XP page file: Should I set it permanently?
This is a discussion about XP page file: Should I set it permanently? in the Customization Tweaking category; I remember reading that for best performance in Win 98 I should make the page file a set number about 2. 5 times the amount of RAM. Does this work in XP? Is there any advantage in it?.
I remember reading that for best performance in Win 98 I should make the page file a set number about 2.5 times the amount of RAM. Does this work in XP? Is there any advantage in it?
Participate in our website and join the conversation
This subject has been archived. New comments and votes cannot be submitted.
Mar 16
Apr 11
0
4 minutes
Responses to this topic
It works the same. The advantage is that it doesn't keep changing size, which is arguably better than Windows doing it for you. It stays the same size or within a certain size, depending on how you set the max and min setting.
My dad as always told me the same, and he did it in XP as well. I've never really understood that however. I don't understand why you would need MORE swap as you got more ram, I would think you would require less.
However, if you look at what windows has set your pagefile too, it is usually larger than your ram. I have 384mb of ram, and XP sets my swap file to 578(?) to 768mb. I usually reset it to 384 to windows old lower limit.
I'm in linux now though and simply gave it a 2gb swap partition and said screw it.
However, if you look at what windows has set your pagefile too, it is usually larger than your ram. I have 384mb of ram, and XP sets my swap file to 578(?) to 768mb. I usually reset it to 384 to windows old lower limit.
I'm in linux now though and simply gave it a 2gb swap partition and said screw it.
You gave Linux a 2 gig swap?!?!
That's like using a small 75mm howitzer to chop down a tree...
But yeah, msft pages stuff out to the swap like it's the 1950's and punch cards are still used. Like the 'Unlimited' Dll cache, web pages in IE cache, and Office executables. You could have 15.5 terabytes of bubble memory, and msft would still use a swap file like it was the only ram that was in your machine. Set the page file up to a permanent amt., and defrag it.
That's like using a small 75mm howitzer to chop down a tree...
But yeah, msft pages stuff out to the swap like it's the 1950's and punch cards are still used. Like the 'Unlimited' Dll cache, web pages in IE cache, and Office executables. You could have 15.5 terabytes of bubble memory, and msft would still use a swap file like it was the only ram that was in your machine. Set the page file up to a permanent amt., and defrag it.
Quote:Set the page file up to a permanent amt., and defrag it.Errrmm... don't you mean defrag then set the swap file, so that the file will be created in 1 contiguous lump? [is easier if you temporarily assign the swap file to a diff. partition or just turn it off, then defrag, then set it to fixed size] I haven't heard of any defraggers that can defrag the swap file in XP - 9x, yes, but not XP.
Quote:
I haven't heard of any defraggers that can defrag the swap file in XP - 9x, yes, but not XP.
That's true.While Norton speed disk defrags the pf in 9x/ME/2000 it does not on XP.Quite strange i would say, unless Symantec is going to release an update that does.Maybe it's because XP have NTFS 5.1?
I haven't heard of any defraggers that can defrag the swap file in XP - 9x, yes, but not XP.
That's true.While Norton speed disk defrags the pf in 9x/ME/2000 it does not on XP.Quite strange i would say, unless Symantec is going to release an update that does.Maybe it's because XP have NTFS 5.1?
Quote:Maybe it's because XP have NTFS 5.1?Which is why I don't understand why it's still crappy with FAT32.
The diskeeper 'boot time' can defrag the MFT and page files.
And yes, you should definately defrag before flipping back on the page file. NT5.1 finally tries to make it a lump sum, but sometimes there's a few K that's spread. Hence, defrag it.
And yes, you should definately defrag before flipping back on the page file. NT5.1 finally tries to make it a lump sum, but sometimes there's a few K that's spread. Hence, defrag it.
I use Diskeeper for boot-time defrag on my NTFS FS, and apparently it is faster to defrag NTFS (with Diskeeper 7) than on FAT32.
And as for pagefile query, if you have 512mb ram, it is kind of stupid to use 2.5x theory. I set my pagefile to min/max of 512mb on another physically hard drive. As to my experience, my system hardly swap memory to hard disk. So that 512mb is just there for backup/just in case.
You don't have to apply 2.5x theory to a system with 512mb ram or above, but you need at least 4mb-16mb pagefile even if you have abundant capacity of ram.
Kel.
And as for pagefile query, if you have 512mb ram, it is kind of stupid to use 2.5x theory. I set my pagefile to min/max of 512mb on another physically hard drive. As to my experience, my system hardly swap memory to hard disk. So that 512mb is just there for backup/just in case.
You don't have to apply 2.5x theory to a system with 512mb ram or above, but you need at least 4mb-16mb pagefile even if you have abundant capacity of ram.
Kel.
want to see performance, move it to another volume. It will remove lag of an overworked operating volume
(trust me ive crashed every ms os in tweaking and this is the best yet)
(trust me ive crashed every ms os in tweaking and this is the best yet)
