Swap area is in 3 parts

  • Thread starter Thread starter William B. Lurie
  • Start date Start date
W

William B. Lurie

I may have asked this before....and my system is working okay,
but a defrag shows the swap area to be split into three non-
contiguous parts. It ain't broke so it don't need to be fixed,
but it would probably be more efficient if it was all one big
swap area. Esthetically, too. Any MVP comments or advice?
 
Will Denny said:
Hi

Have you had a look at the following article by Alex Nichol:

"Virtual Memory in Windows XP"
http://aumha.org/win5/a/xpvm.php

--

Will Denny
MS-MVP - Windows Shell/User
Please reply to the News Groups

Yes William, you have asked this before and it was answered very
satisfactorily. But I am of course biased as I think I gave the best answer!

However I note that you are soliciting answers from MVPs so presumably
solutions put forward by mere mortals are ignored.

AS you may not have even read my suggestion it was to put the swap file in a
separate partition of it's own preferably on a separate HDD. The swap file
will then remain contiguous if that's what floats your boat.

Richard.
 
William B. Lurie said:
I may have asked this before....and my system is working okay,
but a defrag shows the swap area to be split into three non-
contiguous parts. It ain't broke so it don't need to be fixed,
but it would probably be more efficient if it was all one big
swap area. Esthetically, too. Any MVP comments or advice?

In another thread you said, "...I had an I.Q. of 189 in the days
when those tests meant anything..." It appears that you might have omitted
a decimal point between the "8" and "9." Otherwise, you would have known to
search for your previous post on the same subject, which was answered. You
would also know that there is nothing to be gained in performance by
reuniting a swap file that's in only three fragments, and that there's also
nothing to be gained by compulsive defragging. That leaves aesthetics, and
if you are concerned only with appearances, you have a personal problem
that's off-topic here.
 
Richard said:
Yes William, you have asked this before and it was answered very
satisfactorily. But I am of course biased as I think I gave the best answer!

However I note that you are soliciting answers from MVPs so presumably
solutions put forward by mere mortals are ignored.

AS you may not have even read my suggestion it was to put the swap file in a
separate partition of it's own preferably on a separate HDD. The swap file
will then remain contiguous if that's what floats your boat.

Richard.
Sorry, Richard, on several counts. I thought I had saved
the prior responses. MVP as a matter of deference, no
insult intended. I now recall your suggestion......but
on a one-hard-drive system, I might find it awkward to
put .swp in a separate partition....and, to do so I'll
have to open a new thread and inquire how.......
 
In
Richard said:
AS you may not have even read my suggestion it was to put the
swap
file in a separate partition of it's own preferably on a
separate
HDD. The swap file will then remain contiguous if that's what
floats
your boat.


Putting the page file in a separate partition of its own is *not*
a good idea. It puts it farther from the other frequently used
used files on the drive, thereby increasing the time it takes for
head movement to and from it and hurting performance.

Putting the page file on a seomcd *physical* drive, if you have
one, *is* a good idea, since that decreases head movement to and
from it.

Keeping the page file contiguous is not a particularly valuable
thing to do, since access to it is mostly random anyway.
 
William, you could drive at your IQ and never get a ticket. Even in a
school zone.
 
Ken said:
In



Putting the page file in a separate partition of its own is *not*
a good idea. It puts it farther from the other frequently used
used files on the drive, thereby increasing the time it takes for
head movement to and from it and hurting performance.

Putting the page file on a seomcd *physical* drive, if you have
one, *is* a good idea, since that decreases head movement to and
from it.

Keeping the page file contiguous is not a particularly valuable
thing to do, since access to it is mostly random anyway.
Thank you, Ken, for the concise answers, especially since they
include the *why*.
 
Ken Blake said:
In


Putting the page file in a separate partition of its own is *not* a good
idea. It puts it farther from the other frequently used used files on the
drive, thereby increasing the time it takes for head movement to and from
it and hurting performance.


Putting the page file on a seomcd *physical* drive, if you have one, *is*
a good idea, since that decreases head movement to and from it.

Keeping the page file contiguous is not a particularly valuable thing to
do, since access to it is mostly random anyway.
Ken, I agree with you entirely as regarding a swap file in a separate
partition on a single drive. However doing so would definitely fulfil W B
Lurie's requirement of a contiguous swap file for cosmetic reasons, I use
the word cosmetic because even with his towering intelligence W B L might
not recognise the term "aesthetic". With the current performance available
from hard drives and their inbuilt buffering I doubt that there would be any
noticeable change in actual measured performance wherever the swap file was
located or it's degree of fragmentation.

