Can't set pagefile beyond 2047 MB

I have 2 GB worth of physical memory and 1 physical drive with 1 partition
because I don't have the need for a second partition. Windows recommends
using pagefile setting of 3070 MB, but sets my pagefile to 2047 MB regardless
of what setting I use, unless I choose not to have a pagefile. I have tried
using 2048 MB to match the currrent amount of physical RAM, and it still
won't set it to 2047 MB. I have tried using 4096, but Windows won't let me
set it beyond 4095, which I know is a physical limitation on the OS per
partition. I know it's a moot point since I have 2 GB worth of physical
memory, but I'm baffled here and have to have an answer to the equation
before me. I would rather not use anything at all, but I know a pagefile is
necessary. I just keep thinking that I'm missing out on the maximum
performance I can get out of this laptop since my pagefile setting is lower
than what the system is recommending. Any thoughts?

Laurence wrote:

I have 2 GB worth of physical memory and 1 physical drive with 1 partition
because I don't have the need for a second partition. Windows recommends
using pagefile setting of 3070 MB, but sets my pagefile to 2047 MB regardless
of what setting I use, unless I choose not to have a pagefile. I have tried
using 2048 MB to match the currrent amount of physical RAM, and it still
won't set it to 2047 MB. I have tried using 4096, but Windows won't let me
set it beyond 4095, which I know is a physical limitation on the OS per
partition. I know it's a moot point since I have 2 GB worth of physical
memory, but I'm baffled here and have to have an answer to the equation
before me. I would rather not use anything at all, but I know a pagefile is
necessary. I just keep thinking that I'm missing out on the maximum
performance I can get out of this laptop since my pagefile setting is lower
than what the system is recommending. Any thoughts?

In theory, the more memory you have, and you have a lot, the less page
file space will be used. Certain apps take some page file space for
their own use even though they shouldn't, but for the most part this
theory is what works in practice. Read about virtual memory and how to
set the page file in Alex Nichol's excellent article,
http://aumha.org/win5/a/xpvm.htm. XP's recommendations are out of line
when there is a large amount of memory in the system. You might want
start with an initial size of 50 or 100 MB and set a maximum of 1024 or
2048. XP won't take that much. It will just take what it needs.


You can use Bill James' utility to monitor page file use and see how
much your system is actually using. The numbers provided by XP in task
manager are deceptive. Get it from he http://www.dougknox.com/ in
the Win XP Utilities section, "XP Page File Monitoring Utility."

C 'Setting VM with WinXP-2K_Pagefile.vbs' of 12/16/04
for what stats mean & how PF operates & implications to
OS with RAM1GB. NOT lite reading tho

HTH-Larry

On Thu, 23 Dec 2004 20:59:01 -0800, Rock
wrote:

| Read about virtual memory and how to
|set the page file in Alex Nichol's excellent article,
|http://aumha.org/win5/a/xpvm.htm. XP's recommendations are out of line
|when there is a large amount of memory in the system. You might want
|start with an initial size of 50 or 100 MB and set a maximum of 1024 or
|2048. XP won't take that much. It will just take what it needs.
|
|
|You can use Bill James' utility to monitor page file use and see how
|much your system is actually using. The numbers provided by XP in task
|manager are deceptive. Get it from he http://www.dougknox.com/ in
|the Win XP Utilities section, "XP Page File Monitoring Utility."

Any advise is my attempt to contribute more than I have received but I can only assure you that it works on my PC. GOOD LUCK.

Laurence wrote:

I have 2 GB worth of physical memory and 1 physical drive with 1 partition
because I don't have the need for a second partition. Windows recommends
using pagefile setting of 3070 MB, but sets my pagefile to 2047 MB regardless
of what setting I use, unless I choose not to have a pagefile.

