrents. Files that are shared are located on several HDDs.

On the example, 90% of upload bandwidth is used by other torrents and occasional webbrowsing, etc. but one torrent is seeded from GreenPower. 10% of remaining bandwidth is 51 kbit/s =6 KiB/s.

If torrent program used 64 chunks of data requested (or if OS did that sized IOs), it would need to request ever 11 seconds to feed 6 KiB/s to peers over slow internet connection. 11 seconds > 8 seconds. Bad.

So, watching HD video at 300 KiB/s = good.
Downloading/uploading HD video at 6 KiB/s = bad.

This of course assumes that 64 KB chunk size. It been a while since I did some detective work on disk IO (SysInternals' Filemon?) during various tasks (idle, video playback, etc.). If I don't remember incorrectly 64 KiB was a typical request length. Back then I wasn't really interested in chunk size but what caused chunks to be read as I had problems getting HDDs spin down on idle. (NTFS background tasks plagued even USB HDDs. I tracked it to a third-party defragmenter.)

This thread could have been a huge benefit to the internet, had it not been filled with so much unwarranted FUD. Because what applies to GreenPower unloads apply to many laptop HDD unloads as well. And while laptop drives are quiet, unloads are not necessarily quiet at all (especially if rest of system is tuned down to the level of regular HDD noise).

Syph007
Posts: 2
Joined: Sun May 10, 2009 11:42 am
Location: Canada

Post by Syph007 » Sun May 10, 2009 11:49 am

Hey guys, just thought I'd post my experience. I did a lot of research before buying a drive in the 1TB range. What I found is that WD is still the most reliable and lowest energy usage by far.

I did use the Wdidle3 utility to turn off the parking, I mean why not. In my testing it using about 1 extra watt, yes thats it. Even using that extra bit its still more GREEN than the other options.

I wanted a lower power silent file server. This it what I built.

Shuttle K45 with ice genie cooler
E5200 CPU (power sipping)
2x 1TB WD10EADS drives with parking disabled

This setup uses about 75 watts at typical use.

My system is silent and stable and im 100% happy. Dont shy away from an excellent drive. I still dont know if the parking issue will cause a failure, but who cares if you can turn it off for no REAL penalty.

NickBurns23
Posts: 4
Joined: Thu May 07, 2009 2:18 am
Location: NYC

Post by NickBurns23 » Mon May 11, 2009 2:21 am

Syph007 wrote:H
My system is silent and stable and im 100% happy. Dont shy away from an excellent drive. I still dont know if the parking issue will cause a failure, but who cares if you can turn it off for no REAL penalty.
However, hasn't it been established the Wdidle3 utility only works on certain green drives, the 1.5tb not being one of them?

Syph007
Posts: 2
Joined: Sun May 10, 2009 11:42 am
Location: Canada

Post by Syph007 » Mon May 11, 2009 4:37 am

NickBurns23 wrote:
Syph007 wrote:H
My system is silent and stable and im 100% happy. Dont shy away from an excellent drive. I still dont know if the parking issue will cause a failure, but who cares if you can turn it off for no REAL penalty.
However, hasn't it been established the Wdidle3 utility only works on certain green drives, the 1.5tb not being one of them?
Ah if that's the case I guess that could be cause for concern. I only bought the 1TB drives so I can't comment.

NickBurns23
Posts: 4
Joined: Thu May 07, 2009 2:18 am
Location: NYC

Post by NickBurns23 » Wed May 20, 2009 5:23 am

Syph007 wrote:Hey guys, just thought I'd post my experience. I did a lot of research before buying a drive in the 1TB range. What I found is that WD is still the most reliable and lowest energy usage by far.

I did use the Wdidle3 utility to turn off the parking, I mean why not. In my testing it using about 1 extra watt, yes thats it. Even using that extra bit its still more GREEN than the other options.

I wanted a lower power silent file server. This it what I built.

Shuttle K45 with ice genie cooler
E5200 CPU (power sipping)
2x 1TB WD10EADS drives with parking disabled

This setup uses about 75 watts at typical use.

My system is silent and stable and im 100% happy. Dont shy away from an excellent drive. I still dont know if the parking issue will cause a failure, but who cares if you can turn it off for no REAL penalty.
I thought you are only able to increase the idle to 25 seconds from 8? How did you turn it off completely?

If I can turn it off completely on the 1.5TB green drives I'll buy one.

twoscoreandfour
Posts: 37
Joined: Mon Apr 16, 2007 7:33 pm
Location: Sydney

Post by twoscoreandfour » Fri May 22, 2009 2:00 am

Are the people experience issues due to Linux IO running ext3 with journal=ordered? This forces disk flushes fairly frequently, but as I understand file systems like ext4 and XFS flush much less frequently

twoscoreandfour
Posts: 37
Joined: Mon Apr 16, 2007 7:33 pm
Location: Sydney

Post by twoscoreandfour » Sat May 23, 2009 6:30 am

Additionally you could change the commit option when mounting the drive.

However, the vm will still write out dirty pages to disk every 30s, but this can be changed by upping the dirty_exprie_centiseconds value.

m0002a
Posts: 2831
Joined: Wed Feb 04, 2004 2:12 am
Location: USA

Post by m0002a » Sat May 23, 2009 7:40 am

The problem with the WD Green Drives and Linux is that Linux does not access the disk (when "idle") often enough, not that it accesses too often. Windows accesses disk at least every 8 seconds, which is why the drives do not constantly unload/load.

I don't know why everyone has such a hard time understanding this.

glassdarkly
Posts: 5
Joined: Wed Mar 18, 2009 9:41 am
Location: SG

Post by glassdarkly » Sun May 24, 2009 2:56 am

Would this be an issue for OSes like OpenSolaris or FreeBSD, or is it confined to Linux?

Post Reply