Virtual memory question

[Vassik]

I have 2 separate hard drives, Windows XP home version with a Gig of ram and
I am not very clear about the Virtual memory allocation. I have set the
paging file on drive D (system managed size) and my question is shall I
allocate any size on drive C as well and if yes, shall I choose system
managed size or custom size? If custom size how much?

Thanks for any feedback.

 

[Will Denny]

Hi

Please try the following article:

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

--

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


| I have 2 separate hard drives, Windows XP home version with a Gig of ram
and
| I am not very clear about the Virtual memory allocation. I have set the
| paging file on drive D (system managed size) and my question is shall I
| allocate any size on drive C as well and if yes, shall I choose system
| managed size or custom size? If custom size how much?
|
| Thanks for any feedback.
|
|

 

[DILIP]

Remove the PF from the systemroot drive and put it only on the other
physical disk. Actually, there's 1 GB of RAM, so any PF usage will be
minimal - So there will be little or no performance increase. So just keep
the Minimum at 50 MB and allow the maximum to be around 1 GB (if it is ever
required, it should be available).

 

[Vassi]

Thanks very much for the feedback. I iwll follow your advise and just
keep the PF on D and set the PF to none on C.

 

[Alex Nichol]

I have 2 separate hard drives, Windows XP home version with a Gig of ram and
I am not very clear about the Virtual memory allocation. I have set the
paging file on drive D (system managed size) and my question is shall I
allocate any size on drive C as well and if yes, shall I choose system
managed size or custom size? If custom size how much?

You should have a notional amount on C: - say initial 2 max 50 or 100,
or the system is liable to sulk and either have no VM support at all, or
make an enormous file on C:. It will probably never actually come into
existence, but keeps things happy

On D: I would set initial 100 and max plenty high - say 1GB. It will
almost certainly not grow above that initial size, but means that the
'potential' space is available, leaving RAM free. Read more at my page
www.aumha.org/win5/a/xpvm.htm

 

[Ron Martell]

Remove the PF from the systemroot drive and put it only on the other
physical disk. Actually, there's 1 GB of RAM, so any PF usage will be
minimal - So there will be little or no performance increase. So just keep
the Minimum at 50 MB and allow the maximum to be around 1 GB (if it is ever
required, it should be available).

That is less than an optimum situation.

There should be a small page file on the boot drive - perhaps 10 mb
minimum, 50 mb maximum, so as to provide for those special situations,
such as System Failure Memory Dumps, where the page file on the boot
drive is required.


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."

 

[Vassik]

Thanks so much Alex. As usual great information from you and your site. I
appreciate the time you and Dilip took to respond.

 

[DILIP]

While I haven't faced any such problems with a single PF on a second
drive, I don't dispute the information furnished by Alex and Ron - Maybe you
should see how it goes - If you have problems with a single PF, apply the
changes. And get back to us if possible.

--
Replace the obvious with "hotmail"

Thanks so much Alex. As usual great information from you and your site. I
appreciate the time you and Dilip took to respond.

 

[Alex Nichol]

While I haven't faced any such problems with a single PF on a second
drive, I don't dispute the information furnished by Alex and Ron - Maybe you
should see how it goes - If you have problems with a single PF, apply the
changes. And get back to us if possible.

It doesn't always happen. But often enough to make it worth having the
precautionary minimum setting for C:. especially as that will almost
certainly not come into existence, except for exceptional circumstances
like making a dump.

 

[Jim Macklin]

I was browsing in the MSKB and saw an article about stand-by
and hibernation problems if the virtual; memory was not at
least equal to the installed RAM, I would assume that if you
have 1 GB of RAM then you would be advised to,
have 50 MB VM on C: and 1 GB on D:


--
The people think the Constitution protects their rights;
But government sees it as an obstacle to be overcome.


|
| >Remove the PF from the systemroot drive and put it only
on the other
| >physical disk. Actually, there's 1 GB of RAM, so any PF
usage will be
| >minimal - So there will be little or no performance
increase. So just keep
| >the Minimum at 50 MB and allow the maximum to be around 1
GB (if it is ever
| >required, it should be available).
|
| That is less than an optimum situation.
|
| There should be a small page file on the boot drive -
perhaps 10 mb
| minimum, 50 mb maximum, so as to provide for those special
situations,
| such as System Failure Memory Dumps, where the page file
on the boot
| drive is required.
|
|
| 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."

