Fabric Networking

These materials are © 2020 John Wiley & Sons, Inc.
Any dissemination, distribution, or unauthorized use is strictly prohibited.

Fabric
Networking
Extreme Networks Special Edition
by Sara Perrott
These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
Fabric Networking For Dummies®, Extreme Networks
Special Edition
Published by
John Wiley & Sons, Inc.
111 River St.
Hoboken, NJ 07030-5774
www.wiley.com
Copyright © 2020 by John Wiley & Sons, Inc., Hoboken, New Jersey
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any
means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted under Sections
107 or 108 of the 1976 United States Copyright Act, without the prior written permission of the Publisher. Requests
to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111
River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/
permissions.
Trademarks: Wiley, For Dummies, the Dummies Man logo, Dummies.com, and related trade dress are trademarks
or registered trademarks of John Wiley & Sons, Inc. and/or its affiliates in the United States and other countries,
and may not be used without written permission. Extreme Networks and the Extreme logo are trademarks or
registered trademarks of Extreme Networks, Inc. All other trademarks are the property of their respective owners.
John Wiley & Sons, Inc., is not associated with any product or vendor mentioned in this book.
LIMIT OF LIABILITY/DISCLAIMER OF WARRANTY: THE PUBLISHER AND THE AUTHOR MAKE NO

REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE ACCURACY OR COMPLETENESS OF THE
CONTENTS OF THIS WORK AND SPECIFICALLY DISCLAIM ALL WARRANTIES, INCLUDING WITHOUT
LIMITATION WARRANTIES OF FITNESS FOR A PARTICULAR PURPOSE. NO WARRANTY MAY BE CREATED
OR EXTENDED BY SALES OR PROMOTIONAL MATERIALS. THE ADVICE AND STRATEGIES CONTAINED
HEREIN MAY NOT BE SUITABLE FOR EVERY SITUATION. THIS WORK IS SOLD WITH THE UNDERSTANDING
THAT THE PUBLISHER IS NOT ENGAGED IN RENDERING LEGAL, ACCOUNTING, OR OTHER PROFESSIONAL
SERVICES. IF PROFESSIONAL ASSISTANCE IS REQUIRED, THE SERVICES OF A COMPETENT PROFESSIONAL
PERSON SHOULD BE SOUGHT. NEITHER THE PUBLISHER NOR THE AUTHOR SHALL BE LIABLE FOR
DAMAGES ARISING HEREFROM. THE FACT THAT AN ORGANIZATION OR WEBSITE IS REFERRED TO IN
THIS WORK AS A CITATION AND/OR A POTENTIAL SOURCE OF FURTHER INFORMATION DOES NOT MEAN
THAT THE AUTHOR OR THE PUBLISHER ENDORSES THE INFORMATION THE ORGANIZATION OR WEBSITE
MAY PROVIDE OR RECOMMENDATIONS IT MAY MAKE. FURTHER, READERS SHOULD BE AWARE THAT
INTERNET WEBSITES LISTED IN THIS WORK MAY HAVE CHANGED OR DISAPPEARED BETWEEN WHEN
THIS WORK WAS WRITTEN AND WHEN IT IS READ.
For general information on our other products and services, or how to create a custom For Dummies book for your
business or organization, please contact our Business Development Department in the U.S. at 877-409-4177,
contact [email protected], or visit www.wiley.com/go/custompub. For information about licensing the For
Dummies brand for products or services, contact BrandedRights&[email protected].
ISBN 978-1-119-61937-6 (pbk); ISBN 978-1-119-61940-6 (ebk)
Manufactured in the United States of America
10 9 8 7 6 5 4 3 2 1
Publisher’s Acknowledgments
We’re proud of this book and of the people who worked on it. For details on how to
create a custom For Dummies book for your business or organization, contact info@
dummies.biz or visit www.wiley.com/go/custompub. For details on licensing the For
Dummies brand for products or services, contact BrandedRights&[email protected].
Some of the people who helped bring this book to market include the following:
Project Editor: Martin V. Minner Business Development Representative:
Karen Hattan
Acquisitions Editor: Ashley Coffey
Production Editor: Tamilmani Varadharaj
Editorial Manager: Rev Mengle
Table of Contents
INTRODUCTION................................................................................................ 1
About This Book.................................................................................... 1
Icons Used in This Book........................................................................ 2
CHAPTER 1: Recognizing the Need for More Flexible

and Automated Networking............................................... 3
Analyzing the State of Network Technology...................................... 3
Learning About the Types of Fabric Networks.................................. 5
Identifying the Benefits of Fabric Networking................................... 6
Understanding Adjacent Networking Technologies......................... 7
CHAPTER 2: Exploring the Use Cases for

Fabric Networking....................................................................... 9
Understanding Data Center Use Cases for Fabric Networking....... 9
Checking Out Fabric Networking in the Campus............................ 12
CHAPTER 3: Exploring Fabric Networks................................................. 15

Introducing Shortest Path Bridging................................................... 15
Understanding How SPB Works........................................................ 17
Building the SPB fabric.................................................................. 17
Deploying SPB services................................................................. 17
Examining the Security Attributes of SPB........................................ 20
Looking at Complementary Technologies........................................ 21
Dynamic Auto-Attach..................................................................... 21
Seamless extension of SPB........................................................... 22
Designing multi-site networks with mobile endpoints............. 22
Seeing Where SPB Fits in the Network............................................. 22
Introducing TRILL Fabrics................................................................... 23
Seeing Where TRILL Fits in the Network........................................... 24
Delving into IP Fabrics........................................................................ 24
Understanding IP Fabrics................................................................... 24
Building the IP Fabric.......................................................................... 24
Deploying Services.............................................................................. 26
Layer 2 VNI...................................................................................... 26
Layer 3 VNI...................................................................................... 27
Looking at Complementary Features............................................... 27
Seeing Where IP Fabrics Fit in the Network..................................... 28
Table of Contents iii
CHAPTER 4: Examining Key Considerations in Fabric
Technology Evaluations........................................................ 29
Enabling Automation.......................................................................... 30
Network automation..................................................................... 30
Cross-domain automation............................................................ 31
AI/ML driven automation.............................................................. 32
Enhancing Visibility............................................................................. 33
On-box visibility.............................................................................. 34
Network analytics.......................................................................... 34
Integrating Ecosystem Partners........................................................ 35
Improving Security.............................................................................. 35
Integrating Wired/Wireless................................................................ 36
CHAPTER 5: Choosing the Right Fabric Solution

for Your Business........................................................................ 37
Focusing on Your Pain Points............................................................ 37
Evaluating Fabric Networking Technologies.................................... 39
CHAPTER 6: Ten Things to Know About Fabric

Solutions from Extreme Networks.............................. 41
You Have a Choice............................................................................... 41
Single Pane of Glass............................................................................ 42
Visibility from the Edge to Multi-Cloud............................................. 42
Solutions Are Truly Multi-Vendor...................................................... 42
Deployment History............................................................................ 42
Innovation!........................................................................................... 43
Field Proven in a Demanding Environment..................................... 43
Plug and Play........................................................................................ 43
Ecosystem Integrations...................................................................... 43
Unbreakable in Hack-a-thons............................................................ 44
iv Fabric Networking For Dummies, Extreme Networks Special Edition
Introduction
I
n today’s world, businesses grow and change at a pace that can
be difficult to keep up with. This is especially true if you are in
charge of architecting or operating the communications net-
works that enable the business to serve its employees and its
customers.
Networks of the past were designed for simple, static workflows

like connecting an office user to the Internet or to email. Fast
forward to today, and work is something you do, not a place that
you go to. Today’s workplace is dynamic and mobile. It requires a
network that can adapt and change on demand.
Although computing technologies have matched the needs of

modern organizations, networking technology has not progressed
at the same pace. Many organizations still use a legacy network
infrastructure that is inadequate for today’s ever-changing
business.
Fear not, network administrators, the next evolution in network-

ing technology is here! Fabric networking allows you to scale on
demand, create redundant paths for high availability and fault
tolerance, improve your security with network segmentation, and
so much more. You can move from an old legacy network to a
modern, scalable, secure network infrastructure that will enable
your business to succeed and grow. The network is no longer the
bottleneck. It can instead be the enabler for success.
About This Book

