"After following this thread for some time, and checking the info in the main article about WDC's existing firmware fix for (only) the RE2 drive (WD1000FYPS) I've been trying to see if they even have firmware in development to address this issue on the WD10EACS."

How about wdidle3.exe? It has been around longer than RE2 GP firmware fix and some people have reported it being capable of disabling the feature.

"A major reason I bought these drives (I have 6 currently, two of which are RMA replacements, and I'm begin to understand why that might have happened)"

When did the two originals die? If you don't have information on load/unload counts at time of death, knowing the age they died (and assuming rate of cycles has remained the same) would give an approximate number.

I don't think 2 drives failing, even if they were possibly due to "old age" is proof of an issue but it is a start. If we get more reports on them dying after first few months of infant mortality, I would re-evaluate my opinion on seriousness of the problem. Your unload rates are quite higher than average. Even leaving SpeedFan polling once every minute would mean "only" 525600 cycles/year at 24/7 use. You have otherwise non-favourable software conditions for the drive longevity so while don't think monitoring HDDs temperature constantly is necessary, disabling monitoring wouldn't completely solve it either. It could make cut some of the cycles... naturally not full 525600 cycles since some IO will statistically occur during the 8 seconds of heads loaded... cutting for example one third of million cycles per year by disabling SpeedFan monitoring on your GPs could be possible, though.

HDD temperatures are quite predictable so you could just set HDD cooling fan to constant rpm (using SpeedFan or otherwise) and leave the temperature monitoring on for stuff like CPU. You don't have to shut the utility down (like I did until lately, when I noticed just configuring SpeedFan is enough).

__

As for something less problematic, Hitachi 7K1000.B is a good bet. It definitely is not as quiet as GreenPower or Samsung EcoGreen F1 (5400rpm) and probably not even as quiet as Samsung Spinpoint F1 (7200rpm) but in the time when Samsung has/had problems with some controllers, Seagates die like flies as their firmware corrupts at few months of age and WDs increase load/unload count in some software environments, I think Hitachi is probably the only "safe" one. But remember that statistically every drive model has it's share of failures. There's no invincible ones so always back-up anything worth anything. Capacity is cheap.

I have no experience with modern Hitachi. I don't know if someone else has. Not the best rep among silencers, probably mostly due to historical reasons. Today, they are probably a bit better than Seagates when it comes to noise, especially because Hitachi supports AAM and APM.

But I would try wdidle3.exe with your existing drives and see if you can disable unload feature as your software environment no doubt causes excessive cycles. 1M in one year = 5M in five = 17 times the minimum sustained cycle count. Even though it's "minimum" it's still a friggin' 17 times. At least try the utility.

Hi, guys.

I've stumbled upon this thread a while back, and I'm rather interested in it because I'm actually considering WD GP drives for the NAS I am planning.

I have some questions about this issue, though, to which I haven't found answers yet. Since some of you are owners, can you tell me:

1) Which OS are you using? From what I've read, it seems the high load/unload count is more common in Linux, but I can't find much data on Windows OSes, namely WHS.

2) Does the load/unload software "fix" (not the firmware upgrade, just the "sicekick application" WD put up) survive power-cycles? I'm guessing "yes", but I'd like to be sure.

3) Which consumer-grade drives are affected? Every 1st and 2nd-gen GP? Not the WD10EADS? How about the new WD20EADS?

4) Has anyone tried the new firmware on the RE2 GP drives? If so, I'm very much interested in the spin-up power figures with the "low power spin-up". The ~1.7A on the 12V rail seems a little too much for what I want (4 drives + mobo + underclocked/undervolted CPU on a PicoPSU, with Staggered Spin-up still being a big "?" from both the motherboard manufacturer AND WD).

5) As of now, the best choices to avoid this problem altogether are the RE2 GPs with the updated firmware and tweaked unload cycle, right?

Btw, I think I know why WD isn't paying the regular GPs too much attention. They'll probably change the firmware on future iterations of the consumer GP, just in case, but I don't think there will be a firmware release for the older ones.

Why? Well, because those are consumer-grade drives, which means they are not validated, nor expected to, be turned on 24/7, but at most 5 to 6 hours a day. If you pay attention, only GPs used 24/7 (or about that) seem to have insanely large load/unload cycles. Regular users have somewhat high counts, but well within expected parameters for the useful drive life (a.k.a., 5 years or less).

