Tags: 768mb, box, drive, file, home, memory, microsoft, moved, msdn, own, page, partition, ram, secondary, software, starts, stops, system, toits, virtual

system stops using RAM and starts using virtual memory too soon

On Microsoft » Microsoft XP

3,885 words with 2 Comments; publish: Mon, 26 May 2008 15:16:00 GMT; (331109.38, « »)

I have XP Home on a P4 box with 768MB of RAM and have moved the page file to

its own partition on a secondary HD (drive G:\). The page file partition is

1GB and I've currently set the initial size to 500MB and the maximum size to

989MB (i.e., the maximum free space the partition showed as being

available).

I've been using Cacheman v5.50 to monitor the dynamics and have found that

the system begins ignoring the RAM and heavily accessing the HD (especially

whenever an application is launched) as soon as the amount of free RAM drops

to 300MB. Something seems wrong since I have 768MB of total RAM, 300MB of

which show to be free when it decides to start doing this.

I've tried setting the page file to both 0 init./0 max. and 0 init./989

max., but the system still wants to start using virtual memory when the free

RAM gets to about 300MB. Interestingly, from viewing file properties it

appears that the page file size never changes at any time after the virtual

memory has started to become accessed. And I've made sure there is only

*one* pagefile.sys on the system, and it's on drive G:\.

Has anyone else experienced similar behavior? Does anyone know of a way to

get my system to use all of its free RAM before calling upon virtual memory?

TIA

Tony

All Comments

Leave a comment...

  • 2 Comments
    • Did it do this when the page file was on in the original place..?

      "particle" <fake.windows-xp.todaysummary.com.false.net> wrote in message

      news:c797tl$hdd$1.windows-xp.todaysummary.com.nnrp.atgi.net...

      > I have XP Home on a P4 box with 768MB of RAM and have moved the page file

      to

      > its own partition on a secondary HD (drive G:\). The page file partition

      is

      > 1GB and I've currently set the initial size to 500MB and the maximum size

      to

      > 989MB (i.e., the maximum free space the partition showed as being

      > available).

      > I've been using Cacheman v5.50 to monitor the dynamics and have found that

      > the system begins ignoring the RAM and heavily accessing the HD

      (especially

      > whenever an application is launched) as soon as the amount of free RAM

      drops

      > to 300MB. Something seems wrong since I have 768MB of total RAM, 300MB of

      > which show to be free when it decides to start doing this.

      > I've tried setting the page file to both 0 init./0 max. and 0 init./989

      > max., but the system still wants to start using virtual memory when the

      free

      > RAM gets to about 300MB. Interestingly, from viewing file properties it

      > appears that the page file size never changes at any time after the

      virtual

      > memory has started to become accessed. And I've made sure there is only

      > *one* pagefile.sys on the system, and it's on drive G:\.

      > Has anyone else experienced similar behavior? Does anyone know of a way to

      > get my system to use all of its free RAM before calling upon virtual

      memory?

      > TIA

      > Tony

      >

      #1; Mon, 26 May 2008 15:18:00 GMT
    • Yes, it did this with the pagefile on the OS partition too.

      "Nick Burns" <thedoc.windows-xp.todaysummary.com.hotmail.com> wrote in message

      news:uQtsLFlMEHA.3712.windows-xp.todaysummary.com.TK2MSFTNGP10.phx.gbl...[vbcol=seagreen]

      > Did it do this when the page file was on in the original place..?

      > "particle" <fake.windows-xp.todaysummary.com.false.net> wrote in message

      > news:c797tl$hdd$1.windows-xp.todaysummary.com.nnrp.atgi.net...

      file[vbcol=seagreen]

      > to

      > is

      size[vbcol=seagreen]

      > to

      that[vbcol=seagreen]

      > (especially

      > drops

      of[vbcol=seagreen]

      > free

      > virtual

      to

      > memory?

      >

      #2; Mon, 26 May 2008 15:19:00 GMT