Transferring front panel between 3 computers

[ninajos]

I have a virtual lab setup, which is operated remotely. The virtual lab sytem has two PCs - one for authentication (say server) and one on which the experiment is actually running. 

The server PC, has a VI used for authentication. On successful authentication,the front panel of the experiment VI (running on the experiment PC) is opened on the server PC. However, only the authentication VI running on the server is web published. As a result, the experiment VI front panel is not transferred to the remote user, even though it can be controlled locally on the server PC. Is there some way by which the remote user can control the experiment VI, without having to publish that VI? (publishing that VI works but it becomes very easy to bypass the authentication system) 

The easiest solution would be to run the authentication and experiment VIs on the same system, but we have multiple experiment PCs, and the authentication system allots a particular pc based on its availability. 

[tst]

Here are a couple of suggestions. I haven't thought them through and it's been years since I've dealt with remote FPs, so they may not work:

1. Do the remote on the actual VI, but use the auth VI to change something about the connection, so that only after the auth VI confirms, the user gets access. Examples could include passwords, port numbers, or enabling the web server. I believe there's an Application class property you can poll to see when there are no more connections and thus reset the VI.
2. Do move the auth to the local PC. Use the server only to allocate the specific PC and give its connection details.
3. Create a VI server connection from the server to the experiment PC and display the relevant VI in a subpanel on the server. I'm assuming this isn't practical because the VI on the server needs to be available for further requests.

[ninajos]

Thankyou! Will try what you suggested. I think enabling/disabling the webserver through authentication can easily be done. I guess another way could be to use some sort of cookie. Thanks for the idea.

The second suggestion sounds fine too. But I still have to figure out how to do resource allocation in labview.