So, though it sucks big time (it IS an issue, and it should be dealt with), most likely WD will say "heh, that't not a server-validated drive. You shouldn't be using it 24/7 to begin with" to consumers. Which is why I'll probably need to start to ask quotes for the RE version...

In the meanwhile, if someone can prove what I just said wrong, PLEASE do. This is one of those few times I actually WANT to be wrong... hehe

Cheers.

Miguel

I have the 750GB GP model.

My load cycle count is at 20,978 (850 hours), but this is for my main system drive running Windows XP. Since I have had the drive for 8 months already, I do not think this is all that bad.

I may consider using the tool to turn this off, since the noise doesn't matter in my main rig. Thanks for posting this thread.

This is a bit worrysome. I turned the head parking completely off with wdlidle3.exe and still my heads unload at alarming rate:

21st Jan 20:23

28th Jan 13:31


That's around 550 parks per day, so it seems the wdidle3.exe does _not_ completely fix the problem.

zds, is that with an RE2 GP drive AFTER the firmware update?

If so, the problem might be deeper than we thought...

Cheers.

Miguel

No, this is with the original firmware:



Unfortunately the disk is in use in 24/7 server, so I have not had a chance to apply the firmware update yet.

This is more like an acknowledgement that wdidle3.exe is not able to fix a drive with bad firmware. I'll report about the firmware update once I have chance to take a look at it.

Actually, could you elaborate a bit? If there is a firmware update to WD10EACS somewhere, I have no idea where it is. Can you please point me to it? The idle mode update listed on WD site does not cover EACS: http://support.wdc.com/product/download.asp?groupid=609&sid=113&lang=en

please, take a look on this topic:

http://www.silentpcreview.com/forums/vi ... highlight=

That was exactly my question.

Since the update is currently only available for RE2 drives, and you said wdidle3.exe was not working for you, I was afraid the firmware update didn't work all that well.

But since that's a consumer 1st-gen GP, then that explains it.

Though you might want to try trxman's solution, and see if that helps.

Cheers.

Miguel

Unfortunately that trick does not work on Green Power:

Hmmm, that error code seems to indicate the normal firmware doesn't support APM states. Or at least changing them.

Not good, if that's true

Also, though I do understand why WD is keeping quiet about consumer drives - they are not meant to be used 24/7, after all -, it does seem a bit yucky of them to go out of their way and release an updated firmware for the RE2 GP versions of the drive, and not for the consumer-grade ones. The differences between the drives are fundamentally the firmware, after all, there are not that many physical differences that require that much of recoding on the firmware side of things. Sad.

Btw, can anyone confirm this issue also appears with the WD10EADS? Or is that one safe? Even if it doesn't, at least the new 2TB (and most likely the 1.5TB one, too, since they share the same components) already fixes it, and hopefully future smaller models will eventually get a new firmware.

Cheers.

Miguel

Yup. This information, btw, is lacking from the recent SPCR article about these matters: http://www.silentpcreview.com/Terabyte_Drive_Fix

It seems right now only the Raid Edition disks have a fix available, for regular Green Powers there's nothing that can be done.

Right ... I missed that detail. Give me a day or two, I'll update it.

That said ... I still think the evidence that this is a real, genuine problem for most users is very sketchy. This feature has been around in notebook drives for ages, with little impact on reliability that I'm aware of. It would be interesting to hear about the robustness of mobile vs. desktop drives from a vendor, but even then, I would expect the typically tougher handling conditions for notebook drives to have more of an effect than load/unload counts.

From what I know, for the people who travel with laptops the notebook drives die periodically, and there's no great fuss done about it, because, well, the conditions they have to endure are what they are.

So in notebook drive a bit increased risk of head wear and a bit decreased risk of head hitting the platter in a crash is a very good trade-off.

My smart data for my WD drives:
WD6400AAKS(0) Firmware: 01.03B01
Load/Unload Cycle Count Value: 200 Data: 475
Power-on Time: 2650

WD6400AAKS(1) Firmware: 01.03B01
Load/Unload Cycle Count Value: 200 Data: 475
Power-on Time: 2659

WD10EACS(0) Firmware: 01.01B01
Load/Unload Cycle Count Value: 199 Data: 5477
Power-on Time: 539

WD10EACS(1) Firmware: 01.01B01
Load/Unload Cycle Count Value: 200 Data: 1118
Power-on Time: 338


All seems fine, only had the 1TB GP drives fora few months so the value isnt worrying, i expect i'll relace the drives long before they get to worrying levels. Good to know of this problem though, WD would likely be happy not saying anything to warn or calm down customers.

