I'm just asking because I think one of those freeware memory tools that force Windows to create a large contiguous block of memory may actually work on XP. It may be worth an experiment to zero in on the issue. You should be able to find some free ones on Softpedia.
They're lame snake-oil. If you're going to try out a memory manager, find one that uses Windows' SetProcessWorkingSetSize API - that's still a symptomatic-"treatment" band-aid, but at least it's less lame. I don't have any suggestions at hand (though I remember seeing a tool a while ago that outlined the snakeoilness of the other applications and "did things right", so they're out there). It's not the first place I'd look, though.nkormanik
: I think the "JVM may fail if a DLL or some other process has pinned a chunk of memory in such a manner that the JVM cannot reserve the contiguous amount of memory requested
" is a red herring, that would be something happening inside a complex java application interacting with native code - not likely to be something that prevents an application from starting.
How much memory is in the machine that fails, and how much in the machines that work? How much free memory is reported (by a proper tool like Sysinternals' Process Explorer)? Are there differences in the JRE versions installed on the failing and the working machines? What amount of JVM.DLL do you have flying around? (There should really only be one copy for each installed JRE and JDK, unless something is very funky - which something related to SAS might very well be
If it's a JRE issue, you might want to try uninstalling all JREs (and JDKs), then run http://cybernetnews.com/uninstall-java/
, and finally install the same JDK version as on one of the working machines.
Also, almost forgot: what's the symptoms when you try to launch the application? If it's a "Windows failed to load JVM.DLL" kinda message, it might be as simple as a x64 vs x86 JRE version issue.