Paging file and purrformance
Some of you may be aware that I have made available a progran that assists in determing just how big a paging file is really needed.
The program is at http://www.standards.com/index.html?...leUsageMonitor.
Using that program, I determined that the paging file on my system rarely exceed 30MB, so on my multiboot system, I put a fixed size 64MB pagefile on C and a 64-1152MB page file on G.
Dirve 0: Has C-D, with an OS on C.
Drive 1: Has F-H, with separate OS on each of F and G.
Drive 2: Has I-M, with an OS on J.
Each OS has a different version of Office.
J is the main OS. The others are, in effect, used only when I need to test/build with a different version of Office.
Things went smoothly until recently. But I then noticed that the pagefile use when booted to J was increasing and Windows would issue annoying messages informing me that it was increasing the size of the paging file.
To avoid the messages, I changed the page file to 128-1152 on G and eliminated the page file on C. This worked well.
But, recently, I've had a need to do a lot of testing with Word on the other drives.
In particular, when booted to the OS on F or G, I had serious performance problems. My theory was this was due to the pagefile being on the same physical drive.
So I added a 64MB page file to C, retaining the 128-1152MB page file on G, and, wow, did performance ever improve!
So it seems that if there are multiple page files, Windows may be smart enough to use the one not on the OS drive, or maybe it just gives preference to the one on C?
OK, now for the big test.
The OS on C was using only the 128-1152MB pagefile on G. I did not want to include a pagefile on C whilst booted to C.
Well, adding the 64MB pagefile to C, improved performance even when booted to C.
Have others noticed similar behavior?