HD Tune: WDC WD10EACS-00ZJB0
(09) Power On Hours Count 91 91 0 6713 Ok
(C1) Load Cycle Count 1 1 0 1283735 Ok

Been using it 24/7 for about 10 months, as my media files storage, on windows xp. I think my numbers are on the high side... What is the maximum recommended load/unload cycle of this drive for it to still working fine? Should i get another drive immediately to backup files on this drive?

I was thinking of expanding my storage capacity with one or two more WD10EADS (or WD20EADS if it is available within 1-2 months), so is that a bad idea now, seeing the numbers i get with my usage pattern?

WD has designed the drive to withstand 300k loads/unloads during the drive lifetime, so you are at four times that. However, there is no information about how much the risk of failure increases after that point, but I'd at least keep my backups up-to-date very dutifully in your situation.

So yes, you are on the high side.

Oh, s#$%, I thought that value had one less digit, so I though "heh, 100k parks is not all that much, since the drive is rated for 300k".

A bit on the high side, you'd only get about 2.5 years before the 300k ran out at that rate, if that were to be true.

However, one million+ parks in only 10 months is just absurd! I mean, those heads are CONSTANTLY parking and unparking... Sheesh!

One thing you can be sure of: those are high quality motors and bearings inside that HDD. They have endured in 10 months more than 3 times the stated max they were supposed to endure for the whole drive lifetime (at least 2 years, more likely 3 or 5) and seem just fine (provided you aren't starting to experience more rotational noise or odd sounds when parking/unparking).

It has been said already, park values usually don't mean anything per se, but you might start to consider to backup your data, just in case.

As for the EADS versions, I'm still waiting to know if the 1TB version is also affected with this issue (which would mean "no" to the buy if it is, unless you want to change it again in 10 months time). The 2TB drive, according to the release information, seems to have the new firmware version already embedded, because of the features it boasts, at least one of which is supposedly only available for post-firmware update drives.

However, until some of the guys here can actually get their hands on the WD20EADS and test it, it will be difficult to know that for sure. And this kind of problem usually appears in the wild after at least 6 months after release, so that's a pickle too...

Hope this helps on your decision. Cheers.

Miguel

My new 1TB WD EADS GP. Delivered Dec 18 2008.

Note: Start/Stop times is only 3 (Edit: Cycles) more than Power Cycle count.
Ratio seems to be about the same as for my old drive (data below).
Power Cycle time is probably about the same amount of times I have turned my computer on?
I do not think there is any problem with this hard drive.

My "old" 500GB WD RE2 drive.

This older drive is inside a Scythe Quiet Drive.
Since it's not a GP (low power) drive and it's enclosed it gets warmer.
Both are lying on foam in different compartments of my P180B.
My fans blow 400 RPM around them most of the time.

I run Windows Vista.

I get the same data in SpeedFan where the corresponding "Data" values are displayed in the "Raw" column with hexadecimal values.
They are the exact same values (checked by letting my calculator convert hex -> dec on them).

I might post values again after a month if you like.

I don't understand why is that drive stressed so extremely, the figure averages like 3 parks per min! I only store video files in it, just one local user and another user in the network accessing those files (quite frequently). Is it due to the nature of media players using buffering techniques to play video files?

I reckon that it'll be the same if i get another GP... if there is no way to change the park on idle time, or unless the time has been significantly increased in EADS. But i think i'll get one WD10EADS anyway, because my storage capacity is running very low, have to expand now.

zhiyung: "HD Tune: WDC WD10EACS-00ZJB0
(09) Power On Hours Count 91 91 0 6713 Ok
(C1) Load Cycle Count 1 1 0 1283735 Ok

Been using it 24/7 for about 10 months, as my media files storage, on windows xp. I think my numbers are on the high side... What is the maximum recommended load/unload cycle of this drive for it to still working fine?"