 

[Alex Nichol]

I was browsing in the MSKB and saw an article about stand-by
and hibernation problems if the virtual; memory was not at
least equal to the installed RAM, I would assume that if you
have 1 GB of RAM then you would be advised to,
have 50 MB VM on C: and 1 GB on D:

If you have a reference, please let me know. Hibernation needs a file
big enough to hold the content of RAM, but in XP has one of its own -
hiberfil.sys - and does *not* use the page file. I think you must have
seen something related only to one of the earlier versions of windows

 

[Jim Macklin]

I don't remember now, it was something I came across while
looking for something else. If I get time, I'll see if I
can find it again and post back.

Found these, relates to XP...
330909 - Hibernation Problem on Computers with 1 GB of RAM
Hibernation Problem on Computers with 1 GB of RAM. View
products that this article
applies to. This article was previously published under
Q330909. SYMPTOMS. ...
support.microsoft.com/?kbid=330909 - 15k - Cached - Similar
pages

326662 - Hibernation Problem with Computers with One
Gigabyte of ...
Hibernation Problem with Computers with One Gigabyte of
RAM Under High-Stress
Conditions. View products that this article applies to.
....
support.microsoft.com/default.aspx?kbid=326662 - 15k -
Cached - Similar pages
[ More results from support.microsoft.com ]



message | "Jim Macklin" <p51mustang[threeX12]@xxxhotmail.calm>
wrote:
|
| >I was browsing in the MSKB and saw an article about
stand-by
| >and hibernation problems if the virtual; memory was not
at
| >least equal to the installed RAM, I would assume that if
you
| >have 1 GB of RAM then you would be advised to,
| >have 50 MB VM on C: and 1 GB on D:
|
| If you have a reference, please let me know. Hibernation
needs a file
| big enough to hold the content of RAM, but in XP has one
of its own -
| hiberfil.sys - and does *not* use the page file. I think
you must have
| seen something related only to one of the earlier versions
of windows
|
|
| --
| Alex Nichol MS MVP (Windows Technologies)
| Bournemouth, U.K. (e-mail address removed)8E8L.org (remove the D8
bit)

 

[Alex Nichol]

Found these, relates to XP...
330909 - Hibernation Problem on Computers with 1 GB of RAM
Hibernation Problem on Computers with 1 GB of RAM. View
products that this article
applies to. This article was previously published under
Q330909. SYMPTOMS. ...
support.microsoft.com/?kbid=330909 - 15k - Cached - Similar
pages

Those appear to relate to a very specific high-stress condition, and
will I think be purely a matter of the hibernation system. Only way I
can see it might relate to VM is if the stressing processes were causing
so much thrashing of the VM system that Hibernation could not get in to
have a clear cut version of RAM to file away. On the whole I'd suggest
that anyone trying to hibernate in such conditions is asking for trouble
anyway

 

[Jim Macklin]

Probably true, thanks.

Jim
message | "Jim Macklin" <p51mustang[threeX12]@xxxhotmail.calm>
wrote:
|
| >
| >Found these, relates to XP...
| >330909 - Hibernation Problem on Computers with 1 GB of
RAM
| >Hibernation Problem on Computers with 1 GB of RAM. View
| >products that this article
| >applies to. This article was previously published under
| >Q330909. SYMPTOMS. ...
| >support.microsoft.com/?kbid=330909 - 15k - Cached -
Similar
| >pages
|
| Those appear to relate to a very specific high-stress
condition, and
| will I think be purely a matter of the hibernation system.
Only way I
| can see it might relate to VM is if the stressing
processes were causing
| so much thrashing of the VM system that Hibernation could
not get in to
| have a clear cut version of RAM to file away. On the
whole I'd suggest
| that anyone trying to hibernate in such conditions is
asking for trouble
| anyway
|
|
| --
| Alex Nichol MS MVP (Windows Technologies)
| Bournemouth, U.K. (e-mail address removed)8E8L.org (remove the D8
bit)