I know that VBScripts allows to make:

{
FreeMem=Space(256000000)
}

This will free 256 MB of RAM... How can I do this from inside a VBS script, using Object Model + WMI? Could you please an VBS sample?


Thanks in advance.

Regards.

Hi asgarcymed,

There's no scriptable method that I know about that can "free" memory -- the only way to really "free" memory is to stop unneeded processes, and Windows' built-in memory manager handles memory allocation for you. What exactly are you trying to do?

Bill

I am trying to run a script which analyzes thousands of files, and thus, it is very CPU and RAM consumer... When I run it for a small number of files, no problems; however it "eats" all RAM when number of files > 500...

 

Regards. 

Hi asgarcymed,

As I said, there are no script APIs (that I know of, anyway) that allow you to "free" memory. In addition, the notion of "free" memory is not really that straightforward. For example, Windows handles swap file space as an extension of physical memory. I think you're stuck with closing programs you don't need or optimizing your script.

Bill

:(

If you don't mind posting your code, it's possible that we could point out where the problem causing major memory consumption is. In all likelihood, it comes down to one of the following:

1. Recursively enumerating files and folders and keeping a handle to them before performing processing. Rewriting your code to stream the content could improve things substantially. So could off-loading file/folder recursion - that gets really messy in WSH, but a commandline WSH script that reads its input stream can off-load all of the recursive enumeration by using the command shell dir command properly - for example, using
   dir /a /s /b
   enumerates all files and folders recursively and returns them as paths, one per line, and they get used and dumped quickly
2. Another problem might be your string manipulation, if you're actually doing something with text from those files. Although VBScript is fairly fast with small strings and arrays, its approach to dynamic arrays and string-building scales horribly; it actually creates a complete new copy under the covers each time it adds text to a string or a new element to an array. Between increasing data set size and the number of operations it performs, the speed slows exponentially with respect to array/string length.