Like zds said, you are already way over the rating. And if you continue like this, you'd reach 7 million cycles by end of 5 year service life if WD doesn't release a firmware/utility to change the unload behaviour or if you don't tweak your system to reduce number of cycles.

"Should i get another drive immediately to backup files on this drive?"

Yes. But you should have got the back-up drive when you bought it. You should ALWAYS keep a back-up of any important data and that recommendation has NOTHING to do whether WD has a reliability issue or not. ALL HDDs fail sooner or later and you never know when it happens.

But would I be especially scared of your drive failing? No. At 1 year of age, HDDs are less prone to failure than HDDs in early life. Of course high unload counts can cause "old age" failures at a younger age...

"I was thinking of expanding my storage capacity with one or two more WD10EADS (or WD20EADS if it is available within 1-2 months), so is that a bad idea now, seeing the numbers i get with my usage pattern?"

Not necessarily a bad idea but you should try to find a way to reduce number of cycles via some method. Sure, I admit it's not necessarily easy: you can't always just reduce number of IO, and you can't configure the HDD unless manufacturer provides the tools. The fool proof solution would be to use a background task that makes uncached writes to a file at <8 intervals, preventing HDD from ever unloading. Of course you'd lose any power saving benefit from unload feature and these write operations would cause slight reduction in performance and sequential transfers will be occasionally interrupted (with added seeking and noise).

Alex: "Note: Start/Stop times is only 3 times more than Power Cycle count."

Don't say "3 times more" when you mean 3 cycles more. Assume X to be one value, 3 times X is same as 3X. If X was for example 12, it'd be 36. Three times MORE is X+3X =4X (or 48 if we assume X to be 12).

Three times as big is 3X. Three times bigger is 4X. Of course while this is mathematically valid, vast majority of ignorant retards consider three times bigger to be 3X.

"Ratio seems to be about the same as for my old drive (data below).
Power Cycle time is probably about the same amount of times I have turned my computer on?"

Power Cycle Count is the number of times computer has been turned on. Does not include soft reboots or hard resets.

Start/Stop Count = Power Cycle Count + times HDD has spun down due to use of STAND-BY command (Windows power management or firmware implemented HDD spindown timer). This attribute has nothing to do with the issue discussed in this thread.

The attribute we are discussing is
(C1) Load Cycle Count 200 200 0 64 Ok
and in your drive it appears to be the same as Start/Stop Count. This means that either unload feature is disabled in WD10EADS-00L5B1 or (like with WD10EACS-00D6B0) it's simply LYING.

The only way to determine if your drive is affected is to listen for load/unload clicks (they are distinguishable from regular seeks). SMART data is reliable method of determining load/unload count only for WDxxxxCS-00ZJB0.

"My "old" 500GB WD RE2 drive."

That is a CSS based drive. Regarless of what firmware it has, it CANNOT have the problem with unloads. It lacks C0 and C1 attributes in SMART.


______

My current stats:
00ZJB0 #1: 533687
00ZJB0 #2: 313680
00D6B0 #1: 44 (lying)

All of them working normally.

I've lately noticed that some of them start buzzing and continue that for several minutes, not unloading in 8 second idle. I don't think it's software so it must be some self-diagnostic run by the firmware. I don't know if it's just one of them doing the occasional buzz of if they all do it once in a while. It's not a loud sound so it can be easily masked by other noise which is probably why I hadn't noticed it before... that, and the fact that it only occurs a few times a day.

This is one of the nasty things there: the 10EACS actually is *more* likely to park heads if it's not heavily stressed. So media server use might very well be one of the worst usage patterns for the drives with this "thing".

This is also why I don't buy the "but GP is not meant to be used 24/7" argument. It's a desktop drive, and workplace desktop machines tend to be on 24/7, they are just not stressed most of the time, unlike servers. And this particular issue might actually hit *more* the machines that mostly just idle than busy servers that have some load all the time.

I'd say our home (media/email) server drives are actually stressed less than our desktop drives. And that's exactly the problem.

What's ironic is that GP is otherwise a perfect match for media servers - low power, quiet, large storage and speed is not that critical.



I second this. *Any* storage media can fail at any point of time. Never have just one copy of anything of importance to you.

Well, I've just read this whole thread in the vain hope that someone would know a way to adjust diskio's in Linux.

