Small Enterprise Design Profile Reference Guide

Small Enterprise Design Profile Reference Guide
Small Enterprise Design Profile Reference Guide
Last Updated: July 8, 2010
Building Architectures to Solve Business Problems
ii
Small Enterprise Design Profile Reference Guide
About Cisco Validated Design (CVD) Program
The CVD program consists of systems and solutions designed, tested, and documented
to facilitate faster, more reliable, and more predictable customer deployments. For more
information visit www.cisco.com/go/designzone.
ALL DESIGNS, SPECIFICATIONS, STATEMENTS, INFORMATION, AND RECOMMENDATIONS (COLLECTIVELY,
"DESIGNS") IN THIS MANUAL ARE PRESENTED "AS IS," WITH ALL FAULTS. CISCO AND ITS SUPPLIERS DISCLAIM ALL WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT OR ARISING FROM A COURSE OF
DEALING, USAGE, OR TRADE PRACTICE. IN NO EVENT SHALL CISCO OR ITS SUPPLIERS BE LIABLE FOR
ANY INDIRECT, SPECIAL, CONSEQUENTIAL, OR INCIDENTAL DAMAGES, INCLUDING, WITHOUT LIMITATION,
LOST PROFITS OR LOSS OR DAMAGE TO DATA ARISING OUT OF THE USE OR INABILITY TO USE THE
DESIGNS, EVEN IF CISCO OR ITS SUPPLIERS HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
THE DESIGNS ARE SUBJECT TO CHANGE WITHOUT NOTICE. USERS ARE SOLELY RESPONSIBLE FOR
THEIR APPLICATION OF THE DESIGNS. THE DESIGNS DO NOT CONSTITUTE THE TECHNICAL OR OTHER
PROFESSIONAL ADVICE OF CISCO, ITS SUPPLIERS OR PARTNERS. USERS SHOULD CONSULT THEIR
OWN TECHNICAL ADVISORS BEFORE IMPLEMENTING THE DESIGNS. RESULTS MAY VARY DEPENDING
ON FACTORS NOT TESTED BY CISCO.
CCDE, CCENT, Cisco Eos, Cisco Lumin, Cisco Nexus, Cisco StadiumVision, Cisco TelePresence, Cisco WebEx, the Cisco logo, DCE, and
Welcome to the Human Network are trademarks; Changing the Way We Work, Live, Play, and Learn and Cisco Store are service marks; and
Access Registrar, Aironet, AsyncOS, Bringing the Meeting To You, Catalyst, CCDA, CCDP, CCIE, CCIP, CCNA, CCNP, CCSP, CCVP, Cisco, the
Cisco Certified Internetwork Expert logo, Cisco IOS, Cisco Press, Cisco Systems, Cisco Systems Capital, the Cisco Systems logo, Cisco Unity,
Collaboration Without Limitation, EtherFast, EtherSwitch, Event Center, Fast Step, Follow Me Browsing, FormShare, GigaDrive, HomeLink,
Internet Quotient, IOS, iPhone, iQuick Study, IronPort, the IronPort logo, LightStream, Linksys, MediaTone, MeetingPlace, MeetingPlace Chime
Sound, MGX, Networkers, Networking Academy, Network Registrar, PCNow, PIX, PowerPanels, ProConnect, ScriptShare, SenderBase,
SMARTnet, Spectrum Expert, StackWise, The Fastest Way to Increase Your Internet Quotient, TransPath, WebEx, and the WebEx logo are
registered trademarks of Cisco Systems, Inc. and/or its affiliates in the United States and certain other countries.
All other trademarks mentioned in this document or website are the property of their respective owners. The use of the word partner does not
imply a partnership relationship between Cisco and any other company. (0809R)
© 2010 Cisco Systems, Inc. All rights reserved
Small Enterprise Design Profile Reference Guide
iii
About the Authors
Solution Authors
Martin Pueblas, CCIE#2133, CISSP#40844—Technical Leader, CMO Enterprise
Solutions Engineering (ESE), Cisco Systems
Martin Pueblas
Martin is the lead system architect of the Cisco SAFE Security Reference Architecture. He is a network security expert
with over 17 years of experience in the networking industry. He obtained his CCIE certification in 1996 and CISSP in
2004. Martin joined Cisco in 1998 and has held a variety of technical positions. Started as a Customer Support Engineer in Cisco’s Technical Assistance Center (TAC) in Brussels, Belgium. In 1999 moved to the United States where
soon became technical leader for the Security Team. Martin’s primary job responsibilities included acting as a primary
escalation resource for the team and delivering training for the support organization. At the end of 2000, he joined the
Advanced Engineering Services team as a Network Design Consultant, where he provided design and security consulting services to large corporations and Service Providers. During this period, Martin has written a variety of technical documents including design guides and white papers that define Cisco’s best practices for security and VPNs.
Martin joined Cisco’s Central Marketing Organization in late 2001, where as a Technical Marketing Engineer, he
focused on security and VPN technologies. In late 2004, he joined his current position acting as a security technical
leader. As part of his current responsibilities, Martin is leading the development of security solutions for enterprises.
Steve Gyurindak, CCIE#9057, CISSP#61046—Solutions Architect, Enterprise
Solutions Engineering (ESE), Cisco Systems
Steve Gyurindak
Steve is a solutions architect with over 15 years of industry experience. He joined Cisco in 2000 and worked the first 8
and a half years as a Systems Engineer covering the Service Provider, North Florida/Alabama Commercial, Georgia
Enterprise and US Channels sales markets. Steve has been recognized for his work with some of Cisco's most influential customers as well as for his work in South America and Europe. Steve joined ESE in 2009 to lead the development of customer-focused architectures and designs for the Education Market. Steve has a Bachelor of Science
degree in Telecommunications from the State University of New York at Buffalo, and is currently pursuing a Master's of
Science degree in Network Telecommunications at New York University. In addition to a CCIE in Routing and Switching, Steve holds the following certifications: CISSP, CCNP, CCDP, CCNA, CCDA, MCSE, and MCNE.
John Strika, Technical Marketing Engineer, CMO Enterprise Solutions Engineering
(ESE), Cisco Systems
John Strika
iv
John is a Technical Marketing Engineer in Cisco's Public Sector ESE team, with expertise in the areas of mobility and
location-based services. He has coauthored documents on enterprise mobility and Wi-Fi location-based services. As
a member of Cisco's Enterprise Architecture Board, he helps maintain Cisco's vision and architectural direction and
define Cisco's roadmap for context-aware and presence solutions. Previously, John was Cisco's first mobility consulting systems engineer, responsible for architecting creative wireless solutions for large enterprise customers. His 28
years of experience spans network design and implementation, applications development, facilities planning and
management, consulting, and general management. His past roles have included mission-critical telecommunications
design and development at AT&T and systems programming and data communications management with Wall Street
brokerages and commercial banks. Prior to joining Cisco, Strika was at Telxon Corporation (parent of Cisco's Aironet
wireless acquisition) for nine years, reaching the position of Southern Division Vice President of Wireless Technologies and Services. He is a member of the IEEE and has held several Federal Communications Commission licenses in
the use and modification of amateur and commercial radio. His educational background is in electrical engineering
and computer applications programming from Columbia University and in finance from Fordham University's College
of Business Administration, and he holds a masters of communications technology certificate from the American Institute. He was a charter Novell Certified Netware Engineer in the greater New York City area. Always seeking opportuni-
Small Enterprise Design Profile Reference Guide
About the Authors
Solution Authors
ties to use his mobility and advanced communications knowledge to improve public safety as well as
the safety of our public servants, John has served in volunteer search and rescue as well as a
Reserve Deputy.
Rahul Kachalia, CCIE#11740—Technical Marketing Engineer, CMO
Enterprise Solutions Engineering (ESE), Cisco Systems
Rahul is a technical marketing engineer in Cisco's Enterprise Solution Engineering group,
helping to create the design guidance that will help build the 21st century school network
infrastructure. Rahul has more than 14 years of broad engineering experience, primarily
in service provider core and edge focused products and technologies including broadband, MPLS, VPN and managed services. He has led many assurance projects to develop
Rahul Kachalia
solutions that can deliver design guidance and accelerate deployments from traditional
WAN infrastructure to next-generation IP/MPLS managed core networks. In the Enterprise
Solution Engineering group he has also worked on designing next-generation unified virtual campus networks for large enterprise customers. In addition to CCIE, Rahul holds
CCNP, CCNA, MCSE, MCP, and CNE. He holds a bachelor's degree from Mumbai University, India.
Dan Hamilton, CCIE #4080 —Technical Leader, CMO Enterprise Solutions Engineering (ESE), Cisco Systems
Dan Hamilton
Dan has over 15 years experience in the networking industry. He has been with Cisco for 9 years. He
joined Cisco in 2000 as a Systems Engineer supporting a large Service Provider customer. In 2004,
he became a Technical Marketing Engineer in the Security Technology Group (STG) supporting IOS
security features such as infrastructure security, access control and Flexible Packet Matching (FPM)
on the Integrated Security Routers (ISRs), mid-range routers and the Catalyst 6500 switches. He
moved to a Product Manager role in STG in 2006, driving the development of new IOS security features before joining the ESE Team in 2008. Prior to joining Cisco, Dan was a network architect for a
large Service Provider, responsible for designing and developing their network managed service
offerings. Dan has a Bachelor of Science degree in Electrical Engineering from the University of Florida.
Srinivas Tenneti, CCIE#10483—Technical Marketing Engineer, CMO
Enterprise Solutions Engineering (ESE), Cisco Systems
Srinivas is a Technical Marketing Engineer for WAN and branch architectures in Cisco's ESE team.
Prior to joining the ESE team, Srinivas worked two years in Commercial System Engineering team
where he worked on producing design guides, and SE presentations for channel partners and SEs.
Before that, he worked for 5 years with other Cisco engineering teams. Srinivas has been at Cisco for
8 years.
Srinivas Tenneti
Small Enterprise Design Profile Reference Guide
v
C O N T E N T S
CHAPTER
1
Small Enterprise Design Profile (SEDP) Overview
1-1
Small Enterprise Design Profile 1-1
Main Site Design 1-3
Remote Large Site Design 1-3
Remote Small Site Design 1-3
Access Devices 1-3
Network Foundation Design Considerations 1-3
LAN Design Considerations 1-4
High Availability Design Considerations 1-4
Routing Protocol Selection Criteria 1-4
Access Layer Design Considerations 1-5
LAN Foundational Services 1-5
WAN Design Considerations 1-6
WAN Transport 1-6
WAN Foundational Services 1-6
Cisco SAFE Security Architecture Design Considerations
Mobility
1-7
Collaboration Services Design Considerations
1-7
Unified Communications Design Considerations
Call Processing Considerations 1-7
Gateway Design Considerations 1-8
Dial Plan Considerations 1-8
Survivability Considerations 1-8
IP Video Surveillance Design Considerations
Digital Media Systems 1-9
Summary
CHAPTER
2
1-6
1-7
1-8
1-10
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
Hierarchical Network Design 2-2
Collapsed Core Network Design 2-4
Main Site Network Design 2-5
Resilient Distributed System 2-9
2-1
2-1
Small Enterprise Design Profile Reference Guide
i
Contents
Main Site Access-Layer Edge Services 2-9
Access-Layer Network Control Services 2-11
Resilient Access-Layer Network and System 2-12
Main Site Data Center Network Design 2-13
Remote Site Network Design 2-14
Remote Site Collapsed Core Network Design 2-14
Remote Site Access-Layer Design 2-16
Deploying Network Foundation Services 2-16
Implementing EtherChannel in Network 2-16
EtherChannel Load Balancing 2-19
Deploying Core Network Layer 2-21
Routing Protocol 2-21
Routing Protocol Selection Criteria 2-21
Designing End-to-End EIGRP Routing Domain 2-22
Deploying Multi-Layer Network 2-27
Spanning-Tree in Multilayer Network 2-28
Logical Multi-Layer Network 2-29
Flat Logical Network Design 2-29
Segmented Logical Network Design 2-30
Implementing Layer 2 Trunk 2-32
Unidirectional Link Detection 2-33
Deploying Routed-Access Network 2-34
Implementing EIGRP Routing in Access-Distribution Block
Building EIGRP Network Boundary 2-37
EIGRP Adjacency Protection 2-41
Tuning EIGRP Protocol Timers 2-42
Deploying Multicast in Network 2-42
Multicast Routing Design 2-44
Deploying PIM-SM 2-46
Implementing IGMP 2-49
Multicast Security—Preventing Rogue Source 2-49
Multicast Security—Preventing Rogue RP 2-50
Deploying QoS in Network 2-51
QoS in Catalyst Fixed Configuration Switches
QoS in Cisco Modular Switches 2-52
QoS Framework 2-54
QoS Trust Boundary 2-55
Deploying QoS 2-55
Implementing QoS Trust Mode 2-56
Implementing QoS Classification 2-57
Small Enterprise Design Profile Reference Guide
ii
2-51
2-36
Contents
Implementing Ingress Policer 2-60
Implementing Ingress Marking 2-61
Applying Ingress Policies 2-62
Applying Ingress Queueing 2-63
Deploying Egress QoS 2-65
Deploying Network Core QoS 2-69
Deploying Main or Remote Large Site Ingress QoS
Deploying Remote Small Site Ingress QoS 2-71
Deploying Main Site Egress QoS 2-72
Deploying Remote Large Site Egress QoS 2-74
Deploying Remote Small Site Egress QoS 2-76
2-70
Building a Resilient Network 2-78
Redundant Hardware Components 2-79
Redundant Power System 2-80
Redundant Network Connectivity 2-80
Redundant Control-Plane 2-80
Operational Resiliency Strategy 2-83
Deploying High Availability in Network 2-83
Network Resiliency 2-83
Implementing IP Event Dampening 2-83
Device Resiliency 2-85
Summary
CHAPTER
3
2-92
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Design 3-1
MPLS/VPN 3-1
Internet 3-2
Metro Ethernet 3-2
Why Metro Ethernet Service Is Needed 3-2
Types of Services Available in Metro Ethernet
WAN Service Deployed in the Small Enterprise Design
When to Consider WAN Technologies 3-4
Bandwidth Capacity 3-5
Planning 3-5
Implementation 3-7
Aggregation Structure 3-7
Routing for WAN Connections 3-8
QoS Design 3-9
QoS Policy at Main Site 3-9
3-1
3-2
3-3
Small Enterprise Design Profile Reference Guide
iii
Contents
QoS Policy at Remote Site
CHAPTER
4
3-11
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Cisco Unified Wireless Network Architecture
LWAPP Features 4-3
Small Enterprise Design Profile 4-4
4-1
Management 4-5
Connection to the Small Enterprise Design Profile Network
RF Groups and Mobility Groups 4-12
Example WLAN Configurations 4-13
Secured Employee WLAN 4-14
Secured VoWLAN 4-15
AP Deployments Considerations 4-18
AP 1250 4-18
AP 1140 4-18
Coverage and Site Surveys 4-19
Single-Band versus Dual-Band APs 4-19
WLC Discovery 4-20
WLC Failover Options 4-20
Appendix A—Devices and Software Used
CHAPTER
5
4-1
4-9
4-21
Small Enterprise Design Profile(SEDP)—Security Design
5-1
Small Enterprise Network Security Design 5-1
Network Foundation Protection 5-3
Internet Perimeter Protection 5-5
Internet Border Router Security Guidelines 5-6
Internet Firewall Guidelines 5-7
Intrusion Prevention Guidelines 5-10
E-mail Security Guidelines 5-14
Web Security Guidelines 5-19
Serverfarm Protection 5-23
Network Access Security and Control 5-25
Catalyst Integrated Security Features 5-26
Cisco Unified Wireless Network (CUWN) Integrated Security Features
Cisco NAC Appliance 5-27
NAC Appliance Modes and Positioning 5-29
NAC Appliance Deployment in Small Enterprise Networks 5-34
Cisco Identity-Based Network Networking Services (IBNS) 5-35
802.1X and EAP 5-36
Small Enterprise Design Profile Reference Guide
iv
5-26
Contents
Impacts of 802.1X on the Network 5-37
802.1X in Small Enterprise Networks 5-37
NAC 802.1X and CISF in Combination 5-38
Secure Mobility 5-38
Threats Mitigated 5-40
Network Security Deployment 5-41
Internet Border Router Deployment 5-41
Internet Firewall Deployment 5-42
Firewall Hardening and Monitoring 5-43
Network Address Translation (NAT) 5-45
Firewall Access Policies 5-45
Botnet Traffic Filter 5-47
Firewall Redundancy 5-51
Routing 5-52
Intrusion Prevention Deployment 5-55
Deploying IPS with the Cisco ASA 5-55
IPS Global Correlation Deployment 5-55
Web Security Deployment 5-60
Initial System Setup Wizard 5-61
WCCP Transparent Web Proxy 5-65
Web Access Policies 5-69
CISF Protected Ports 5-70
NAC Appliance Deployment 5-70
Adding a CAS to the CAM 5-71
Managing the CAS 5-72
Clean Access Roles 5-77
Layer 2 OOB Example 5-77
Availability Considerations 5-79
Basic Clean Access Switch Configuration 5-80
Basic Clean Access Out-of-Band Switch Configuration
NAC CAS Connection 5-80
Basic 802.1X Switch Configuration 5-81
CHAPTER
6
5-80
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Introduction 6-1
What Are Context-Aware Services? 6-1
Why Use Context-Aware Services? 6-2
Is It Here?—Zone or Inventory Management
Where Is It?—Asset Tracking 6-3
6-1
6-3
Small Enterprise Design Profile Reference Guide
v
Contents
What Is Its Condition?—Condition Tracking 6-3
What Is Its Status?—Status Monitoring 6-4
Where Is It in My Network?—Network Location Services
Cisco Context-Aware Components 6-5
Wired and Wireless Client Devices 6-5
Cisco Unified Network 6-6
Management and Applications 6-7
Context-Aware Component Interaction 6-8
Network Mobility Services Protocol (NMSP) 6-11
6-4
Context-Aware Services in Small Enterprise Designs 6-12
Component Capacities 6-15
Mobility Services Engine 6-15
WLAN Controllers 6-16
Wireless Control System (WCS) 6-16
Context-Aware Engine for Tags (AeroScout) 6-17
Integration within Small Enterprise Designs 6-18
MSE Connection to the Network 6-18
Clock Synchronization 6-18
SEDP Wireless Control System (WCS) Context-Aware Considerations 6-22
WLAN Controller and Cisco Catalyst Switch Definition and Synchronization 6-36
Wireless Client Context-Aware Considerations 6-37
RFID Asset Tag Context-Aware Considerations 6-41
Rogue Device Context-Aware Considerations 6-45
Context-Aware Considerations for Wired Device Tracking 6-48
Classification and Marking of NMSP Sessions 6-55
NMSP Traffic Flows Originating At The MSE 6-57
NMSP Traffic Generated By WLAN Controllers 6-58
NMSP Traffic Generated By Context-Aware Switches 6-59
Hardware/Software Releases
6-60
Context-Aware Services—General Best Practice References
CHAPTER
7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Media System Overview
7-1
Video Surveillance Overview 7-1
Cisco Digital Media System Architecture
DMS Solution for Small Enterprise 7-3
7-2
Desktop Video Application Overview 7-5
Desktop Video Components 7-5
Publishing Live and Video On-Demand Content
Small Enterprise Design Profile Reference Guide
vi
6-61
7-5
7-1
Contents
Enterprise TV Application Overview 7-6
Enterprise TV Components 7-7
Broadcasting Live TV or Video On-Demand Content
7-7
Digital Signage Application Overview 7-8
Digital Signage Components 7-9
Video Surveillance System Architecture 7-9
Cisco Video Surveillance Media Server 7-11
Cisco Video Surveillance Operations Manager 7-11
Cisco Video Surveillance IP Cameras 7-12
Deploying Digital Signage in Small Enterprise Network 7-14
Centralized Management Model 7-14
Distributed Content Storage Model 7-16
Validated Content Distribution Model 7-17
Implementing Network Services for Digital Signage 7-18
Deploying Cisco DMP in the Access Layer Network 7-19
Manual Deployment 7-19
Applying Ingress QoS Policy on Access Layer 7-21
Applying Egress QoS Policy on Access Layer 7-23
Auto Smartport Macro Deployment 7-23
Implementing Auto SmartPort Macro 7-25
Tuning Auto SmartPort Macro 7-25
Implementing Cisco Digital Media Player 7-26
Assigning IP Address to Cisco DMP 7-26
Registering Cisco DMP to Cisco DMM Database 7-27
Summary
7-36
Small Enterprise Design Profile Reference Guide
vii
Contents
Small Enterprise Design Profile Reference Guide
viii
CH A P T E R
1
Small Enterprise Design Profile (SEDP) Overview
The Enterprise Design Profile delivers the foundational network design that all enterprise services,
applications, and solutions use to interact and communicate with one another. The Enterprise Design
Profile is constructed in a fashion that supports all the applications and services that will ride on it.
