Reducing hard drive access times (myth or fact)

The following quote is from a forum discussing reducing access times by reducing the size of a partition:

So is short-stroking a myth or fact?
The definition of short-stroking from an online dictionary:

What are your opinions on this?

 

I believe I read that last year but have not done. Sure sounds good on paper. Can you link to original artical if this is where from?

 

Isn't that just saying to partition your HDD so that the OS is on one partition and file storage on the other to make the HDD more "effective"?

 

This is a good hypothesis and article...would like to see more data/proof though!!

 

sure:
http://forums.slickdeals.net/showthread.php?sduid=0&t=759070&page=5

 

I dont think its possible, because you cant speed up a mechanical drive. You would have to partition the Operating system to only use the inner part of the disk, where the sectors are closest together, and being able to do that is probably impossible.
The only way to know is to test yourselves, good luck and let us know what you find

K-TRON

 

K-T the outer part of circular drives moves faster than the inner part in any speed measurement other than RPM's for example ft/s MPH. So I think you have it reversed. All other comments I must read more before I can respond.

And thanks for the link jisaac but well I read the article that is this guys source as parts are word for word. That was by a computer scientist or engineer and in a publication which adds credibility.


NBR Disclaimer- NBR and it's parent company TechTarget are not responsible for the views and recommendations contained in powerpack's posts. The views expressed are solely powerpack's and therefor his and only his responsibility. If said advice should lead to damages we at NBR/TechTarget encourage you to sue powerpack. Hopefully to the point he loses his notebook or at a minimum his Internet connection.

 

right you are about bandwidth/speed, but the access time should be faster closer to the motor, since the head does not have to move very far to write on the inner parts of the drive.

K-TRON

 

I've always done this by default, for other reasons (multiboot, reduce fragmentation). Always knew there'd be some kind of speed advantage, didn't think it'd be faster than a raptor though. My room mate has a computer he really doesn't take care of and it is substantially slower, but it's hard to say if it's because the disk is thrashed or the os is not thrashed and not optimized, probably both.

 

I might be wrong because well I have zero computer education, but with that said please try and understand my example and understand I know I could be wrong best I can do.

What do you mean the head/actuator does not have to move as far? While technically not correct for this situation you need to think in 2D not 3D. By that I mean the actuator travels the same distance "north and south/up and down.

Let us imagine a round concert hall? The 1st row is 100 seats (data) the 2nd row can fit more say 125? Simply put the linear distance increases as you go further out on any round/circular object? I think you would accept that without argument?

So we have established greater data as you expand on a circle? Why do you insist on making these concert seats farther apart as they expand? There is no logic for that? The seats in the front are the same distance from each other ( very critical to understand, I am saying in front or behind, not next to)as the ones in front from front of stage to back have no distance differential? So where is the perceived benefit you are talking about.

I believe you are falling into a common misconception having to do with in most of the real world which I might add is almost always linear indicates to the man of logic that proximity is king? And it is, but look at the formula for pie (no special key sorry) and consider.

I learned this as a logical question having to do with LP's, we called them records? And the dilemma the appearance of being counter intuitive was. What spins faster the inside or outside of a record? Very important to understand if one would like to understand the complexity of the Universe?

Well clearly they travel the same speed? Record 33RPM's HDD's 7200. But if you measured distance traveled the outer travel more?

Anyway I hope I explained and was not jerky I would more than happy to expound or accept errors on my part. I could be incorrect but I am not seeing it.

 

Storagereview about Short Stroking http://www.storagereview.com/short_stroke.sr

 

Very interesting article their philflow,

So I guess it is possible to decrease seek times, but the amount decrease may be less on a 7200rpm or 5400rpm drive.

PP, I understand your explanation, but I think the head may regardless of the partition, scroll to the inner of the drive, resulting in larger seek times. If the data was all by the motor, the head would have no reason to move to the outer part of teh drive. I may be wrong, cause I am not sure that one can force the partition to just be on the outer part of the drive, or just the inner part of the drive.

K-TRON

 

Nice article. But how would you partition a drive to make sure your OS and system files reside on the fastest part of the drive? Is there a way to set a partitioning program to partition it and tell it to put the OS and system files on the outer edge/fastest part of the disk?

 

From extensive google searching, I've read that the first partition is placed on the outside of the drive which is the faster part of the drive, so that would be good for data while 2nd and 3rd and so on would be good for data. However everything I've read online still shows no good evidence and too much debate, so I think I'll pass. Other opinions appreciated though. Here's one link that I find a good explanation of why it isn't really worth it:

http://forums.cnet.com/5208-7588_102-0.html?hhTest=1&forumID=70&threadID=260071&messageID=2561094

This is also interesting:

http://www.hardforum.com/showthread.php?t=1057604

 

They asked some of the same questions in this forum
http://club.cdfreaks.com/f7/hd-platters-location-data-238801/

 

OK, i've tried short stroking, and at least in my case it did not work. I created a 20gb partition at the start of my 640gb hd. If there are any effects, they are neglidgible IMO.

 

Someone told me that if you have a large hd just partitioning it into smaller parts will give faster data access.

 

Thanks for the update. Did you mean 200GB? Or did you reall mean 20GB if so why so small?

What did you use to test?

 

yes I made a 20gb partition, because i wanted to reduce the amount of time spent by the drive actuator seeking sectors on a platter by as much as possible. I used hdtune and hdtach, I probably should have used something like pcmark for more definitive testing though.

 

I have concern you picked too small of partition to get the advantage?

 

well I may try quarter stroking as in my OP, when i next reinstall my os, which is generally every week....