Fabric Networking For Dummies, Extreme Networks Edition, is your
introduction to the world of fabric networking and the underlying
protocols and technologies that support it.
Fabric networking enables you to scale with demand and create

highly available and fault-tolerant networks. This book explains
what fabric networking is, including its history, use cases, current
technologies available to you, and some of the key points to think
about when you consider fabric networking.
Introduction 1
Icons Used in This Book
As you read this book, notice the icons in the margins. These
indicate information that may be of interest. The material that
accompanies the icons can enrich your understanding of fabric
networking. I highly recommend reading them!
Here’s what the icons mean:
Tips provide guidance that may save you time and effort. These
are typically based on real-world experience and are there to help
you hit the ground running with fabric networking.
The Remember icon indicates information that deserves your

special attention.
2 Fabric Networking For Dummies, Extreme Networks Special Edition
IN THIS CHAPTER
»» Analyzing the state of networking
technology
»» Introducing fabric networking
»» Learning about the types of fabric

networks
»» Understanding adjacent networking

technologies
Chapter 1
Recognizing the Need
for More Flexible and
Automated Networking
T
echnology is advancing at a rapid pace. Robotics, Internet of
Things (IoT) devices, virtual reality, and augmented reality
are all making their way onto your network. Bound by com-
plexity, traditional networks can no longer keep pace with the
speed of innovation. What’s needed is a simpler, automated, and
secure network environment.
This chapter introduces the challenges faced today with tradi-

tional networking technology and how fabric networking can
help. It shows you what fabric networking is, how it advances
networking, and the significant benefits it provides.
Analyzing the State of

Network Technology
As organizations rapidly adopt new technologies, applications,
and delivery models, they’re rendering traditional network design
obsolete. From the influx of a wide variety of IoT and “bring your
CHAPTER 1 Recognizing the Need for More Flexible and Automated Networking 3
own device” (BYOD) devices that require secure connectivity, to
the requests for support of more real-time, bandwidth-intensive
traffic such as video, to the rapid growth of hybrid cloud delivery
models — all of these demands are having a significant impact on
the network.
The challenges with traditional networks include:
»» Outdated designs: Traditional network design has remained

consistent for decades, with many of the common network
protocols originating almost 30 years ago. Back then network-
ing environments were static — applications were tied to
servers and employees to their desktop computers. With server
virtualization, mobility, IoT, and cloud computing more com-
mon, networks are being asked to support dynamic, mobile
environments, and are therefore falling short.
»» Manual configuration: Traditional networking is largely
configured manually, switch by switch, through a command-
line interface (CLI). This method was adequate in a static
networking environment. However, in today’s environment
where new devices and applications are moved, added, or
changed frequently, manual configuration is time consuming.
It also introduces the risk of an outage or even a security
breach resulting from human error during a change.
»» Too much complexity: Traditional network design requires
network administrators to understand and configure many
interdependent protocols. These include Spanning Tree
Protocol (STP), Open Shortest Path First (OSPF), Protocol-
Independent Multicast (PIM), and Border Gateway Protocol
(BGP). All this complexity can also slow network recovery
because each layer is reliant on the layer below it to
re-establish connectivity.
»» Vulnerability to breaches: Security concerns are front and
center in businesses today. For example, you don’t want
guest Wi-Fi devices to communicate over the same network
as human resources systems. In legacy networks, this issue
has traditionally required multiple virtual local area networks
(VLANs) with access controls and firewalls to separate the
traffic. However, if you are sharing a routing table, your IP
network is flat and if someone breaches your network, they
might make their way to sensitive data such as customer
payment information or patient health records.
Learning About the Types
of Fabric Networks
Fabric networks get their name from the diagram of their compo-
nent connectivity, which resembles a piece of fabric. The network
is woven together into a connectivity mesh.
Today, you’re likely to encounter two types of fabric networks:
»» Ethernet fabrics are based on industry standard protocols

such as Shortest Path Bridging (SPB) or Transparent
Interconnection of Lots of Links (TRILL) and use Ethernet-
switched paths to forward traffic. They also use a link state
protocol such as Intermediate System to Intermediate System
(IS-IS) or OSPF as their control plane to bring more determin-
istic, carrier-grade functionality to Ethernet. Ethernet fabrics
overcome the issues associated with traditional Ethernet
network designs. They allow all links to be active with multiple
equal cost paths. They eliminate hop-by-hop provisioning
and they enable very large-scale Ethernet networks to be
implemented, getting past the limitations of traditional VLAN
and Media Access Control (MAC) scaling. Most importantly,
they enable much faster reconvergence times, generally
achieving sub-second network-wide recovery.
Ethernet fabrics define the type of fabric; however, they do
not dictate the type of services or traffic running across it.
Most support full Layer 2 and Layer 3 services within the
fabric for attached devices. Certain Ethernet fabrics even
offer extensions to support integrated virtual routing and
forwarding (VRF) and IP Multicast routing capability. This
enables them to not only be a replacement for STP, but also
other protocols such as OSPF, BGP, PIM, and even
Multiprotocol Label Switching (MPLS).
»» IP fabrics are based on industry standard protocols such as
BGP and Ethernet Virtual Private Network (EVPN). Like
Ethernet fabrics, IP fabrics use equal cost multi-pathing to
improve efficiency. They also support Layer 2 and Layer 3
services for end devices, embrace server virtualization, and
provide automation for complete plug-and-play provisioning.
The main differences between the Ethernet and IP fabrics are
in the control and data planes used to construct the fabric. In
most cases, BGP is used for the underlay (the control plane
network), and BGP/EVPN with Virtual Extensible LAN (VXLAN)
tunnels are used for the overlay (the data plane).
The driver behind Layer 3 fabrics was the use of a single
protocol stack and single virtualization technology. By
leveraging BGP as the underlay, the network can deliver
massive scalability. These fabrics all offer the benefit of full
vendor interoperability for both the underlay and overlay
networks, allowing for vendor-agnostic network
implementations.
Although fabric technology is not new (initial deployments started

as early as 2009), the capabilities have been greatly expanded
upon, giving it renewed relevance when you’re designing a next-
generation network.
Because networking standards continue to evolve, its best to stay

up to date with relevant standards-bodies such as the Institute
of Electrical and Electronics Engineers (IEEE) and the Inter-
net Engineering Task Force (IETF). In fact, the IETF is currently
working on additional IP fabric technologies such as Routing in
Fat Trees (RIFT) and link state vector routing (LSVR).
Identifying the Benefits

of Fabric Networking
When you consider the advantages of fabric networking, these
benefits stand out and make the business case for switching from
traditional to fabric networking:
»» Improved time-to-service: Likely, you’re being asked to

deliver more connectivity and enable network changes at an
increasing pace. Many of these changes require a mainte-
nance window, thus introducing delays into accommodating
the requests. Fabric networking can automate network
changes, allowing you to make them in minutes, rather than
days or weeks.
»» No downtime from human error: Because adding and
removing network services in a fabric network only needs to
be done at the edge, and is typically done from a centralized
management system, mistakes are far less likely and the
network is much more stable.
»» Better resiliency: Fabric networking provides the opportu-
nity to simplify your network by reducing the number of
protocols in use. With a simpler network, recovery times are
faster. Also, more interconnectivity means the loss of a link
or even a network device will not impact your applications.
»» Enhanced security: Fabric networks allow you to easily
implement network segmentation with the creation of
secure zones. This feature prevents lateral movement across
the network. For example, if a system is infected by a virus,
the infected system is not allowed to communicate outside
of its zone or segment.
»» Integrated wired and wireless: With certain fabric tech-
nologies you can fully integrate your wired and wireless
networks to create a unified campus network. This unifica-
tion leads to greater simplicity in deploying wireless APs and
providing a consistent quality of service across both wired
and wireless access.
»» Enhanced quality of experience (QoE): Fabric technologies
use algorithms to calculate the shortest path between any
source and destination. This capability ensures that in any
network, users and devices are connected to their applica-
tions by the shortest and most efficient path to reduce
latency.
Understanding Adjacent
Networking Technologies
This section examines important technologies that are synergistic
to fabrics:
»» Software-defined networking (SDN): SDN is a concept that

is loosely defined in the industry. The original approach
involved an open protocol (OpenFlow) to program network
switches and routers from a centralized controller. Today,
the industry has largely moved away from this approach and
more toward application programming interfaces (APIs) and
automation.
Fabric networking can be deployed independently or as part
of an SDN solution. Independently, fabrics can deliver
inherent automation capabilities. Alternatively, they can be
deployed with a management system or controllers for
centralized control and automation.
»» Network functions virtualization (NFV): NFV virtualizes
network services like routing, switching, load balancing,
security, and wide area network (WAN) optimization so they
can be deployed on commodity hardware. Because fabric
networking is a feature commonly available on routing and
switch OSs, it is synergistic to NFV, since the fabric OS can
run within a hypervisor or a container-based framework
alongside other required network functions like security or
WAN optimization.
»» Multi-Chassis Link Aggregation (MLAG): MLAG logically
aggregates two or more switches to form one logical entity. It
provides link-level and device-level network resiliency and
eliminates single points of failure. Originally designed to
enable an STP-free core network, MLAG can be used in
conjunction with fabric networking for a higher degree of
resiliency, offering the ability to take nodes out of service for
software upgrades and patches without any impact to
availability. It can also provide enhanced resiliency for end
devices such as servers, firewalls, and load balancers.
»» Port extender technology: Port extender technology
replaces traditional fully-featured access layer switches with
simple port extenders that are fully managed and controlled
by an aggregation switch (typically called a controlling bridge).
Port extender technology can be used in conjunction with
fabric networking to provide a high fan-out of ports that are
controlled, operated, and managed by a fabric-enabled
aggregation switch.
IN THIS CHAPTER
»» Understanding data center use cases for
fabric networking
»» Examining campus use cases for fabric