Additionally, these profiles must be aware of the type of traffic traversing and treat each application or
service with the correct priority based on the needs and importance of that application.
The Small Enterprise Design Profile is made up of the following four distinct components:
•
Network Foundation guidance
•
Cisco SAFE Security Architecture guidance
•
Mobility guidance
•
Collaboration Services guidance such as Unified Communications
Each of these critical foundation components have been carefully designed and tuned to allow for a
secure environment that provides for business continuity, service awareness and differentiation, as well
as access flexibility. See Figure 1-1.
Figure 1-1
Small Enterprise Design Profile Design Components
Network Foundation
Guidance
Cisco SAFE Security Architecture
Guidance
Enterprise Design
Profiles
Collaboration Services
Guidance
229341
Mobility
Guidance
Small Enterprise Design Profile
The design used for the Small Enterprise Design Profile is intended to represent as many small-size
Enterprise network environments as possible. To accomplish this, a modular design is used representing
sites of varying sizes (see Figure 1-2). The Small Enterprise Design Profile is built upon a network
foundation consisting of a main site, where the majority of the critical applications reside.
Small Enterprise Design Profile Reference Guide
1-1
Chapter 1
Small Enterprise Design Profile (SEDP) Overview
Small Enterprise Design Profile
Connected through a Metro Ethernet WAN are remote sites of varying sizes. The remote small site is
designed to support up to 100 employees. The remote large site is designed to support up to 500
employees. Each site can coexist in a small Enterprise network or can be treated as separate modules.
Design guidance for remote sites of varying sizes provides flexibility, modularity, and scalability as the
Enterprise grows. Additionally, it is expected that half of all network can be accessed wired and
wirelessly.
Figure 1-2
Small Enterprise Design Profile Design
HDTV
IP
User Devices
M
M
M
M
M
Services Block
Serverfarm
Internet Edge
Main Site
PSTN
Internet
HDTV
IP
User Devices
Remote Large Site
Small Enterprise Design Profile Reference Guide
1-2
HDTV
IP
User Devices
Remote Small Site
229342
WAN
Chapter 1
Small Enterprise Design Profile (SEDP) Overview
Network Foundation Design Considerations
Main Site Design
The main site design represents a centralized services site that interconnects the remote sites regardless
of size. The main site consists of a main building core connected to a serverfarm and service block
design. The remote large site and the remote small site connect to the main site via a 100Mbps Metro
Ethernet link. Because many Enterprise services and applications are centrally located within the main
site rather than each remote site, high network availability must be maintained. The main site is
connected to outside entities such as the partners, vendors, customers, and the Internet using the Internet
edge components. All the remote sites within the small enterprise system also connect to the main site
and use the main site to connect outside the Enterprise.
Remote Large Site Design
The remote large site design has been developed for Enterprise sites that have up to 500 employees. The
remote large site is connected to the main site via a 100Mbps Metro Ethernet link. While this site
connects to the main site where most Enterprise services and applications are centralized, this site also
uses resilient application services features to maintain critical services such as Digital Media Signage,
Video Surveillance, and SRST in case of a WAN failure. For resiliency, this site is designed with a dual
switch and dual aggregation links.
Remote Small Site Design
The remote small site profile represents a site supporting up to 100 employees. The core and distribution
layers in the remote site network are collapsed into one and use Cisco’s Catalyst Stackwise switching
technology for redundancy. The remote small site is connected to the main site via a 10Mbps Metro
Ethernet link and like the remote large site, resilient application service features are used to maintain
critical services in case of a WAN failure.
Access Devices
The devices that connect to the Small Enterprise Design Profile network include phones, cameras,
displays, laptops, desktops, mobile phones, and personal devices (iPod, MP3, etc). Half of all the devices
are expected to connect to the network using 802.11 ABGN wireless access.
The Small Enterprise Design Profile consists of four major components. The sections below provide a
brief description of each of these components.
Network Foundation Design Considerations
The Small Enterprise Design Profile Network Foundation guidance is made up of routers and switches
deployed in a three-tier hierarchical model that uses Cisco IOS to provide foundational network
technologies needed to provide a highly available, application-aware network with flexible access. LAN
and WAN guidance is provided under this component.
Small Enterprise Design Profile Reference Guide
1-3
Chapter 1
Small Enterprise Design Profile (SEDP) Overview
Network Foundation Design Considerations
LAN Design Considerations
Hierarchical network design model components:
•
Core layer—The site backbone consisting of a Layer-3 core network interconnecting to several
distributed networks and the shared services block to access local and global information.
•
Distribution layer—The distribution layer uses a combination of Layer-2 and Layer-3 switching to
provide for the appropriate balance of policy and access controls, availability, and flexibility in
subnet allocation and VLAN usage.
•
Access layer—The Demarcation point between network infrastructure and access devices. Designed
for critical network edge functionality to provide intelligent application and device aware services.
High Availability Design Considerations
To ensure business continuity and prevent catastrophic network failure during unplanned network
outage, it is important to identify network fault domains and define rapid recovery plans to minimize the
application impact during minor and major network outages.
The Small Enterprise Design Profile design ensures network survivability by employing three major
resiliency methods that serve to mitigate most types of failures. The appropriate resiliency option should
be selected given the network system tier, role, and network service type:
•
Link resiliency—Provides redundancy during physical link failures (i.e., fiber cut, bad transceivers,
incorrect cablings, etc.)
•
Device resiliency—Protects network during abnormal node failure triggered by hardware or
software (i.e., software crashes, non-responsive supervisor etc.)
•
Operational resiliency—Enables higher level resiliency capabilities, providing complete network
availability even during planned network outage conditions.
Routing Protocol Selection Criteria
Routing protocols are essential for any network, because they allow for the routing of information
between buildings and sites. Selecting the right routing protocol can vary based on the end-to-end
network infrastructure. The routers and switches support many different routing protocols that will work
for Small enterprise network environments. Network architects must consider all the following critical
design factors when selecting the right routing protocol to be implemented throughout the internal
network:
•
Network design—Proven protocol that can scale in full-mesh site network designs and can optimally
function in hub-and-spoke WAN network topologies.
•
Scalability—Routing protocol function must be network and system efficient that operates with a
minimal number of updates, recomputation independent of number of routes in the network.
•
Rapid convergence—Link state versus DUAL recomputation and synchronization. Network
reconvergence also varies based on network design, configuration, and a multitude of other factors
which are beyond the routing protocol.
•
Operational considerations—Simplified network and routing protocol design that can ease the
complexities of configuration, management, and troubleshooting.
Small Enterprise Design Profile Reference Guide
1-4
Chapter 1
Small Enterprise Design Profile (SEDP) Overview
Network Foundation Design Considerations
Access Layer Design Considerations
The access layer represents the entry into the network, consisting of wired and wireless access from the
client to the network. The switch that the client connects to will ultimately connect to the network
distribution layer of and the method communication here must be considered in any design. Traditional
Layer 2 connectivity is prevalent in most networks today; however, it comes at some cost in
administration, configuration, and timely resiliency. The emerging method of connectivity is a Layer 3
connection, commonly referred to as routed-access.
Performing the routing function in the access-layer simplifies configuration, optimizes distribution
performances, and allows for the use of well known end-to-end troubleshooting tools. Implementing a
Layer 3 access-layer in lieu of the traditional Layer 2 access replaces the required Layer 2 trunks with a
single point-to-point Layer 3 link. Pushing Layer 3 routing functionality one tier down on Layer 3 access
switches changes traditional multilayer network topology and the forwarding path. Implementing a
routed access layer does not require any physical or logical link reconfiguration or changes.
See Figure 1-3.
Figure 1-3
Control Function in Multi-Layer and Routed-Access Network Design
VSL
VSL
Core
Core
Routing
Routing
VSL
VSL
Layer 3
Distribution
Layer 3
Distribution
STP
Routing
Layer 2
Access
Access
Layer 2
Library
VLAN
20
Arts
VLAN
30
Multi-Layer Network
Admin
VLAN
10
Library
VLAN
20
Arts
VLAN
30
Routed-Access Network
228467
Admin
VLAN
10
At the network edge, Layer 3 access switches provides an IP gateway function and serve as a Layer-2
demarcation point to locally connected endpoints that can be logically segmented into multiple VLANs.
LAN Foundational Services
The Small Enterprise Design Profile uses essential foundational services to efficiently disseminate
information that is used by multiple clients, as well as identify and prioritize different applications traffic
based on their requirements. Designing the foundational services in a manner consistent with the needs
of the Small enterprise network system is paramount. Some of the key foundational services discussed
include the following:
Small Enterprise Design Profile Reference Guide
1-5
Chapter 1
Small Enterprise Design Profile (SEDP) Overview
Cisco SAFE Security Architecture Design Considerations
•
Multicast routing protocol design considerations
•
Designing QoS in the site network
WAN Design Considerations
WAN Transport
In order for sites to communicate with one another and/or to communicate outside the Enterprise
network system, network traffic must traverse over a WAN. WAN transport differs greatly from LAN
transport due to variables such as the type of connection used, the speed of the connection, and the
distance of the connection. The service fabric design model covers the following WAN transport design
considerations:
•
Internet
•
Metro Ethernet
WAN Foundational Services
Similar to the LAN, the WAN must employ essential foundational services to ensure the proper transport
and prioritization of community college services. WAN foundation services considered are as follows:
•
Routing protocol design
•
Quality-of-service (QoS)
•
WAN resiliency
•
Multicast
Cisco SAFE Security Architecture Design Considerations
Security of Small Enterprise Design Profile is essential. Without it, Enterprise solutions, applications,
and services are open to be compromised, manipulated, or shut down. The Small Enterprise Design
Profile was designed with built-in security by leveraging the proven design and deployment guidelines
of the Cisco SAFE security architecture. The following are the primary security design considerations:
•
Network Foundation Protection (NFP)—Ensuring the availability and integrity of the network
infrastructure, protecting the control and management planes.
•
Internet perimeter protection— Ensuring safe connectivity to the Internet, and protecting internal
resources and users from malware, viruses, and other malicious software. Protecting users from
harmful content. Enforcing E-mail and web browsing policies.
•
Serverfarm protection—Ensuring the availability and integrity of centralized applications and
systems. Protecting the confidentiality and privacy of user information and records.
•
Network access security and control—Securing the access edges. Enforcing authentication and
role-based access for employees and users residing at the main and remote sites. Ensuring systems
are up-to-date and in compliance with the network security policies.
•
Network endpoint protection—Protecting servers and Enterprise-controlled from viruses, malware,
botnets, and other malicious software. Enforcing E-mail and web browsing policies for users.
Small Enterprise Design Profile Reference Guide
1-6
Chapter 1
Small Enterprise Design Profile (SEDP) Overview
Mobility
Mobility
Mobility is an essential part of the Small Enterprise Design Profile. Most users will connect wirelessly
to site networks and other devices will also rely on the mobile network. In designing the mobility portion
of the service fabric, the following design criteria were used:
•
Accessibility—Enables employees and guests to be accessible and productive, regardless of where
they meet. This design element provides for easy, secure guest access to guests such as contractors,
vendors and other visitors.
•
Usability—In addition to extremely high WLAN transmission speeds made possible by the current
generation of IEEE 802.11n technology, latency sensitive applications (such as IP telephony and
video-conferencing) are supported over the WLAN using appropriately applied QoS. This gives
preferential treatment to real-time traffic, helping to ensure that video and audio information arrives
on time.
•
Security—Segment authorized users and block unauthorized users. Extend the services of the
network safely to authorized parties. Enforce security policy compliance on all devices seeking to
access network computing resources. Employees enjoy rapid and reliable authentication through
IEEE 802.1x and Extensible Authentication Protocol (EAP), with all information sent and received
on the WLAN being encrypted.
•
Manageability—Network administrators must be able to easily deploy, operate, and manage
hundreds of access points within multiple Enterprise network site deployments. A single, easy to
understand WLAN management framework is desired to provide small and large Enterprise systems
with the same level of wireless LAN management scalability, reliability and ease of deployment that
is demanded by traditional enterprise business customers.
•
Reliability—Provide adequate capability to recover from a single-layer fault of a WLAN
accessibility component or controller wired link. Ensure that wireless LAN accessibility is
maintained for employees and guest visitors in the event of common failures.
Collaboration Services Design Considerations
Adoption of IP technology has led to a fundamental change in designing networks. No longer are
networks used solely to provide data communication between computers and servers. IP technology has
extended beyond the data network and is now used extensively for Unified Communications and Video
communication as well. Unified Communications, IP Video Surveillance and Digital Media systems
were validated in the Small Enterprise Design Profile.
Unified Communications Design Considerations
Call Processing Considerations
How calls are processed in the Small Enterprise Design Profile environment is an important design
consideration. Guidance in designing scalable and resilient call processing systems is essential for the
successful deployment of a unified communications system. Considerations include the following:
•
Scale—The number of users, locations, gateways, applications, and so forth
•
Performance—The call rate
Small Enterprise Design Profile Reference Guide
1-7
Chapter 1
Small Enterprise Design Profile (SEDP) Overview
IP Video Surveillance Design Considerations
•
Resilience—The amount of redundancy
Gateway Design Considerations
Gateways provide a number of methods for connecting an IP telephony network to the Public Switched
Telephone Network (PSTN). Design considerations for gateways include the following:
•
PSTN trunk sizing
•
Traffic patterns
•
Interoperability with the call processing system
Dial Plan Considerations
Dial plan is one of the key elements of a unified communications system, and an integral part of all call
processing agents. Generally, the dial plan is responsible for instructing the call processing agent on how
to route calls. Specifically, the dial plan performs the following main functions:
•
Endpoint addressing
•
Path selection
•
Calling privileges
•
Digit manipulation
•
Call coverage
Survivability Considerations
Voice communications are a critical service that must be maintained in the event of a network outage and
for this reason the service fabric must take survivability into consideration. Guidance on how the Small
Enterprise Design Profile is equipped and designed to keep voice communications active in the event of
an outage is provided.
IP Video Surveillance Design Considerations
Video surveillance systems have proven their value in a wide range of applications. Video documentation
of critical incidents enhances employee safety and better protects valuable assets. However, traditional
analog Closed-circuit TeleVision (CCTV) surveillance systems have many limitations—they are unable
to store recorded video in local and remote locations or provide video access to mobile or remote users.
Network-centric video surveillance components include the following:
•
Cisco Video Surveillance Manager—Enables IT administrators and security personnel to view,
manage and record video locally and remotely using the IP network and a standard Internet browser.
Video can be securely accessed anywhere, at any time, enabling faster response, investigation and
resolution of incidents. Video can be recorded and stored locally off and at the main site allowing it
to be managed and aggregated with video from multiple locations.
•
Cisco Video Surveillance Media Server—A highly scalable and reliable video management platform
that manages, replicates, distributes and archives video systems.
Small Enterprise Design Profile Reference Guide
1-8
Chapter 1
Small Enterprise Design Profile (SEDP) Overview
IP Video Surveillance Design Considerations
•
Cisco Video Surveillance Operations Manager—A web-based user interface that authenticates and
manages access to video feeds. It is a centralized administration tool for the management of Media
Server hosts, Virtual Matrix hosts, cameras, encoders, and viewers.
•
Cisco Video Surveillance Media Virtual Matrix—Monitors video feeds in command center and other
24-hour monitoring environments. It allows operators to control the video being displayed on
multiple local and remote digital monitors.
Digital Media Systems
•
The Cisco Digital Media Suite—A comprehensive portfolio of digital signage, desktop video, and
enterprise TV components and applications that can be centrally managed. The Cisco Digital Media
Suite is comprised of three distinct subsystems.
•
Cisco Digital Signs—Provides scalable centralized management and publishing of compelling
digital media to networked, on-premise digital signage displays. It enables the disseminations news
and emergency information to large screens connected to the Enterprise existing network. The same
content to all signs in the network can be delivered.
•
Cisco Cast—Uses the same hardware as Cisco Digital Signs, but has different usage models. With
Cast, all control switches to the end user via a remote control—and with the latest release, IP phones,
smartphones, and touch screens—can be used to control what content comes to the screen. Cast has
three user interfaces, one to access VoDs, one for scrolling through live channels, and a channel
guide,
•
Cisco Show and Share—Enables employees to create, capture, and receive live and pre-recorded
video on their desktop computers.Digital media can be browsed, searched, and viewed over the
network through a unique, easy-to-use Cisco video portal experience—anywhere, anytime.
The components of the Cisco Digital Media Suite include the following:
•
Cisco Digital Media Manager (DMM)—The central management application for all Cisco Digital
Media Suite products. It is used to manage, schedule, and publish compelling digital media for Cisco
Digital Signs, Cisco Cast and Cisco Show and Share. As an integrated part of the Cisco Digital
Media Suite, this web-based media management application enables content owners to easily
upload, catalogue, edit, package, and publish digital media content for live or on-demand playback.
•
Scientific Atlanta Encoder—Encodes live video input into a MPEG-2 or MPEG-4 multicast stream
for Cisco Digital Signs and Cisco Cast.
•
Digital Media Encoders (DME) for Cisco Show and Share—Register with the DMM and broadcast
live video to the Streaming Server using RTSP. Recorded video may be archived to the Content
Repository.
•
Streaming Server for Cisco Show and Share—Provides stream splitting capabilities, allowing many
clients to view a single live stream from a DME or pre-recorded source (live rebroadcast).
•
Show and Share User Interface—Provides the web-based interface for clients. All navigation and
authentication is completed through this server.
•
Web Server/Content Repository—Holds all VoDs referenced by the Show and Share server. All VoD
streaming requests to the Show and Share server are redirected to this server. The web server/content
repository is also the component that holds all VoDs referenced by the DMM for Cisco Digital Signs
and Cisco Cast. For Digital Signs and Cast, all VoD streaming requests issued to the DMP are
serviced from this server.
•
Digital Media Player (DMP) for Cisco Digital Signs and Cisco Cast—Decodes and displays unicast
(prerecorded content streamed over http or RTSP) and multicast streamed video as well as flash
content (live content from the Scientific Atlanta encoder or other multicast encoder).
Small Enterprise Design Profile Reference Guide
1-9
Chapter 1
Small Enterprise Design Profile (SEDP) Overview
Summary
•
Cisco LCD Professional Series Displays—An integral part of the Digital Media System (DMS) suite
of products and are used to display information, Cisco LCD displays are available in different sizes
and models and offer full 1080p resolution.
Summary
The Small Enterprise Design Profile delivers validated network design and deployment best practices to
evolve small enterprise networks into Borderless networks. Enterprise Design Profile applies the best
practices from a collection of Cisco Validated Designs (CVD) and integrates these best practices into a
customer-based design profile.
For the existing network, the Small Enterprise Design Profile provides guidance for the evolution of the
Enterprise network to a Borderless Network. For new networks, the Small Enterprise Design profile
steps you from planning your Enterprise network to using technology to enable and solve your business
needs.
Small Enterprise Design Profile Reference Guide
1-10
CH A P T E R
2
Small Enterprise Design
Profile(SEDP)—Network Foundation Design
This chapter describes the Small Enterprise Design Profile network design, which is a well designed and
validated network architecture that is flexible, and cost effective to support a wide range of network
foundational services. Key features of this network design include the following:
•
High availability
•
Single fabric, multi services
•
Differentiated services
•
Layer 2 and Layer 3 access
This chapter provides design guidance to build a highly resilient, manageable, and cost-effective small
enterprise network that provides a solid foundation for seamless integration and operation of
applications and network services. The network has been specifically designed to meet the challenges of
the small enterprise environment.
Building Unified Small Enterprise Network Infrastructure
Cisco has years of experience developing high performance, highly available, multi service networks.
The key to developing a robust design is applying a proven methodology. The following design
principles were applied to develop the Small Enterprise Network Design architecture:
•
Hierarchy
– Clarifies the role of each device in each tier
– Simpler to deploy, operate, and manage the network
– Reduces fault domains at every tier
•
Modularity
– Enables growing the network on demand basis
•
Resiliency
– Meet users expectation of network always being available.