It is absurdly too much. You need a page file - principally to give a
'home' to pages of memory that have been allocated to programs but never
brought into use. I suggest an initial size 100 MB, Max maybe 1024 to
cover such contingencies. You will probably find if you check on the
super-hidden pagefile.sys in explorer that it never gets bigger than the
100. Read more at my page www.aumha.org/win5/a/xpvm.htm


--
Alex Nichol MS MVP (Windows Technologies)
Bournemouth, U.K. (remove the D8 bit)

Alex Nichol Wrote:

It is absurdly too much. You need a page file - principally to give a
'home' to pages of memory that have been allocated to programs but
never
brought into use. I suggest an initial size 100 MB, Max maybe 1024 to
cover such contingencies. You will probably find if you check on the
super-hidden pagefile.sys in explorer that it never gets bigger than
the
100. Read more at my page www.aumha.org/win5/a/xpvm.htm


--
Alex Nichol MS MVP (Windows Technologies)
Bournemouth, U.K. (remove the D8 bit)
what a ridiculous statement...in every regard.

Alex is under the impression that since he doesn't put his memory under
pressure, nobody does.

foremost, there is NO penalty for having a pagefile that might be too
big...there is a HUGE penalty for having one too small

plenty of people need MORE then two gig page file..ESPECIALLY people
that have over a gig of memory

second, the pagefile IS NOT a "'home' to pages of memory that have been
allocated to programs but never brought into use."

the pagefile is an area on the disk that allows xp to make modified
pages part of the memory management strategy

pages that haven't been modified are simply paged using the file whence
they originally came

DO NOT USE THE SETTINGS ALEX SUGGESTS, they are counterproductive

if you are not short on hardrive space NEVER LOWER THE DEFAULT SETTINGS
OF THE PAGEFILE

period

now, if you'd like a proper explanation of the pagefile,
http://www.osnn.net/articles.php?act...showarticle=99 is a good
one

you'll need to have a basic understanding of the pagefile in the first
place to follow that one along


--
perris
------------------------------------------------------------------------
perris's Profile: http://forum.osnn.net/member.php?userid=17
View this thread: http://forum.osnn.net/showthread.php?t=46009

In news perris typed:

Alex Nichol Wrote:

It is absurdly too much. You need a page file - principally
to give
a 'home' to pages of memory that have been allocated to
programs but
never
brought into use. I suggest an initial size 100 MB, Max maybe
1024
to cover such contingencies. You will probably find if you
check on
the super-hidden pagefile.sys in explorer that it never gets
bigger
than the
100. Read more at my page www.aumha.org/win5/a/xpvm.htm


--
Alex Nichol MS MVP (Windows Technologies)
Bournemouth, U.K. (remove the D8 bit)

what a ridiculous statement...in every regard.

Alex is under the impression that since he doesn't put his
memory
under pressure, nobody does.


No, Alex's statement is completely accurate.


foremost, there is NO penalty for having a pagefile that might
be too
big...


The only penalty is that it's waste of disk space. In these days
of very cheap large drives, that's a much smaller penalty than it
used to be, but it is a penalty.


there is a HUGE penalty for having one too small


Only if you prevent it from growing larger as needed.


plenty of people need MORE then two gig page file..ESPECIALLY
people
that have over a gig of memory


That's backwards. The more memory you have, the *less* page file
you need.

--
Ken Blake - Microsoft MVP Windows: Shell/User
Please reply to the newsgroup

my statemensts stand

you wrote;
The only penalty is that it's waste of disk space"

that's correct and we agree on that...as I allready stated, my points are
for those that do not have a storage issue.

if there is no hardrive issue, there is absolutely no lowering the setting
from the default

you wrote;
[there is a penalty]""Only if you prevent it from growing larger as needed."

this is oncorrect in every sense...expansion is a hit a; by, when it was
invoked, and obviously since it might need to be invoked

in addition, by haing a smaller pagefile then the amount of memory in use, a
user will at times take private writable memory, (modified pages) out of
concideration for the memory manager.

obvioulsy, dlls and exe's would be unloaded in these situations instead of
the best candidate

you wrote;
[the idea that the more memory the bigger the pagefile]"That's backwards.
The more memory you have, the *less* page file
you need.

absolutely incorrect.

every bit of memory a user has in use needs backing on the hardrive.

the only users that don't need to increase the size of the pagefile when
they have more memory are the users that don't use the extra memory they
installed.

of course, there are users that install more memory then they use..these
users won't suffer a hit by lowering the default pagefile settings...but they
get absolutely no gain by doing it..,.and they DO risk a slow down in the
future

pretty simple;

since there is no penalty what so ever if the pagefile might be too big, DO
NOT MAKE THE SETTINGS LOWER THEN THE DEFAULT, unless you have a storage issue

pretty self evident

perris wrote:

[the idea that the more memory the bigger the pagefile]"That's backwards.
The more memory you have, the *less* page file
you need.

absolutely incorrect.

