I see my post is confusing --- I'll try again...
Windows manages swap file sizes quite efficiently. But while the swap file size is shrunk, other data can get written to the disk's outer tracks, which is prime disk real estate. To prevent this happening, set a swap file minimum that is large enough that the swap file never finds itself occupying disk space anywhere other than those highly desirable and speedy outside tracks. Even if the space isn't being used at the moment, the good seats are being held for the swap file's return. When needed again, the swap file simply resumes residence in the preferred location. Less critical data is forced to reside elsewhere.
We're looking a a trade off. Some disk space in desirable locations remain unused, forcing data to seek residence in less attractive locations, while swap file data gets the red-carpet treatment, front-row seats, and royal service. Since in most systems, and most cases, swap file operations cause annoying delays, while non-swap-file operations are less disruptive and less noticeable, the user will enjoy a more favourable computing experience when the swap-file is given priority over other data. This is the foundation upon which my decision structure has been built, and leads to the choices I have made. This approach works for me. When I find a better one, I'll adopt it, test it, and post the results. (Change is inevitable: I wish improvement was!)
Cheers, PW. |
