2TB WD green questions

[vibratingdonkey]

I'm planning to pick up one of these drives for storage in the next few days and have a couple of quick questions.

-does disabling the head parking make the drive function like any other drive without this eco feature? the head will still park when the computer shuts down or windows puts the drive to sleep correct?

-any negatives to disabling head parking other than power savings?

-is it recommend to run a program to thoroughly check drive health before copying over my files and if so which? I plan to keep a backup of vital files on my existing drives but i rather not find out the drive was bad after spending time copying stuff over.

thanks in advance

 

[Soulkeeper]

I wouldn't disable head parking, didn't even know that was possible.
you might be thinking of the issue they had with excessive head parking on the green drives ... I believe the newer firmware fixed this issue.
Think of it as a safety feature also, you want the head off the platters during idle periods incase you bump the computer or something.

As far as checking your drives, i'd partition/format then run a long smart self test (can take an hour or two). You can still use the drives while the test is taking place, just don't reboot.
This should be all you need to do, the low level or slow formats aren't really needed anymore as smart checks have essentially replaced this.

 

[Vinwiesel]

You don't need to disable head parking, except in some linux situations where it can constantly park the heads and wreck a drive. You probably do want to disable hdd hibernation/sleep in windows though, as this causes the annoying lag when your system tries to access a sleeping drive.
I have a WD10EACS with 160k+ load/unload cycles which performs fine, but was only ever used in windows. Oddly, my WD10EADS has very few load/unload cycles and is nearly the same age. The EACS drive originally came from an external hdd, used internally for storage, and then was partitioned for a windows XP OS drive for part of its life.

 

[murphyc]

As far as checking your drives, i'd partition/format then run a long smart self test (can take an hour or two). You can still use the drives while the test is taking place, just don't reboot.
This should be all you need to do, the low level or slow formats aren't really needed anymore as smart checks have essentially replaced this.

This is not entirely true. If a sector goes "bad" there are a number of reallocation types possible, including the kind which isn't reallocated at all except when the sector is written to. Zeroing a drive (or ATA Secure Erase) will remove such bad sectors, whereas the extended offline SMART test will not.

 

[nervx]

You don't need to disable head parking, except in some linux situations where it can constantly park the heads and wreck a drive. You probably do want to disable hdd hibernation/sleep in windows though, as this causes the annoying lag when your system tries to access a sleeping drive.

I have two hard drives right now with the win7 power setting at balanced and the sleep option set to never. If I access the d drive drive after say 30 minutes it takes a couple seconds before I can browse the files. Why does it do this? Is it somehow going into sleep mode or is the drive simply spinning down?

Also with hdtune it shows a massive downward spike down to 10MB/s right near the beginning of the drive even thought the smart settings claim its healthy. Any idea whats causing the spike?

 

[murphyc]

Also with hdtune it shows a massive downward spike down to 10MB/s right near the beginning of the drive even thought the smart settings claim its healthy. Any idea whats causing the spike?

If there are bad sectors on the disk, the drive ECC will try to re-read and even reconstruct data in those sectors. This takes time, and reduces performance, and does not change the health of the drive in SMART.

You have to run an extended offline SMART test (smartmontools), and then look at the complete SMART attribute listing to see what specific issues the drive may be having. The health status of a disk as PASS, FAILING, or FAILED are next to useless.

 

[Smoblikat]

Only disabling head parking if its a boot drive 9it shouldnt be)

I always do an HDtune full scan on a drive if its for storage, let it run overnight while you sleep. If its all green then proceed with copying your files, ANY red and you send it back.

 

[Soulkeeper]

This is not entirely true. If a sector goes "bad" there are a number of reallocation types possible, including the kind which isn't reallocated at all except when the sector is written to. Zeroing a drive (or ATA Secure Erase) will remove such bad sectors, whereas the extended offline SMART test will not.

I don't see any reason to "remove bad sectors" on a brand new drive.
If I see many reallocations or uncorrectable errors in my smart log for a brand new drive it's usually a sign it needs to be RMA.
In the old days we did not have this functionality.

 

[coffeejunkee]

You don't need to disable head parking, except in some linux situations where it can constantly park the heads and wreck a drive.

Are you sure? I have a WD20EARS under Windows and the head parking value was already at a very high number after a few days of use.

Also, you don't really disable the parking, you set it to happen after a longer time than the default 8 seconds. Search for WDIdle if you want to do this.

 

[murphyc]

I don't see any reason to "remove bad sectors" on a brand new drive.