For a given workload it is entirely correct.

You seem unable to understand that nothing is written into the page file
(other than a small amount the system seems to park there for
contingencies) until there is insufficient room in RAM. It is then used
for overflow. On large RAM this may be never - or it may be quite early
if the workload is very heavy. Hence the advice (which you will also
find in Ed Bott's 'Windows XP Inside out') to set initial size at 100
*in the first instance*. If the size of the file grows, update the
initial size accordingly, to cover all normal use.

every bit of memory a user has in use needs backing on the hardrive.

Now that *is* incorrect. The total VM allocation in use must be
contained in the sum of RAM and page file. As above - *nothing* is
written to page file until RAM overflows. If there is a significant page
file actual use, then RAM is marginally adequate at best - get more. Or
do not load so many things to sit around doing nothing.

Also realise that there is a total limit of virtual memory space set but
the underlying system memory model; RAM plus page file of more than 4GB
is certainly no use (apart from the case of Fast User Switching), and
more than 3 GB very dubious


the only users that don't need to increase the size of the pagefile when
they have more memory are the users that don't use the extra memory they
installed.

And so is that

And that is all I am going to say.

--
Alex Nichol MS MVP (Windows Technologies)
Bournemouth, U.K. (remove the D8 bit)

You can actually monitor & test Pagefile usage with a couple of XP
tools. One, Perfmon.Msc will show you in %, total Pagefile usage.
You do have to add counters for Pagefile ( Usage, Peak Usage).
By using TaskMgr, Performance - you can monitor Memory usage
(The PF meter).
My system has 1.0 Gigabytes of memory, so take that into account
when reading the values presented below.
On my System I have a 128 Min Pagefile. When no applications
are running my PF meter runs at 187 Megabytes and the Pagefile %
will be at 11%. Under heavy loading (VPC, Word, Outlook, OE &
Streets & Trips) memory usage climbs to 527 Megabytes and the
Pagefile percentage climbs to 27% or 34 Megabytes of the Pagefile.
My VPC uses 256 megabytes for the Windows 2000 instance.
The point of all this is that as long as available Physical memory is
available to XP, it won't make extensive use of the Pagefile.
Because of memory requirements, I now recommend that all my
customers do not purchase a PC that has less than 512 Megabytes
of memory.
If you spend 30-45 minutes testing your own system, you can see
in real world terms how XP utilizes memory.

"Alex Nichol" wrote in message
...
perris wrote:

[the idea that the more memory the bigger the pagefile]"That's backwards.
The more memory you have, the *less* page file
you need.

absolutely incorrect.

For a given workload it is entirely correct.

You seem unable to understand that nothing is written into the page file
(other than a small amount the system seems to park there for
contingencies) until there is insufficient room in RAM. It is then used
for overflow. On large RAM this may be never - or it may be quite early
if the workload is very heavy. Hence the advice (which you will also
find in Ed Bott's 'Windows XP Inside out') to set initial size at 100
*in the first instance*. If the size of the file grows, update the
initial size accordingly, to cover all normal use.

every bit of memory a user has in use needs backing on the hardrive.

Now that *is* incorrect. The total VM allocation in use must be
contained in the sum of RAM and page file. As above - *nothing* is
written to page file until RAM overflows. If there is a significant page
file actual use, then RAM is marginally adequate at best - get more. Or
do not load so many things to sit around doing nothing.

Also realise that there is a total limit of virtual memory space set but
the underlying system memory model; RAM plus page file of more than 4GB
is certainly no use (apart from the case of Fast User Switching), and
more than 3 GB very dubious


the only users that don't need to increase the size of the pagefile when
they have more memory are the users that don't use the extra memory they
installed.