networking
Chapter 2
Exploring the Use Cases
for Fabric Networking
D
igital transformation — and the influx of advanced tech-
nology that it brings onto your network — requires net-
working that easily adapts to changing business needs.
A step in the right direction is the use of fabric networking.
In this chapter, you learn about using fabric networking in cam-

pus and data center environments.
Understanding Data Center Use

Cases for Fabric Networking
Data centers and the technologies housed within them are rapidly
evolving as virtualization, hyperconverged infrastructure (HCI),
containers, and the cloud are becoming more prevalent. Tradi-
tional networks haven’t kept pace with the changing ecosystem in
the data center. With all the manual processes needed to scale up
networking resources, traditional networks tend not to be easily
scalable, agile, or flexible.
CHAPTER 2 Exploring the Use Cases for Fabric Networking 9
This section shows how fabric networking can address the
challenges of the modern-day data center and enable your IT
department to work at cloud speed while becoming more agile
and responsive to business needs.
Identifying the need for a data center

fabric network
Here are a few of the technologies that benefit greatly from the
implementation of fabric networking in the data center:
»» Server virtualization has become common and is the best

way to ensure that you are utilizing your server hardware to
its fullest. One of the initial use cases for fabric networking
was to solve the challenge of stretching virtual LANs (VLANs)
or subnets within and between data centers so that the IP
address of the virtual machine could be maintained regard-
less of what server it moved to. Today, fabric networking
makes server and virtual machine (VM) deployments and
migrations much quicker by offering any service — for
example, VLAN — on any port within the data center.
Automated top-of-rack switch provisioning addresses new
VMs being brought up as well as moved within the network
with no manual network configurations.
»» Hyperconverged infrastructure (HCI) is the convergence
of virtualized compute, storage, and networking services
onto a standard commercial off-the-shelf (COTS) server.
Having a truly integrated solution provides significant value
in capital expenditures as well as ongoing operations and
maintenance. Fabric networking fits within HCI perfectly,
providing an adaptable infrastructure that scales out in
conjunction with compute and storage. Certain fabrics also
inherently support advanced storage technologies to ensure
optimal performance and availability.
»» Containers are one of the newest technologies within the
data center. They have gained significant popularity because
of their ease of use and quick provisioning. Containers,
however, do not have long life spans. They must be spun up
and shut down quickly as workloads change. The network
that supports containers, as well as container hosts, must
be able to adapt and respond to changing needs within
seconds. Fabric networks enable a more agile infrastructure
through automation. When containers are created on a
container host, fabric networks allow you to generate the
resources the containers demand to support their
workloads.
»» IP storage connectivity requires a network that delivers the
bandwidth, performance, and reliability needed in today’s
demanding environment. Fabric networks that support
storage technologies such as Internet Small Computer
Systems Interface (iSCSI), Network File System (NFS), or
Non-Volatile Memory Express (NVMe), either in a dedicated IP
storage network or in a hyperconverged environment, allow
network administrators to architect a flexible, robust network
that scales easily to align with storage expansion needs.
»» The hybrid cloud is becoming a common way for busi-
nesses to expand their data centers by providing additional
capacity for bursting, as well as by hosting applications that
can be easily housed off premises. This is often done by
extending a virtualized infrastructure into a cloud provider’s
environment. Certain fabric technologies can extend into a
cloud provider’s environment to enable a robust, flexible,
and scalable network that works seamlessly on and off
premises — with the ability to manage both through a single
pane of glass.
»» Data Center Interconnect (DCI) connects multiple data
centers together, as shown in Figure 2-1. With multiple data
centers it is critical to have active-active connectivity to ensure
business continuity. Ideally, if all data centers have fabric
networking implemented, and you have dark fiber intercon-
necting them, it’s simple to extend the fabric between sites.
However, in most cases, there is a wide area network (WAN)
network to interconnect the data centers. In this scenario, you
may have to insert a DCI solution. This allows sites to become
one logical data center that has applications and services
dispersed between or across locations. This can be done as
an overlay to the WAN infrastructure you have in place.
When you use a DCI, the technologies in play will most likely
encapsulate traffic across the WAN in Virtual Extensible Local
Area Network (VXLAN), for example. This requires support of
a larger maximum transmission unit (MTU) of 1600 bytes, so
ensure that your WAN and WAN provider will support this before
implementing your DCI.
FIGURE 2-1: A data center interconnect.
Checking Out Fabric Networking

in the Campus
Fabric networks represent the next evolutionary step for cam-
pus networking by offering simplicity, flexibility, and resiliency
with inherent security. The right fabric technology will allow
you to easily automate the secure attachment of IoT devices and
enhance security by reducing the attack surface and preventing
lateral movements. It will also increase agility and streamline
operations with automation of network services and zero-touch
provisioning.
Here are a few of the technologies that benefit greatly from the
implementation of fabric networking in the campus:
»» IoT: One of the largest challenges you will face as a network

operator is IoT. You need to onboard and extend connectiv-
ity to these devices and at the same time ensure security.
Lacking in robust security features, many IoT devices rely on
the network for protection.
IoT is the extension of Internet connectivity into physical
devices and everyday objects. A few examples include
security cameras, medical devices like magnetic resonance
imaging (MRI) machines and infusion pumps, intelligent
lighting systems and other smart building systems, and
programmable logic controllers and sensors.
In addition to being able to easily extend connectivity for these
devices with little to no manual configuration, fabric network-
ing enables you to separate groups of IoT devices into their
own secure zones or network segments, isolating them from
the rest of the network far more easily than traditional VLAN,
firewall, or virtual routing and forwarding (VRF) segmentation.
»» Network segmentation: The increase in frequency and
sophistication of cyber-attacks, combined with new attack
vectors, such as IoT and cloud, requires reexamination of
network security. In fact, companies now assume there will
be a breach, and focus on minimizing the damage that can
occur as a result. Network segmentation ensures that if a
breach takes place, it is contained to where it occurred.
Fabric networking allows you to segment the network at
scale. Segments can be used to isolate IoT devices or groups
of users, and segregate critical or sensitive information to
assist with compliance and regulatory requirements.
Security professionals agree that network segmentation is a
must for modern networks. By segmenting your network,
you provide a greater degree of protection to your business’s
most valuable assets.
»» Integrated wired and wireless: In the era of BYOD, you
need to enable your workers to do their jobs while you also
keep the network secure. Fabric networks allow you to easily
segment your network and create a special segment for
untrusted devices, like personal cell phones and laptops, that
is completely isolated from the rest of the enterprise network.
Furthermore, some fabric technologies are designed to
extend to the wireless network to provide a more unified
wired and wireless network. This enables the unified and
dynamic attachment of users and devices to fabric services to
dramatically simplify management and operations.
»» IP multicast: Enterprise applications that rely on IP multicast
can be a challenge to network managers. Applications, such
as video streaming, IP television (IPTV), digital signage,
software distribution, and others, rely on IP multicast to
distribute traffic from a single source to multiple destinations.
The technologies required to make multicast work over a
traditional network are complicated, involving protocol
overlays that must be kept meticulously in sync. These
protocols are difficult to configure and troubleshoot,
reconvergence times can be slow, and scalability can often
be limited. Some fabric technologies excel in areas such as
multicast and dramatically simplify deployment by making it
easier to configure, faster to reconverge, and easier to scale.
»» IP video surveillance: IP video surveillance is an application
that is transitioning to IP multicast. The challenge is that IP
multicast was designed for applications like IPTV, where a
single source sends traffic to multiple destinations. IP video
surveillance, on the other hand, typically involves many
sources (IP cameras) sending traffic to just a few destina-
tions. Since IP multicast wasn’t specifically designed to
address this scenario, implementation can be even more
challenging.
The right fabric networking solution can support even
this very complex form of multicast, making the network
scalable, resilient, and far simpler to deploy and operate so
that a surveillance deployment works seamlessly.
»» Edge computing: Edge computing is a development on the
rise due to the growth of IoT and sensors. It is the migration
of compute to the edge of the network, away from central-
ized data centers, in order to reduce the distance that the
data must travel.
Fabric networking can deliver on the network being more
agile, reliable, and flexible for edge computing.
Figure 2-2 shows an example of campus fabric networking.
FIGURE 2-2: Campus fabric networking.
IN THIS CHAPTER
»» Understanding Shortest Path Bridging
(SPB)
»» Looking at complementary technologies
»» Understanding IP fabrics
Chapter 3
Exploring Fabric
Networks
C
hapter 1 introduces the two main types of fabrics in the
industry today: Ethernet-based fabrics, normally built with
either Shortest Path Bridging (SPB) or Transparent
Interconnection of Lots of Links (TRILL), and IP-based fabrics,
normally built with Border Gateway Protocol (BGP) and Ethernet
Virtual Private Network (EVPN).
In this chapter, you learn about the key aspects of each of these
types of fabrics. After completing this chapter, you will under-
stand how these fabrics work, what the main values of them are,
and where they fit in the network.
Introducing Shortest Path Bridging