•
Flexibility
– Allows intelligent traffic load-sharing by using all network resources
Small Enterprise Design Profile Reference Guide
2-1
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
The Unified Campus network is designed to be highly available, and cost effective, while delivering
capabilities necessary to enable advanced services, such as IP telephony, video, security, wireless LANs.
The network design includes the following key features;
•
Hierarchical design with collapsed Core
•
Quality-of-service (QoS) to ensure real-time data (telephony, video) are given higher priority
•
Application of resilient design principles
•
Multi cast
•
Routed access
•
Redundancy
Hierarchical Network Design
The three-tier hierarchical model (see Figure 2-1) is the approach typically employed to achieve a high
performance, highly available, scalable network design. This design employs the four key design
principles of hierarchy, modularity, resiliency and flexibility.
Figure 2-1
Three-Tier Hierarchical Model
Core
Distribution
227529
Access
Each layer in the three-tier hierarchical model has a unique role to perform:
•
Access Layer—The primary function of an access-layer is to provide network access to the end user.
This layer often performs OSI Layer-2 bridge function that interconnects logical Layer-2 broadcast
domains and provides isolation to groups of users, applications, and other endpoints. The
access-layer interconnects to the distribution layer.
•
Distribution Layer—Multi-purpose system that interfaces between access layer and core layer.
Some of the key function for a distribution layer include the following:
– Aggregate and terminate Layer-2 broadcast domains
– Provide intelligent switching, routing, and network access policy function to access the rest of
the network.
Small Enterprise Design Profile Reference Guide
2-2
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
– Redundant distribution layer switches provides high availability to the end-user and equal-cost
paths to the core. It can provide differentiated services to various class-of-service applications
at the edge of network.
•
Core Layer—The core-layer provides high-speed, scalable, reliable and low-latency connectivity.
The core layer aggregates several distribution switches that may be in different buildings. Backbone
core routers are a central hub-point that provides transit function to access the internal and external
network.
Table 2-1 lists the key functions of each layer.
Table 2-1
Key Functions of Hierarchical Network Layer Devices
Key Function
Access
Distribution
Core
Network Transit
Rest of the network.
Internal and External network
Intelligent Services
PoE, IEEE 802.1AD,
Mobility, AutoQoS,
Auto-SmartPort Macro(ASP)
Route optimization
Network and System Virtualization
Layer-2 Interconnect
Forwarding Decision
Layer 2/Layer 3
Layer 3
Security Services
CISF, 802.1x, NAC, ACL etc. CISF, ACL, Route Filter,
CoPP etc.
ACL, Route Filter, CoPP etc.
QoS Services
Classification, Marking,
Policer and Queueing
Classification, Marking, and Queueing
To learn more about typical network designs, refer to the following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Campus/campover.html
Figure 2-2 illustrates a sample network diagram for a multi-building small enterprise network design.
Small Enterprise Design Profile Reference Guide
2-3
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
Figure 2-2
Multi Building Small Enterprise Network Design
Building B –
Marketing
and Sales
Building C –
Engineering
Access
Distribution
Building A –
Management
Core
Distribution
Building D –
Research and
Development
Building E –
Information
Technology
Building F –
Serverfarm
229273
Access
Collapsed Core Network Design
The three-tier hierarchical design maximizes performance, network availability, and the ability to scale
the network design. Most small enterprise campus' do not grow significantly larger over time, and most
small enterprise campus are small enough to be well served by a two-tier hierarchical design, where the
core and distribution layers are collapsed into one layer. The primary motivation for the collapsed core
design is reducing network cost, while maintaining most of the benefits of the three-tier hierarchical
model.
Deploying a collapsed core network results in the distribution layer and core layer functions being
implemented in a single device. The collapsed core/distribution device must provide the following:
Note
•
High speed physical and logical paths connecting to the network
•
Layer-2 aggregation and demarcation point
•
Define routing and network access policies
•
Intelligent network services—QoS, Network virtualization, etc.
If the main site or a remote site campus has multiple buildings, and is expected to grow over time, then
implementing the three-tier hierarchical model is a better choice.
Figure 2-3 illustrates a sample network diagram for a single main site building.
Small Enterprise Design Profile Reference Guide
2-4
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
Figure 2-3
Main Site—Collapsed Core Network Design
Access
Floor 6 –
Research and Development
Floor 5 –
Engineering
WAN
Floor 4 –
Serverfarm
Floor 2 –
Management
PSTN
Collapsed
Distribution/
Core
229274
Floor 3 –
Information Technology
Main Site Network Design
If the main site has multiple buildings and it is expected to grow significantly over time, then
implementing the three-tier hierarchical model is a good choice. For small size main sites that are
unlikely to grow significantly, the collapsed core model is more cost effective. The Small Enterprise
Design Profile uses the collapsed core network design in the main site.
The collapsed core network (see Figure 2-4) may be deployed with redundant core/distribution router,
or consolidated core/distribution router.
Small Enterprise Design Profile Reference Guide
2-5
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
Figure 2-4
Small Enterprise Design —Collapsed Core Network Models
Campus Deployment Design – 1
Internet
WAN
PSTN
Campus Deployment Design – 2
Internet
WAN
Distribution/Core
PSTN
Distribution/Core
Access
227532
Access
The redundant design is more complex, because all of the core/distribution functions must be
implemented on two routers in a complimentary fashion. To learn more about the redundant designs,
refer to High Availability Campus Recovery Analysis Design Guide at the following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Campus/HA_recovery_DG/campusRecovery.ht
ml
The main site is designed with a consolidated core/distribution router to maximize performance, while
keeping costs affordable (design 2). The consolidated collapsed core model has the following benefits:
•
Simplifies network protocols (eases network operations)
•
Enables symmetric forwarding paths
•
Delivers deterministic network recovery performance
With this design, the default behavior of Layer-2 and Layer-3 network control protocols is to create a
redundant view between two systems. The core router builds a ECMP routing topology which results in
symmetric forwarding paths beyond the main site.
Default Layer-2 configuration eliminates the need for FHRP, automatically eliminating the asymmetric
forwarding behavior which causes unicast flooding in the network. This simplifies the network
operation, since there is no need to configure or tune FHRP protocols.
The disadvantage of this Layer-2 network design is that the network is under-utilized. This is due to the
way Layer-2 protocols are designed to build loop-free network topologies. When two Layer-2 bridges
are directly connected, the STP protocol will block low-priority STP physical port in the forwarding
table. Figure 2-5 illustrates the control-plane, and the forwarding-plane for this design.
Small Enterprise Design Profile Reference Guide
2-6
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
Figure 2-5
Design Model 2 – Developing Control and Forwarding Paths
Control-Plane
Internet
WAN
Forwarding-Path
PSTN
Internet
WAN
PSTN
Per Physical
Port Layer 3
IGP adjacency
STP Primary
Root
STP Primary
Root
Per Physical
Port Layer 2
STP operation
VLAN
10
VLAN
20
VLAN
30
Layer 2 Trunk Port
Layer 3 Rounded Port
Standby Firewall Port
STP Block Port
VLAN
10
VLAN
20
Bi-directional Traffic Port
Non-Forwarding Port
VLAN
30
227536
Per Physical
Port Layer 3
IGP adjacency
This design suffers from two challenges:
•
Multiple-routing adjacencies between two Layer-3 systems. This configuration doubles the
control-plane load between each of the Layer-3 devices. It also uses more system resources like CPU
and memory to store redundant dynamic-routing information with different Layer-3 next-hop
addresses connected to same router.
•
As depicted in Figure 2-5, STP protocol blocks one of the physical ports in the Layer-2 network.
Since this design employs point-to-point links between the collapsed core and peer devices, the
solution is to tune the network to enable a single control plane, to improve forwarding efficiency and
resource utilization. The recommendation is to aggregate all physical ports into a single logical
channel-group. This logical aggregated Ethernet bundle interface is known as EtherChannel.
EtherChannel Fundamentals
EtherChannel provides inverse-multiplexing of multiple ports into a single logical port to a single
neighbor. This technique increases bandwidth, link efficiency, and resiliency. EtherChannel technology
operates on the MAC layer. Upper layer protocols require a single instance to operate over the logical
interface. EtherChannel provides efficient network operation and graceful recovery to higher layer
protocols during bundle port failure and restoration.
Small Enterprise Design Profile Reference Guide
2-7
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
The control-plane depicted in Figure 2-5 builds redundant Layer- 2 or Layer-3 network information over
each physical links. Each device builds common network prefix entries with a different next-hop path
pointing to same next hop device. Implementing EtherChannel results in a network topology with a
single destination entry for single next-hops, via the egress logical EtherChannel port. EtherChannel
reduces storing redundant network entries in the database and forwarding tables, which automatically
improves network convergence times and system resources utilization.
EtherChannel helps improve the overall network stability and availability. Failure of individual physical
link will cause network topology recomputation, restoration, and may be rerouted. Such process requires
CPU interruption that could impact the overall application performance. EtherChannel significantly
simplifies the network response to a individual link failure. If an individual link in EtherChannel fails,
the interface will not trigger any network topology changes. All underlying hardware changes remain
transparent to higher-layer protocols, thus minimizing impact to network and application performance,
and improving network convergence. Figure 2-6 illustrates how enabling EtherChannel in Layer-2 and
Layer-3 network simplifies control-plane and forwarding-plane.
Figure 2-6
Design Model 2 – Optimized Control and Forwarding Paths with EtherChannel
Control-Plane
Internet
WAN
Forwarding-Path
PSTN
Internet
WAN
PSTN
Single Layer 3
IGP adjacency
STP Primary
Root
STP Primary
Root
Single Layer 2
STP operation
VLAN
10
VLAN
20
VLAN
30
Layer 2 Trunk Port
Layer 3 Rounded Port
Standby Firewall Port
Small Enterprise Design Profile Reference Guide
2-8
VLAN
10
VLAN
20
Bi-directional Traffic Port
VLAN
30
227537
Single Layer 3
IGP adjacency
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
Resilient Distributed System
The Small Enterprise Design Profile uses the Cisco Catalyst 4500 with next-generation Supervisor-6E
in the consolidated core/distribution layer. It is chosen for its price performance, and the high availability
features within the device. The Cisco Catalyst 4500 switch supports redundant supervisor engines and
provides Stateful Switchover (SSO) and Non-Stop Forwarding (NSF) capabilities. SSO ensures the
Layer-2 and Layer-3 protocol state-machines and network forwarding entries on the standby supervisor
engine are maintained, and can quickly assume control-plane responsibilities and gracefully restore the
control-plane in the event of a primary supervisor failure. While the control-plane is gracefully
recovering, the NSF function continues to switch traffic in hardware.
The Cisco Catalyst 6500 platform is an enterprise-class system providing integrated network services
for large scale and high-speed networks. For large, multi building sites, or in situations where future
scalability is important, the Catalyst 6500 is a better choice for core/distribution layer switch. The design
principles remain the same when deploying a Catalyst 6500.
Main Site Access-Layer Edge Services
The access layer is the first tier or edge of the network. It is the layer where end-devices (PCs, printers,
cameras, etc.) attach to the small enterprise network. It is also the layer where devices that extend the
network out one more level are attached; IP phones and wireless access points (APs) are examples of
devices that extend the connectivity out from the access switch. The wide variety of devices that can
connect and the various services and dynamic configuration mechanisms required, make the access layer
the most feature-rich layer of the small enterprise network. Figure 2-7 illustrates a main site network
deployment with various types of trusted and untrusted endpoints.
Small Enterprise Design Profile Reference Guide
2-9
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
Figure 2-7
Access-Layer Trust Boundary and Network Control Services
Core
Distribution/Core
Network
Control
Services
Access
Trust
Boundary
IP
227538
Wired/Wireless
Trusted/Untrusted
Endpoints
Table 2-2 examples of the types of services and capabilities that need to be defined and supported in the
access layer of the network.
Table 2-2
Access-layer Services and Capabilities
Service Requirements
Service Features
Discovery and Configuration Services
802.1AF, CDP, LLDP, LLDP-MED
Integrated Security Services
IBNS (802.1X), CISF – Port-Security, DHCP Snooping, DAI and IPSG
Network Identity and Access
802.1X, MAB, Web-Auth
Application Recognition Services
QoS marking, policing, queueing, deep packet inspection NBAR
Intelligent Network Control Services
PVST+, Rapid PVST+, EIGRP, OSPF, DTP, PAgP/LACP, UDLD, FlexLink,
Portfast, UplinkFast, BackboneFast, LoopGuard, BPDUGuard, Port Security,
RootGuard
Energy Efficient Services
Power over Ethernet, EnergyWise, Energy efficient systems
Management Services
Auto-SmartPort Macro, Cisco Network Assistant
The access layer provides the intelligent demarcation between the network infrastructure and the
computing devices that use the infrastructure. It provides network edge security, QoS, and policy trust
boundary. It is the first point of negotiation between the network infrastructure and the end devices
seeking access to the network.
Small Enterprise Design Profile Reference Guide
2-10
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
A flexible network design, and the demand for mobility are two requirements which drive the access
layer design. A flexible network design allows any legitimate device to be connected anywhere in the
network (eg IP Phone, printer, video surveillance camera, digital signage, etc). Network users expect to
be able to move around their devices (laptops, PDAs, printers, etc) and gain network access wherever
necessary.
In order to allow devices to be moved within the network and ensure they associate with the correct
network policies and services; the following access services are integrated into the small enterprise
architecture:
•
Ability to physically attach to the network and be associated with or negotiate the correct Layer-1
and Layer-2 network services—PoE, link speed and duplex, subnet (VLAN or SSID)
•
Ability to provide device identification and, where needed, perform network access authentication
•
Ability for the network to apply the desired QoS policies for the specific user, device or traffic flow
(such as RTP streams)
•
Ability for the network to apply the desired security policies for the specific user or device
•
Ability for the network and device to determine and then register the location of the attaching device
•
Ability for the device to negotiate and register the correct end station parameters (such as DHCP),
as well as register for any other necessary network services (such as register for Unified
Communications presence and call agent services)
The basic steps for deploying edge access switch features are as follows:
Step 1
Configure the baseline switching foundation
Step 2
Protect the network infrastructure
Step 3
Protect the end devices and their application data flows
Step 4
Apply the necessary network policies (QoS) to provide for the required service levels.
Step 5
Create the final template macro to allow for simplified configuration
Access-Layer Network Control Services
Properly designing the distribution block ensures the stability of the overall architecture. In the collapsed
core model, the access-distribution block includes the access and distribution layers. Each of these layers
has specific service and feature requirements. The network control plane choice (i.e., routing or spanning
tree protocols) are central to determining how the distribution block fits within the overall architecture.
The Small Enterprise Design Profile includes two designs for configuring the access-distribution block:
multi-layer and routed-access. See Figure 2-8.
Small Enterprise Design Profile Reference Guide
2-11
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
Figure 2-8
Access-Distribution Deployment Model
Multi-Layer
Routed-Access
Distribution/Core
Distribution/Core
Layer 3
Access
VLAN
10
VLAN
20
Layer 2 Trunk Port
VLAN
10
VLAN
20
Layer 3 Rounded Port
227539
Layer 2
Access
While both of these designs use the same basic physical topology and cabling plant, there are several key
differences:
•
Where the Layer-2 and Layer-3 boundaries exist
•
How the network redundancy is implemented
•
How load-balancing works
A complete configuration description of each access-distribution block model is provided in “Logical
Multi-Layer Network” section on page 2-29 and the “Deploying Routed-Access Network” section on
page 2-34 of this document.
Resilient Access-Layer Network and System
The access-layer provides endpoint connectivity to the rest of the network. Typical access switches like
the Cisco Catalyst 2900 Series and Cisco Catalyst 3500 Series switches becomes single-point-of-failure
(SPOF), if the hardware fails or if there is a software upgrade. Disrupting communication to mission
critical endpoints (e.g., physical security camera) may be unacceptable.
The Small Enterprise Design Profile is designed with 2 to 4 uplink ports for each access switch,
providing link-failure protection. For mission critical endpoints, this design employs the Cisco
StackWise Plus and FlexStack solution in the access. It is designed to physically stack and interconnect
multiple Layer-2 or Layer-3 switches using special cables. Stacking multiple switches into a logical ring
creates a single unified and resilient access-layer network topology (see Figure 2-9). The
next-generation Cisco Catalyst 2960-S FlexStack can be deployed in Layer-2 network domain and the
Cisco Catalyst 3750-X StackWise Plus is deployed for routed access implementations.
Small Enterprise Design Profile Reference Guide
2-12
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
Figure 2-9
Resilient, Scalable and Efficient Access-Layer Network Design
Core
Distribution/Core
Link Protection –
Resilient Logical Link
Node Protection –
Catalyst 3750-X/2960S
with StackWise Plus/
Flex Stack
229275
Access
Distributed Layer 2 EtherChannel Uplink
Cross-StackWise Ring link
Main Site Data Center Network Design
The serverfarm is a central location which houses servers and storage. These resources must be available
to users throughout the small enterprise network. The serverfarm may be collocated at the small
enterprise network, or in a nearby site. Typically, small enterprise network are unable to afford
high-speed redundant WAN links between the serverfarm and the remote sites. This makes the design
vulnerable to service outage at the remote sites, in the event of WAN link failure. The Small Enterprise
Design Profile recommends which services should be placed in the centralized serverfarm, and which
services should be distributed (i.e., hosted at each remote site). The key criteria to consider when making
this decision include:
•
Scalability—The compute capacity and storage capacity of the centralized serverfarm must be
sufficient to handle peak loads.
•
Network Load—The overall network design, from serverfarm to remote site must have enough
capacity to carry the anticipated traffic (data and control traffic) to ensure good application
performance.
•
Redundancy—A WAN link failure will result in service outage between serverfarm and the remote
site. This can impact network data services and can expose security issue.
•
Synchronization—Some applications which are hosted locally will require content synchronization
with a centralized server.This can often be scheduled for off hours, to avoid adding traffic to WAN
links during normal working hours.
Table 2-3 provides a sample list of centralized and distributed servers.
Small Enterprise Design Profile Reference Guide
2-13
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
Table 2-3
Sample List of Centralized and Distribution Servers
Data Center Model
Server Function
Deployment Location
Centralized
Database server(i.e., Oracle, Sybase, etc)
Main Site Data Center
Cisco Unified Call Manager
Cisco Presence Server
Cisco Digital Media Manager (DMM)
E-mail Messaging Server
Distributed
Hosted services – Web, FTP, DHCP, DNS, NTP
Access-Control – Cisco Access Control Server
Main and Remote Site Data
Center
Cisco Video Surveillance Operation Management
Media Storage Server
Remote Site Network Design
The Small Enterprise Design Profile includes two remote site designs. One for larger remote site, and
another for medium to smaller remote sites. The typical remote site is a single building with a limited
population which makes the collapsed core network design a suitable choice.
Remote Site Collapsed Core Network Design
The key criteria to consider when designing a remote site network are the network size, bandwidth
capacity and high-availability requirements. The Small Enterprise Design Profile includes two models:
one for smaller remote site and another for larger remote site. Both designs offer high capacity,
performance, and availability. Figure 2-10 illustrates the two remote site network design models.
Small Enterprise Design Profile Reference Guide
2-14
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building Unified Small Enterprise Network Infrastructure
Figure 2-10
Remote Site—Collapsed Core Network Models
Small Enterprise
Campus Design Model – 1
WAN
PSTN
Cisco
Catalyst
4507-E
Distribution/Core
Small Enterprise
Campus Design Model – 2
WAN
PSTN
Cisco Catalyst
3750-X
StackWise Plus
Distribution/Core
Access
Access
VLAN
10
VLAN
10
VLAN
20
Layer 2 Trunk Port
Layer 3 Rounded Port
VLAN
20
VLAN
30
VLAN
30
Layer 2 Trunk Port
Layer 3 Rounded Port
229276
Chapter 2
Design Model-1 is for a larger remote site. The network design is the same as the main site network
design, with the same performance capabilities, scalability options, and high availability features.
Design Model-2 is for a medium to small remote site. The primary difference is the use of the Cisco
Catalyst 3750-X Stack Wise Plus switch in the collapsed core/distribution layer. The 3750-X Stack Wise
Plus deploys up to nine 3750-X switches in a ring topology as a single virtual switch. Each chassis
replicates the control functions, and provides packet forwarding. If the master switch fails, another
switch will quickly assume the control plane 'master' function. This results in a cost effective, high
performance, scalable solution, with built in resiliency.