And so is that

And that is all I am going to say.

--
Alex Nichol MS MVP (Windows Technologies)
Bournemouth, U.K. (remove the D8 bit)

Alex wrote;
(the more ram, the smaller the pagefile) "for a given workload it is
entirely correct."

well, this is like saying "when it doesn't rain, it doesn't
percipitate."isn't it...this is like saying "if I don't use my memory,
then I won't use my memory"...that's all youve said.

for those of you that want best performance, every bit of memory in use
needs an area on the hardrive, otherwise, memory that MIGHT have been
written to the pagefile will not be considered for the memory
manager...other candiates will be concidered instead...pretty simple

therefore, for whatever amount of memory you have, the pagefile needs
to be AT LEAST as big as that...in the event YOU MIGHT use the memory
you have installed...that's the purpose of memory in the first place.

as Alex himself has stated, "free memory is wasted memory"...he and I
agree on that point

Alex wrote;
"
You seem unable to understand that nothing is written into the page
file"

rediculous...the pagefile ALWAYS has information written to it, the
caveat you mention is hardly all that's written to the pagefile.

on the other hand, what YOU definitely don't understand is that it
matters not one stitch how much is ACTUALLY written or isn't written to
the pagefile

all that matters is that there is an area on the hardrive for modified

pages

are you under the impression that modified data can share the same area
in the pagefile, simply because nobody yet is written to the area?..that
is incorrect

if I work at night and you work in the day, we still can't share the
same bed can we

self evident once told

again, ALL data in memory needs IT'S OWN hardrive area...it doesn't
"share" the address.

you also seem to think that there is no performance hit "until" the os
requests more virtual memory.

this is incorrect...if there is private writable that is has no area on
the hardrive for backing store, the memory manager goes to the next
best candidate

there are hit's long before expansion is invoked...and then obviously
WHEN expansion is invoked

Alex wrote';
"nothing* is written to page file until RAM overflows."

this is incorrect...pages are written long before "ram overflows"...xp
writes information of most likely pages before memory is under pressure
so there is less overhead at the time of pressure

IN ADDITION;

all memory in use NEED it's own address, whether or not it's yet
written..Data doesn't share it's bed simply because no other data is in
the bed at the moment

Alex wrote;
"If there is a significant page file actual use, then RAM is marginally
adequate at best - get more. Or do not load so many things to sit around
doing nothing."

you can't be serious!...you are trying to say that to accommodate your
preference in how big a pagefile should be, users have to load fewer
programs?...or buy more ram?

umm...no, all most users need do is NOT lower the default
settings...you've made my point for me

in the end, you've made every case...you agree that for people that do
not have a storage issue, there is ABSOLUTELY NO penalty for a pagefile
that might be too big

you also correctly site the instances of fast users on the box need a
bigger pagefile

you also agree there is a hit for people that have a pagefile too
small

so

exactly what is your point?

why on earth would a person want to lower the default settings when
there is nothing to gain and possibly performance to suffer?

you tell me what is the point of your settings for people that don't
have a storage issue?

for users that want best performance on their box, DO NOT lower the
settings of the pagefile


--
perris
------------------------------------------------------------------------
perris's Profile: http://forum.osnn.net/member.php?userid=17
View this thread: http://forum.osnn.net/showthread.php?t=46009

there is obviously a missunderstanding of what the pagefile does, and
what it
's there for.


the pagefile is not where most of paging happens.

the pagefile is only for the data that doesn't look the same as when it
was first brought into use

the pagefile is for "modified pages"

any information that looks the same as it did before you started
working simply gets "released" from memory.

for instance, if you launch a program...if memory needs to be claimed,
and this program that you launched will be least likely missed, then
the most rarely used features of this program are simply taken out of
physical memory.

if you happen to reference this rarely used feature once it's been
released, the operating system just goes back to the original program
to retrieve that information

there is no pagefile used for this paging, and this is the majority of
paging on your computer

now, suppose the feature that is the least likely to be used happens to
not look the same as it did when it was loaded?

