Our I.T. organization supports two independent companies. From a user's point of view the two companies are not connected. They are on independent networks with separate servers and no crossover. The only exception is that a few members of I.T. have networking rules in place that give them access to both networks. All I.T. employees have their computer's on corporation "A"s network. There is no name resolution between "A" and "B". We currently have Service Desk set up for corporation "A" that works for those users. The Web, Application, and Database portions of Service Desk have been split up between 3 servers for performance reasons. All users in corporation "A" access service desk via the web portal. We were wondering if it was possible to set up a server at corporation "B", that was identical to our current web server, and link it back to the database and application servers at corporation "A". This way all the users at corporation "B" will be able to use the web portal to, without requiring us to poke holes between the networks for all the users. Only the web server would need to have a new hole poked. The other benefits would be that we would not need to duplicate work when it comes to administrating the program, and our help desk employees would not need to remote into the other network to access Service Desk. Is this possible? Will it create issues for us? Will Service Desk allow us the tools too isolate the two sets of users from each other? Is there a better way? Having public facing portals is not an option at this point due to security concerns.
Sounds reasonable. Bung network cards in the right servers and away you go. The web server for B will either need a direct connection to your existing database server or firewall rules, whichever is used. I suspect from what you have said a direct connection will be the preferred option with a pair of static IP addresses.
There are bound to be some features that will need tinkering with routing. For example if you are using knowledge base or service catalog the data for these is typically held in a share and access to that will be needed for B. Any analyst might want to print reports so will need access to the share that holds those, that sort of thing.