•
Performance—Provides wire-rate network connection to access switches
•
Scalable—May deploy up to 9 switches in a stack to aggregate a reasonable number of access
switches
•
High Availability—Stack provides a virtual switch, with distributed control plane, delivering
subsecond system failure recovery times
The Cisco 3750-X StackWise Plus delivers high performance routing and switching capability and
robust IOS feature support. The control-plane and forwarding paths functions for the Cisco 3750-X
StackWise Plus in the collapsed core network design remain the same. However, the switching
architecture of the Cisco 3750-X StackWise differs significantly from the high-end distributed and
modular switch platforms like Cisco Catalyst 4500-E and 6500-E Series switches.
Small Enterprise Design Profile Reference Guide
2-15
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
For more information about the Cisco 3750-X StackWise-Plus architecture, refer to the following URL:
http://www.cisco.com/en/US/partner/prod/collateral/switches/ps5718/ps5023/prod_white_paper09186
a00801b096a_ps7077_Products_White_Paper.html
Remote Site Access-Layer Design
The access-layer network designate the remote site is the same as at the main site. The same devices are
available, and the same design choices may be deployed to achieve a high performance, secure and
resilient access layer. To simplify the overall system design, and network operations, it is recommended
to use consistent design and platform selections in the access-layer role, at the main and remote sites.
This will allow a common configuration template and simplify operations and troubleshooting
procedures.
Deploying Network Foundation Services
The two-tier hierarchical design delivers a reliable and resilient, scalable, and manageable foundation
network design. This subsection provides design and deployment guidelines for the small enterprise
campus core layer, and access-distribution block.
The access-distribution block, as described in the “Main Site Data Center Network Design” section on
page 2-13, uses a combination of Layer-2 and Layer-3 switching to provide a balance of policy and
access controls, availability, and flexibility in subnet allocation and VLAN usage. Deployment
guidelines are provided to implement multi-layer, and routed access designs in the access-distribution
block.
Implementing EtherChannel in Network
Etherchannel is used throughout the network design, and the implementation guidelines are the same for
multi-layer, and routed-access models, and in the WAN edge design. As recommended in the
“EtherChannel Fundamentals” section on page 2-7, there should be single logical point-to-point
EtherChannel deployed between collapsed core and access-layer. The EtherChannel configuration on
each end of the link in the access-distribution block must be consistent to prevent a link bundling
problem. EtherChannels use link bundling protocols to dynamically bundle physical interfaces into a
logical interface.
The following are the benefits of building EtherChannel in dynamic mode:
•
Ensure link aggregation parameters consistency and compatibility between switches.
•
Ensure compliance with aggregation requirements.
•
Dynamically react to runtime changes and failures on local and remote Etherchannel systems
•
Detect and remove unidirectional links and multi-drop connections from the Etherchannel bundle.
EtherChannels can be deployed in dynamic or static modes. Both EtherChannel modes can coexist in a
single system; however, the protocols (PagP, LACP) can not interoperate with each other.
•
Cisco proprietary link aggregation—Cisco's implementation of Port Aggregation group Protocol
(PAgP) is supported on all the Cisco Catalyst platforms. The PAgP protocol is not supported when
the Cisco Catalyst 2960-S or 3750-X Series switches are deployed in FlexStack or StackWise mode.
The PAgP protocol can operate in the different channel-group modes shown in Table 2-4 to initialize
link bundling process.
Small Enterprise Design Profile Reference Guide
2-16
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Table 2-4
channel-group
mode
Cisco’s Proprietary PAgP Channel-Group Mode
Distribution
Access-switch and WAN Edge
EtherChannel State
auto
auto
Non-Operational
desirable (recommended)
desirable (recommended)
Operational State
desirable
auto
auto
desirable
•
Table 2-5
IEEE 802.3ad link aggregation—Link Aggregation Control Protocol (LACP) is based on IEEE
802.3ad specification to operate in vendor-independent network environment. LACP link bundling
protocol is developed with same goal as Cisco's PAgP. Cisco Catalyst switches in FlexStack and
StackWise Plus mode must use LACP to dynamically bundle. LACP can operate in the following
different channel-group modes to initialize the link bundling process. See Table 2-5.
IEEE 802.3ad LACP channel-group Mode
channel-group mode
Distribution
Access-switch and WAN Edge
EtherChannel State
passive
passive
Non-Operational
active
active
Operational State
(recommended)
(recommended)
active
passive
passive
active
•
Static Mode—Each system statically bundles selected physical ports into a logical port-channel. In
static mode, Etherchannel consistency check is not performed between two switches, which may
lead to network protocol instability or network outage due to mis-configuration. This mode is not
recommended and should only be considered when EtherChannel is required but side of the link
does not support PAgP or LACP link aggregation protocol.
Small Enterprise Design Profile Reference Guide
2-17
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Figure 2-11
Implementing EtherChannel in Main Site Network
WAN
PSTN
Cisco Catalyst
3750-ME
Cisco
2800
Collapsed
Distribution/Core
Cisco Catalyst
4507R-E
Access
L2 EtherChannel
Cisco Catalyst
3560-X/2960
L3 EtherChannel
229277
Cisco Catalyst
3750-X
StackWise Plus
Cisco
Catalyst
2960-S
The following sample configuration shows how to build Layer-2 and Layer-3 EtherChannel
configuration and bundling physical ports into appropriate logical EtherChannel-group:
cr24-4507-1#config t
Enter configuration commands, one per line. End with CNTL/Z.
cr24-4507-1(config)#interface Port-channel1
cr24-4507-1(config-if)# description Connected to cr24-3750ME-1
cr24-4507-1(config-if)#
cr24-4507-1(config-if)#interface Port-channel11
cr24-4507-1(config-if)# description Connected to cr24-2960-1
cr24-4507-1(config-if)# switchport
cr24-4507-1(config-if)#
cr24-4507-1(config-if)#interface Port-channel16
cr24-4507-1(config-if)# description Connected to cr25-3750s-1
cr24-4507-1(config-if)# switchport
cr24-4507-1(config-if)#
cr24-4507-1(config-if)#interface range Gig 3/3 , Gig 4/3
cr24-4507-1(config-if-range)# description Connected to cr24-3750ME-1
cr24-4507-1(config-if-range)# channel-protocol pagp
cr24-4507-1(config-if-range)# channel-group 1 mode desirable
cr24-4507-1(config-if-range)#
cr24-4507-1(config-if-range)#interface range Gig 1/1 , Gig 2/1
cr24-4507-1(config-if-range)# description Connected to cr24-2960-1
cr24-4507-1(config-if-range)# channel-protocol pagp
cr24-4507-1(config-if-range)# channel-group 11 mode desirable
cr24-4507-1(config-if-range)#
cr24-4507-1(config-if-range)#interface range Gig 1/6 , Gig 2/6
cr24-4507-1(config-if-range)#description Connected to cr26-3750s-1
cr24-4507-1(config-if-range)# channel-protocol lacp
cr24-4507-1(config-if-range)# channel-group 16 mode active
Small Enterprise Design Profile Reference Guide
2-18
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Enabling EtherChannel on each switch endpoint will automatically form a logical connection and can be
verified using following CLI command:
cr24-4507-1#show etherchannel summary | inc Po
Group Port-channel Protocol
Ports
1
Po1(RU)
PAgP
Gi3/3(P)
Gi4/3(P)
11
Po11(SU)
PAgP
Gi1/1(P)
Gi2/1(P)
16
Po16(SU)
LACP
Gi1/6(P)
Gi2/6(P)
EtherChannel Load Balancing
EtherChannel load-sharing is based on a polymorphic algorithm. On per protocol basis, load sharing is
done based on source XOR destination address or port from Layer 2 to 4 header and ports. For higher
granularity and optimal utilization of each member-link port, an EtherChannel can intelligently
load-share egress traffic using different algorithms. EtherChannel load balancing method support varies
on Cisco Catalyst platforms. Table 2-6 summarizes the currently supported EtherChannel
load-balancing methods.
Table 2-6
EtherChannel Load Balancing Support Matrix
Packet Type
Classification
Layer
Load Balancing Mechanic
Supported Cisco Catalyst Platform
Non-IP
Layer 2
src-dst-mac
29xx, 35xx, 3750, 4500
IP
src-mac
dst-mac
src-dst-mac
IP
Layer 3
src-ip
dst-ip
src-dst-ip
IP
Layer 4
src-port
4500
dst-port
src-dst-port
EtherChannel load-balancing mechanisms function on a per-system basis. By default, EtherChannel will
use the hash computation algorithm. The network administrator can globally configure the load
balancing mechanism. In Cisco Catalyst platforms, EtherChannel load balancing is performed in
hardware and it cannot perform per-packet-based load balancing among different member links within
EtherChannel. Bandwidth utilization of each member-link may not be equal in default load balancing
mode. The Ether Channel load balancing method should be changed to source and destination IP
address-based throughout the main and remote site network for the following reasons:
•
One cannot optimize load balancing using hash tuning in a general network deployment model. This
is due to variations in application deployment and usage patterns.
•
EtherChannel does not take into account the bandwidth of each flow. Instead, it relies on the
statistical probability that the load is equally distributed across the links of the port-channel group,
given a large number of flows of relatively equal bandwidths. However, this may not always be true.
Small Enterprise Design Profile Reference Guide
2-19
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Tuning the load-balancing to source-and-destination IP address allows for statistically-superior
load-distribution. When loads are balanced in this manner, packets belonging to a single flow will
retain their packet order. See Figure 2-12.
Figure 2-12
EtherChannel Load-Balance Method
WAN
PSTN
EtherChannel Load
Balance Hash
src-dst-ip
Edge
src-dst-ip
Distribution/Core
227543
Access
src-dst-ip
The following output provides sample configuration guideline for changing the default port-channel
load-balance setting to source-destination-ip based. Aside from Layer-2 or Layer-3 EtherChannel
mode, similar configuration must be applied on each system in the access-distribution block and WAN
edge.
cr24-4507-1(config)#port-channel load-balance src-dst-ip
cr24-4507-1#show etherchannel load-balance
EtherChannel Load-Balancing Configuration:
src-dst-ip
EtherChannel Load-Balancing Addresses Used Per-Protocol:
Non-IP: Source XOR Destination MAC address
IPv4: Source XOR Destination IP address
IPv6: Source XOR Destination IP address
The following additional EtherChannel design and configuration must be taken into consideration for an
optimal EtherChannel design:
•
Enable single EtherChannel between access-layer and distribution system. Enabling more than a
single Ether Channel in a collapsed core network design imposes the same limitations as discussed
in non-EtherChannel scenario in Figure 2-5.
•
For optimal load sharing and hashing computation, it is recommended to bundle the number of
physical ports in powers of 2 (i.e., 2, 4, and 8).
•
EtherChannel is a logical interface in Cisco Catalyst platform. EtherChannel scalability in collapsed
core and distribution must be taken into account. The Cisco Catalyst 4500 can support up to 64
EtherChannels, whereas the Cisco Catalyst 3750 StackWise can support up to 48 EtherChannels
per-system.
Small Enterprise Design Profile Reference Guide
2-20
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Deploying Core Network Layer
This section provides implementation and best practice guidelines for deploying the core-layer in both
the main and remote enterprise sites. Proper design of the core network layer ensures reachability,
transparency and availability. This section focuses on building a unicast routing topology.
Routing Protocol
Enabling routing in the small enterprise network is a simple task. However, the network physical layout
must be carefully planned and designed to ensure flexible, stable and efficient routing. Developing a
hierarchical network addressing scheme enables a stable, efficient and scalable design.
•
Hierarchical network addressing—Structured IP network addressing in small enterprise LAN/WAN
network is a must to make network scalable, stable.
•
Routing protocol—Cisco IOS supports wide range of Interior Gateway Protocol (IGP). It is
recommended to deploy a single choice of routing protocol across the network infrastructure. This
solution guide does not recommended any particular IGP to deploy in the small enterprise network
architecture as it significantly varies based on different network infrastructure. However it will
provide some key points to be considered when selecting unicast routing protocol.
•
Hierarchical routing domain—Routing protocols must be designed in a hierarchical model that
allows network to scale and operate with greater stability. Building routing boundaries and
summarizing the network addresses minimizes topology size and synchronization procedure, which
improves the overall network resource utilization and reconvergence.
Routing Protocol Selection Criteria
•
Efficient address allocation—Hierarchical addressing enables efficient use of address space, since
groups are contiguous.
•
Improves routing efficiency—Using contiguous ip addresses enables efficient route summarization.
Route summarization simplifies the routing database, and computations during topology changes.
This reduces the network bandwidth used by the routing protocol, and improves routing protocol
performance by reducing network convergence time.
•
Improves system performance—Hierarchical, contiguous ip addressing reduces router memory
usage by eliminating dis-contiguous and non-summarized route entries. It saves on CPU cycles
needed to compute the routing database during topology changes. This contributes to a more stable
routing network, and simplifies the task of network operations and management.
Cisco IOS supports many Interior Gateway Protocols (IGP), including EIGRP and OSPF, either of which
are suitable for large network deployments. While OSPF is capable of greater scale, it is also more
complex, and hence more difficult to configure, operate and manage. The Small Enterprise Design
Profile is designed and validated using EIGRP, since it is a stable, high performance, efficient protocol,
which is simple to implement and manage. The same design principles apply whether using EIGRP or
OSPF.
Table 2-7 lists some of the EIGRP and OSPF side-by-side feature comparison information.
Small Enterprise Design Profile Reference Guide
2-21
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Table 2-7
EIGRP and OSPF feature comparison chart
Feature
EIGRP
OSPF
Classless Routing
Both routing protocols support classless routing that
allows to partition networks into VLSM.
Loop Prevention
Built-in mechanic to prevent routing loop in network.
Robust metric
Aggregated link Bandwidth + Delay
Efficient routing
Partial update
Multi-access routing
adjacency
Full-mesh
Hierarchical Routing
No. All routers considered in backbone. Non-backbone or OSPF area is divided in multiple
routers in non-transit path can be deployed in Stub role.
routing domains. Backbone area
maintains complete summarized
network topology; non-backbone
area can be transit or non-transit
OSPF routers.
Network convergence
Both routing protocol offers rapid network recovery during
link failure.
Graceful-Restart
Support
Yes
Yes. Cisco and IETF based
Route Summarization
Flexibility to manual summarized on any routing node.
Can only be performed on ABR or
ASBR
Load-Balancing
Support equal and un-equal cost load balancing
Equal-cost path only.
Standard
Cisco proprietary
IETF standard
Aggregated link Bandwidth
Hub-n-spoke
Designing End-to-End EIGRP Routing Domain
EIGRP is a balanced hybrid routing protocol that builds neighbor adjacency and a flat routing topology
on a per-autonomous-system (AS) basis. The LAN/WAN infrastructure of Small Enterprise Design
Profile should be deployed in a single EIGRP AS to prevent route redistribution, loops, and other
problems that may occur due to misconfiguration. See Figure 2-13.
Small Enterprise Design Profile Reference Guide
2-22
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Figure 2-13
End-to-End EIGRP Routing Design in Small Enterprise Architecture
Remote Site
EIGRP AS – 100
Remote Site
EIGRP AS – 100
EIGRP AS – 100
PSTN
WAN
PSTN
Main Site
EIGRP AS – 100
229278
Internet
Implementing EIGRP Routing
The main site is the central hub in the network. Each remote site is connected to the main site over the
WAN infrastructure. The main site network includes the Internet gateway and provides access to the
central data-center. Since main and remote sites use the collapsed core design, the routing configuration
of the core routers is the same.
The following is a sample configuration to enable EIGRP routing process at the edge of the main site
collapsed core network. EIGRP is enabled in the remote sites network with the same configuration:
cr24-4507-1(config)#interface Loopback0
cr24-4507-1(config-if)# ip address 10.125.100.1 255.255.255.255
cr24-4507-1(config-if)#interface Port-channel1
cr24-4507-1(config-if)# description Connected to cr24-3750ME-1
cr24-4507-1(config-if)#no switchport
cr24-4507-1(config-if)# ip address 10.125.32.4 255.255.255.254
cr24-4507-1(config-if)#interface Port-channel2
cr24-4507-1(config-if)# description Connected to cr24-2851-1
cr24-4507-1(config-if)#no switchport
cr24-4507-1(config-if)# ip address 10.125.32.6 255.255.255.254
Small Enterprise Design Profile Reference Guide
2-23
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
cr24-4507-1(config)#interface Vlan200
cr24-4507-1(config-if)# description Connected to cr24_ASA_Inside_Port
cr24-4507-1(config-if)# ip address 10.125.33.9 255.255.255.0
cr24-4507-1(config)#router eigrp 100
cr24-4507-1(config-router)# no auto-summary
cr24-4507-1(config-router)# eigrp router-id 10.125.100.1
cr24-4507-1(config-router)# network 10.125.0.0 0.0.255.255
cr24-4507-1#show ip eigrp neighbor port-channel 13
EIGRP-IPv4:(100) neighbors for process 100
H
Address
Interface
Hold Uptime
(sec)
1
10.125.33.10Vl200111d00h
1
200 0 171
0
10.125.32.7Po2161d02h
1
200 0 304
2
10.125.32.5Po1141d02h
2
200 0 25038
SRTT
RTO
(ms)
Q Seq
Cnt Num
EIGRP Adjacency Protection
Implementing summarization in the EIGRP routing process automatically enables EIGRP routing
process on each interface that is summarized. By default, the router transmits and accept EIGRP hello
messages from remote device to form an adjacency on all EIGRP enabled interfaces. This behavior needs
to be modified to ensure a secure, efficient and stable routing design:
•
System efficiency—There is no need to send EIGRP hellos on an interface where there is no trusted
EIGRP neighbor. In a large network, sending EIGRP hello messages periodically to such interfaces
consumes unnecessary CPU resource. EIGRP route processing should only be enabled on interfaces
where trusted network devices are connected. All other interfaces can be suppressed in passive
mode. The following configuration shows how to automatically disable EIGRP processing on all the
Layer-3 interfaces and only enable on the trusted interface. This design principle must be applied
on each EIGRP router, including distribution and core routers:
cr24-4507-1(config)#router eigrp 100
cr24-4507-1(config-router)# network 10.125.0.0 0.0.255.255
cr24-4507-1(config-router)# passive-interface default
cr24-4507-1(config-router)# no passive-interface Port-channel1
cr24-4507-1(config-router)# no passive-interface Port-channel2
cr24-4507-1(config-router)# no passive-interface Vlan200
cr24-3560r-1#show ip eigrp interface
EIGRP-IPv4:(100) interfaces for process 100
Interface
Vl2001
Po1 1
Po2 1
Xmit
Queue
Peers
Un/Reliable SRTT
0/01 0/1
50
0
0/02 0/1
50
0
0/04 0/1
50
0
Mean
Pacing Time
Multicast Pending
Un/Reliable
Flow Timer Routes
cr24-4507-1#show ip protocols | inc Passive|Vlan
Passive Interface(s):
Vlan1
Vlan101
Vlan102
Small Enterprise Design Profile Reference Guide
2-24
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Vlan103
Vlan104
•
Network Security—Sending unnecessary EIGRP Hello messages opens a security vulnerability in
two ways. An attacker can detect EIGRP operation and send flood of EIGRP hello messages to
destabilize the network. Or an attacker could establish a “fake” EIGRP adjacency and advertise a
best metric default-route into the network to black hole and compromise all critical traffic. Each
EIGRP system should implement MD5 authentication, and each EIGRP neighbor should validate
MD5 authentication is enabled on adjacent systems. This provides a secure method of transmitting
and receiving routing information between devices in the network. Following is a sample
configuration to enable EIGRP neighbor authentication using MD5:
– Distribution
cr24-4507-1(config)#key chain eigrp-key
cr24-4507-1(config-keychain)# key 1
cr24-4507-1(config-keychain-key)#
key-string <password>
cr24-4507-1(config)#interface Port-channel1
cr24-4507-1(config-if)# description Connected to cr24-3750ME-1
cr24-4507-1(config-if)# ip authentication mode eigrp 100 md5
cr24-4507-1(config-if)# ip authentication key-chain eigrp 100 eigrp-key
– WAN Aggregation
cr24-3750ME-1(config)#key chain eigrp-key
cr24-3750ME -1(config-keychain)# key 1
cr24-3750ME -1(config-keychain-key)#
key-string <password>
cr24-3750ME
cr24-3750ME
cr24-3750ME
cr24-3750ME
•
-1(config)#interface Port-channel1
-1(config-if)# description Connected to cr24-4507-1
-1(config-if)# ip authentication mode eigrp 100 md5
-1(config-if)# ip authentication key-chain eigrp 100 eigrp-key
System Stability—As mentioned in Table 8, EIGRP allows network administrator to summarize
multiple individual and contiguous networks into a single summarized network before advertising
to neighbors. Route summarization improves performance, stability, and convergence times, and it
makes the network easier to manage operate and troubleshoot.
