Configuration Guide

©2021 Extreme Networks, Inc. All rights reserved
October 2021
19
1.2 SPB Benefits
The benefits that SPB brings to the Enterprise network can be listed as follows.
 Backbone provisioning simplicity
Provisioning an SPB core is as simple as enabling SPB and IS-IS globally on all the nodes and on the core
facing links. The IS-IS protocol operates at layer 2, it does not need IP addresses configured on the links
to form IS-IS adjacencies with neighboring switches (like OSPF does). Hence there is no need to configure
any IP addresses on any of the core links.
 Natively provides virtualized Layer 2 services
Layer 2 virtualization is handled by the Backbone Edge Bridges (BEBs) where the end-user VLAN is
mapped into a Backbone Service Instance Identifier (ISID) by local provisioning. Any BEB that has the
same ISID configured can participate in the same L2 virtual services network (VSN). IS-IS within the SPB
backbone is used as the Layer 2 routing protocol to forward traffic between the BEB and Provider Backbone
Core Bridges (BCBs). Only the BEB has knowledge of the L2 VSN and corresponding MAC addresses.
The BCB only has knowledge of each Backbone MAC address (B-MAC) used to send traffic across an SPB
network.
 Natively provides virtualized routing services
Layer 3 virtualized routing is handled by the Backbone Edge Bridges (BEBs) where the end-user IPv4
enabled VLAN or VLANs are mapped to a Virtualized Routing and Forwarding (VRF) instance. The VRF in
turn is mapped into a Backbone Service Instance Identifier (ISID) by local provisioning. Any BEB that has
the same ISID configured can participate in the same L3 virtual service network (VSN). IS-IS within the
SPB backbone is used as the Layer 2 routing protocol to forward traffic between the BEB and Backbone
Core Bridges (BCB). Only the BEB has knowledge of the L3 VSN and corresponding IP/ARP/MAC
addresses. The BCB only has knowledge of each Backbone MAC address (B-MAC) used send traffic
across an SPB network.
 Adapts to any physical layer / fibre plant
IS-IS is a link-state protocol which will compute the shortest open path just like OSPF does. It can therefore
be deployed on any regular (e.g. square or fully meshed core-to-distribution topologies) or irregular (e.g.
ring topologies) fibre plants.
Whereas OSPF computes the shortest path to destination subnets and then populates the IP routing table
with the results, IS-IS (as used with SPB) computes the shortest path to backbone node MAC addresses
(B-MACs) and then populates the backbone MAC tables.
 Robust/Scalable link-state routing applied to MAC tables
With SPB, the MAC table is now only populated by the IS-IS control plane. The conventional Ethernet
bridging behavior which consisted of (a) “learning” the MAC tables with the source MAC address of packets
seen arriving on local ports and (b) flooding unknown and broadcast traffic to all ports no longer apply in an
SPB backbone.
Furthermore, with SPB, IS-IS is leveraged to build source based forwarding trees for the delivery of
multicast and broadcast traffic across the SPB backbone in such a way that the replication of
broadcast/multicast traffic within the core is optimized to follow the shortest path from source to leaf nodes.
 Separation between Services and Backbone
Since SPB leverages the MACinMAC encapsulation of 802.1ah (BCB) only the nodes at the edge of the
SPB backbone (the Backbone Edge Bridges - BEBs) need to learn the MAC addresses (C-MACs) used
within the transported Customer VLANs (L2VSNs). These same nodes, when forwarding traffic into the
SPB core will always re-encapsulate the service traffic in a Backbone MAC header with a destination B-