We are having two different servers - one Prod and one DR Server?
We are running Tuxedo 11gR1 in both the servers.
How can we achieve seamless fail-over in case the network connectivity of Prod goes down?
Can you provide a bit more detail about your application and environment? How do requests enter the Tuxedo system? By way of a workstation client, a web service client, from a JEE application server such as WLS, etc.? When you say a server, do you literally mean a single host machine? Are you running Tuxedo in a clustered environment with multiple machines at each site? Is there a database involved? Are you using distributed transactions?
Without knowing how clients are connected and some of the other information above, it's really hard to give much guidance. There is a white paper on configuring Tuxedo for high availability that may address some of your concerns at: http://www.oracle.com/technetwork/middleware/tuxedo/overview/tuxedo-ha-wp-2008-130747.pdf
Oracle Tuxedo Chief Architect