SPBM has a straightforward architecture that simply forwards encapsulated C-MACs across the backbone. Because the B-MAC header stays the same across the network, there is no need to swap a label or perform a route lookup at each node. This architecture allows the frame to follow the most efficient forwarding path from end to end.
The following reference architectures illustrate SPBM with multiple switches in a network.
For information about solution-specific architectures like Video Surveillance or Data Center implementation using the switch, see Solution-Specific Reference Architectures.
The following figure shows the MAC-in-MAC SPBM domain with BEBs on the boundary and BCBs in the core.
The following figure illustrates an existing edge that connects to an SPBM core.
The boundary between the MAC-in-MAC SPBM domain and the 802.1Q domain is handled by the BEBs. At the BEBs, VLANs or VRFs are mapped into I-SIDs based on the local service provisioning. Services (whether Layer 2 or Layer 3 VSNs) only need to be configured at the edge of the SPBM backbone (on the BEBs). There is no provisioning needed on the core SPBM nodes.
Provisioning an SPBM core is as simple as enabling SPBM and IS-IS globally on all the nodes and on the core facing links. To migrate an existing edge configuration into an SPBM network is just as simple.
All BEBs that have the same I-SID configured can participate in the same VSN. That completes the configuration part of the migration and all the traffic flows return to normal operation.
For Layer 3 virtualized routing (Layer 3 VSN), map IPv4-enabled VLANs to VRFs, create an IP VPN instance on the VRF, assign an I-SID to the VRF, and then configure the desired IP redistribution of IP routes into IS-IS.
For Layer 2 virtualized bridging (Layer 2 VSN), identify all the VLANs that you want to migrate into SPBM and assign them to an I-SID on the BEB.