Shortest Path Bridging (SPB) is a widely adopted fabric technol-
ogy. Although two variants of SPB exist (SPBV and SPBM), SPBM
is where the current momentum lies and is what this book dis-
cusses. As shown in Figure 3-1, SPB is an Ethernet-based fabric
where all networking services, whether Layer 2, Layer 3, IPv4,
CHAPTER 3 Exploring Fabric Networks 15
IPv6, or multicast, are virtualized and decoupled from the physi-
cal infrastructure.
FIGURE 3-1: Shortest Path Bridging (SBP).
SPB enables businesses to implement flexible, scalable networks.

Here are some of the characteristics of a fabric network that uses
SPB:
»» Multiple physical topologies are supported, including mesh,

partial-mesh, and rings. There are no blocked paths; thus all
links are always forwarding.
»» Logical loops in the network are eliminated.
»» The control plane calculates the shortest path through the
fabric and therefore removes the flooding and learning of a
typical Ethernet network.
»» SPB consistently delivers sub-second failover and recovery
for all network services with a single network protocol.
»» SPB supports edge-only provisioning that is done in real
time with minimal configuration with dynamic auto-attach,
enabling authenticated endpoints to connect seamlessly to
fabric-based services.
»» SPB enables all network services with a single control plane.
Understanding How SPB Works
This section introduces some of the key concepts that explain how
SPB works.
Building the SPB fabric

SPB uses Intermediate System to Intermediate System (IS-IS) as
the control plane to create a “stateful” network topology. With
IS-IS, each switch advertises itself to all other nodes in the fabric
area so that each has a complete network topology map.
The standard defines a 24-bit service ID (I-SID). An I-SID is a

unique service identifier, used within the SPB fabric, that can
be extended to any service whether it is Layer 2 or Layer 3. It
overcomes traditional virtual LAN (VLAN) scaling limitations and
enables the creation of secure network segments.
SPB packet headers are based on Ethernet Media Access Control

(MAC) layer addresses. However, when bridging traffic, it does
not use typical flooding and learning mechanisms. Instead, it uses
a routing protocol (IS-IS) to populate all forwarding entries. This
results in a predictable and robust fabric infrastructure that is not
prone to network loops.
Packets are forwarded through the fabric using the backbone MAC
addresses. The user packet header is encapsulated. It is used only
for forwarding, and only at the fabric boundary.
SPB is standardized by both the Institute of Electrical and Elec-

tronics Engineers (IEEE) and the Internet Engineering Task Force
(IETF). In the IEEE, it was originally standardized as IEEE 802.1aq.
IEEE 802.1aq was then added to IEEE 802.1Q as an amendment.
The Layer 3 extensions to SPB are defined in RFC 6329.
Deploying SPB services

Typically, network operators think in terms of constructs like
VLANs, IP subnets, and virtual routing and forwarding (VRF).
Those connectivity configurations are treated as network ser-
vices in an SPB infrastructure and are named virtual service net-
works (VSNs) and are comparable with virtual network identifiers
(VNIs) in an IP fabric network. VLAN extensions across a network
infrastructure are called L2 VSNs, and VRFs extended across the
network are called L3 VSNs. The ease of extending VSNs across an
SPB network, the use of a single protocol, and the clear separation
of the services or VSNs from the underlying fabric infrastructure
ensure the simplicity and agility of the SPB solution.
Layer 2 virtualized network services (L2 VSNs) allow you to take

Layer 2 VLANs and extend them anywhere in the network — even
across geographical distances. These Layer 2 segments can either
be E-Line (point-to-point) services, E-Tree (private LAN) ser-
vices, or E-LAN (any-to-any) services.
At the source edge nodes of the network, VLANs are mapped to

SPB I-SIDs, as shown in Figure 3-2. The I-SID provides the back-
bone connectivity to interconnect the L2 VSN service end points.
Then at the terminating edge node or nodes, the I-SID is mapped
to the VLAN.
FIGURE 3-2: SBP Layer 2 service.
Layer 3 virtualized network services (L3 VSNs) enable enter-

prises to separate portions of the network into logical segments,
restricting communications to those members of that segment.
Layer 3 virtual private networks have their own routing topology,
providing a high degree of isolation. They are also scalable and
easy to deploy. Layer 3 VSNs are useful in multi-tenant environ-
ments as well as environments where IoT devices, applications,
user groups, or critical information must be in its own secure seg-
ment for security purposes.
At the source edge nodes of the network, Layer 3 VRFs are mapped
to I-SIDs, as shown in Figure 3-3. IS-IS then advertises the ser-
vice and IP routes only where the VRF IP routes are needed.
These routes are installed only on nodes that contain the same
I-SID. This is all done natively within the SPB fabric without
requiring any additional routing protocols.
FIGURE 3-3: SPB Layer 3 service.
Any SPB service is enabled simply by provisioning at the edge of

the fabric, and IS-IS dynamically establishes the shortest path
between the service endpoints.
IP shortcuts leverage the global routing table to forward IP pack-

ets directly over the SPB network. Conceptually they are similar
to Layer 3 VSNs; however, the difference is that rather than being
mapped to an I-SID, the IP encapsulation is mapped directly to
an Ethernet header, as shown in Figure 3-4. A route look-up is
done on the edge to determine the packet’s destination. The edge
node adds the Ethernet header that contains the destination MAC
address. The fabric then efficiently cut-through switches to the
destination node without any hop-by-hop route look-ups.
FIGURE 3-4: IP shortcuts.
Layer 3 VSNs and IP shortcuts support both IPv4 and IPv6 routing.
IP multicast services bring complexity with all the configu-

ration that goes along with setting up Protocol-Independent
Multicast (PIM) or Distance Vector Multicast Routing Protocol
(DVMRP). With SPB, you don’t have to worry. You get high-
performance, multicast functionality but without the complex
configuration overhead. Multicast over SPB also offers supe-
rior performance for any type of multicast deployment whether
it is one-to-many (IPTV), many-to-few (video surveillance),
or many-to-many. When a multicast stream is received, it is
mapped to a dedicated multicast service identifier (or I-SID), as
shown in Figure 3-5. IS-IS then advertises the I-SID to the rest
of the fabric and forwards it only to nodes that register to receive
it through IGMP. If a node does not request the stream, it is not
forwarded, enabling far more efficient distribution of multicast
traffic than in a traditional network.
IP multicast can be enabled over both L3 and L2 VSNs or over the

global routing table as IP multicast shortcuts.
FIGURE 3-5: Multicast services.
Examining the Security Attributes of SPB

One of the biggest challenges with traditional IP network design is
that the entire routing table is exposed to every device and every
endpoint in the network. This characteristic leads to the ability
for a malicious attacker to gain entry to the network, discover the
network topology using common IP scanning mechanisms, then
move through the network to access your valued data.
In an SPB network, this vulnerability is eliminated in many ways.

