Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I know that pagefile questions and advice litter the internet, but this question doesn't seem to have been asked very often.

I've read that you shouldn't put a pagefile on a different partition of the same drive (you only have one physical hard disk). However, I've also read that you can do this by creating two pagefiles. A small pagefile on the partition that contains your Windows install, and another (bigger) pagefile on another partition.

Is this true? Did I imagine reading this or misread something? DO I sound crazy?

share|improve this question
add comment

1 Answer 1

up vote 3 down vote accepted

It is true.

There are some very specific instances when it can improve performance (by keeping the swap file closer on the physical disk to the most used parts of the other data) but in general it will make either no difference or sometimes negative difference so you should only do it if you have split your drive(s) such that the system partition has grown too full and you need to free a little space by moving the pagefile elsewhere.

Using this technique to move the pagefile onto another drive that is otherwise hardly ever accessed is much more often a good idea, though you'll only see the benefit when the system is a fair amount of paging to/from disk activity.

Another reason often cited for moving the pagefile to another drive is that it may be less likely to fragment on another drive that hardly ever sees other write activity, but this can also be achieved by simply setting the page file to either a fixed size or set it such that the minimum size is fairly large - that way it won't grow (much) over time so is not in danger of fragmenting badly.

People often misunderstand the common practise of keeping swap in separate partitions on Unix-al-like operating systems as meaning having the swap space in a separate part of the system is more efficient when in fact what makes this more efficient is that there isn't a filesystem layer between the rest of the kernel and page/swap areas. You can in fact define swap areas in files on most unix-a-like OSs if you want to and don't mind the reduced efficiency introduced by working through the filesystem layer when pages need to be moved between RAM and disk, but very few people do because it isn't the default and there is very little (if anything) to gain by it in most, if not all, common use cases.

share|improve this answer
    
can somebody explain what it means by "there isn't a filesystem layer between the rest of the kernel and page/swap areas" and "You can in fact define swap areas in files on most unix-a-like OSs if you want to"? –  動靜能量 Dec 19 '13 at 21:33
    
Instead of using a pure partition you can use a file on any filesystem for swap space, but if you do, the OS needs to do more work to find anything in the swap space as the file could be spread all over the disk instead of being in a contiguous block as it will be with a dedicated swap partition. This makes dedicated swap partitions more efficient. –  David Spillett Dec 19 '13 at 22:15
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.