In today's high performance networks, organizations need the freedom to implement packet forwarding and routing according to their own defined policies in a way that goes beyond traditional routing protocol concerns. While static and dynamic routing focus on the traffic destination for routing, policy-based routing provides a mechanism to mark packets so that certain kinds of traffic receive differentiated routing. Destination-based routing techniques make it difficult to change the routing behavior of specific traffic. Also known as intelligent routing, policy-based routing allows you to dictate the routing behavior based on a number of different criteria other than destination network, including source interface, source or destination address, or service type.

Consider a company that has two links between locations, one a high bandwidth, low delay expensive link and the other a low bandwidth, higher delay lower expense link. Using traditional routing protocols, the higher bandwidth link would get most if not all of the traffic sent across it based on the metric savings obtained by the bandwidth and/or delay (using EIGRP or OSPF) characteristics of the link. Policy-based routing can route higher priority traffic over the high bandwidth/low delay link while sending all other traffic over the low bandwidth/high delay link.

With policy-based routing, Deep Edge can route traffic from multiple ISPs and WANs. The following illustration shows how to configure Deep Edge for two ISPs using an L2 switch.