First, routing is enabled only at the network edges. It doesn’t
reside in the network aggregation or core. Rather than forwarding
based on IP look-ups, traffic is forwarded using Ethernet switched
paths. Therefore, anyone running an IP scan against an SPB envi-
ronment gets a list of IP subnets all showing just a single hop to the
egress of the network. Everything in between is “dark.” This inabil-
ity to discover the topology of the network makes it nearly impos-
sible for hackers to laterally move to sensitive areas of the network.
The other important piece to security is the ability to segment

the routing tables through Layer 3 VSNs. These secure segments
run as ships in the night without any awareness of each other
and without allowing any access, in or out, unless otherwise con-
figured. Having a highly segmented environment ensures that
if the network is breached, that breach is contained to where it
occurred, thus minimizing damage.
This inherent security extends to Layer 2 network design. In

traditional Ethernet networks, endpoint MAC addresses such as
those belonging to a user’s computer or to an IoT device are visi-
ble throughout the network. In SPB, endpoint MAC addresses are
only visible on the fabric edge switches.
The final way SPB enhances an organization’s security posture is

by eliminating back-door entry points to the network. SPB VSNs
can be elastic in nature; they can extend and retract as authen-
ticated users and devices connect to the network and disconnect.
When a user disconnects from a switch port and access to the seg-
ment is no longer required, residual configuration is automati-
cally deleted on the edge switches. This practice not only removes
the delays and risks associated with manually configured conven-
tional networks — it also eliminates the risk of a back-door entry
point to the network.
Looking at Complementary Technologies

Once an SPB core is in place, you can deploy additional features to
increase the reach and the value of the deployment.
Dynamic Auto-Attach
Auto-Attach (IEEE draft P802.1Qcj) provides for automatic attach-
ment of users, devices, and virtual machines (VMs) to connect to SPB
services or I-SIDs. It uses extensions to the IEEE802.1AB Link Layer
Discovery Protocol (LLDP) to automatically attach network devices
to I-SIDs or VSNs in an SPB network. This auto-attach capabil-
ity can be deployed on endpoints, such as IP surveillance cameras;
wireless APs; and/or non-SPB-compliant access layer switches so
that seamless communication with the SPB fabric is possible.
Seamless extension of SPB
Fabric Extend is a feature that enables SPB to be extended over
third-party IP networks whether they are IP cores or service pro-
vider wide area networks (WANs). Any SPB service, whether it is
L2, L3, or multicast, can be seamlessly extended across the WAN or
across the IP core easily — without the WAN or IP core having vis-
ibility to those services.
Fabric Extend supports an IP mode, an L2 mode or an encrypted

mode. It works by creating an overlay of point-to-point tun-
nels using either Virtual Extensible Local Area Network (VXLAN)
native L2 tunnels for IP mode or encrypted IPSec tunnels.
Designing multi-site networks with

mobile endpoints
Distributed Virtual Routing (DVR) is an SPB feature that improves
the scale and performance of multi-site networks with mobile
end points. These can be either VMs or wireless clients at the net-
work edge.
DVR’s main value is that it eliminates the “tromboning” problem

that can occur when a user’s or device’s IP subnet is stretched far
away from its default gateway. DVR distributes the routing func-
tion to all switches that have a presence in the IP subnet so that
the default gateway is always available at the first network hop.
At the same time, it keeps provisioning simple because the rout-
ing configuration is performed only on centralized controllers or
spine/ aggregation nodes.
Seeing Where SPB Fits in the Network

The earliest implementations of SPB were focused on Metro
Ethernet services. The technology then had uptake within the
data center to solve the challenges with VM mobility.Today, most
of the momentum for SPB is for campus networking and collapsed
or converged data center/campus environments.
SPB’s dynamic auto-attach capability, inherent network secu-

rity, and integrated high-performance multicast make it ideal for
these environments.
Though multi-vendor interoperability tests have been done and
are publicly documented, not all implementations of SPB are
the same. This chapter provides a guide as to what the technol-
ogy is capable of; however, due diligence must be done to ensure
that the vendor you are working with supports a comprehensive
implementation of SPB.
Introducing TRILL Fabrics

If you’re a Star Trek fan, I know you got excited when you saw this
section. But if you were expecting a discussion on fashion trends
concerning your favorite humanoid aliens with spots, you are in
the wrong place. TRILL stands for Transparent Interconnection of
Lots of Links. The TRILL algorithm was written by Radia Perl-
man, who also wrote the algorithm for Spanning Tree Protocol
(STP). Like SPB, TRILL was designed to address the shortcomings
of STP; specifically, to remove blocked links and to be topology
agnostic without having logical loops in the network.
Another major advantage of TRILL technology is simplicity. This

technology abstracted away the protocols and the need for intense
manual configurations. These advantages helped this new way
of networking gain traction quickly. Switches and routers that
implement TRILL are referred to as routing bridges (Rbridges).
When connected, they automatically form a fabric. As services
(such as VLANs) are provisioned at the edge ports, they are auto-
matically learned across the fabric, with no need to tag links
between switches. When a new end device is attached to the fab-
ric, the entire fabric is made aware of the device’s MAC along with
which switch and ports it resides on. Any other device needing
to communicate with it already knows where it is attached. Thus
there is no need to flood traffic looking for it.
Like SPB, TRILL allows you to take advantage of the simple con-
figuration that you get with an Ethernet network, but it uses a
link-state routing protocol (control plane) to avoid traditional
Ethernet-based flooding and learning. The link-state proto-
col determines the shortest available path to the destination and
routes packets accordingly. The most common link-state proto-
cols used with TRILL fabrics include IS-IS, Open Shortest Path
First (OSPF), and Fabric Shortest Path First (FSPF).

Seeing Where TRILL Fits in the Network
Over the past several years, the concept of any service on any port,
the ability to provide a reliable infrastructure for storage technol-
ogies, and ecosystem integrations with hypervisor vendors such
as VMware, make TRILL a fit for the virtualized data center. This
technology continues to be a viable fabric technology solution for
the data center today; however, there is now an alternative solu-
tion with IP fabrics.
Delving into IP Fabrics

As data center networking continues to evolve, there is a growing
need to implement fabric technology that is not only standards-
based, but also fully interoperable among vendors. At the same
time, providers who already had an Multiprotocol Label Switching
(MPLS)-based IP-VPN solution wanted to extend this technology
into the data center without the complexity of MPLS. This was
the birth of the IP fabric, based on BGP as the control plane (now
called the underlay) and BGP/EVPN as the data plane (now called
the overlay) using VXLAN.
The underlay can use other routing protocols such as OSPF or

IS-IS if desired.
Understanding IP Fabrics
This section goes into the details of how the IP fabric is imple-
mented and how services are provisioned in and across the fabric.
Although IP fabrics are based on BGP, it is only a small subset of

the protocols that are necessary for the fabric.
Building the IP Fabric

The industry has moved into a common architecture for data cen-
ter fabric networking by using the spine/leaf topology. This can
take the form of a three-stage Clos (pronounced “clo” from its
originator Frederick Clos), which is a simple spine and leaf, as
shown in Figure 3-6. Scaling this design out, the topology can
move into a five-stage Clos with the addition of a super-spine
that connects three-stage Clos points of delivery (PoDs). Moving
to a consistent architecture makes traffic flow across the network
predictable and much easier to troubleshoot.
FIGURE 3-6: Three-stage and five-stage Clos designs.
One key aspect of the IP fabric is its use of BGP, which is a well-
known and well-understood technology. BGP forms the underlay
network for the fabric. This is how all switches and routers con-
nect with each other and share routing and topology informa-
tion. In the fabric, every leaf connects to every spine and forms a
BGP neighborship. The fabric uses equal-cost multi-path routing
(ECMP) to distribute traffic across all links, which provides the
bandwidth and resiliency required in the modern data center.
Deploying Services
Once the fabric infrastructure is in place, the next step is to ena-
ble services on the fabric to support connections from end devices
such as servers, storage, and appliances. The overlay is used to
extend services across the fabric. BGP/EVPN using VXLAN makes
this extension simple and easy, as shown in Figure 3-7. The default
gateway for every end device is at the leaf (top of rack) using a
static anycast gateway. When VLANs or VRFs need to extend beyond
the leaf, a VXLAN tunnel is used. Each leaf or leaf pair becomes a
VXLAN tunnel endpoint (VTEP). This arrangement allows the leaf
to encapsulate VLANs to VXLAN tunnels and do the reverse, taking
VXLAN tunnels and breaking out the VLANs. As per the standard,
the VTEPs use an auto-discovery mechanism to create tunnels
across the fabric, thus eliminating manual configuration.
FIGURE 3-7: BGP-EVPN (VXLAN) design.
The creation of a Layer 2 or Layer 3 VNI facilitates the transport of

traffic as described in the next two sections. Both of these concepts
use Integrated Routing and Bridging (IRB). Asymmetric IRB is the
case where the same VLANs exist on both ends of the VXLAN tunnel.
Routing is done on the ingress leaf and then traffic is sent across an
L2 VNI to the destination leaf (route first, bridge second). Symmetric
IRB is used when the same VLANs don’t exist on both ends of the
VXLAN tunnel. On ingress, the VLAN routes to its destination via the
L3 VNI to communicate to a different VLAN across the VXLAN tunnel.
Layer 2 VNI
VLANs entering the leaf from the edge device are mapped into
a VXLAN tunnel using a VNI. In most cases, it’s easy to map the
VLAN ID to the VNI ID without creating a complicated mapping
structure. The VLANs are placed into an EVPN instance on the
leaf and VXLAN tunnels are automatically created to all other leaf
switches in the fabric where that VLAN exists. You don’t need
an external controller or manual mapping of tunnels; it’s all
automatic.
Layer 3 VNI
When there is a need to extend VRFs between leaf switches,
a Layer 3 VNI is created. This is a dedicated VLAN with routing
enabled on it that is used as the “connector” between leaf switches
for that VRF. This allows VLANs within VRFs to communicate
across the fabric and eliminates the need for any additional rout-
ing protocol for this purpose.
Looking at Complementary Features