For some time now I have been fairly sure that WBL is nothing more than a
reasonably successful troll. For someone with his claimed I Q score of 189
he regularly shows a remarkable lack of insight and finds it very hard to
follow detailed instructions through to a successful conclusion. Perhaps he
should think carefully on the wisdom of claiming such a high score. It puts
him in the top rank of all people currently living. Hmm I don't think so !

Long live the MVPs and all the freely given advice that they provide.

Richard.
 
William B. Lurie said:
I may have asked this before....and my system is working okay,
but a defrag shows the swap area to be split into three non-
contiguous parts. It ain't broke so it don't need to be fixed,
but it would probably be more efficient if it was all one big
swap area. Esthetically, too. Any MVP comments or advice?

Swap file fragmentation is right up there with Santa Claus and the
Easter Bunny - lots of hype and very little substance.

The only possible circumstance where this could have any performance
impact would be if a single swap file action - read or write -
involved accessing two or more of the parts of the swap file for items
that would be contiguous if the file was all in one place. As swap
file read/write actions are normally done in quite small amounts, most
often only a few kilobytes, the prospects of this happening are quite
remote.

The best solution for all performance related issues regarding the
swap file is to have sufficient RAM so as to eliminate the need for
Windows to actually move active memory content to and from the swap
file.

You can check to see how much active memory content is actually
contained in the swap file by using a free utility written by MVP Bill
James. Get it from
http://www.dougknox.com/xp/utils/xp_pagefilemon.htm or from
http://billsway.com/notes_public/WinXP_Tweaks/

Good luck



Ron Martell Duncan B.C. Canada
--
Microsoft MVP
On-Line Help Computer Service
http://onlinehelp.bc.ca

"The reason computer chips are so small is computers don't eat much."
 
Swap file fragmentation is right up there with Santa Claus and the
Easter Bunny - lots of hype and very little substance.

The only possible circumstance where this could have any performance
impact would be if a single swap file action - read or write -
involved accessing two or more of the parts of the swap file for items
that would be contiguous if the file was all in one place. As swap
file read/write actions are normally done in quite small amounts, most
often only a few kilobytes, the prospects of this happening are quite
remote.

The best solution for all performance related issues regarding the
swap file is to have sufficient RAM so as to eliminate the need for
Windows to actually move active memory content to and from the swap
file.

You can check to see how much active memory content is actually
contained in the swap file by using a free utility written by MVP Bill
James. Get it from
http://www.dougknox.com/xp/utils/xp_pagefilemon.htm or from
http://billsway.com/notes_public/WinXP_Tweaks/

Good luck



Ron Martell Duncan B.C. Canada

Here is a free tool to defragment your swap and other system files:

http://www.sysinternals.com/ntw2k/freeware/pagedefrag.shtml
 
In
William B. Lurie said:
Ken Blake wrote:
Thank you, Ken, for the concise answers, especially since they
include the *why*.


You're welcome, William. Glad to help.
 
In
Richard said:
Ken, I agree with you entirely as regarding a swap file in a
separate
partition on a single drive. However doing so would definitely
fulfil
W B Lurie's requirement of a contiguous swap file for cosmetic
reasons,


OK. It sounded to me like you were recommending it. If not, then
I'm glad we agree.

I use the word cosmetic because even with his towering
intelligence W B L might not recognise the term "aesthetic".
With the
current performance available from hard drives and their
inbuilt
buffering I doubt that there would be any noticeable change in
actual
measured performance wherever the swap file was located or it's
degree of fragmentation.


You're right that it's entirely possible that any performance
difference might be so slight as to be unnoticeable. Many people
these days have enough RAM so that the page file is hardly used
anyway.
 
Ken said:
In



OK. It sounded to me like you were recommending it. If not, then
I'm glad we agree.






You're right that it's entirely possible that any performance
difference might be so slight as to be unnoticeable. Many people
these days have enough RAM so that the page file is hardly used
anyway.
By the way, Ken, as a parting note: I added a 2nd(Slave) drive
and BIOS-booted to its OS. Swap file remained on the Master. I shut
down, disconnected the second drive, fired up the Master again...
and now the swap file is all in one piece. Some people may feel
otherwise, but I feel that the inquiry elicited some answers
with useful, educational content.
 
Back
Top