By far the simplest way to allow remote users to run processes on your server, would be to give them ids on the server.

These IDs can be locked down tight so that they can only see whatever you chose to let them see; and run whatever you choose to let them run.

If they are other windows users on the same LAN/WAN then you only need add their domain IDs to an appropriate group on your server and the job is pretty much done. After all, this is what multi-user OSs are all about. If you don't know how to go about setting up an appropriately locked-down group, consider talking to your local systems administrator.

Alternatively, if the users are getting to your server via the internet; or from machines running other OSs, then perhaps setting up a simple webserver which authenticates the users, and then presents them with a menu of tasks they can run would be your best option.

Please don't be offended, but when you use phrases like "how do I run the socket", it indicates a lack of knowledge on your behalf that means that you trying to replicate even a small part of existing OS or webserver functionality yourself, in Perl, really isn't a good idea.