The following features and functions are typical in IP fabric
implementations:
»» EVPN Instance (EVI) is a construct used for extending VLANs

or VRFs across the IP fabric. When a VLAN is added to the
EVI, this triggers a routing update to be sent, and VXLAN
tunnels are automatically created to all other leaf switches
where the VLAN exists. When the VLAN is deleted, the tunnel
is taken down automatically.
»» Automatic VTEP discovery happens during the EVI process.
There is no need to configure any static tunnels because the
creation and deletion are automatic. The VTEPs talk to each
other over the underlay network that is providing IP connec-
tivity across the fabric.
»» VXLAN traffic distribution is achieved by changing the
source User Datagram Protocol (UDP) port in the VXLAN
packet. This is part of the VXLAN standard and happens
automatically. By changing the source UDP port, the hashing
algorithm sends traffic across different ECMP links in the
fabric. This ensures that one link will not get saturated while
other links remain idle.
»» Static Anycast Gateway is the default gateway for the end
devices. This gateway will exist on every leaf switch where
the VLAN is present. No control protocols are needed
because this is a static configuration across the fabric.
»» Address Resolution Protocol (ARP) suppression is
accomplished by keeping ARP information in a suppres-
sion cache on each leaf, thus reducing the amount of
traffic across the fabric for end station discovery.
»» MAC/IP Learning is done via BGP/EVPN, which eliminates
nearly all flooding of traffic to find end stations. As soon as a
MAC is learned on the leaf, a routing update is sent to all leaf
switches telling each of them the MAC just learned and the
VTEP IP where that MAC is attached.
Seeing Where IP Fabrics

Fit in the Network
IP fabrics have gained significant momentum in the data center.
This success can be attributed to the simple scaling that the fab-
ric provides, its virtualization capabilities to have any VLAN on
any port in any data center, and the inherent automation. Because
deterministic behavior is so critical in the data center, the IP fab-
ric and its spine/leaf topology are the perfect fit for this.
The IP fabric is also seeing traction outside of the data center

in uses for data center interconnects as well as campus imple-
mentations. The attraction of having a standards-based and fully
interoperable fabric that uses a simple underlay/overlay con-
cept is piquing the interest of network administrators looking for
ways to streamline operations with automation and provide ser-
vices quickly and easily across an infrastructure that is adaptable
through virtualization.
Every use case and need is different; therefore, picking the right
fabric for the right use case will vary. As IP fabrics continue to
gain momentum, their use will continue to diversify.
IN THIS CHAPTER
»» Enabling automation
»» Gaining operational efficiency through

artificial intelligence and machine
learning
»» Enhancing visibility
»» Integrating ecosystem partners
»» Improving security
»» Integrating wired and wireless
Chapter 4
Examining Key
Considerations in Fabric
Technology Evaluations
I
f you think that fabric networking sounds like the coolest thing
in the world . . . you’re right. Your next step is to choose a solu-
tion that will meet your needs. As with any solution, you should
take certain key considerations into account. New infrastructure
is an investment after all, so you want to choose a solution that
meets your needs now but can grow with you as your business
needs evolve.
In this chapter, you learn about the key considerations that you
need to review with any fabric networking solution. These traits
far outweigh price and product feeds and speeds in the evaluation
process.
CHAPTER 4 Examining Key Considerations in Fabric Technology Evaluations 29
Enabling Automation
Automation is the new buzzword in information technology circles.
With all the manual processes that are done daily, it’s easy to see
why automation is so critical. It allows manual operational tasks
to be offloaded from the network administration team, leaving
network administrators to focus on strategic projects.
There are different stages of automation, so think of it as a jour-

ney. Where you start and where you stop is your choice and based
on the needs of the business.
Network automation
Automating network tasks falls into two areas:
»» Infrastructure provisioning entails all the configuration

tasks to instantiate the fabric. These extend from the point
where switches and routers are powered on and cabled to
the point where the fabric is ready to have end devices
attached to it.
»» Tenant/services provisioning includes creating all the
necessary configuration for end devices to attach and use
the fabric. Over the lifecycle of a network, this is where most
configuration tasks occur.
With regard to infrastructure provisioning, once the devices are

powered on and cabled, it should be simple plug and play to build
the underlying fabric topology.
Once the fabric network has been built, it is time to build the net-
work services. Being able to leverage the power of automation is
where you can recognize significant time savings. Some fabric
technologies have embedded automation features and can estab-
lish network connectivity services on demand without relying on
any external controllers or management systems. An example
is the dynamic establishment of a guest Wi-Fi service as a per-
sonal iPad is detected through Institute of Electrical and Electron-
ics Engineers (IEEE) auto-attach features and on-boarded onto
the network.
Some fabric technologies work synergistically with central-

ized management or controller-based systems that can deliver
automated provisioning of the network based on an external
event or trigger. An example is the dynamic provisioning of top-
of-rack switch ports as a new virtual machine (VM) is turned up,
through integrations between the hypervisor and the network
environment.
Differing approaches to automation exist. One is for the protocol

to deliver inherent automation capabilities so that services can be
dynamically established. The other is for the fabric infrastruc-
ture to be programmable, meaning that it supports application
programming interfaces (APIs), or it supports open configuration
through programming languages such as NetConf, RESTConf,
Ansible, or Python. An external controller or management system
would then be used to automate the network infrastructure.
What’s key is not the underlying technology used to automate —

but the result! That means what’s important is that the technol-
ogy enables you to achieve better agility. This, in turn, enables
you able to respond quickly to business requirements while elimi-
nating time-consuming and risky manual provisioning.
Cross-domain automation
Cross-domain automation is a step forward from network auto-
mation because you are automating components from other IT
domains to complete the entire service chain. This process auto-
mation can include the provisioning of servers, storage, and net-
working, as well as network functions such as firewalls and load
balancers, as shown in Figure 4-1.
FIGURE 4-1: An overview of cross-domain automation.
This type of automation is important because most organizations
have silos of teams that need to work together in order to bring a
new application or service online. There is inherent delay because
of manual hand-offs between teams and the back-and-forth to
clarify what’s needed and when. Cross-domain automation sim-
plifies this process and ensures that teams are working in the
most efficient way possible.
This stage of automation is not solely about provisioning; it also

includes full lifecycle management. It allows you to automate
any manual process that is repeated on a regular basis, includ-
ing processes related to validation, troubleshooting, and remedi-
ation. An example is the creation of a workflow for your help desk
that automates the gathering of information when a user calls in,
improving the speed and efficiency of that process.
A key requirement of a cross-domain automation tool is that it

is customizable so that it can fit into your existing IT processes
and workflows. Generally, these tools are DevOps inspired and
can be programmed easily through common tools such as Python.
Another consideration is to ensure that the tool has ties to an open
source community, such as Github. When you are looking to auto-
mate a particular process, you can see what others have done and
simply modify existing workflows to fit your environment.
Take the micro-services approach to process automation by cre-

ating several small workflows and tying them together into a
larger workflow. This approach makes the creation of workflows
easier and enables you to reuse the small workflows many times
in different larger workflows, thus getting the benefits of reuse.
Cross-domain automation is a big step forward from network

automation, but you can ease into it by creating small workflows
and building from there. This is a great way to dip your toe in the
water.
AI/ML driven automation

Wouldn’t it be great if your fabric networking solution could use
artificial intelligence or machine learning? I know, it sounds
like science fiction, but it’s worth your consideration. AI/ML
technologies hold the promise of achieving true autonomous
networking — where the network becomes self-learning, self-
healing, self-securing, and self-optimizing.

