Configuration Guide

©2021 Extreme Networks, Inc. All rights reserved
October 2021
17
1. Overview
1.1 Evolution of Ethernet Bridging
The evolution of Ethernet technologies continues with the IEEE 802.1aq standard of Shortest Path Bridging.
This next generation virtualization technology will revolutionize the design, deployment and operations of
the Enterprise Campus core networks along with the Enterprise Data Centre. The benefits of the technology
will be clearly evident in its ability to provide massive scalability while at the same time reducing the
complexity of the network. This will make network virtualization a much easier paradigm to deploy within
the Enterprise environment.
Shortest Path Bridging brings the features and benefits required by Carrier grade deployments to the
Enterprise market without the complexity of alternative technologies traditionally used in Carrier
deployments (typically MPLS).
The IEEE has been working on Layer 2 virtualization techniques over the last decade. It had standardized
a set of solutions that built on each other and continuously addressed the predecessor’s disadvantages.
In 1998, IEEE 802.1Q provided a simple way to virtualize Layer 2 broadcast domains by using VLAN
tagging to form Virtual LANs. The 12 bits that are available in the 802.1Q defined header provided the ability
to separately transport 4096 individual virtual LANs.
The loop free topology had been provided through IEEE 802.1D spanning tree and later rapid spanning
tree (RSTP) and multiple spanning tree (MSTP) extensions. However, spanning tree is not the technology
of choice for large network deployments.
Carrier deployments wanted to leverage the cost points of Ethernet and wanted to use the virtual LAN
technology. In order to improve scalability, the IEEE introduced the QinQ approach, where the header had
been extended to provide a carrier tag attached to a customer tag (QinQ). This allowed the carrier to
transport customer tagged traffic over its Ethernet based 802.1ad backbone. However in large deployments
this technology did not scale well, because the carrier’s backbone still “saw”, and thus learned, all the end-
customer MAC addresses (C-MAC).
In order to overcome this scaling limitation, the IEEE standardized 802.1ah (also known as Provider
Backbone Bridging – BCB) in 2008 which introduced a new header encapsulation to hide the customer
MAC addresses inside an additional backbone MAC header (MACinMAC encapsulation).
In addition to this, the new header also includes a service instance identifier (ISID) with a length of 24 bits.
This ISID can be used to identify any virtualized traffic across an 802.1ah encapsulated frame. In 802.1ah,
these ISIDs are used to virtualize VLANs across a BCB network. The “hiding/encapsulating” of customer
MAC addresses in backbone MAC addresses greatly improves network scalability (no end-user C-MAC
learning required in the core) and also significantly improves network robustness (loops have no effect on
the backbone infrastructure).
So BCB addressed the scaling issues of virtualizing and transporting VLANs across a provider backbone.
Yet, within that backbone, even with BCB, the loop free topology still had to be provided by 802.1D
Spanning Tree (or RSTP or MSTP).
With the latest 802.1aq Shortest Path Bridging MacInMac (SPBM) standard this final limitation is being lifted
via the development of a new link-stated based technology.