Use Multitopology Routing (MTR) to configure service differentiation through class-based forwarding. Two primary components comprise MTR configuration: independent topology configuration and traffic classification configuration.
A topology is defined as a subset of devices and links in a network for which a separate set of routes is calculated. The entire network itself, for which the usual set of routes is calculated, is known as the base topology. The base topology (or underlying network) is characterized by the Network Layer Reachability Information (NLRI) that a device uses to calculate the global routing table to make routing and forwarding decisions. The base topology is the default routing environment that exists prior to enabling MTR.
Any additional topologies are known as class-specific topologies and are a subset of the base topology. Each class-specific topology carries a class of traffic and is characterized by an independent set of NLRI that is used to maintain a separate Routing Information Base (RIB) and Forwarding Information Base (FIB). This design allows the device to perform independent route calculation and forwarding for each topology.
MTR creates a selection of routes within a given device upon which to forward to a given destination. The specific choice of route is based on the class of the packet being forwarded, a class that is an attribute of the packet itself. This design allows packets of different classes to be routed independently from one another. The path that the packet follows is determined by classifiers configured on the devices and interfaces in the network. The figure below shows a base topology, which is a superset of the red, blue, and green topologies.
Figure 1. MTR Base Topology
The figure below shows an MTR-enabled network that is configured using the service separation model. The base topology (shown in black) uses NLRI from all reachable devices in the network. The blue, red, and purple paths each represent a different class-specific topology. Each class-specific topology calculates a separate set of paths through the network. Routing and forwarding are independently calculated based on individual sets of NLRI that are carried for each topology.
Figure 2. Defining MTR Topologies
The figure below shows that the traffic is marked at the network edge. As the traffic traverses the network, the marking is used during classification and forwarding to constrain the traffic to its own colored topology.
Figure 3. Traffic Follows Class-Specific Forwarding Paths
The same topology can have configured backup paths. In the figure below, the preferential path for the voice topology is represented by the solid blue line. In case this path becomes unavailable, you can configure MTR to choose the voice backup path represented by the dotted blue line. Both of these paths represent the same topology and none overlap.
Figure 4. MTR Backup Contingencies Within a Topology
The figure below shows the MTR forwarding model at the system level. When a packet arrives at the incoming interface, the marking is examined. If the packet marking matches a topology, the associated topology is consulted, the next hop for that topology is determined, and the packet is forwarded. If there is no forwarding entry within a topology, the packet is dropped. If the packet does not match any classifier, it is forwarded to the base topology. The outgoing interface is a function of the colored route table in which the lookup is done.
Figure 5. MTR Forwarding at the System Level
MTR is implemented in Cisco software according to a address family and subaddress family basis. MTR supports up to 32 unicast topologies (including the base topology) and a separate multicast topology. A topology can overlap with another or share any subset of the underlying network. You configure each topology with a unique topology ID. You configure the topology ID under the routing protocol, and the ID is used to identify and group NLRI for each topology in updates for a given protocol.