Downtime—planned or unplanned—is generally considered a bad thing, especially when the network is large, and the traffic through the network is business critical. There are two ways to reduce the impact of a network topology change: route around the change as quickly as possible, or try to find some way to not report the change to the rest of the network through continuous forwarding and quick recovery.
Nonstop Forwarding and Fast Convergence
These two options are, generally, mutually exclusive, or complementary. Routing around a topology change very quickly typically involves discovering the topology change as soon as possible and then propagating information about the change to the other intermediate systems in ...