EIGRP provides the flexibility to summarize at any point in the network. Proper design requires
determining which routers will serve as Aggregators, and advertise summarized network information to
peers. Routers which connect multiple access devices, or connect to the WAN edge should be made
Aggregators. Figure 2-14 provides an example small enterprise network with route aggregator devices
identified with the direction of route summarization illustrated.
Small Enterprise Design Profile Reference Guide
2-25
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Figure 2-14
Route Aggregator and Summary Route Advertisement Direction
Aggregator
Aggregator
WAN
Aggregator
227545
Aggregator
The following sample configuration shows EIGRP route summarization. In this example, the entire
access-layer network is summarized into a single classless network and advertised to the WAN edge, the
ASA firewall and the PSTN gateway:
•
Distribution
cr24-4507-1(config)#interface Port-channel1
cr24-4507-1(config-if)# description Connected to cr24-3750ME-1
cr24-4507-1(config-if)# ip summary-address eigrp 100 10.125.0.0 255.255.0.0
cr24-4507-1(config-if)#interface Port-channel2
cr24-4507-1(config-if)# description Connected to cr24-2851-1
cr24-4507-1(config-if)# ip summary-address eigrp 100 10.125.0.0 255.255.0.0
cr24-4507-1(config-if)#interface Vlan200
cr24-4507-1(config-if)# description Connected to cr24_ASA_Inside_Port
cr24-4507-1(config-if)# ip summary-address eigrp 100 10.125.0.0 255.255.0.0
cr24-4507-1#show ip protocols | inc Address|10.125.0.0
Address Family Protocol EIGRP-IPv4:(100)
Address Summarization:
10.125.0.0/16 for Port-channel1, Vlan200, Port-channel2
Small Enterprise Design Profile Reference Guide
2-26
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
•
WAN Aggregation
Verifying main site EIGRP summarized route status at WAN aggregation layer as follows:
cr24-3750ME-1#show ip route 10.125.0.0 255.255.0.0
Routing entry for 10.125.0.0/16
Known via "eigrp 100", distance 90, metric 1792, type internal
Redistributing via eigrp 100
Last update from 10.125.32.4 on Port-channel1, 1d04h ago
Routing Descriptor Blocks:
* 10.125.32.4, from 10.125.32.4, 1d04h ago, via Port-channel1
Route metric is 1792, traffic share count is 1
Total delay is 20 microseconds, minimum bandwidth is 2000000 Kbit
Reliability 255/255, minimum MTU 1500 bytes
Loading 1/255, Hops 1
Tuning EIGRP Protocol Timers
EIGRP uses Hello messages to form adjacencies and determine if neighbors are alive. EIGRP adjacency
is declared down if it fails to receive Hello messages within the Hold down timer interval. All the
prefixes discovered from a dead neighbor are removed from the routing table. By default, EIGRP
transmits a Hello message every 5 seconds to notify neighbors that it is still alive. The EIGRP hold-down
timer gets reset each time the router receives a EIGRP Hello message. Default EIGRP adjacency
hold-down timer is 15 seconds.
Lowering EIGRP hello and hold-down timer intervals improves network convergence times (i.e. time to
detect and respond to an outage). For small enterprise network design it is recommended to use the
default EIGRP Hello and Hold timer values for the following reasons:
•
EtherChannel benefits—EIGRP operates over the Layer-3 EtherChannel. In the event of a single
member-link failure condition, layer 2 will respond more quickly than the routing protocol, and
switchover traffic from the impacted link to an alternate member link. EIGRP routing is not
impacted by individual link member and no change in the routing table is required. Thus reducing
the EIGRP timers will not result in quicker convergence, and may adversely impact system stability.
•
High availability—The Cisco Catalyst 4507R-E and 3750-X Stack Wise Plus layer 3 switches
support graceful-restart protocol extensions which enables a redundant module or member switch
to gracefully assume the active role while maintaining adjacency with neighbors, during a active
supervisor failure condition. The backup supervisor requires sufficient time to detect a failure and
initiate graceful recovery with neighbors. Implementing aggressive timers may abruptly terminate
adjacency and cause network outage before a stateful switch over is accomplished. Thus, default
EIGRP Hello and Hold timers are recommended on Cisco Catalyst 4507R-E and 3750-X Stackwise
Plus Series Layer-3 platforms.
Deploying Multi-Layer Network
Multilayer design is one of the two access-distribution block designs included in the Small Enterprise
Design Profile. This section provides implementation and best practices guidelines the multi-layer
design. The deployment and configuration guidelines for the multi-layer access-distribution block are
the same for both main and remote site networks.
Small Enterprise Design Profile Reference Guide
2-27
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Spanning-Tree in Multilayer Network
Spanning Tree (STP) is a Layer-2 protocol that prevents logical loops in switched networks with
redundant links. The Small Enterprise Design Profile uses Etherchannel (point-to-point logical Layer-2
bundle) connection between access-layer and distribution switch which inherently simplifies the STP
topology and operation. In this design, the STP operation is done on a logical port, therefore, it will be
assigned automatically in forwarding state.
Over the years, the STP protocols have evolved into the following versions:
•
Per-VLAN Spanning Tree Plus (PVST+)—Provides a separate 802.1D STP for each active VLAN in
the network.
•
IEEE 802.1w – Rapid PVST+—Provides an instance of RSTP (802.1w) per VLAN. It is easy to
implement, proven in large scale networks that support up to 3000 logical ports and greatly improves
network restoration time.
•
IEEE 802.1s – MST—Provides up to 16 instances of RSTP (802.1w) and combines many VLANs
with the same physical and logical topology into a common RSTP instance.
Following is the example configuration to enable STP protocol in multi-layer network:
Distribution
cr24-4507-1(config)#spanning-tree mode rapid-pvst
cr24-4507-1#show spanning-tree summary | inc mode
Switch is in rapid-pvst mode
Access-Layer Switch
cr24-2960-1(config)#spanning-tree mode rapid-pvst
Default STP parameters optimize the network for packet forwarding. Best practice design includes
hardening STP parameters in the access and distribution switch to protect against STP misconfiguration,
or malicious user by deploying spanning-tree toolkit in the access-distribution block. See Figure 2-15.
Figure 2-15
Hardening Spanning-Tree Toolkit in Multi-Layer Network
Spanning-Tree
Root Switch
Distribution/Core
Root-Guard
Access
Edge-Port
Layer 2 Trunk Port
Small Enterprise Design Profile Reference Guide
2-28
227546
BPDU-Guard
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
The following is the configuration deploys spanning-tree toolkit in the access-distribution block:
Distribution
cr24-4507-1(config)#spanning-tree vlan 1-4094 root primary
cr24-4507-1(config)#interface range Gig 1/1 - 2 , Gig 2/1 - 2
cr24-4507-1(config)#spanning-tree guard root
Access
cr26-2960S-1(config)#interface GigabitEthernet1/0/1
cr26-2960S-1(config-if)#description CONNECTED TO UNTRUSTED-PC
cr26-2960S-1(config-if)#spanning-tree bpduguard enable
Other STP Toolkit Consideration
When the access-distribution block multi-layer design is deployed using the recommended best
practices, it automatically minimizes the need for deploying the following additional spanning-tree
toolkit technologies:
•
UplinkFast—Improves the network convergence time by providing direct access to the root switch
link failure. UplinkFast is not necessary in this design, because there is no alternate STP path and
RSTP protocol natively includes rapid recovery mechanism.
•
BackBone Fast—Provides rapid convergence from indirect Layer-2 link failures in a redundant
distribution switch configuration. This is feature is not necessary for the same reason as stated for
UplinkFast.
•
LoopGuard—Protects Layer-2 networks from loops that occur due to any malfunction that prevents
normal BPDU forwarding. A STP loop is created when a blocking port in a redundant topology
erroneously transitions to the forwarding state. This usually happens because one of the ports in a
physically redundant topology (not necessarily the blocking port) stopped receiving BPDUs.
Because there is single point-to-point STP forwarding port in this design, enabling Loopguard does
not provide any additional benefit. UDLD protocol must be implemented to prevent STP loop that
may occur in the network due to network malfunction, mis-wiring, etc.
Logical Multi-Layer Network
VLAN assignment can have a significant impact on network performance and stability. There are three
basic ways to assign VLANs within the access-distribution block.
Flat Logical Network Design
Spanning a single VLAN across multiple access-layer switches is much simpler with a single collapsed
core-distribution device versus a design with redundant distribution devices. The flat multi-layer design
has a single VLAN across multiple access devices, as shown in See Figure 2-16.
Small Enterprise Design Profile Reference Guide
2-29
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Figure 2-16
Multi-Layer Flat Network Design
Multi-Layer – Flat Network
Distribution/Core
VLAN
10
Marketing
VLAN
10
Sales
VLAN
10
Engineering
229279
Access
A flat multi-layer network deployment introduces the following challenges:
•
Scalability—Spanning the same VLAN in different access-layer switches will create a large Layer-2
broadcast domain that dynamically discovers and populates MAC address entries for endpoints that
may not need to communicate. In a large network, this may become a scalability issue (i.e. memory
required to hold large CAM table).
•
Performance—In a large network, spanning a large number of broadcast domains will impact the
performance of all network devices in the access-distribution block, because the switch will have to
process many more broadcast packets such as ARP.
•
Security—The flat muli0layer design widens the fault domain which increases possible attacks to a
larger number of users.The number of users is not necessarily due to the number switches spanned
and applications during DoS or viruses attack.
Segmented Logical Network Design
Best practice design includes identifying meaningful groups within the user community, and assigning
a unique VLAN to each group. These groups may be departments, user groups, or any other logical
grouping of users. Enabling a unique VLAN for each group will segment the network and build a logical
network structure. All network communication between groups will pass through the routing and
forwarding policies defined at the distribution layer. See Figure 2-17.
Small Enterprise Design Profile Reference Guide
2-30
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Figure 2-17
Multi-Layer Segmented Network Design
Multi-Layer – Segmented Network
Distribution/Core
Access
VLAN
20
Sales
VLAN
30
Engineering
229280
VLAN
10
Marketing
A segmented VLAN design is the solution to the challenges described in the flat network design. VLAN
segmentation improves the scalability, performance, and security of the network.
Hybrid Logical Network Design
The segmented logical network design improves scalability, performance and security, and addresses the
challenges of a flat network design. In real world deployments, there is usually a need for some users or
applications to communicate with all users (eg system administrator). The hybrid network design is the
segmented design, with the addition of a exceptional VLAN which spans the entire access-distribution
block. See Figure 2-18.
Figure 2-18
Multi-Layer Hybrid Network Design
Multi-Layer – Hybrid Network
Distribution/Core
Access
Sales
VLAN
20
Engineering
VLAN
30
VLAN 900
Network Managment
229281
Marketing
VLAN
10
Cisco recommends the segmented VLAN network design and optionally hybrid network for centralized
users or applications that requires distributed function across the access-layer network.
Following are the sample VLAN configuration steps in the access and the distribution layer switches.
Small Enterprise Design Profile Reference Guide
2-31
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Distribution
VLAN Trunking Protocol (VTP) is a Cisco proprietary Layer -messaging protocol that manages the
addition, deletion, and renaming of VLANs on a network-wide basis. Cisco's VTP simplifies
administration in a switched network. VTP can be configured in three modes: server, client, and
transparent. Set the VTP domain name and change the mode to the transparent mode as follows:
cr24-4507-1(config)#vtp domain campus
cr24-4507-1(config)#vtp mode transparent
cr24-4507-1(config)#vlan 10
cr24-4507-1(config-vlan)#name
cr24-4507-1(config-vlan)#vlan
cr24-4507-1(config-vlan)#name
cr24-4507-1(config-vlan)#vlan
cr24-4507-1(config-vlan)#name
cr24-3750-Mktg-Dept
20
cr24-3560-Sales-Dept
30
cr24-2960-Engg-Dept
Access
Set VTP domain name and change the mode to the transparent mode as follows:
cr24-3750-1(config)#vtp domain campus
cr24-3750-1(config)#vtp mode transparent
cr24-3750-1(config)#vlan 10
cr24-3750-1(config-vlan)#name cr24-3750-Sales-Dept
Implementing Layer 2 Trunk
In a typical network design, a single access switch will have more than one VLAN, for example a Data
VLAN and a Voice VLAN. The network connection between Distribution and Access device is a trunk.
VLAN's tag their traffic to maintain separation between VLANS across the trunk. By default on Cisco
Catalyst switches, the native VLAN on each layer 2 trunk port is VLAN 1, and cannot be disabled or
removed from VLAN database. The native VLAN remains active on all access switches layer 2 ports.
There are two choices for encapsulating the tagged VLAN traffic on the trunk: IEEE 802.1Q or Cisco
ISL. It is recommended to implement trunk encapsulation in static mode instead of negotiating mode, to
improve the rapid link bring-up performance. Not all Cisco Catalyst platforms support ISL
encapsulation; therefore IEEE 802.1Q is recommended, and validated in the access and distribution
switches.
Enabling the Layer-2 trunk on a port-channel, automatically enables communication for all of the active
VLANs between the access and distribution. This means an access-switch which has implemented, for
example, VLANs 10 to 15, will receive flood traffic destined for VLANs 20 to 25, which are
implemented on another access switch. RPVST +, using logical ports, operates on a per-VLAN basis to
load balance traffic. In a large network, it is important to limit traffic on Layer-2 trunk ports to only the
assigned VLANS, to ensure efficient and secure network performance. Allowing only assigned VLANs
on a trunk port automatically filters rest.
The default native VLAN must be properly configured to avoid several security risks—Attack, worm and
virus or data theft. Any malicious traffic originated in VLAN 1 will span across the access-layer
network. With a VLAN-hopping attack it is possible to attack a system which does not reside in VLAN
1. Best practice to mitigate this security risk is to implement a unused and unique VLAN ID as a native
VLAN on the Layer-2 trunk between the access and distribution switch. For example, configure VLAN
Small Enterprise Design Profile Reference Guide
2-32
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
802 in the access-switch and in the distribution switch. Then change the default native VLAN setting in
both the switches. Thereafter, VLAN 802 must not be used anywhere for any purpose in the same
access-distribution block.
Following is the configuration example to implement Layer-2 trunk, filter VLAN list and configure the
native-VLAN to prevent attacks on port channel interface. When the following configurations are
applied on port-channel interface (i.e., Port-Channel 11), they are automatically inherited on each
bundled member-link (i.e., Gig1/1 and Gig2/1):
Distribution
cr24-4507-1(config)#vlan 802
cr24-4507-1(config-vlan)#name Admin-Hopping-VLAN
cr24-4507-1(config)#interface Port-channel 11
cr24-4507-1(config-if)# description Connected to cr24-3750-1
cr24-4507-1(config-if)# switchport
cr24-4507-1(config-if)# switchport mode trunk
cr24-4507-1(config-if)# switchport trunk allowed vlan 101-110,900
cr24-4507-1(config-if)# switchport trunk native vlan 802
cr24-4507-1#show interface port-channel 11 trunk
Port
Po11
Mode
on
Encapsulation
802.1q
Status
trunking
Native vlan
802
Port
Po11
Vlans allowed on trunk
101-110,900
Port
Po11
Vlans allowed and active in management domain
101-110,900
Port
Po11
Vlans in spanning tree forwarding state and not pruned
101-110,900
Access-switch
cr24-3750-1(config)#vlan 802
cr24-3750-1(config-vlan)#name Admin-Hopping-VLAN
cr24-3750-1(config)#interface Port-channel 1
cr24-3750-1(config-if)# description Connected to cr24-4507-1
cr24-3750-1(config-if)# switchport
cr24-3750-1(config-if)# switchport mode trunk
cr24-3750-1(config-if)# switchport trunk allowed vlan 101-110,900
cr24-3750-1(config-if)# switchport trunk native vlan 802
Unidirectional Link Detection
UDLD is a Layer 2 protocol that works with the Layer 1 features to determine the physical status of a
link. At Layer 1, auto-negotiation takes care of physical signaling and fault detection. UDLD performs
tasks that auto-negotiation cannot perform, such as detecting the identity of neighbors and shutting down
Small Enterprise Design Profile Reference Guide
2-33
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
misconnected ports. When both auto-negotiation and UDLD are enabled, Layer 1 and Layer 2 detection
works together to prevent physical and logical unidirectional connections and the malfunctioning of
other protocols.
Copper media ports use Ethernet link pulse as a link monitoring tool and are not susceptible to
unidirectional link problems. Because one-way communication is possible in fiber-optic environments,
mismatched transmit/receive pairs can cause a link up/up condition even though bidirectional
upper-layer protocol communication has not been established. When such physical connection errors
occur, it can cause loops or traffic black holes. UDLD functions transparently on Layer-2 or Layer-3
physical ports. UDLD operates in one of two modes:
•
Normal mode—If bidirectional UDLD protocol state information times out; it is assumed there is
no-fault in the network, and no further action is taken. The port state for UDLD is marked as
undetermined. The port behaves according to its STP state.
•
Aggressive mode—If bidirectional UDLD protocol state information times out, UDLD will attempt
to reestablish the state of the port, if it detects the link on the port is operational. Failure to
reestablish communication with UDLD neighbor will force the port into the err-disable state. That
must be manually recovered by user or the switch can be configured for auto recovery within
specified interval of time.
Following is the configuration example to implement UDLD protocol:
Distribution
cr24-4507-1(config)#interface range Gig 1/2 , Gig 2/2
cr24-4507-1(config-int)#udld port
cr24-4507-1#show udld neighbor
PortDevice Name
Device ID
Port IDNeighbor State
-------------------------------- -------- -------------Gi1/2
FOC1318Y06V
1
Gi1/0/49Bidirectional
Gi2/2
FOC1318Y06J
1
Gi3/0/49Bidirectional
Access
cr26-2975-1(config)#interface Gig 1/0/49 , Gig 3/0/49
cr26-2975-1(config-if)#description Connected to cr24-4507-1
cr26-2975-1(config-if)#udld port
cr26-2975-1#show udld neighbor
PortDevice Name
Device ID
Port IDNeighbor State
-------------------------------- -------- -------------Gi1/0/49 FOX1216G8LT
1
Gi1/2
Bidirectional
Gi3/0/49 FOX1216G8LT
1
Gi2/2
Bidirectional
Deploying Routed-Access Network
This section provides implementation and best practices guidelines to deploy routed-access in the
access-distribution block. The routed access design moves the boundary between Layer 2 and Layer 3
from the distribution layer to the access layer as seen in Figure 2-19.
Small Enterprise Design Profile Reference Guide
2-34
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Figure 2-19
Control Function in Multi-Layer and Routed-Access Network Design
Routing
Routing
Layer 3
Distribution/Core
Layer 3
Distribution/Core
STP
Routing
Layer 2
Access
Access
Layer 2
Sales
VLAN
20
Engineering
VLAN
30
Multi-Layer Network
Marketing
VLAN
10
Sales
VLAN
20
Engineering
VLAN
30
Routed-Access Network
229282
Marketing
VLAN
10
Routing in the access-layer simplifies configuration, optimizes distribution performance, and improves
end-to-end troubleshooting tools. Implementing routing in the access-layer replaces Layer-2 trunk
configuration with single point-to-point Layer-3 interface in distribution layer. Placing Layer-3 function
one tier down on access-switches, changes the multilayer network topology and forwarding path.
Implementing Layer-3 function in the access-switch does not require a physical or logical link
reconfiguration; the same EtherChannel in access-distribution block can be used.
At the network edge, Layer-3 access-switches provides an IP gateway and become the Layer-2
demarcation point to locally connected endpoints that could be logically segmented into multiple
VLANs. Following are the benefits of implementing routed-access in the access-distribution block:
•
Eliminates the need to implement STP and the STP toolkit in the distribution layer. As a best
practice, STP toolkit must be hardened at the access-layer.