Pretty rare indeed. And the Google Failure Trends in a Large Disk Drive Population study indicates that once sector reallocation events start to occur, they're more likely to occur, and the drive is also more likely to fail.

But short of a disk being dropped from a crane I don't see how you get a disk upon arrival that has failing sectors. That'll happen in days or weeks later, and that just requires vigilance (preferably smartd running regular extended offline tests). The SMART conveyance test is specific to that of indicating transport related damage so that should be run for sure. And if you're a bit neurotic you could do several "burn-in" extended offline tests, and write zeros, before committing data to a drive. If there are any reallocation events in SMART attributes return the disk - if the manufacturer or retailer will accept it. I'm only suggesting they're not obligated to because the disk is still functioning fine until it's not. That there is a statistical probability of failure, that there are a tiny fraction of bad sectors, is not a warrantyable offense. But if you can get a drive swapped, great.

If I see many reallocations or uncorrectable errors in my smart log for a brand new drive it's usually a sign it needs to be RMA.

I think it depends on the error and magnitude. Again from a consumer point of view I'm inclined to agree, but the manufacturer has a different point of view. A disk with a dozen reallocated sectors will still have a PASSED status in SMART.

In the old days we did not have this functionality.

We had bad block maps supported by some file systems. NTFS still has this legacy feature, and it can still be leveraged with the full long format and by chkdsk with the /r flag. But with a modern disk, that check is sorta useless because the drive firmware will mask any sector error except the ones it can't fix and also submits a sector read error back to the file system. At that point the data is simply corrupt - the /r flag will cause the file system to recover what it can, and move it to another sector (LBA changes, physical sector changes). You'll end the delays resulting from disk firmware trying to recover data from that bad sector, over and over, but the data is unreliable.

With RAID, even software mdraid on Linux, the data from a bad sector is automatically reconstructed, and overwritten to the bad sectors causing the disk firmware to reallocate (LBA stays the same, physical sector on disk changes).

 

[thelastjuju]

The WD Greens still head park?

There seems to be all sorts of contradictory information regarding this dreaded "feature" (I call it more of a flaw)

Some say its only the older generation of WD Greens that do this, while others claim that even newer WD Greens AND some WD Blues do it too.



All I can say is that the whole "energy saving" stuff is pure marketing gimmickry. Hard drives idle at like 5 watts. That's not even $5 worth of energy A YEAR if ran 24/7, 365... its almost laughable. Parking the heads will do nothing but age the drive at a faster rate.. explaining why only a select few people have WD Greens that lasted over the 2 year period they are warrantied for.

 

[Smoblikat]

The WD Greens still head park?

There seems to be all sorts of contradictory information regarding this dreaded "feature" (I call it more of a flaw)

Some say its only the older generation of WD Greens that do this, while others claim that even newer WD Greens AND some WD Blues do it too.



All I can say is that the whole "energy saving" stuff is pure marketing gimmickry. Hard drives idle at like 5 watts. That's not even $5 worth of energy A YEAR if ran 24/7, 365... its almost laughable. Parking the heads will do nothing but age the drive at a faster rate.. explaining why only a select few people have WD Greens that lasted over the 2 year period they are warrantied for.

There are no WD blues that do it, thats fact. And head parking can be disabled, but if the drive is being used for its intended purpose then head parking is desirable. It puts less strain on the moving parts and reduces the chance that microscopic bits of platter get scratched off and break the drive.

 

[murphyc]

I've yet to hear of a definitive case where a drive died as a result of actuator failure. WDC support also told me this. The click being annoying is a real problem, but so far the notion that it's bad for the drive appears to be completely imaginary.

 

[StarTech]

I have a system with 2x1TB and 2x2TB greens that has been running with no problems for a couple of years. The W7 partition is on one of the 1GB drives which probably horrifies some people. On the other hand I have SSDs on my notebook.

 

[coffeejunkee]

The WD Greens still head park?

There seems to be all sorts of contradictory information regarding this dreaded "feature" (I call it more of a flaw)

Some say its only the older generation of WD Greens that do this, while others claim that even newer WD Greens AND some WD Blues do it too.

I have a 1TB EVDS, a 1TB EARS and a 2TB EARS and they all did the headparking thing. Can't speak for the latest EARX model though.

 

[weevilone]

I have a 1TB EVDS, a 1TB EARS and a 2TB EARS and they all did the headparking thing. Can't speak for the latest EARX model though.

I just received a brand new 2TB EARX directly from WD yesterday. It was manufactured 8/15/12 and it most definitely was set to park the heads at 8 seconds.