What’s the difference between automation and true autonomous
networking? Automation streamlines manual processes, but it
isn’t yet intelligent. When you combine automation with AI/ML
within a reliable network infrastructure (like a fabric), you have a
truly autonomous network.
Although truly autonomous networking will develop gradually

over time, AI/ML technology is here today and can make your life
easier. For example, you can troubleshoot performance problems
by extracting meaning from large amounts of data in network
logs and databases. Another example is 24/7 monitoring of IoT
devices, where normal behavior is learned through AI/ML, and
anything that deviates from that behavior is remediated against
quickly and efficiently.
In general, AI and ML close the feedback loop between large

amounts of data/analytics and automation tools. This form of
intelligence can help reduce costs while enhancing reliabil-
ity, availability, overall performance, and the productivity of
network-operations teams and end-users alike.
When evaluating fabric networking solutions, you should consider

the vendor’s strategy regarding automation and AI/ML. Because
AI/ML is poised to be a significant part of next-generation net-
working, it is recommended that you start evaluating AI/ML tech-
nologies for targeted use cases.
Enhancing Visibility
The fabric solution you choose should allow for an in-depth view
of the network, its services, its applications, and its connected
users and devices.
Having the right visibility can help avoid business disruptions

through the proactive monitoring of application performance
and the network. It facilitates troubleshooting by giving a deep,
real-time view into the network and its traffic, and it can improve
network security through extensive visibility into unapproved
applications, unusual traffic, and shadow IT.
On-box visibility
On-box visibility is the ability to view traffic that traverses a partic-
ular fabric-enabled switch or router. Previously through features
such as Switched Port Analyzer (SPAN) or Remote SPAN (RSPAN),
a traffic sniffer or protocol analyzer would be plugged into the
switch or router, either locally in the case of SPAN, or remotely in
the case of RSPAN, for traffic on multiple virtual LANs (VLANs) or
ports to be analyzed.
Now, with many switches using x86 hardware, it is possible for the
traffic analyzer function to be run directly on the switch within a
guest VM. This approach reduces the mean time to repair. Another
big advantage is that multiple types of analysis tools can be run
on the VM. Examples include Wireshark, TCPDump, Splunk, and
PerfSONAR.
Network analytics
In addition to on-box visibility, having an end-to-end view of
the network, its services, its applications, and its connected users
and devices is essential. Seeing application and network response
time, top talkers, and top applications provide actionable insights
into the overall IT infrastructure. Having the right analytics data
can give you a better understanding of user behavior on the net-
work. It can also help with enhancing security by being able to
pinpoint unusual traffic and unauthorized applications.
Some of the key criteria in an analytics platform include having

full Layer 4-7 visibility end-to-end across the network — from
the user or end device all the way to the data center and out to
the cloud. It’s important to consider a tool that extends into the
hypervisor environment in the data center for visibility into traf-
fic between VMs that never touch a physical switch.
Another criterion is how easily the data is collected from the

network devices. With some solutions, the data can be collected
directly from the network switches and routers in an efficient and
cost-effective way through embedded telemetry features. In other
scenarios, external probes or sensors might be required. These
add significant cost and complexity when altering the physical
cabling environment.
You should deploy a solution that has a robust catalog of applica-

tion fingerprints for detecting and measuring the performance of

the applications that your business uses today. Having the ability
to easily customize the solution to include any custom applica-
tions is also a huge benefit to ensure full coverage of applications
across the infrastructure.
Integrating Ecosystem Partners

A fabric networking solution should not create vendor lock-in.
Instead, it should provide automated, simplified network connec-
tivity throughout your overall IT ecosystem.
With so many solutions available across the campus and data

center, it is virtually impossible for a fabric solution to sup-
port every ecosystem partner. However, there are high runners
that all should and likely do support. Examples of these include
hypervisors such as VMware, Microsoft, and KVM. These might
also include a multitude of storage options that utilize Internet
Small Computer Systems (iSCSI), Network File System (NFS), or
Non-Volatile Memory Express (NVMe); either as stand-alone
storage arrays or hyperconverged infrastructure (HCI).
In the campus, many IoT and security solutions are available that
should provide some level of integration with the fabric or its
associated management platform. These include IoT devices like
IP cameras, third-party switches, and security products like fire-
walls and access control systems. This capability enables the infra-
structure to be managed holistically, allowing the network to react
in real time to alerts received by third-party solutions. It is also
important for fabric-based switches to support traditional proto-
cols so that interoperability with existing networks is not an issue.
With your fabric solution, APIs should be available to ensure smart

integrations between your networking environment and the rest
of your IT infrastructure.
Improving Security
A fabric networking solution should make securing your network
simpler. Many methods of security are available including macro-
segmentation, micro-segmentation, hyper-segmentation, and
virtual routing and forwarding (VRF). All of these accomplish
the segmentation of traffic across the network to add levels of
security.
However, having a network that can easily be segmented at scale

allows you to improve your overall security posture by dramati-
cally reducing the attack surface and preventing lateral move-
ment to more sensitive areas of the network. A stealth network
prevents malicious actors from discovering the network topology.
In addition, service elasticity removes potential back-door entry
points to the network by removing residual configuration.
Are you in a network that has strong regulatory requirements?

Through fabric security features like segmentation and stealth,
compliance with the European Union’s General Data Protection
Regulation (GDPR), the Health Insurance Portability and Account-
ability Act (HIPAA), and the Payment Card Industry Data Security
Standard (PCI-DSS) can be facilitated.
Integrating Wired/Wireless
You need a fabric solution that can work across wired or wireless
networks. For maximum simplicity, the solution should offer the
ability to converge the wired and wireless networks. Having these
tightly integrated provides a seamless user experience no matter
what device or devices your users attach to. It also significantly
eases the administration of the network by eliminating the need
to manage two disparate network environments.
Being able to configure network services from a single pane of

glass — regardless of whether they are wired or wireless — is a
huge win for your business as well as for your network adminis-
tration team. It is also important that common policies be applied
to users and devices across wired and wireless, and that common
telemetry information be collected from the network switches as
well as the APs so that you can gain a clear view of network ser-
vices and application performance to simplify troubleshooting.
IN THIS CHAPTER
»» Focusing on your pain points
»» Evaluating fabric networking

technologies
Chapter 5
Choosing the Right
Fabric Solution for
Your Business
C
hapter 3 introduces the different fabric technologies and
the relative fit of each technology in the network. As that
chapter explains, Shortest Path Bridging (SPB) is well posi-
tioned for the campus or metropolitan area network (MAN) as
well as collapsed data center/campus cores, while Transparent
Interconnection of Lots of Links (TRILL) and now more signifi-
cantly IP fabrics are well positioned for the data center with
increased traction in the campus.
With the lines blurred between the different fabrics, where do you
start with the evaluation process?
Focusing on Your Pain Points

A logical way to figure out what fabric technology is best suited
for your business is to determine what your biggest networking
pain points are and what technology might be the best fit to solve
those.
CHAPTER 5 Choosing the Right Fabric Solution for Your Business 37
Here’s an overview of the top values for each technology:
»» SPB
• Faster time-to-service by eliminating core and aggrega-
tion node reconfiguration (provisioning at the edges only)
• Dynamic auto-attach and segmentation for IoT devices,

users, and virtual machines
• Simplification of the network with the ability to use just a

single control plane for all Layer 2 and Layer 3 services
• Unification of wired and wireless networking

• Inherent security through network segmentation at scale,
stealth topology, and service elasticity
• Less complex administration of IP multicast

• High-performance and scaling for IP video surveillance
• Topologies supporting any service, any port, any place
• Ecosystem integrations, virtual machines (VMs), Internet
of Things (IoT), security
»» TRILL
• Simple plug-and-play data center networking
• Built to support server and network virtualization
• Embedded automation to simplify provisioning and
operations
• Fully programmable with application programming

interfaces (APIs)
• Topology agnostic supporting any virtual LAN (VLAN),

any port, any place
• Simplified convergence of IP storage

• Ecosystem integrations, VMs, storage, security
»» IP fabrics
• Vendor-agnostic deployments with full interoperability
• Scale from the very small to the very large
• Simple plug-and-play data center networking
• Built to support server and network virtualization
• Embedded automation to simplify provisioning
• Fully programmable with APIs
• Topologies supporting any VLAN, any port, any place
Evaluating Fabric Networking

Technologies
To help you navigate the confusing process of evaluating fabric
solutions, Chapter 4 provides guidance on what to look for in your
fabric technology. This section builds on that discussion, pro-
viding simple checklists that you can use and customize. These
checklists identify the important features to look for to help you
modernize campus and data center networks:
»» Data center networking checklist

• Must be easy to deploy and operate
• Vendor-agnostic, industry standard with full
interoperability
• Embedded automation to simplify provisioning

• Scales from the very small to the very large
• Topologies supporting any VLAN, any port, any place
• Supports the ability to add new infrastructure (links and
nodes) with no service interruption
• Designed from the ground up to support virtualization

• Fully programmable with APIs
• Decouples connectivity services provisioning from
infrastructure topology configuration
• Provides rapid time-to-service for provisioning in real

time, and must not require device-by-device, hop-by-hop
configuration

• Enables the scalable deployment of network segments to
isolate and protect critical information
CHAPTER 5 Choosing the Right Fabric Solution for Your Business 39
• Must provide on-box, application telemetry, and analytics
capabilities to provide visibility into the network, users,
traffic, and devices.
• Simple and resilient interconnect capabilities