•
Shrinks the Layer-2 fault domain, which minimizes the number of endpoints affected by a
DoS/DDoS attack.
•
Improves Layer-3 uplink bandwidth efficiency by suppressing Layer-2 broadcasts at the access edge
port.
•
Improves performance by reducing resource utilization in collapsed core-distribution layer. In a
large multilayer network, the aggregation layer may consume more CPU cycles due to the large
number of MAC and ARP discovery and processing and storing required for each end-station.
Routed-access reduces the load of this Layer-2 processing and storage in the distribution layer, by
moving the load to layer-3 access-switches. Figure 2-20 illustrates where Layer-2 and Layer-3
forwarding entry processing and storage takes place when access-distribution block is implemented
as multi-layer versus routed-access network.
Small Enterprise Design Profile Reference Guide
2-35
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Forwarding Entry Development in Multi-tier Network
Distribution/Core
Distribution/Core
MAC
ARP
Route
Route
Access
Access
MAC
Marketing
VLAN
10
Sales
VLAN
20
Engineering
VLAN
30
MAC
ARP
Route
Multi-Layer Network
Marketing
VLAN
10
Sales
VLAN
20
Engineering
VLAN
30
Routed-Access Network
229283
Figure 2-20
While the routed access design is appropriate for many small enterprise networks it is not suitable for
all environments. Routed access does not allow a VLAN to span multiple access switches. Refer to
following URL for detailed design guidance for the routed access distribution block design:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Campus/routed-ex.html
Implementing EIGRP Routing in Access-Distribution Block
The Small Enterprise Design Profile uses EIGRP routing protocol, and all the devices in the LAN and
WAN sub-networks are deployed in a single AS. This subsection focuses on implementing EIGRP in the
access-distribution block. All the deployment and configuration guidelines in this section are the same
for deploying in the main or remote site network.
Following is the example configuration to enable basic EIGRP routing in the distribution layer and in
the access layer:
Distribution
cr24-4507-1(config)#interface Port-channel13
cr24-4507-1(config-if)# description Connected to cr24-3560r-1
cr24-4507-1(config-if)#no switchport
cr24-4507-1(config-if)# ip address 10.125.32.0 255.255.255.254
cr24-4507-1(config)#router eigrp 100
cr24-4507-1(config-router)# no auto-summary
cr24-4507-1(config-router)# eigrp router-id 10.125.100.1
cr24-4507-1(config-router)# network 10.125.0.0 0.0.255.255
cr24-4507-1#show ip eigrp neighbor port-channel 13
EIGRP-IPv4:(100) neighbors for process 100
H
Address
Interface
Hold Uptime
SRTT
(sec)
(ms)
3
10.125.32.1
Po13
14 00:02:14
2
Small Enterprise Design Profile Reference Guide
2-36
RTO
Q
Seq
Cnt Num
200
0 385
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Access
cr24-3560r-1(config)#interface Loopback0
cr24-3560r-1(config-if)# ip address 10.125.100.4 255.255.255.255
cr24-3560r-1(config-if)#
cr24-3560r-1(config-if)#interface Port-channel1
cr24-3560r-1(config-if)# description Connected to cr24-4507-1
cr24-3560r-1(config-if)# no switchport
cr24-3560r-1(config-if)# ip address 10.125.32.1 255.255.255.254
cr24-3560r-1(config)#ip routing
cr24-3560r-1(config)#router eigrp 100
cr24-3560r-1(config-router)# no auto-summary
cr24-3560r-1(config-router)# eigrp router-id 10.125.100.4
cr24-3560r-1(config-router)# network 10.125.0.0 0.0.255.255
cr24-3560r-1#show ip eigrp neighbor port-channel 1
EIGRP-IPv4:(100) neighbors for process 100
H
Address
Interface
Hold
Uptime
SRTT
RTO
(sec)
(ms)
0
10.125.32.0
Po1
13
00:10:00
1
Q
Seq
Cnt Num
200
0 176
Building EIGRP Network Boundary
EIGRP creates and maintains a single flat routing network topology between EIGRP peers. Building a
single routing domain enables complete network visibility and reach ability between all of the elements
within the network.(access, distribution, core, serverfarm, WAN, etc)
In a tiered design, the access layer always has a single physical or logical forwarding path to the
distribution layer. The access switch will build a forwarding topology pointing to same distribution
switch as a single Layer-3 next-hop. Since the distribution switch provides a gateway function to the
access switch, the routing design can be optimized with the following two techniques to improve
performance and network convergence in the access-distribution block:
•
Deploy Layer 3 access-switch in EIGRP stub mode
•
Summarize network view to Layer-3 access-switch for intelligent routing function
Deploy Layer 3 Access-Switch in EIGRP Stub Mode
The Layer-3 access switch can be deployed to announce itself as a stub router that acts as a non-transit
router and does not connect any other Layer-3 stub or non-stub routers. Announcing itself as a
non-transit stub Layer-3 router is one way to notify the distribution router that it should not include the
Layer-3 access switch in the EIGRP topology recomputation process. This optimized recomputation
process will prevent unnecessary EIGRP network queries, which reduces network traffic, and simplifies
the route computation.
As illustrated in Figure 2-21, implementing EIGRP stub function in the access switches, greatly reduces
the number of EIGRP network queries.
Small Enterprise Design Profile Reference Guide
2-37
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Figure 2-21
EIGRP Query Path with and without Stub Implementation
EIGRP Query Path without Stub
EIGRP Query Path with Stub
10.100.1.0/24
10.100.1.0/24
Marketing
VLAN
10
Sales
VLAN
20
Engineering
VLAN
30
Routed-Access Network
Marketing
VLAN
10
Sales
VLAN
20
Engineering
VLAN
30
Routed-Access Network
229284
EIRGP
Stub
EIGRP stub router in Layer-3 access-switch can announce routes to a distribution-layer router with great
flexibility.
EIGRP stub router can be deployed to announce routes dynamically discovered or statically configured.
Best practice design is to deploy EIGRP stub router to announce locally learned routes to aggregation
layer.
Following is the example configuration to enable EIGRP stub routing in the Layer-3 access-switch, no
configuration changes are required in distribution system:
Access
cr24-3560r-1(config)#router eigrp 100
cr24-3560r-1(config-router)#eigrp stub connected
cr24-3560r-1#show eigrp protocols detailed
Address Family Protocol EIGRP-IPv4:(100)
EIGRP metric weight K1=1, K2=0, K3=1, K4=0, K5=0
EIGRP maximum hopcount 100
EIGRP maximum metric variance 1
EIGRP NSF-aware route hold timer is 240
EIGRP stub, connected
Topologies : 0(base)
Distribution
cr24-4507-1#show ip eigrp neighbors detail port-channel 13
EIGRP-IPv4:(100) neighbors for process 100
H
Address
Interface
Hold Uptime
SRTT
(sec)
(ms)
1
10.125.32.1
Po13
13 00:19:19
16
Version 12.2/3.0, Retrans: 0, Retries: 0, Prefixes: 11
Topology-ids from peer - 0
Small Enterprise Design Profile Reference Guide
2-38
RTO
Q Seq
Cnt Num
200 0 410
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Stub Peer Advertising ( CONNECTED ) Routes
Suppressing queries
Summarizing Stub Routed-Access Network
Enabling the EIGRP stub function on the access switch does not change the distribution router behavior
of forwarding the full EIGRP topology table. The Distribution router must be configured to advertise
summarized routes that do not compromise end-to-end reach ability, and help access switches maintain
minimal routing information. In a network with a well designed IP addressing scheme, the aggregation
system can advertise summarized routes in a classless address configuration, that reduce individual
network advertisements, improve network scalability and network convergence. The distribution router
must have full network topology information to ensure efficient reachability paths. Therefore, it is
recommended to summarize at the distribution router, and not summarize at the access-layer.
Route summarization must be implemented on the distribution layer of main and each remote site
network. This includes devices such as the WAN aggregation in the main site. The distribution router
must advertise the following summarized network information to Layer 3 access-switch:
•
Local Network—Distribution router can be implemented in hybrid access-distribution configuration
that interconnects several multi-layer or routed-access enabled access-layer switches. Independent
of route origination source (connected or dynamic route) and network size within the
access-distribution block, the distribution router in main and remote site network must advertise a
single, concise and summarized Layer 3 network to each Layer 3 access-switch and to core devices.
•
Remote Network—Summarized network will be propagated dynamically across the network. Single
summarization of all remote networks may be advertised to local Layer 3 access-switches, since it
improves bandwidth efficiency. During a network outage, Layer 3 access-switch may drop traffic at
the network edge instead of transmitting it to the distribution router to black hole traffic.
•
WAN Network—Announcing a single summarized WAN network provides flexibility to troubleshoot
and verify network availability.
•
Default Network—When Layer 3 access-switch receives un-known destination traffic from the edge
that does not match any of the above mentioned summarized networks, then it is sent to the
distribution router to make a forwarding decision. The distribution router performs a forwarding
table lookup and may forward to appropriate path or black hole the traffic. In a typical small
enterprise network environment, a default route is announced by an Internet edge system, to forward
all internet traffic. Distribution router must propagate this default route to the Layer 3 access-switch.
Figure 2-22 illustrates a summarized EIGRP network advertisement, by route aggregation system, that
provides end-to-end internal and external network reachability.
Small Enterprise Design Profile Reference Guide
2-39
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Figure 2-22
End-to-End Routed-Access Network
10.127.0.0/21
10.125.0.0/16
10.126.0.0/16
10.127.0.0/16
0.0.0.0/0
Remote Site
10.127.112.0/21
10.125.0.0/16
10.126.0.0/16
10.127.0.0/16
0.0.0.0/0
Internet
WAN
Remote Site
10.126.0.0/16
0.0.0.0/0
Main Site
10.125.0.0/16
229285
10.125.0.0/16
10.126.0.0/16
10.127.0.0/16
0.0.0.0/0
Following is configuration example to deploy summarized and filtered Layer-3 network information to
Layer-3 access-switch.
Distribution
interface Port-channel13
description Connected to cr24-3560r-1
dampening
ip address 10.125.32.0 255.255.255.254
ip authentication mode eigrp 100 md5
ip authentication key-chain eigrp 100 eigrp-key
ip summary-address eigrp 100 10.125.0.0 255.255.0.0 5
load-interval 30
carrier-delay msec 0
!
!configure ACL and route-map to allow summarized route advertisement to Layer 3 accessswitch
!
access-list 1 permit 0.0.0.0
access-list 1 permit 10.126.0.0
access-list 1 permit 10.127.0.0
access-list 1 permit 10.125.0.0
Small Enterprise Design Profile Reference Guide
2-40
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
!
route-map EIGRP_STUB_ROUTES permit 10
match ip address 1
!
router eigrp 100
distribute-list route-map EIGRP_STUB_ROUTES out Port-channel13
cr24-4507-1#show ip protocols | inc Outgoing|filtered
Outgoing update filter list for all interfaces is not set
Port-channel13 filtered by
Access
cr24-3560r-1#show ip route eigrp
10.0.0.0/8 is variably subnetted, 15 subnets, 4 masks
D
10.126.0.0/16 [90/3328] via 10.125.32.0, 01:37:21, Port-channel1
D
10.127.0.0/16 [90/3584] via 10.125.32.0, 01:37:21, Port-channel1
D
10.125.0.0/16 [90/1792] via 10.125.32.0, 01:34:29, Port-channel1
D*EX 0.0.0.0/0 [170/515072] via 10.125.32.0, 00:03:15, Port-channel1
cr24-3560r-1#
EIGRP Adjacency Protection
EIGRP adjacency protection guidelines discussed earlier for the core network, apply equally to routed
access in the access-distribution block. The two challenges, system efficiency, and network security also
apply equally to the routed access design, and the same solution is applied:
•
System efficiency—EIGRP hello transmission must be blocked on an interface where there are no
trusted EIGRP neighbors, to reduce CPU utilization and prevent network attacks. EIGRP routing
process should only be enabled on interfaces where trusted enterprise devices are connected. All
other interfaces can be suppressed in passive mode.
Following is the example configuration on Layer-3 access-switch that advertises networks enabled
on SVI interfaces; however, keeps them in passive mode and explicitly allows EIGRP function on
uplink port-channel to distribution router. Same configuration principle must be applied on each
EIGRP router including distribution and core routers:
cr24-3560r-1(config)#router eigrp 100
cr24-3560r-1(config-router)# network 10.125.0.0 0.0.255.255
cr24-3560r-1(config-router)# passive-interface default
cr24-3560r-1(config-router)# no passive-interface Port-channel1
cr24-3560r-1#show ip eigrp interface
EIGRP-IPv4:(100) interfaces for process 100
Xmit
Interface
Po1
Peers
1
Un/Reliable
0/0
Queue
SRTT
1
Mean
Pacing Time
Multicast Pending
Un/Reliable
Flow Timer Routes
0/1
50
0
cr24-3560r-1#show ip protocols | inc Passive|Vlan
Passive Interface(s):
Vlan1
Vlan11
Vlan12
Vlan13
Vlan14
Small Enterprise Design Profile Reference Guide
2-41
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
•
Network Security—EIGRP adjacency between distribution and Layer-3 access-switch must be
secured. Following is the example configuration to enable EIGRP neighbor authentication using
MD5:
Distribution
cr24-4507-1(config)#key chain eigrp-key
cr24-4507-1(config-keychain)# key 1
cr24-4507-1(config-keychain-key)#
key-string <password>
cr24-4507-1(config)#interface Port-channel13
cr24-4507-1(config-if)# description Connected to cr24-3560r-1
cr24-4507-1(config-if)# ip authentication mode eigrp 100 md5
cr24-4507-1(config-if)# ip authentication key-chain eigrp 100 eigrp-key
Access
cr24-3560r-1(config)#key chain eigrp-key
cr24-3560r-1(config-keychain)# key 1
cr24-3560r-1(config-keychain-key)#
key-string <password>
cr24-3560r-1(config)#interface Port-channel1
cr24-3560r-1(config-if)# description Connected to cr24-4507-1
cr24-3560r-1(config-if)# ip authentication mode eigrp 100 md5
cr24-3560r-1(config-if)# ip authentication key-chain eigrp 100 eigrp-key
Tuning EIGRP Protocol Timers
EIGRP protocol functions the same in routed-access as it does in the core network. It is highly
recommended to retain default EIGRP hello and hold timers on distribution and Layer 3 access-switch
and rely on EtherChannel and SSO-based recovery mechanisms, that offers sub-second network
convergence, during individual link or supervisor failure scenarios.
Deploying Multicast in Network
Communications in a IP network can be:
•
Unicast—One source sends a message to one destination
•
Broadcast—One source sends a message to all destinations
•
Multicast—One source sends a message to a subset of destinations
IP multicast allows a source to transmit a message as a group transmission to a subset of hosts on the
network. Many collaboration applications, such as video conferencing, distance learning, software
distribution, utilize multicast techniques. IP multicast improves network bandwidth utilization, by
reducing un necessary duplicate traffic. Multicast improves efficiency by reducing data processing on
the source server, and sending a single flow into the network. Multicast packets are replicated in the
network where paths diverge, by Protocol Independent Multicast (PIM)-enabled routers, and other
supporting multicast protocols. See Figure 2-23.
Small Enterprise Design Profile Reference Guide
2-42
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Figure 2-23
Unicast versus Multicast Communication in Network
Unicast Communication
Remote Site
Remote Site
Main Site
229286
Main Site
Multicast Communication
Multicast IP Addressing
The Internet Assigned Numbers Authority (IANA) controls the assignment of IP multicast addresses. A
range of class D address space is assigned for IP multicast applications. All multicast group addresses
fall in the range of 224.0.0.0 through 239.255.255.255. In IP multicast packets, the destination IP
address is in the multicast group range, while the source IP address is always in the unicast address
range. The multicast IP address space is further divided into several pools for well-known multicast
network protocols, and inter-domain multicast communications as shown in Table 2-8.
Small Enterprise Design Profile Reference Guide
2-43
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Table 2-8
Multicast Address Range Assignments
Application
Address Range
Reserved – Link Local Network Protocols
224.0.0.0/24
Globally Scope – Group communication between organization and Internet
224.0.1.0 – 238.255.255.255
Source Specific Multicast (SSM) – PIM extension for one-to-many unidirectional multicast 232.0.0.0/8
communication
GLOP – Inter-domain Multicast group assignment with reserved global Autonomous System 233.0.0.0/8
(AS)
Limited Scope – Administratively scope address that remains constrained within local
organization or AS. Commonly deployed in enterprise and other organization.
239.0.0.0/8
For the Schools SRA network design, the multicast IP addresses must be selected from the Limited
Scope pool (239.0.0.0/8).
Multicast Routing Design
Each device between a multicast source and receiver must enable dynamic multicast. The technique for
creating a multicast forwarding table is different than unicast routing and switching techniques.
Multicast requires Multicast Routing Protocol (MRP) and Dynamic Group Membership (DGM) to
enable communication.
Multicast Routing Protocol
IP multicast delivers source traffic to multiple receivers using the least amount of network resources,
without placing additional burden on the source or the receivers. Multicast packet replication in the
network is performed by Cisco routers and switches enabled with Protocol Independent Multicast (PIM)
and other multicast routing protocols.
The network must build a packet distribution tree that specifies a unique forwarding path between the
source subnet and each multicast group members subnet. A primary goal for the tree is to ensure that
only one copy of each packet is forwarded on each branch of the tree. The two basic types of multicast
distribution trees are source trees and shared trees:
•
Source trees—The simplest form of a multicast distribution tree is a source tree, with the source at
the root and the receivers at the branches. Because this tree uses the shortest path through the
network, it is also referred to as a shortest path tree (SPT).
•
Shared trees—A shared tree uses a single common root placed at a chosen point in the network. This
shared root is called a Rendezvous Point (RP).
PIM protocol has two modes which support both types of multicast distribution trees:
•
Dense Mode—This mode assumes that most routers in the network will distribute multicast traffic
to each multicast group. PIM-DM builds distribution trees by initially flooding the entire network
and then pruning back the small number of paths without receivers.
•
Sparse Mode—This mode assumes that relatively few routers in the network will be involved in each
multicast group. The hosts belonging to the group are usually widely dispersed, as would be the case
for most multicast over the WAN. PIM-SM begins with an empty distribution tree and adds branches
only as the result of explicit IGMP requests to join.
Small Enterprise Design Profile Reference Guide
2-44
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
It is recommended to deploy multicast in PIM-SM in the small enterprise network design. All the
recommended platforms in this design support PIM-SM mode on physical or logical (SVI and
EtherChannel) interfaces.
Dynamic Group Membership
Multicast receiver registration and deletion is done via Internet Group Management Protocol (IGMP)
signaling. IGMP operates between a multicast receiver in the access-layer and a collapsed core router at
the distribution layer in the main or the remote site.
In a multi-layer design, the layer 3 boundary is at the distribution switch. Multi-layer access-switches
do not run PIM, and therefor flood the traffic on all ports. This multi-layer access-switch limitation is
solved by using IGMP snooping feature, which is enabled by default. Best practice is to not disable
IGMP snooping feature.
In a routed-access network design, the Layer-3 boundary is at the access-layer and IGMP
communication is between receiver and access-switch. Along with unicast routing protocol, PIM-SM
must be enabled on the Layer 3 access-switch to communicate with RP in the network.
Figure 2-24 demonstrates multicast source and receiver registration procedure and how shared-tree is
dynamically developed for multicast data delivery.
Multicast Source and Receiver Registration Procedure
Distribution/Core
Serverfarm
Source
Access
Receiver
PIM-SM
IGMP
229430
Figure 2-24
Small Enterprise Design Profile Reference Guide
2-45
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Deploying PIM-SM
PIM-SM distributes information about active sources by forwarding data packets on the shared tree.
Because PIM-SM uses shared trees initially, it requires the use of a RP. It is recommended to deploy the
RP close to the multicast source (collapsed core-distribution router in the main site is a good choice).
Multicast sources centrally deployed in main site will register themselves with the RP and then data is
forwarded down the shared tree to the receivers that could be located anywhere in the network.
PIM-SM Rendezvous Point
PIM-SM distributes information about active sources by forwarding data packets on the shared tree.
Because PIM-SM uses shared trees initially, it requires the use of a RP. It is recommended to deploy the
RP close to the multicast source (collapsed core-distribution router in the district office is a good choice).
