When I was working with Silk Test, it was possible to switch execution to an 'agent' computer and back again, allowing one script to simulate both the client and the server in a client-server model. I seem to recall that I installed the agent on the 2nd PC, where it waited for the driver to contact it. Once the driver made contact, the agent started running the script in place until it reached to point where it was told to surrender control back to the driver. This was very useful when doing things like making changes that should be reflected in a log file, switching to the machine having the log file, checking it, and switching back to the 1st computer to make more changes.

I have searched the QTP documentation, the MI website's knowledge base and user discussion forum and qaforums and can't find any reference to anyone doing this with QTP. Does anybody know how to do what I'm describing?

As a fallback, I also tried using VMware, running QTP outside the virual machine and trying to record inside the virtual machine. That didn't work - QTP couldn't recognize the browser or my app within the virtual machine. I've searched all of the above forums for VMware discussion and it seems like an instance of QTP will run within a VMware image, but can't communicate across the image to the 'primary' machine.

What I'm really trying to do is create a script that will launch my app and log in, then launch a second instance of my app and verify that the same user can't have 2 concurrent logins. I can't launch 2 instances of my app on the same computer, which is what led me to exploring remote execution or VMware use. Can anybody think of another way to accomplish this?