Silencing hard drives, optical drives and other storage devices
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
-
dan
- Posts: 1243
- Joined: Sun Jan 25, 2004 2:01 am
-
Contact:
Post
by dan » Thu Sep 08, 2011 2:30 pm
Western Digital Caviar Green WD20EARS 2TB 64MB Cache
got 2-3 stars out of 5
http://www.newegg.com/Product/Product.a ... 6822136514
Cons: Where do I start? So, here are the problems. 1) The drive uses a non standard 4K block size vs. 512 bytes, this throughs off many OSes like OpenSolaris 2) The drives "park" the heads every 8 seconds, so they park, unpark, park, unpark over 1,000 times per day in a normal environment. I had to write a program to ping the drives every 6 seconds to stop this... 3) Failure rate 3 out of 8 have failed in 18 months, this is too high... 4) performance sucks, 2 of mine read/write at 1MB/sec, thats the worst I've seen since 1990.
What is this park head the reviewer is complaining about and is it a serious problem? Do other HD have this problem? Why does WD engineer it to park the head so often?
-
m1st
- Posts: 132
- Joined: Sun Jan 31, 2010 6:43 pm
- Location: US
Post
by m1st » Thu Sep 08, 2011 4:48 pm
Simply, the head parking problem is an artifact of one of the features of the Western Digital Green line. Basically, every time a hard drive takes its reading arms off the platters, it ticks a counter. On normal hard drives, the only normal time the arms were taken off the platters was during power down and power up, so head parking readings became a good indicator of how many times the hard drive has been powered on and off. Counting how many times a hard drive powers on or off is important, as the most stressful time for a hard drive is during this operation.
With the Green line, Western Digital started parking the heads after a period of inactivity in order to save power. It still counted these head parking activities as before, however, so the Green hard drives showed way too many head parks, and many S.M.A.R.T. programs threw up red flags.
There have been plenty of theoretical discussions whether these head parks are actually causing undue wear on the hard drives or if it is just bad reporting, and the arguments go both ways. However, I haven't seen a large-scale study done to show either case. Maybe others can chime in with (admittedly very unscientific) anecdotal evidence? I have owned one WD Green and haven't had any problems with it.
-
nzdcoy
- Posts: 36
- Joined: Tue Feb 08, 2011 12:44 am
Post
by nzdcoy » Thu Sep 08, 2011 5:32 pm
1) The drive uses a non standard 4K block size vs. 512 bytes, this throughs off many OSes like OpenSolaris
afaik this is normal for all (currect) 2tb+ drives
2) The drives "park" the heads every 8 seconds, so they park, unpark, park, unpark over 1,000 times per day in a normal environment. I had to write a program to ping the drives every 6 seconds to stop this...
I remember reading that the problem lies mainly with *nix based systems because they only write to the drives every 10seconds or so-meaning it parks then unparks throughout the day
3) Failure rate 3 out of 8 have failed in 18 months, this is too high...
no comment.
4) performance sucks, 2 of mine read/write at 1MB/sec, thats the worst I've seen since 1990.
could misaligned partitions or the start-stop in *nix be making this problem? Ive never seen this low on mine
-
Dirge
- Posts: 111
- Joined: Sun Apr 11, 2004 8:55 pm
- Location: New Zealand
Post
by Dirge » Fri Sep 09, 2011 12:11 am
Personally I wouldn't use a WD Caviar Green as my main system drive due to the 8 second head parking. I have heard they may be dropped from RAID arrays if the controller thinks the drive has become unresponsive.
If you are using it for a data drive or as an external drive then you should be A OK.
The reviewers degraded performance is probably due to missaligned sectors. You can read about the 4k sector 'Advanced Format' on Western Digitals website
here.
-
quest_for_silence
- Posts: 5275
- Joined: Wed Jun 13, 2007 10:12 am
- Location: ITALY
Post
by quest_for_silence » Fri Sep 09, 2011 2:49 am
dan wrote:What is this park head the reviewer is complaining about and is it a serious problem? Do other HD have this problem? Why does WD engineer it to park the head so often?