Multicast sources centrally deployed in district office will register themselves with the RP and then data
is forwarded down the shared tree to the receivers that could be located anywhere in the network.
PIM-SM supports RP deployment in the following three different modes in the network:
•
Static—As the name implies, RP must be statically identified and configured on each PIM router in
the network. RP load-balancing and redundancy can be achieved using Anycast RP.
•
Auto-RP—Dynamic method to discover and announce RP in the network. Auto-RP implementation
is beneficial when there are multiple RPs and groups that often change in the network. To prevent
network reconfiguration during change, RP mapping agent router must be designated in the network
to receive RP group announcements and arbitrate conflicts. This capability is part of PIM version 1
specification.
•
BootStrap Router (BSR)—Performs same task as Auto-RP but different mechanism. This capability
is pare of PIM version 2 specification. Auto-RP and BSR cannot coexist or interoperate in the same
network.
In a small to mid-size multicast network, static RP configuration is best overall, due primarily to the
amount of administrative overhead that Auto-RP or BSR introduce. Static RP implementation offers
same RP redundancy and load sharing and a simple ACL can be applied to deploy RP without
compromising multicast network security. See Figure 2-25.
Small Enterprise Design Profile Reference Guide
2-46
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Figure 2-25
PIM-SM Network Design in Network Infrastructure
PIM-SM
PIM-SM
PIM-SM
PIM-SM
PIM-SM
PIM-SM
WAN
PIM-SM
PIM-SM
Serverfarm
Source
PIM-SM
RP
229431
Chapter 2
PIM-SM
Following is an example configuration to deploy PIM-SM RP in the small enterprise network. Similar
static PIM-SM configuration must be enabled on each Layer-3 PIM router or an access-switch in the
remote sites:
Distribution - RP
cr24-4507-1(config)#interface Loopback1
cr24-4507-1(config-if)# description RP
cr24-4507-1(config-if)# ip address 10.125.100.100 255.255.255.255
cr24-4507-1(config)#ip multicast-routing
cr24-4507-1(config)#ip pim rp-address 10.125.100.100
Layer 3 Access
cr24-3560r-1(config)#ip multicast-routing distributed
cr24-3560r-1(config)#ip pim rp-address 10.125.100.100
Remote Site Core
cr36-3750s-1(config)#ip multicast-routing distributed
cr36-3750s-1(config)#ip pim rp-address 10.125.100.100
Small Enterprise Design Profile Reference Guide
2-47
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Upon successful PIM-SM RP implementation throughout the enterprise network, PIM-SM must be
enabled on Layer-3 edge and core network-facing ports. The following sample configuration provides a
simple PIM-SM implementation guideline to be implemented on every intermediate Layer-3 systems
between receiver and source:
Distribution - RP
! Main Site - Access Network
cr24-4507-1(config)#interface range Vlan101 - 140
cr24-4507-1(config-if-range)# ip pim sparse-mode
! Main Site - Data Center Network
cr24-4507-1(config)#interface range Vlan141 - 150
cr24-4507-1(config-if-range)# ip pim sparse-mode
! Layer 3 Core and Routed-Access Port-Channel
cr24-4507-1(config)#interface range Port-channel 1, Port-channel 13, Port-channel 15
cr24-4507-1(config-if-range)# ip pim sparse-mode
cr24-4507-1#show ip pim interface
Address
Interface Ver/
Nbr
10.125.32.4
<omitted>
10.125.1.1
Port-channel1v2/S
Vlan101
Query DR
Mode
Count
30
1
1
v2/S 0
30
DR
Intvl Prior
10.125.32.4
1
10.125.1.1
cr24-4507-1#show ip mroute sparse
(*, 239.192.51.8), 02:33:37/00:03:12, RP 10.125.100.100, flags: SJC
Incoming interface: Null, RPF nbr 0.0.0.0
Outgoing interface list:
Vlan111, Forward/Sparse, 02:04:33/00:02:44, H
Vlan101, Forward/Sparse, 02:04:59/00:02:58, H
Port-channel15, Forward/Sparse, 02:04:59/00:02:47, H
Vlan131, Forward/Sparse, 02:04:59/00:02:32, H
Port-channel13, Forward/Sparse, 02:04:59/00:03:12, H
Vlan121, Forward/Sparse, 02:04:59/00:02:14, H
Vlan146, Forward/Sparse, 02:21:26/00:02:01, H
Layer 3 Access
! Main Site - Layer 3 Access Network
cr24-3560r-1(config)#interface range Vlan11 - 20
cr24-3560r-1(config-if-range)# ip pim sparse-mode
! Routed-Access Port-Channel
cr24-4507-1(config)#interface Port-channel 1
cr24-4507-1(config-if)# ip pim sparse-mode
cr24-3560r-1#show ip pim interface
Address
Interface Ver/
Nbr
10.125.32.1Port-channel1v2/S
10.125.11.1
Vlan11v2/S
Small Enterprise Design Profile Reference Guide
2-48
1
0
Query DR
DR
Mode
Count Intvl Prior
30
1
10.125.32.0
30
1
10.125.11.1
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
Implementing IGMP
By default the Layer-2 access-switch will dynamically detect IGMP hosts and multicast-capable Layer-3
routers in the network. The IGMP snooping and multicast router detection functions on a per VLAN
basis, and is globally enabled by default for all the VLANs. The IGMP configuration can be validated
using the show command on the Layer-2 access-switch:
cr24-2960-1#show ip igmp snooping
Global IGMP Snooping configuration:
------------------------------------------------IGMP snooping
: Enabled
IGMPv3 snooping (minimal)
: Enabled
Report suppression
: Enabled
TCN solicit query
: Disabled
TCN flood query count
: 2
Robustness variable
: 2
Last member query count
: 2
Last member query interval
: 1000
cr24-2960-1#show ip igmp snooping mrouter
Vlan
ports
----------101
Po1(dynamic)
102
Po1(dynamic)
cr24-2960-1#show ip igmp snooping group
Vlan
GroupType
Version
Port List
----------------------------------------------------------------------101
239.192.51.1igmp
v2
Fa0/1, Po1
101
239.192.51.2igmp
v2
Fa0/2, Po1
Multicast routing function changes when the access-switch is deployed in routed-access mode. PIM
operation is performed at the access layer, therefore multicast router detection process is eliminated. The
following output from a Layer-3 switch verifies that the local multicast ports are in router mode, and
provide a snooped Layer-2 uplink port-channel which is connected to the collapsed core router, for
multicast routing:
cr24-3560r-1#show ip igmp snooping mrouter
Vlan
ports
----------11
Router
12
Router
cr24-3560r-1#show ip igmp membership | inc Channel|Vl
Channel/Group
Reporter
Uptime
Exp.FlagsInterface
*,239.192.51.8
10.125.11.2000:17:52 02:45 2A
Vl11
*,239.192.51.9
10.125.11.13100:17:52 02:43 2A
Vl12
Multicast Security—Preventing Rogue Source
This section provides basic multicast security configuration guidelines to prevent an unauthorized host
in the network from acting like a rogue source in the network and sending multicast traffic.
In a PIM-SM network, an unwanted traffic source can be controlled with the pim accept-register
command. When the source traffic hits the first-hop router, the first-hop router (DR) creates (S,G) state
and sends a PIM Source Register message to the RP. If the source is not listed in the accept-register filter
list (configured on the RP), then the RP rejects the Register and sends back an immediate Register-Stop
Small Enterprise Design Profile Reference Guide
2-49
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying Network Foundation Services
message to the DR. The drawback with this method of source-filtering is that the pim accept-register
command on the RP, PIM-SM (S,G) state is still created on the source's first-hop router. This can result
in traffic reaching receivers local to the source and located between the source and the RP. Furthermore,
the pim accept-register command works on the control plane of the RP, which could be used to overload
the RP with “fake” register messages, and possibly cause a DoS condition.
Best practice is to apply the pim accept-register command on the RP in addition to other edge-filtering
methods, such as simple data plane ACLs on all DRs and on all ingress points into the network. While
ingress ACLs on the DR are sufficient in a perfectly configured and operated network, best practice
includes configuring the pim accept-register command on the RP in the main site as a secondary security
mechanism in case of misconfiguration on the edge routers.
Following is the sample configuration with a simple ACL which has been applied to the RP to filter only
on the source address. It is also possible to filter the source and the group with the use of an extended
ACL on the RP:
Distribution-RP
cr24-4507-1(config)#ip access-list extended PERMIT-SOURCES
cr24-4507-1(config-ext-nacl)# permit ip 10.125.31.80 0.0.0.15 239.192.0.0 0.0.255.255
cr24-4507-1(config)#ip pim accept-register list PERMIT-SOURCES
Multicast Security—Preventing Rogue RP
Any router can be misconfigured or maliciously advertise itself as a multicast RP in the network. with
the valid multicast group address. With a static RP configuration, each PIM-enabled router in the
network can be configured to use the static RP for the multicast source and ignore any Auto-RP or BSR
multicast router announcement.
Following is the sample configuration that must be applied to each PIM-enabled router in the main and
remote sites, to accept PIM announcements only from the static RP and ignore dynamic multicast group
announcement from any other RP:
Distribution-RP
cr24-4507-1(config)#ip access-list standard Allowed_MCAST_Groups
cr24-4507-1(config-std-nacl)# permit 224.0.1.39
cr24-4507-1(config-std-nacl)# permit 224.0.1.40
cr24-4507-1(config-std-nacl)# permit 239.192.0.0 0.0.255.255
cr24-4507-1(config)#ip pim rp-address 10.125.100.100 Allowed_MCAST_Groups override
cr24-4507-1#show ip pim rp mapping
PIM Group-to-RP Mappings
Acl: Allowed_MCAST_Groups, Static-Override
RP: 10.125.100.100 (?)
Small Enterprise Design Profile Reference Guide
2-50
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
Deploying QoS in Network
IP networks forward traffic on a best-effort basis by default. The routing protocol forwards packets over
the best path, but offers no guarantee of delivery. This model works well for TCP-based data applications
that adapt gracefully to variations in latency, jitter, and loss. The Small Enterprise Design Profile is a
multi-service network design which supports voice and video as well as data traffic on a single network.
Real time applications (such as voice, video) require packets delivered with in specified loss, delay and
jitter parameters. Quality-of-Service (QoS) is a collection of features which allows the network to
dedicate network resources for higher priority real time applications, while reserving sufficient network
resources to service lower priority traffic. QoS accomplishes this by providing differentiated services,
depending on the traffic type. For a detailed discussion of QoS, refer to the Enterprise QoS SRND at the
following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/WAN_and_MAN/QoS_SRND/QoS-SRND-Bo
ok.html
While design principles are common, QoS implementation varies between fixed-configuration switches
and the modular switching platforms like the Cisco Catalyst 4500-E/6500-E. This section discusses the
internal switching architecture and the differentiated QoS structure on a per-hop-basis.
QoS in Catalyst Fixed Configuration Switches
The QoS implementation in Cisco Catalyst 2960, 2960S, 3560-X and 3750-X Series switches is similar.
There is no difference in ingress or egress packet classification, marking, queuing and scheduling
implementation among these Catalyst platforms. The Cisco Catalyst switches allow users to create a
policy-map by classifying incoming traffic (Layer 2 to Layer 4). Catalyst switches allow attaching the
policy-map to an individual physical port or to logical interfaces (SVI or port-channel). This creates a
common QoS policy which may be used in multiple networks. To prevent switch fabric and egress
physical port congestion, the ingress QoS policing structure can strictly filter excessive traffic at the
network edge. All ingress traffic from edge ports passes through the switch fabric and congestion may
occur at the egress ports. Congestion in access-layer switch can be prevented by tuning queuing
scheduler and Weighted Tail Drop (WTD) drop parameters. See Figure 2-26.
Receive
Fixed Configuration Catalyst QoS Architecture
Policer
Marker
Policer
Marker
Internal
Ring
Ingress
Queue
NormalQ
SRR
Classify
Priority-Q
Policer
Marker
Policer
Marker
Ingress QoS
Egress
Queue
Q1
Q2
Q3
Q4
SRR
Transmit
Egress QoS
227557
Figure 2-26
The main difference between these platforms is the switching capacity which ranges from 1G to 10G.
The switching architecture and some of the internal QoS structure differs between these switches also.
Following are some important differences to consider when selecting the access switch:
•
The Cisco Catalyst 2960 does not support multilayer switching and does not support per-VLAN or
per-port/per-VLAN policies.
Small Enterprise Design Profile Reference Guide
2-51
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
•
The Cisco Catalyst 2960 can police to a minimum rate of 1 Mbps; all other switches including
next-generation Cisco Catalyst 2960-S Series within this product family can police to a minimum
rate of 8 kbps.
•
Only the Cisco Catalyst 3560-X and 3750-X support IPv6 QoS.
•
Only the Cisco Catalyst 3560-X and 3750-X support policing on 10-Gigabit Ethernet interfaces.
•
Only the Cisco Catalyst 3560-X and 3750-X support SRR shaping weights on 10-Gigabit Ethernet
interfaces.
The next-generation Cisco Catalyst 2960-S Series platform introduces modified QoS architecture. To
reduce the latency and improve application performance, the new Cisco 2960-S platform does not
support ingress queueing and buffer function in harware. All other ingress and egress queuing, buffer
and bandwidth sharing function remain consistent as the Cisco Catalyst 2960 platform. Each physical
ports, including StackPort, have 2 MB buffer capacity to prevent traffic drop during congestion. This
buffer allocation is static and cannot be modified by the user. However, when the Cisco Catalyst 2960-S
is deployed in FlexStack configuration mode, there is a flexibility to assign different buffer size on egress
queue of StackPort. Figure 2-27 illustrates QoS architecture on Catalyst 2960-S Series platform.
Figure 2-27
QoS Implementation in Catalyst 2960-S Switches
Policer
Marker
Policer
Marker
Egress
Queues
Q1
Receive
Q2
Classify
SRR
Transmit
Q3
Policer
Marker
Q4
Marker
Ingress QoS
Egress QoS
229372
Policer
QoS in Cisco Modular Switches
Cisco Catalyst 4507R-E and 6500-E are high density, resilient switches for large scale networks. The
Small Enterprise Design Profile uses the Cisco Catalyst 4507R-E in the main and larger remote site
designs, hence all the QoS recommendations in this section will be based on 4500-E architecture. Cisco
Catalyst 4500-E Series platform are widely deployed with classic and next-generation supervisors.
The classification function in the classic supervisor module is based on incoming DSCP or CoS setting
in the pack, which was assigned by the access-layer switch. Catalyst 4500-E with classic supervisor
performs ingress and egress QoS function based on internal mapping table that performs DSCP, ToS, or
CoS interworking. Classic supervisor relies on trust model configuration; redirection of ingress traffic
to an appropriate queue is based on the trust model defined on the edge port. See Figure 2-28.
Small Enterprise Design Profile Reference Guide
2-52
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
Figure 2-28
Catalyst 4500-E– Classic Supervisor QoS Architecture
Egress
Queue
Policer
Marker
DBL
Ingress QoS
Queuing/ Transmit
Shaping
Egress QoS
227558
Classify
Q1
Q2
Q3
Q4
The Cisco Catalyst 4500-E with next generation Sup-6E (see Figure 2-29) is designed to offer better
differentiated and preferential QoS services for various class-of-service traffic. New QoS capabilities in
the Sup-6E enable administrators to take advantage of hardware-based intelligent classification and take
action to optimize application performance and network availability. The QoS implementation in Sup-6E
supports Modular QoS CLI (MQC) as implemented in IOS-based routers that overall enhances QoS
capabilities and eases implementation and operations. Following are some of the key QoS features which
differentiate the Sup-6E versus classic supervisors:
•
Trust and Table-Map—MQC based QoS implementation offers a number of implementation and
operational benefits over classic supervisors that rely on Trust model and internal Table-map as a
tool to classify and mark ingress traffic.
•
Internal DSCP—The queue placement in Sup-6E is simplified by leveraging the MQC capabilities
to explicitly map DSCP or CoS traffic in hard-coded egress Queue structure,. For example, DSCP
46 can be classified with ACL and can be matched in PQ class-map of an MQC in Sup-6E.
•
Sequential vs Parallel Classification—With MQC-based QoS classification, the Sup6-E provides
sequential classification rather than parallel. Sequential classification method allows the network
administrator to classify traffic at egress based on the ingress markings.
Figure 2-29
Catalyst 4500-E—Supervisor 6-E QoS Architecture
Ingress QoS
Policer
Receive
Marking
Forwarding
Lookup
Classify
Unconditional
Marking
Egress QoS
Policer
Marking
Classify
DBL
Unconditional
Marking
Q1
Q2
Q3
Q4
Q5
Q6
Q7
Q8
Queuing/ Transmit
Shaping
227559
Receive
Netflow
Features
SupV-10G
Small Enterprise Design Profile Reference Guide
2-53
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
QoS Framework
QoS needs to be designed and implemented considering the entire network. This includes defining trust
points, and determining which policies to enforce at each device within the network. Developing the trust
model, guides policy implementations for each device.
Figure 2-30 depicts QoS trust model that guides QoS policy implementation in the main and remote site
networks.
Figure 2-30
Small Enterprise Network QoS Framework
Classification,
Marking and
Queueing
Trust (classic sup)
Trust (classic sup)
Classification,
Marking and
Queueing
Queueing and WTD
Trust
Trust, Classification,
Marking, Policing
and Queueing
Queueing and WTD
IP
Trust Boundary
Ingress QoS Ploicy
Egress QoS Ploicy
227560
Trusted, Conditional-Trusted or Un-Trusted Endpoints
The devices (routers, switches) within the internal network are managed by the system administrator,
and hence are classified as trusted devices. Access-layer switches communicate with devices that are
beyond the network boundary and within the internal network domain. QoS trust boundary at the
access-layer communicates with various devices that could be deployed in different trust models
(Trusted, Conditional-Trusted, or Un-Trusted). This section discusses the QoS policies for the traffic that
traverses access-switch QoS trust boundary. The QoS function is unidirectional; it provides flexibility to
set different QoS polices for traffic entering the network versus traffic that is exiting the network. See
Figure 2-31.
Small Enterprise Design Profile Reference Guide
2-54
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
Figure 2-31
Small Enterprise Network Edge QoS Boundary
Distribution/Core
Ingress QoS Ploicy
Egress QoS Ploicy
227561
Access
QoS Trust Boundary
The access-switch provides the entry point to the network for end devices. The access-switch must
decide whether to accept the QoS markings from each endpoint, or whether to change them. This is
determined by the QoS policies, and the trust model with which the endpoint is deployed.
End devices are classified into one of three different trust models; each with its own unique security and
QoS policies to access the network:
•
Untrusted—An unmanaged device that does not pass through the network security policies. For
example, employee-owned PC or network printer. Packets with 802.1p or DSCP marking set by
untrusted endpoints are reset to default by the access-layer switch at the edge. Otherwise, it is
possible for an unsecured user to take away network bandwidth that may impact network availability
and security for other users.
•
Trusted—Devices that passes through network access security policies and are managed by network
administrator. For example, secure PC or IP endpoints (i.e., servers, cameras, DMP, wireless access
points, VoIP/video conferencing gateways, etc). Even when these devices are network administrator
maintained and secured, QoS policies must still be enforced to classify traffic and assign it to the
appropriate queue to provide bandwidth assurance and proper treatment during network congestion.
•
Conditionally-Trusted—A single physical connection with one trusted endpoint and a indirect
untrusted endpoint must be deployed as conditionally-trusted model. The trusted endpoints are still
managed by the network administrator, but it is possible that the untrusted user behind the endpoint
may or may not be secure. For example, Cisco Unified IP Phone + PC. These deployment scenarios
require hybrid QoS policy that intelligently distinguishes and applies different QoS policy to the
trusted and untrusted endpoints that are connected to the same port.
Deploying QoS
The ingress QoS policy at the access-switches needs to be established, since this is the trust boundary,
where traffic enters the network. The following ingress QoS techniques are applied to provide
appropriate service treatment and prevent network congestion:
•
Trust—After classifying the endpoint the trust settings must be explicitly set by a network
administrator. By default, Catalyst switches set each port in untrusted mode when QoS is enabled.
Small Enterprise Design Profile Reference Guide
2-55
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
•
Classification—IETF standard has defined a set of application classes and provides recommended
DSCP settings. This classification determines the priority the traffic will receive in the network.