Running Conky I find that my drive is hit every ten seconds, all the time. (Intrepid) But what's hitting it? I'm not running indexing, so it might be write caching, but how to confirm?

I am in the process of ditching my rotten Seagate for a 2TB, but I want my system to run quiet and calm. Even if the 2TB has a workaround to this problem I'd like to stop my system accessing every ten seconds. I'm getting suspicious of ext3 journalling.

sda1 is /
sda2 is swap
sda3 is for Myth-recorded videos (NTFS)

This problem (with WD GP drives) was discussed on linux-kernel mailing list before this thread ever started. Try asking there.

Hi to everybody. I'm Italian, then sorry for my (bad) English.
I read that in this article is written that, for WD, the only models affected by the problem are RE2 GP. But this is not real! The Caviar Green are affected too!
I have a WD Elements in which there is a WD10EAVS-00D7B1 (a Caviar Green of 1 TB), and I hear the head that shut off after 10 seconds of inactivity. Also by a WD My Book Home that I tried (and in which there was a WD10EAVS-00D7B1 too) I heard the same head "clunks".
But is not ended here... I have also a WD6400AACS-00G8B1 which produces the same sounds.
For all these hard disks, the SMART shows the number of Load/Unload Head Cycle Count the same than the Start/Stop Count. I think so that ALL Caviar Green park the heads after 8-10 seconds, and, from a determinate firmware, WD have "fixed" the matter camouflaging the SMART data and doing to appear the number of Load/Unload Head Cycle Count the same than the Start/Stop Count.
However, I'm going to buy another hard disk of 1 TB, and, maybe cause I'm a bit masochist, it will be a WD10EADS.
Most Samsung's Terabytes have bad sectors, Seagate's Terabytes have a bricking firmware, and WD's Green Terabytes have the head problem: I think that is anyway the less bad choice (I excluded WD Caviar Blacks cause the noise).

any info about WD20EADS ?

is it affected with this horible WD shame too?

I have two WD10EACS 00D6B0
1. 4317 Hours on, 17 Load Cycle Count
2. 4316 Hours on, 15 Load Cycle Count


Also have 8 other drives, but unfortunately they are connected to RAID card that cannot read SMART data.

Don't know if they are lying as whiic is trying to suggest.

WDC WD10EACS


ID Descrizione attributo Limite Valore Peggiore Dati Stato
01 Raw Read Error Rate 51 200 200 0 OK: il valore è normale
03 Spinup Time 21 173 171 8341 OK: il valore è normale
04 Start/Stop Count 0 97 97 3240 OK: sempre riuscito
05 Reallocated Sector Count 140 200 200 0 OK: il valore è normale
07 Seek Error Rate 51 200 200 0 OK: il valore è normale
09 Power-On Time Count 0 98 98 2013 OK: sempre riuscito
0A Spinup Retry Count 51 100 100 0 OK: il valore è normale
0B Calibration Retry Count 51 100 100 0 OK: il valore è normale
0C Power Cycle Count 0 100 100 606 OK: sempre riuscito
C0 Power-Off Retract Count 0 200 200 65 OK: sempre riuscito
C1 Load/Unload Cycle Count 0 195 195 17776 OK: sempre riuscito
C2 Temperature 0 107 91 45 OK: sempre riuscito
C4 Reallocation Event Count 0 200 200 0 OK: sempre riuscito
C5 Current Pending Sector Count 0 200 200 0 OK: sempre riuscito
C6 Offline Uncorrectable Sector Count 0 200 200 0 OK: sempre riuscito
C7 Ultra ATA CRC Error Rate 0 200 200 69 OK: sempre riuscito
C8 Write Error Rate 51 200 200 0 OK: il valore è normale

All ok???

If the power-on time is correct (2013 hours ie. about 84 days), then the 18k loads/unloads sounds good. Would take almost 4 years at that speed to reach the 300k mark. Well, 4 years is still a bit short lifespan, but we are talking about increased risk here, not sure death. Definitely a lot better than my unit anyway.

Has your drive been used under kind of load that causes EACS drives to unload/load all the time? Ie. has it been under Linux, or some media server drive?

My western digital driver is on WDC Mybook Home edition.

It is connect to pc via firewire, on Vista x64.

If I connect this drive to esata port after 10 minutes don't work anymore.

OK ok 4 years! I have good sleep!