re: Swapfile placement.
Monday, January 6, 2003 at 1:42 pm
Windows Me Annoyances Discussion Forum
Posted by Mac
(2831 messages posted)
I hear what the pundits say BUT my first drive was a single platter 128Mb drive whereas
this one is 60Gb.
With several platters the outside edge of any one of them is faster than further
towards the axis. As the controllers search with multiple contacts, one to each platter,
one is finding data for the application whilst another is searching the swap-file
on a DIFFERENT platter.
I find that this works faster than it did before, but if I really wanted speed I
would use a 2.53GHz P4 and 15,000 rev/min HDD some DDRAM, or whatever they use now
that is really fast, and different bus construction with dual processors. Windows
still manages the swap-file ... inside ... the paging file allocation, but this way
at least the swap-file is in one piece instead of in all sorts of bits of red & white
files spread all over the disk!
Before we get any real advance NTFS systems are better and more stable and when and
if they opt for engineering standard 64-bit working and Itanium processors we are
stuck with a very unstable ME operating system, that USED to need constant defragmentating.
Certainly it would be a LOT better with two physical drives but when you only have
ONE you can only do it this way.
Windows managed swap-file has a zero minimum and a whole disk maximum and yet there
are thise who say that 2.5 times the RAM i.e. 256 x 2.5 will suffice which is already
640 Mb with 512Mb RAM that would be a swap-file 1280 Mb in size.
The real advantage is in using D: for storage.
Happy New Year Jack, and best regards to you.
Iain.
- Written in response to:
- re: Swapfile placement. (Jack Gulley: Monday, January 6, 2003 at 9:19 am)
There are presently no replies to this message.
|
|