 

[bryanW1995]

The head on my wd20ears has been parked for about 5 months. At least, I assume that the head is parked when they're unplugged...

 

[Vinwiesel]

I have two hard drives right now with the win7 power setting at balanced and the sleep option set to never. If I access the d drive drive after say 30 minutes it takes a couple seconds before I can browse the files. Why does it do this? Is it somehow going into sleep mode or is the drive simply spinning down?

The way I understand it, if there is a lag to access the drive it is because it was off and has to spin back up. If they are idle but still spinning, access is near instant. My older system seemed to still turn off the hard drives despite being set to never after I changed from nvidia drivers to the generic windows sata drivers. (added an SSD) I would try setting the time to 120 minutes or so, and see if they still turn off shorter than that. In my case the "never" setting didn't take effect until after a reboot.

 

[vibratingdonkey]

can anyone with a western digital RED drive verify that you can disable TLER and if it stays off after restarting? I've read conflicting reports.

I just received a brand new 2TB EARX directly from WD yesterday. It was manufactured 8/15/12 and it most definitely was set to park the heads at 8 seconds.

what model is your drive? rumor is some greens are showing up with 1TB platters.

 

[StarTech]

I have a system with 2x1TB and 2x2TB greens that has been running with no problems for a couple of years. The W7 partition is on one of the 1GB drives which probably horrifies some people. On the other hand I have SSDs on my notebook.

FWIW they are WD10EADS and WD20EARS. Never done anything to alter head parking.

 

[weevilone]

what model is your drive? Rumor is some greens are showing up with 1tb platters.

wd20earx-008fb0

Why is the forum zapping that to all lowercase when I post or try to edit it?

Anyways, this was an RMA replacement so I was relieved to get new and not refurb. Now if I wasn't still trying to recover a 1.5TB iTunes library from a cloud backup I'd be a happy man.

 

[Soulkeeper]

Pretty rare indeed. And the Google Failure Trends in a Large Disk Drive Population study indicates that once sector reallocation events start to occur, they're more likely to occur, and the drive is also more likely to fail.

But short of a disk being dropped from a crane I don't see how you get a disk upon arrival that has failing sectors. That'll happen in days or weeks later, and that just requires vigilance (preferably smartd running regular extended offline tests). The SMART conveyance test is specific to that of indicating transport related damage so that should be run for sure. And if you're a bit neurotic you could do several "burn-in" extended offline tests, and write zeros, before committing data to a drive. If there are any reallocation events in SMART attributes return the disk - if the manufacturer or retailer will accept it. I'm only suggesting they're not obligated to because the disk is still functioning fine until it's not. That there is a statistical probability of failure, that there are a tiny fraction of bad sectors, is not a warrantyable offense. But if you can get a drive swapped, great.



I think it depends on the error and magnitude. Again from a consumer point of view I'm inclined to agree, but the manufacturer has a different point of view. A disk with a dozen reallocated sectors will still have a PASSED status in SMART.



We had bad block maps supported by some file systems. NTFS still has this legacy feature, and it can still be leveraged with the full long format and by chkdsk with the /r flag. But with a modern disk, that check is sorta useless because the drive firmware will mask any sector error except the ones it can't fix and also submits a sector read error back to the file system. At that point the data is simply corrupt - the /r flag will cause the file system to recover what it can, and move it to another sector (LBA changes, physical sector changes). You'll end the delays resulting from disk firmware trying to recover data from that bad sector, over and over, but the data is unreliable.

With RAID, even software mdraid on Linux, the data from a bad sector is automatically reconstructed, and overwritten to the bad sectors causing the disk firmware to reallocate (LBA stays the same, physical sector on disk changes).

thanks for posting most of a wiki page for us
I don't have the energy to continue your conversation.

 

[StinkyPinky]

I just bought a WD 2TB Green drive for an external USB3 backup drive. IMO that's what they are best used for.

 

[Dessert Tears]

I have two hard drives right now with the win7 power setting at balanced and the sleep option set to never.

There's a separate Hard disk setting under Change advanced power settings. It's in minutes, but it seemed to accept the "Never" I typed in. You could try opening your case and listening to the drive. My storage drive makes a distinctive sound when it spins up.

 

[Joofarc2]

i have a 2 tb wd eacs ( western digitaL ) GREEN
ITS A SATA , MANY HAD POWER SUPPLY ISSUE
DRIVE GAVE UP LESS THAN 2 YEARS AFTER BUY...

DOENT PLUG ( NOT RECOGNIZE FROM USB TO IDE )
NOT RECOGNIZE DIRECTLY PLUG IN MOTHER BOARD