As it says on the label, the SNAP PAC Redundancy Option is for controller redundancy not network redundancy. If you are using SNAP-PAC-R1/R2 controllers as I/O brains, different IP networks can be created for ENET1 and ENET2 ports to create a pseudo redundant network. Note this is not possible if SNAP-EB1/2 are used as I/O brains as their second ENET port is simply a switch connection for the single IP address of the brain.
You will also need to add dual network NIC cards to any PC's as they need to be on both networks and all Opto controllers and brains need to be cabled to both switches at each location.
True network redundancy can only be created with the use of special (expensive) switches at each I/O location.
You may like to consider the use of a "ring" topology for the "main trunk" using suitable switches that detect any cable fault in the main communications link and send all traffic the other way round the "ring". Its not network redundancy, but does improve fault tolerance. Please do not try this with normal Ethernet switches, as it is not possible and will shut the communications down.:mad: