I have two mcd's networked together via ip trunking but not in a cluster. Switch A's primary outgoing route is its T1 and back up route is switch B's T1. Switch B's primary route is switch A's T1' and backup is its own T1. This is all programmed, of course in route lists. How can I prevent the systems from entering a loop and allow a call to go out over switch B's T1 upon a failure on switch A's T1 without changing up the order of preference of routes?