or suppose it wasn't even on the hardrive in the first place...like a
document you were working on an hour ago

well, those pieces of information that have been modified need to have
a place to be backed in case they are the best candidate for memory to
be claimed

the pagefile is where the operating system will back modified
information ....nothing else gets backed to the pagefile , information
that hasn't been modified gets backed to the file, dll, exe from whence
it first loaded

if the pagefile is not there, or is too small, then modified
information won't be a candidate for the memory manager

how will this manifest?

suppose you've minimized something you're working on...like a
document;

you haven't worked on that document for say a half hour or whatever.

if there is no area for that information to get backed, then if your
memory goes under pressure, instead of using this document for physical
memory, the os will go to something you've worked on more recently

that's no good

important;

when the operating system needs to reclaim memory, it will do it
regardless of the size of the pagefile...it will simply reclaim where
it can reclaim, instead of where it should reclaim memory

having the pagefile smaller then the default in no fashion whatsoever
keeps xp from reclaiming memory

having the pagefile too small will actually cause MORE paging, not less
paging!

since the os will at times page features that are are referenced more
recently then what would have been in the pagefile, there is more
swapping, not less swapping if the pagefile is too small.

there is never more swapping if the pagefile is too big

a performance hit of this nature won't even be recognized by most
users, or they won't attribute the hit to a pagefile that's too small

it will manifest as a little hitch or bog as you try to do something
that"s become unloaded that wouldn't have been unloaded if the best
candidate were available


--
perris
------------------------------------------------------------------------
perris's Profile: http://forum.osnn.net/member.php?userid=17
View this thread: http://forum.osnn.net/showthread.php?t=46009

and this just because I'm in the mood;

in a normal configuration, xp will allow each program to allot virtual
address memory up to 4 gigs

how many people does anyone know that hs 4 gigs of memory installed?

obviously, with correct configuration, programs can allocate virtual
memory that will easily exceed the physical memory installed on the
box....although, with some of the suggestions on this thread, that
won't be easy.

obviously again, the operating system needs a memory manager that'll
distribute physical memory as virtual need presents itself.

the memory manager starts out by assigning each process a portion of
the physical memory...this amount of physical memory is called the
"working set" of that process

most programs are written with the "90/10 rule"...that is they spend
90% of the time accessing 10% of their code...thus a working set
doesn't need to be nearly as large as a programs features would imply.

obviously, hardly anybody has enough physical memory to survive without
virtual memory.

The memory manager of xp will actually expand and contract a working
set according to the users need...again, this strategy is impeded when
the kernel teams recomendations are circuvented.

at times a process will request a "page" that has not been represented
in the "working set"

everyone knows, when that happens, a "page fault" is generated....if
there is enough physical memory available, the memory manager simply
assigns some of the memory from the available pool...now the next part
is sweet;..if there is enough physical memory to allow it, then this
process's "working set" is simply increased at no price to any other
process ... nice.

However, if there isn't enough physical memory available, then this new
page will have to replace a page that's somewhere in physical
memory...the memory manager will use the page that hasn't 't been
referenced in the longest period of time...in most cases, this
particular page that hasn't been referanced in the longest while will
most likely not get referenced again, and it is the safest candidate to
take out of physical memory.

so now that the Memory Manager removes a page from a process working
set, it has to decide what to do with the info that was in that page.

If the data has been modified, the Memory Manager will put it on the
modified page list, ( a list of pages that eventually will be written
to the paging file) or if it's not been modified, then back to the
memory-mapped files from whence those pages correspond.

From the modified page list, the Memory Manager moves pages to a pool
called the "standby list".

Unmodified pages go directly to the standby list.!!! ( you can view the
standby list as a cache of file data)

The "stand by list" is one of the sweetest policies of memory
management;...this is a list of physical memory that is available for
anything at all, but it still has the data that was at one time being
used somewhere!...so if that data does happen to get referanced before
this physical memory is claimed, the page comes streight from
ram...very very nice

The standby list is memory that's also considered by the memory manager
as "available memory"...there are other pools that'll contribute to
available memory...pages that contain info which belonged to data
that's been dealocated...for instance, pages that once belonged to
processes you've closed down...also, pages that were freed and filled
with zero data by the Memory Manager's "low-priority zero page".

