Can anyone explain the Disable Paging Exec. tweak? 1

[jsteph]

Hi all,
I tried the aforementioned tweak, along with the related LargeSystemCache tweak...these are supposed to keep the XP kernel in memory so they don't get paged out.

However, looking at Task Manager, it shows under the Kernel Memory section
Total: 150,704
Paged: 101,004
NonPaged: 49,700

So my surprise is that there is *anything* in the "Paged" section--that should be zero, unless the size of the kernel processes get larger than the LargeSystemCache value, which I think is 1 Gig.

Can anyone explain why any of the kernel would show as paged when this tweak is set?
Thanks,

 

[linney]

I am not a great tweaker but you can have a look at the first two links here, see if they throw any light on your question?

Does the DisablePagingExecutive registry change have any actual effect?

http://serverfault.com/questions/12...cutive-registry-change-have-any-actual-effect

LargeSystemCache and DisablePagingExecutive in Registry

http://www.pcmech.com/article/largesystemcache-and-disablepagingexecutive-in-registry/

RAM, Virtual Memory, PageFile and all that stuff.

http://members.shaw.ca/bsanders/WindowsGeneralWeb/RAMVirtualMemoryPageFileEtc.htm

 

[jsteph]

linney,
Thanks, the links were dead on!
In a nutshell, it looks like it's a 'training issue' with the semantics of the Task Manager data--the section where it says:

"Total Paged" and "Total NonPaged" should really be:
"Total Pageable" and "Total NonPageable"

Which, in this setting, makes that a meaningless pane of data because what I (and I assume most people) really would want to see there is what is *actually* paged, not what might or could be paged in some theoretical circumstance.

And while the DisablePagingExec setting may be in effect and working, we really can't verify it via the Task manager both because the semantic issue noted and also the "exec" is just a part of the kernel, and there are other kernel files that are not part of the 'executive' which wouldn't be affected by this setting anyway.

Also, the part about the LargeSystemCache has convinced me to turn that off...bigger is not necessarily better in this case.

So, all in all, it's good information to know, and thanks again for the links.
--Jim