DSM-G600, DNS-3xx and NSA-220 Hack Forum

Very good question and one that is apparently overlooked by many, many folks - and incidentally there is a whole debate over the same issue, from a slightly different angle over at the D-Link forum.

Every time a disk stops spinning, the heads come to rest on the disk surface so that stopping & starting the disk causes abrasion to both the head and the platter, so yes, if the disks are stopping and starting every few minutes, they "wear" faster, potentially reducing the life of the drive.  Each start/stop is known as a load cycle and a drive might have a design load cycle life of 300,000~600,000 cycles.

Now - fire up your favourite search engine and research WD (Western Digital) and load cycle count - and see what you find.  The reason why I'm specifying Western Digital is not because there drives are better or worse in this regard, but because I know you'll find a lot of reading material.

Drives have been designed with head landing zones for years.  I don't concern myself with drive spin-up spin-down wear.

If I used my DNS-323 as primary data storage and accessed it constantly, I'd probably just let it spin.  I use it more for backup, archives, and long term storage.  So I will have it spin down the drives after 20 minutes of inactivity.  This allows the drives to remain up and running most of the time we're periodically accessing it, but it goes to sleep at night.

I'm sure I'll outgrow the drive's capacity before it fails due to wear and tear.

abubin wrote:

Hmm...this is something new to me. I always thought that when the HDD go to sleep, the head will always go back to lock state in sector 0 or some place where no data is there. If I remember correctly, the head mechanism is designed such that when drive spin down, the head will go back to rest. Just like your inkjet printer go back to it's resting place out of harm's way when there are no printing.

Oh..well, time to do some googling and reading then.

Aaah - so easy to get confused - on most disks in current production, the heads come to rest on the platter when the platter stops rotating.  Where the head comes to rest and whether or not there is data there is of no relevance - what I'm describing is the wear that occurs on the heads them selves each time they land and take off.

Imagine an aeroplane with no wheels, it's coming in to land at a specially prepared airport where there are no buildings to get damaged when the plane lands - this is the equivalent of your disk head coming to a stop in an area where there is no data.  Just as the aeroplane without wheels to rotate when it strikes the airport runway is likely to be damaged, the disk head will sustain some wear with each take off and landing.

mixle wrote:

I was just reading about the Western Digital Green 1TB drives I'm using and it appears not to be affected by cycle wear.

I'm sure other drives will start using this also...

"NoTouch ramp load technology -  The recording head never touches the disk media ensuring significantly less wear to the recording head and media as well as better drive protection in transit."


Jason

What model of WD Green 1TB are you using?  As far as I'm aware the Western Digital drives at the center of the load cycle count controversy are 1TB green drives.

Look here.

I'm running two 500GB's of these and it runs cool and almost no noise.
Perfect for a NAS if you ask me.

fordem,

What you describe is a 'head crash', which is VERY bad.  The smallest of dust particles fit between the head and the platter much like you can fit a bowling ball under a door.  There is a very slim margin there, and if the head contacts the drive, it's essentially toast.  Maybe not right away, but impending doom is imminent.

For this reason, when a drive hibernates or the host system is otherwise shut down, the heads are parked off to the side as in the printer analogy above.

More info:
http://en.wikipedia.org/wiki/Head_crash

Here is where a drive will park the heads (sometimes referred to as the landing zone):
http://upload.wikimedia.org/wikipedia/c … g_zone.jpg

Notice that the landing zone is WELL AWAY from the platter, so the head will not touch the platter.

Older drives would physically rest the head on the platter, but no one has manufactured a drive that behaves that way since like 1995.

Yet more info:
http://www.storagereview.com/guide2000/ … eight.html

DasCrushinator, set hibernation to whatever you wish - the concept here is such that the faster the time-out, the more power you will conceivably save in the long run, but you'll likely have to wait more frequently for the drive to come OUT of hibernation.  Personally, I set this to 30 minutes, since once I get going with something, I usually don't stop and don't like lag.  That's just me though

Hope that helps!

Last edited by DigitalJer (2009-06-19 00:52:46)

DigitalJer wrote:

fordem,

What you describe is a 'head crash', which is VERY bad.  The smallest of dust particles fit between the head and the platter much like you can fit a bowling ball under a door.  There is a very slim margin there, and if the head contacts the drive, it's essentially toast.  Maybe not right away, but impending doom is imminent.

For this reason, when a drive hibernates or the host system is otherwise shut down, the heads are parked off to the side as in the printer analogy above.

More info:
http://en.wikipedia.org/wiki/Head_crash

Here is where a drive will park the heads (sometimes referred to as the landing zone):
http://upload.wikimedia.org/wikipedia/c … g_zone.jpg

Notice that the landing zone is WELL AWAY from the platter, so the head will not touch the platter.

Older drives would physically rest the head on the platter, but no one has manufactured a drive that behaves that way since like 1995.

Yet more info:
http://www.storagereview.com/guide2000/ … eight.html

DasCrushinator, set hibernation to whatever you wish - the concept here is such that the faster the time-out, the more power you will conceivably save in the long run, but you'll likely have to wait more frequently for the drive to come OUT of hibernation.  Personally, I set this to 30 minutes, since once I get going with something, I usually don't stop and don't like lag.  That's just me though

Hope that helps!

DigitalJer...

Are you going by what you read on the interwebs?  Because I'm not.
Have you ever heard/seen a head crash?  Because I have - and I've replaced heads and platters and made those drives work again.

In short - I don't need to be told what a head crash is or what causes it.

Here's a little experiment you might like to try - find an old disk drive, the capacity doesn't matter, but a working one would be preferable.  Maybe you have one that you've outgrown, your first laptop maybe had a 6GB drive and now you have an 80GB, but the old drive still works and you have no further use for it - because although it will work (if you're careful) after you take it apart, it won't last long.

If you don't have a working one - that's OK - as long as the you have hasn't has a catastrophic failure, you can still use it.

Open it - take a look inside and see where the heads are parked.  Just in case you have a difficulty in finding them, the head is that little black dot at the end of the pointed actuator arm - so to speak, the tip of the arrow.  You'll find that they are sitting on the disk surface, and oh - you're not going to find any "unloading ramps" either.  Yes, the heads are parked to one side, as in the printer analogy - BUT - they are sitting on the surface of the stationary disk platter - and they come into contact with that surface every time the disk stops spinning - unless there is an unloading ramp  (which is not very common in current disk production).

Now power the disk up and watch - you'll see the platters start to rotate, and if it is a laptop disk you may see the little white plastic latch retract unlocking the head actuator assembly and allow it to swing, you should see the arm traverse in an arc and settle with the heads over track zero (near the outer edge of the disk) - now power it off and watch it park, can you see where the heads stop.

Oh wait - you have nice pictures - I didn't see those - when you open your drive, tell me if you find the "landing zone" shown in those pictures, because the vast majority don't.

Last edited by fordem (2009-06-19 16:17:28)