All of this goes on dynamically, and the memory manager examines
working sets once a second..when memory is under pressure, the memory
manager will pro-actively remove pages from those working sets which
haven't encountered a page fault in a certain time frame....now, when
the memory manager pro-actively removes a page from a working set, that
page simply goes into the standby list!!! In this fashion, the data
isn't lost to the hardrive at all, though the system has prepared for a
page fault before it happens!

It's important to remember that the pages on the standby list are
considered as available memory, and equally important to realize that
they still retain the data whence they came.

here's something nice; what this tuning mechanism does for idle
threads, is it will take pages from those idle processes a little at a
time, and the working sets on idle processes eventually
disappear...processes that remain idle for a length of time eventually
consume no physical memory !

OK, now if a process needs a new page of physical memory, the memory
manager first looks to see if that page is on the standby or modified
page list. It will be here if the page will be in one of these lists if
the page was removed from that working set and it wasn't claimed for
another purpose....this operation is called a "soft page fault" since
it doesn't involve a read from the hardrive.

if a page that's requested isn't on one of the available memory lists,
a "Balance Set Manager" is triggered, and it'll trim the process
working sets in order to populate the list that makes up available
memory.

If the memory manager has to remove a page from available memory, it'll
read the data from somewhere on the hardrive...the paging file or an
executable, whatever.

if the data that the memory manager wanted to remove from available
memory happens to be a page who's virtual address isn't availble due to
incorrect settings, the memory manager will go to the next candidate
that can be backed.


--
perris
------------------------------------------------------------------------
perris's Profile: http://forum.osnn.net/member.php?userid=17
View this thread: http://forum.osnn.net/showthread.php?t=46009

R. McCarty wrote:

You can actually monitor & test Pagefile usage with a couple of XP
tools. One, Perfmon.Msc will show you in %, total Pagefile usage.
You do have to add counters for Pagefile ( Usage, Peak Usage).
By using TaskMgr, Performance - you can monitor Memory usage
(The PF meter).

It gives very misleading results. It gives the commit - all VM space
that might go in the file if it all came into use - which it almost
certainly won't. As such the values can easily by much larger than the
physical size of the file if you start with an initial value enough for
normal usage. Use instead the Page File Monitoring program, originally
by Bill James MVP and redone in a complied version by Doug Knox, at
http://www.dougknox.com/xp/utils/index.html. near bottom. which show the
amount of file that is actually doing something.

--
Alex Nichol MS MVP (Windows Technologies)
Bournemouth, U.K. (remove the D8 bit)

I see Alex is basing his opinon on how much data is actually written to
the pagefile

this in no way represents how big the pagefile NEEDS to be

the tool Alex is referring only shows what is already written to the
pagefile...it DOES NOT show how much of the pagefile is being used by
the os for address translation

memory usage with perfuming will show the amount of hardrive area that
is set aside for use.

do not go lower then this EVEN IF YOU HAVE A STORAGE ISSUE

if you have a storage issue, you can safely lower the default settings
to about double your peak pagefile useage...then you won't suffer a hit
as I spoke about in my previous posts.

if you have no storage issue, in the best case scenario is a waste of
time...in plenty of scenarios presents a performance liability

don't do it


--
perris
------------------------------------------------------------------------
perris's Profile: http://forum.osnn.net/member.php?userid=17
View this thread: http://forum.osnn.net/showthread.php?t=46009

I'm waiting for Alex to tell everyone exactly what he's accomplishing by
telling even people with no storage issue to lower the default settings
of the pagefile

if he can come up with one scenario that it gives some kind of gain,
then you could weigh that gain against the "possible" liability that I
speak about

if he can't come up with some kind of gain, then he just wants you to
be pro active for no reason what so ever


--
perris
------------------------------------------------------------------------
perris's Profile: http://forum.osnn.net/member.php?userid=17
View this thread: http://forum.osnn.net/showthread.php?t=46009