I don't think it's a real problem, however, for a tad more information, give a look
here.
-
dan
- Posts: 1243
- Joined: Sun Jan 25, 2004 2:01 am
-
Contact:
Post
by dan » Fri Sep 09, 2011 7:17 am
quest_for_silence wrote:dan wrote:What is this park head the reviewer is complaining about and is it a serious problem? Do other HD have this problem? Why does WD engineer it to park the head so often?
I don't think it's a real problem, however, for a tad more information, give a look
here.
yeah that's it.
does samsung have this problem?
-
quest_for_silence
- Posts: 5275
- Joined: Wed Jun 13, 2007 10:12 am
- Location: ITALY
Post
by quest_for_silence » Fri Sep 09, 2011 1:31 pm
dan wrote:does samsung have this problem?
Again, it's not a "problem". At anyway, about Samsung, I dunno.
-
markanini
- Posts: 163
- Joined: Tue Jan 21, 2003 9:59 am
- Location: Malmo, Sweden
Post
by markanini » Fri Sep 09, 2011 9:51 pm
Here's a WD knowledge base article regarding the issue:
http://wdc.custhelp.com/app/answers/det ... X1ZJRGs%3D
Basically check the 'Load unload cycles' with a SMART monitoring utility. If it's in the several thousands you might want to disable the head parking feature with the 'wdidle3' DOS program linked in the article. I did that on my WD15EADS just to be on the safe side but it seems hard to say whether it's really necessary or not.
-
xan_user
- *Lifetime Patron*
- Posts: 2269
- Joined: Sun May 21, 2006 9:09 am
- Location: Northern California.
Post
by xan_user » Sat Sep 10, 2011 7:25 am
For me the bothersome thing about head parks is the noise it makes while parking.... It sounds kinda like scissors snipping up a manila folder. it was the loudest part of my system till i disabled the feature.
-
loise1996
- Posts: 1
- Joined: Tue Feb 07, 2017 3:17 am
Post
by loise1996 » Tue Feb 07, 2017 3:26 am
m1st wrote:Simply, the
head parking problem is an artifact of one of the features of the Western Digital Green line. Basically, every time a hard drive takes its reading arms off the platters, it ticks a counter. On normal hard drives, the only normal time the arms were taken off the platters was during power down and power up, so head parking readings became a good indicator of how many times the hard drive has been powered on and off. Counting how many times hard drive power on or off is important, as the most stressful time for a hard drive is during this operation.
With the Green line, Western Digital started parking the heads after a period of inactivity in order to save power. It still counted these head parking activities as before, however, so the Green hard drives showed way too many head parks, and many S.M.A.R.T. programs threw up red flags.
There have been plenty of theoretical discussions whether these head parks are actually causing undue wear on the hard drives or if it is just bad reporting, and the arguments go both ways. However, I haven't seen a large-scale study done to show either case. Can maybe others chime in with (admittedly very unscientific) anecdotal evidence? I have owned one WD Green and haven't had any problems with it.
Hello m1st,
I agree with you.But no, an expansive number of drives stop their heads consequently paying little heed to what you set in the power choices, and they utilize pointlessly forceful conduct like stopping after just 8 seconds of idleness. I have a major issue with this in light of the effect it has on the drives acoustics and responsiveness. I'm attempting to make sense of if there is any issue with steady head stopping in applications where those two elements are not an issue.
-
quest_for_silence
- Posts: 5275
- Joined: Wed Jun 13, 2007 10:12 am
- Location: ITALY
Post
by quest_for_silence » Tue Feb 07, 2017 11:24 am
loise1996 wrote:Hello m1st,
I agree with you.
Are you aware you're bumping up a six (6!) years old thread, and you're answering to a user whose last visit here is dated three (3!) years ago?
That's rather pointless...