»» Campus networking checklist
• Must be easy to deploy and operate
• Supports small to very large campus environments
• Provides design flexibility allowing for hierarchical,
full-mesh, partial-mesh, and ring-based topologies
• Decouples connectivity services provisioning from the

fabric infrastructure
• Supports the ability to add new links and nodes without

network downtime
• Provides rapid time-to-service for provisioning without

requiring device-by-device, hop-by-hop configuration
• Provides virtualized IP multicast at scale without requiring

the use of Protocol-Independent Multicast (PIM)
• Provides a secure and robust IP-based video surveillance

service infrastructure that scales to thousands of sources
seamlessly
• Provides a dynamic auto-attach feature to facilitate the

connectivity of users and devices
• Enables the scalable deployment of network segments to

isolate and protect critical information, groups of users,
and IoT devices
• Limits lateral movement by making it difficult for hackers

to discover the network topology using typical IP scanning
techniques
• Eliminates residual configuration on edge switch ports

as users and devices disconnect from the network
dynamically
• Is built on industry standards

• Provides on-box, application telemetry and analytics
capabilities to provide visibility into the network, users,
traffic, and devices
• Delivers wired and wireless integration

• Delivers simple and resilient interconnect capabilities
IN THIS CHAPTER
»» Exploring the breadth of solutions
»» Managing through a single pane of glass
»» Supporting multi-vendor solutions
Chapter 6
Ten Things to Know
About Fabric Solutions
from Extreme Networks
E
xtreme Networks delivers world-class fabric networking
solutions to organizations needing scalability, flexibility,
and ease of administration. Industry leading and differenti-
ating technology will help you transform the data center and
campus network into an infrastructure that operates at cloud
speed with efficiency and operational simplicity.
You Have a Choice

With Extreme, you can choose the fabric solution that best suits
your use case and network requirements. If you are looking for
campus or data center transformation, Extreme can solve your
networking pain points with a flexible portfolio of solutions that
includes Ethernet and IP fabrics. Extreme can work with you to
ensure that you have the right technology for your environment.
CHAPTER 6 Ten Things to Know About Fabric Solutions from Extreme Networks 41
Single Pane of Glass
Extreme’s single operational model extends from the wired/
wireless edge to the data center and multi-cloud, providing
comprehensive management, policy, and analytics for fabric

networking solutions. All Extreme fabric solutions for the campus
and the data center use the same operational model with visibility
and policies consistent end to end.
Visibility from the Edge to Multi-Cloud

Extreme provides a unique end-to-end view of application flows
that extend from the campus edge to the data center, the vir-
tualized environment, and even into multi-cloud environments.
Regardless of the chosen fabric solution, you have the option of
complementing agile, automated networking with deep insights
into the network, applications, and users through application
telemetry and analytics capabilities.
Solutions Are Truly Multi-Vendor

The solutions from Extreme Networks are industry aligned and
based on industry standard technology. Extreme’s solutions are
designed to work seamlessly with whatever you have installed
in your network today. This capability allows you to continue to
get value from devices you have already purchased while gain-
ing flexibility, agility, and scalability from the fabric networking
solutions.
Deployment History
With thousands of fabric deployments worldwide, Extreme Net-
works has the knowledge and experience to provide a solution
that will meet your needs today and in the future. Extreme has
deployed solutions that span sizes and segments of both service
provider and enterprise markets and have deployed fabric solu-
tions in the most demanding, highly sensitive, and highly secure
environments.
Innovation!
Extreme Networks holds more than 500 patents related to fab-
ric networking. In addition to developing many of the drafts
related to Shortest Path Bridging (SPB), the company continues
to invest in research and development to continually improve its
products and increase the functionality of its solutions. You can
view Extreme Networks’ patents at www.extremenetworks.com/
company/legal/patents/.
Field Proven in a Demanding

Environment
Extreme’s fabric solutions are deployed worldwide and are at the
core of most Internet exchange providers, along with deployments
among cloud service providers, content delivery networks, and
enterprises across all verticals and all sizes. This includes mission
critical environments like hospitals, emergency response, and
national defense as well as high-profile events such as the Sochi
Olympic Winter Games.
Plug and Play

Fabric solutions from Extreme Networks can be deployed and
fully operational within minutes, rather than days or weeks. They
can self-provision rapidly, ensuring that you are up and running
quickly with your next-generation networking solution. In addi-
tion, through network service automation and even cross-domain
automation, you can reduce operational expenses and do more
with less.
Ecosystem Integrations
Extreme Networks provides the networking foundation for your
business. This is more than just plumbing; it’s the enabler for
services and applications that the business runs on. Having the
CHAPTER 6 Ten Things to Know About Fabric Solutions from Extreme Networks 43
fabrics ready and able to integrate ecosystem partners is criti-
cal to success. By providing seamless integration with virtual
machine (VM) providers, storage solutions, security partners, and
IoT vendors, Extreme provides you with an integrated network
that can be managed holistically.
Unbreakable in Hack-a-thons
The secure fabric solutions from Extreme Networks have been
part of multiple private and public hack-a-thons. To date, not one
has been breached. This is because of the inherent security of the
solution with features such as hyper-segmentation, which iso-
lates your virtualized networks; stealth networking, which hides
your network’s topology from prying eyes; and service elasticity,
which allows your segmented network to extend and retract when
devices connect or disconnect.
WILEY END USER LICENSE AGREEMENT
Go to www.wiley.com/go/eula to access Wiley’s ebook EULA.

Fabric Networking

Uploaded by

Copyright:

Available Formats

Fabric Networking

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fabric Networking

Uploaded by

Copyright:

Available Formats

These materials are © 2020 John Wiley & Sons, Inc.

Any dissemination, distribution, or unauthorized use is strictly prohibited.

LIMIT OF LIABILITY/DISCLAIMER OF WARRANTY: THE PUBLISHER AND THE AUTHOR MAKE NO

ISBN 978-1-119-61937-6 (pbk); ISBN 978-1-119-61940-6 (ebk)

Manufactured in the United States of America

CHAPTER 1: Recognizing the Need for More Flexible

CHAPTER 2: Exploring the Use Cases for

CHAPTER 3: Exploring Fabric Networks................................................. 15

Table of Contents iii

CHAPTER 5: Choosing the Right Fabric Solution

CHAPTER 6: Ten Things to Know About Fabric

iv Fabric Networking For Dummies, Extreme Networks Special Edition

Networks of the past were designed for simple, static workflows

Although computing technologies have matched the needs of

Fear not, network administrators, the next evolution in network-

About This Book

Fabric networking enables you to scale with demand and create

Here’s what the icons mean:

The Remember icon indicates information that deserves your

2 Fabric Networking For Dummies, Extreme Networks Special Edition

»» Introducing fabric networking

»» Learning about the types of fabric

»» Understanding adjacent networking

This chapter introduces the challenges faced today with tradi-

Analyzing the State of

The challenges with traditional networks include:

»» Outdated designs: Traditional network design has remained

4 Fabric Networking For Dummies, Extreme Networks Special Edition

Today, you’re likely to encounter two types of fabric networks:

»» Ethernet fabrics are based on industry standard protocols

Although fabric technology is not new (initial deployments started

Because networking standards continue to evolve, its best to stay

Identifying the Benefits

»» Improved time-to-service: Likely, you’re being asked to

6 Fabric Networking For Dummies, Extreme Networks Special Edition

»» Software-defined networking (SDN): SDN is a concept that

8 Fabric Networking For Dummies, Extreme Networks Special Edition

»» Examining campus use cases for fabric

In this chapter, you learn about using fabric networking in cam-

Understanding Data Center Use

CHAPTER 2 Exploring the Use Cases for Fabric Networking 9

Identifying the need for a data center

»» Server virtualization has become common and is the best

10 Fabric Networking For Dummies, Extreme Networks Special Edition

CHAPTER 2 Exploring the Use Cases for Fabric Networking 11

Checking Out Fabric Networking

»» IoT: One of the largest challenges you will face as a network

12 Fabric Networking For Dummies, Extreme Networks Special Edition

CHAPTER 2 Exploring the Use Cases for Fabric Networking 13

Figure 2-2 shows an example of campus fabric networking.

FIGURE 2-2: Campus fabric networking.

14 Fabric Networking For Dummies, Extreme Networks Special Edition

»» Looking at complementary technologies

Introducing Shortest Path Bridging

CHAPTER 3 Exploring Fabric Networks 15

FIGURE 3-1: Shortest Path Bridging (SBP).

SPB enables businesses to implement flexible, scalable networks.

»» Multiple physical topologies are supported, including mesh,

16 Fabric Networking For Dummies, Extreme Networks Special Edition

Building the SPB fabric