Using the IETF standard, simplifies the classification process and improves application and network
performance.
•
Policing—To prevent network congestion, the access-layer switch limits the amount of inbound
traffic up to its maximum setting. Additional policing can be applied for known applications, to
ensure the bandwidth of an egress queue is not completely consumed by one application.
•
Marking—Based on trust model, classification, and policer settings the QoS marking is set at the
edge before approved traffic enters through the access-layer switching fabric. Marking traffic with
the appropriate DSCP value is important to ensure traffic is mapped to the appropriate internal
queue, and treated with the appropriate priority.
•
Queueing—To provide differentiated services internally in the Catalyst switching fabric, all
approved traffic is queued into priority or non-priority ingress queue. Ingress queueing architecture
assures real-time applications, like VoIP traffic, are given appropriate priority (eg transmitted before
data traffic).
Implementing QoS Trust Mode
By default, QoS is disabled on all Catalyst switches and must be explicitly enabled in global
configuration mode. The QoS configuration is the same for a multilayer or routed-access deployment.
The following sample QoS configuration must be enabled on all the access-layer switches deployed in
main and remote sites.
cr24-2960-1(config)#mls qos
cr24-2960-1#show mls qos
QoS is enabled
QoS ip packet dscp rewrite is enabled
Upon enabling QoS in the Catalyst switches, all physical ports are assigned untrusted mode. The network
administrator must explicitly enable the trust settings on the physical port where trusted or conditionally
trusted endpoints are connected. The Catalyst switches can trust the ingress packets based on 802.1P
(CoS-based), ToS (ip-prec-based) or DSCP (DSCP-based) values. Best practice is to deploy DSCP-based
trust mode on all the trusted and conditionally-trusted endpoints. This offers a higher level of
classification and marking granularity than other methods. The following sample DSCP-based trust
configuration must be enabled on the access-switch ports connecting to trusted or conditionally-trusted
endpoints.
Access (Multilayer or Routed-Access)
Trusted Port
cr24-2960-1(config)#interface FastEthernet0/5
cr24-2960-1(config-if)# description CONNECTED TO IPVS 2500 - CAMERA
cr24-2960-1(config-if)# mls qos trust dscp
cr24-2960-1#show mls qos interface f0/5
FastEthernet0/5
trust state: trust dscp
trust mode: trust dscp
trust enabled flag: ena
COS override: dis
default COS: 0
DSCP Mutation Map: Default DSCP Mutation Map
Trust device: none
qos mode: port-based
Small Enterprise Design Profile Reference Guide
2-56
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
Conditionally-Trusted Port
cr24-2960-1(config)#interface FastEthernet0/3
cr24-2960-1(config-if)# description CONNECTED TO PHONE
cr24-2960-1(config-if)# mls qos trust device cisco-phone
cr24-2960-1(config-if)# mls qos trust dscp
cr24-2960-1#show mls qos interface f0/3
FastEthernet0/3
trust state: trust dscp
trust mode: trust dscp
trust enabled flag: ena
COS override: dis
default COS: 0
DSCP Mutation Map: Default DSCP Mutation Map
Trust device: cisco-phone
qos mode: port-based
UnTrusted Port
As described earlier, the default trust mode is untrusted when globally enabling QoS function. Without
explicit trust configuration on Fas0/1 port, the following show command verifies current trust state and
mode:
cr24-2960-1#show mls qos interface f0/1
FastEthernet0/1
trust state: not trusted
trust mode: not trusted
trust enabled flag: ena
COS override: dis
default COS: 0
DSCP Mutation Map: Default DSCP Mutation Map
Trust device: none
qos mode: port-based
Implementing QoS Classification
When creating QoS classification policies, the network administrator needs to consider what
applications are present at the access edge (in the ingress direction) and whether these applications are
sourced from trusted or untrusted endpoints. If PC endpoints are secured and centrally administered,
then endpoint PCs may be considered trusted endpoints. In most deployments, this is not the case, thus
PCs are considered untrusted endpoints for the remainder of this document.
Not every application class, as defined in the Cisco-modified RFC 4594-based model, is present in the
ingress direction at the access edge; therefore, it is not necessary to provision the following application
classes at the access-layer:
•
Network Control—It is assumed that access-layer switch will not transmit or receive network control
traffic from endpoints; hence this class is not implemented.
•
Broadcast Video—Broadcast video and multimedia streaming server are centrally deployed at the
main and multicast traffic is originated from trusted serverfarm servers and is unidirectional to
remote site endpoints (and should not be sourced from remote site endpoints).
•
Operation, Administration and Management—Primarily generated by network devices (routers,
switches) and collected by management stations which are typically deployed in the trusted
serverfarm network, or a network control center.
Small Enterprise Design Profile Reference Guide
2-57
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
All applications present at the access edge need to be assigned a classification, as shown in Figure 34.
Voice traffic is primarily sourced from Cisco IP telephony devices residing in the voice VLAN
(VVLAN). These are trusted devices, or conditionally trusted, if users also attach PC's, etc to the same
port. Voice communication may also be sourced from PC's with soft-phone applications, like Cisco
Unified Personal Communicator (CUPC). Since such applications share the same UDP port range as
multimedia conferencing traffic (UDP/RTP ports 16384-32767) this soft-phone VoIP traffic is
indistinguishable, and should be classified with multimedia conferencing streams. See Figure 2-32.
QoS Classes
Application
PHB
Application Examples
Network Control
CS6
EIGRP, OSPF, HSRP, IKE
VoIP
EF
Cisco IP Phone
Present at Campus
Access-Edge (Ingress)?
Trust
Boundary
Yes
Trusted
Cisco IPVS, Enterprise TV
Broadcast Video
Realtime Interactive
CS4
Cisco TelePresence
Yes
Trusted
Multimedia Conferencing
AF4
Cisco CUPC, WebEx
Yes
Untrusted
Multimedia Streaming
AF3
Cisco DMS, IP/TV
Signaling
CS3
SCCP, SIP, H.323
Yes
Trusted
Transactional Data
AF2
ERP Apps, CRM Apps
Yes
Untrusted
OAM
CS2
SNMP, SSH, Syslog
Bulk Data
AF1
Email, FTP, Backup
Yes
Untrusted
Best Effort
DF
Default Class
Yes
Untrusted
Scavenger
CS1
YouTube, Gaming, P2P
Yes
Untrusted
227562
Figure 2-32
MQC offers scalability and flexibility in configuring QoS to classify all 8 application classes by using
match statements or an extended access-list to match the exact value or range of Layer-4 known ports
that each application uses to communicate on the network. The following sample configuration creates
an extended access-list for each application and then applies it under class-map configuration mode.
cr24-3560r-1(config)#ip access-list extended MULTIMEDIA-CONFERENCING
cr24-3560r-1(config-ext-nacl)# remark RTP
cr24-3560r-1(config-ext-nacl)# permit udp any any range 16384 32767
cr24-3560r-1(config-ext-nacl)#!
cr24-3560r-1(config-ext-nacl)#ip access-list extended SIGNALING
cr24-3560r-1(config-ext-nacl)# remark SCCP
cr24-3560r-1(config-ext-nacl)# permit tcp any any range 2000 2002
cr24-3560r-1(config-ext-nacl)# remark SIP
cr24-3560r-1(config-ext-nacl)# permit tcp any any range 5060 5061
cr24-3560r-1(config-ext-nacl)# permit udp any any range 5060 5061
cr24-3560r-1(config-ext-nacl)#
cr24-3560r-1(config-ext-nacl)#ip access-list extended TRANSACTIONAL-DATA
cr24-3560r-1(config-ext-nacl)# remark HTTPS
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 443
cr24-3560r-1(config-ext-nacl)# remark ORACLE-SQL*NET
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 1521
Small Enterprise Design Profile Reference Guide
2-58
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
cr24-3560r-1(config-ext-nacl)# permit udp any any eq 1521
cr24-3560r-1(config-ext-nacl)# remark ORACLE
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 1526
cr24-3560r-1(config-ext-nacl)# permit udp any any eq 1526
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 1575
cr24-3560r-1(config-ext-nacl)# permit udp any any eq 1575
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 1630
cr24-3560r-1(config-ext-nacl)#
cr24-3560r-1(config-ext-nacl)#ip access-list extended BULK-DATA
cr24-3560r-1(config-ext-nacl)# remark FTP
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq ftp
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq ftp-data
cr24-3560r-1(config-ext-nacl)# remark SSH/SFTP
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 22
cr24-3560r-1(config-ext-nacl)# remark SMTP/SECURE SMTP
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq smtp
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 465
cr24-3560r-1(config-ext-nacl)# remark IMAP/SECURE IMAP
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 143
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 993
cr24-3560r-1(config-ext-nacl)# remark POP3/SECURE POP3
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq pop3
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 995
cr24-3560r-1(config-ext-nacl)# remark CONNECTED PC BACKUP
cr24-3560r-1(config-ext-nacl)# permit tcp any eq 1914 any
cr24-3560r-1(config-ext-nacl)#
cr24-3560r-1(config-ext-nacl)#ip access-list extended DEFAULT
cr24-3560r-1(config-ext-nacl)# remark EXPLICIT CLASS-DEFAULT
cr24-3560r-1(config-ext-nacl)# permit ip any any
cr24-3560r-1(config-ext-nacl)#
cr24-3560r-1(config-ext-nacl)#ip access-list extended SCAVENGER
cr24-3560r-1(config-ext-nacl)# remark KAZAA
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 1214
cr24-3560r-1(config-ext-nacl)# permit udp any any eq 1214
cr24-3560r-1(config-ext-nacl)# remark MICROSOFT DIRECT X GAMING
cr24-3560r-1(config-ext-nacl)# permit tcp any any range 2300 2400
cr24-3560r-1(config-ext-nacl)# permit udp any any range 2300 2400
cr24-3560r-1(config-ext-nacl)# remark APPLE ITUNES MUSIC SHARING
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 3689
cr24-3560r-1(config-ext-nacl)# permit udp any any eq 3689
cr24-3560r-1(config-ext-nacl)# remark BITTORRENT
cr24-3560r-1(config-ext-nacl)# permit tcp any any range 6881 6999
cr24-3560r-1(config-ext-nacl)# remark YAHOO GAMES
cr24-3560r-1(config-ext-nacl)# permit tcp any any eq 11999
cr24-3560r-1(config-ext-nacl)# remark MSN GAMING ZONE
cr24-3560r-1(config-ext-nacl)# permit tcp any any range 28800 29100
cr24-3560r-1(config-ext-nacl)#
Creating class-map for each application services and applying match statement:
cr24-3560r-1(config)#class-map match-all VVLAN-SIGNALING
cr24-3560r-1(config-cmap)# match ip dscp cs3
cr24-3560r-1(config-cmap)#
cr24-3560r-1(config-cmap)#class-map match-all VVLAN-VOIP
cr24-3560r-1(config-cmap)# match ip dscp ef
cr24-3560r-1(config-cmap)#
cr24-3560r-1(config-cmap)#class-map match-all MULTIMEDIA-CONFERENCING
cr24-3560r-1(config-cmap)# match access-group name MULTIMEDIA-CONFERENCING
cr24-3560r-1(config-cmap)#
cr24-3560r-1(config-cmap)#class-map match-all SIGNALING
cr24-3560r-1(config-cmap)# match access-group name SIGNALING
cr24-3560r-1(config-cmap)#
cr24-3560r-1(config-cmap)#class-map match-all TRANSACTIONAL-DATA
Small Enterprise Design Profile Reference Guide
2-59
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
cr24-3560r-1(config-cmap)# match access-group name TRANSACTIONAL-DATA
cr24-3560r-1(config-cmap)#
cr24-3560r-1(config-cmap)#class-map match-all
cr24-3560r-1(config-cmap)# match access-group
cr24-3560r-1(config-cmap)#
cr24-3560r-1(config-cmap)#class-map match-all
cr24-3560r-1(config-cmap)# match access-group
cr24-3560r-1(config-cmap)#
cr24-3560r-1(config-cmap)#class-map match-all
cr24-3560r-1(config-cmap)# match access-group
BULK-DATA
name BULK-DATA
DEFAULT
name DEFAULT
SCAVENGER
name SCAVENGER
Implementing Ingress Policer
It is important to limit how much bandwidth each class may use at the ingress to the access-layer for two
primary reasons:
•
Bandwidth Bottleneck—To prevent network congestion, each physical port at trust boundary must
be rate-limited. The rate-limit value may differ based on several factors—end-to-end network
bandwidth capacity, end-station, and application performance capacities, etc.
•
Bandwidth Security—Well-known applications like Cisco IP telephony, use a fixed amount of
bandwidth per device, based on codec. It is important to police high-priority application traffic
which is assigned to the high-priority queue, otherwise it could consume too much overall network
bandwidth and impact other application performance.
In addition to policing, the rate-limit function also provides the ability to take different actions on the
excess incoming traffic which exceeds the established limits. The exceed-action for each class must be
carefully designed based on the nature of application to provide best effort service based on network
bandwidth availability. Table 2-9 provides best practice policing guidelines for different classes to be
implemented for trusted and conditional-trusted endpoints at the network edge.
Table 2-9
Best Practice Policing Guidelines
Application
Policing Rate
Conform-Action
Exceed-Action
VoIP Signaling
<32 kbps
Pass
Drop
VoIP Bearer
<128 kbps
Pass
Drop
Pass
Drop
1
Multimedia Conferencing
<5Mbps
Signaling
<32 kbps
Pass
Drop
<10 Mbps
1
Pass
Remark to CS1
<10 Mbps
1
Pass
Remark to CS1
Best Effort
<10 Mbps
1
Pass
Remark to CS1
Scavenger
<10 Mbps 1
Pass
Drop
Transactional Data
Bulk Data
1.
Rate varies based on several factors as defined earlier. This table depicts sample rate-limiting values.
As described in the “QoS in Catalyst Fixed Configuration Switches” section on page 2-51, the policer
capabilities differ in Cisco Catalyst switching platforms. When deploying policer policies on the
access-layer switches the following platform limitations must be taken into consideration:
•
The Catalyst 2960 can only police to a minimum rate of 1 Mbps; all other platforms, including
next-generation Cisco Catalyst 2960-S Series, within this switch-product family can police to a
minimum rate of 8 kbps.
Small Enterprise Design Profile Reference Guide
2-60
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
•
Only the Cisco Catalyst 3560-X and 3750-X support policing on 10 Gigabit Ethernet interfaces.
The following sample configuration shows how to deploy policing for multiple classes on trusted and
conditionally-trusted ingress ports in access-layer switches.
Trusted or Conditionally-Trusted Port
cr24-3560r-1(config)#policy-map Phone+PC-Policy
cr24-3560r-1(config-pmap)# class VVLAN-VOIP
cr24-3560r-1(config-pmap-c)# police 128000 8000 exceed-action drop
cr24-3560r-1(config-pmap-c)# class VVLAN-SIGNALING
cr24-3560r-1(config-pmap-c)# police 32000 8000 exceed-action drop
cr24-3560r-1(config-pmap-c)# class MULTIMEDIA-CONFERENCING
cr24-3560r-1(config-pmap-c)# police 5000000 8000 exceed-action drop
cr24-3560r-1(config-pmap-c)# class SIGNALING
cr24-3560r-1(config-pmap-c)# police 32000 8000 exceed-action drop
cr24-3560r-1(config-pmap-c)# class TRANSACTIONAL-DATA
cr24-3560r-1(config-pmap-c)# police 10000000 8000 exceed-action policed-dscp-transmit
cr24-3560r-1(config-pmap-c)# class BULK-DATA
cr24-3560r-1(config-pmap-c)# police 10000000 8000 exceed-action policed-dscp-transmit
cr24-3560r-1(config-pmap-c)# class SCAVENGER
cr24-3560r-1(config-pmap-c)# police 10000000 8000 exceed-action drop
cr24-3560r-1(config-pmap-c)# class DEFAULT
cr24-3560r-1(config-pmap-c)# police 10000000 8000 exceed-action policed-dscp-transmit
All ingress traffic (default class) from untrusted endpoint be must be policed without explicit
classification that requires differentiated services. The following sample configuration shows how to
deploy policing on untrusted ingress ports in access-layer switches:
UnTrusted Port
cr24-3560r-1(config)#policy-map UnTrusted-PC-Policy
cr24-3560r-1(config-pmap)# class class-default
cr24-3560r-1(config-pmap-c)# police 10000000 8000 exceed-action drop
Implementing Ingress Marking
Accurate DSCP marking of ingress traffic at the access-layer switch is critical to ensure proper QoS
service treatment as traffic traverses through the network. All classified and policed traffic must be
explicitly marked using the policy-map configuration based on an 8-class QoS model as shown in
Figure 2-32.
Best practice is to use a explicit marking command (set dscp) even for trusted application classes (like
VVLAN-VOIP and VVLAN-SIGNALING), rather than a trust policy-map action. A trust statement in a
policy map requires multiple hardware entries, while the use of an explicit (seemingly redundant)
marking command, improves the hardware efficiency.
The following sample configuration shows how to implement explicit marking for multiple classes on
trusted and conditionally-trusted ingress ports in access-layer switches:
Trusted or Conditionally-Trusted Port
cr24-3560r-1(config)#policy-map Phone+PC-Policy
cr24-3560r-1(config-pmap)# class VVLAN-VOIP
cr24-3560r-1(config-pmap-c)# set dscp ef
cr24-3560r-1(config-pmap-c)# class VVLAN-SIGNALING
cr24-3560r-1(config-pmap-c)# set dscp cs3
cr24-3560r-1(config-pmap-c)# class MULTIMEDIA-CONFERENCING
cr24-3560r-1(config-pmap-c)# set dscp af41
Small Enterprise Design Profile Reference Guide
2-61
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
cr24-3560r-1(config-pmap-c)#
cr24-3560r-1(config-pmap-c)#
cr24-3560r-1(config-pmap-c)#
cr24-3560r-1(config-pmap-c)#
cr24-3560r-1(config-pmap-c)#
cr24-3560r-1(config-pmap-c)#
cr24-3560r-1(config-pmap-c)#
cr24-3560r-1(config-pmap-c)#
cr24-3560r-1(config-pmap-c)#
cr24-3560r-1(config-pmap-c)#
class SIGNALING
set dscp cs3
class TRANSACTIONAL-DATA
set dscp af21
class BULK-DATA
set dscp af11
class SCAVENGER
set dscp cs1
class DEFAULT
set dscp default
All ingress traffic (default class) from an untrusted endpoint must be marked without a explicit
classification. The following sample configuration shows how to implement explicit DSCP marking:
Untrusted Port
cr24-3560r-1(config)#policy-map UnTrusted-PC-Policy
cr24-3560r-1(config-pmap)# class class-default
cr24-3560r-1(config-pmap-c)# set dscp default
Applying Ingress Policies
After creating a complete policy-map with all the QoS policies defined, the service-policy must be
applied on the edge interface of the access-layer to enforce the QoS configuration. Cisco Catalyst
switches offer three simplified methods to apply service-policies. Depending on the deployment model,
any of these methods may be used:
Figure 2-33
•
Port-based QoS—Applying service-policy on a per physical port basis will force traffic to
pass-through the QoS policies before entering the network. Port-based QoS functions on a
per-physical port basis even if the port is associated with a logical VLAN.
•
VLAN-based QoS—Applying service-policy on per VLAN basis requires the policy-map to be
attached to a logical Layer-3 SVI interface. Every physical port associated with the VLAN will
require an extra configuration to enforce the QoS policies defined on a logical interface.
•
Per-Port/Per-VLAN-based QoS—Not supported on all the Catalyst platforms and the configuration
commands are platform-specific. Per-port/per-VLAN-based QoS creates a nested hierarchical
policy-map that operates on a trunk interface. A different policy-map can be applied on each logical
SVI interface that is associated to a single physical port.
Depicts All Three QoS Implementation Method
Port-Based QoS
VLAN-Based QoS
VLAN Interface
VLAN Interface
VLAN 10
VLAN 20
VLAN 10
Per-Port/ Per-VLAN Based QoS
VLAN 20
VLAN Interface
VVLAN 10
VVLAN 20
DVLAN 100
DVLAN 200
Physical port attached
with single service-policy
Single Logical port attached
with single service-policy
Multiple Logical ports attached
with different service-policy
227563
Physical Ports
The following sample configuration shows how to deploy port-based QoS on the access-layer switches:
cr24-3560r-1(config)#interface fastethernet0/4
cr24-3560r-1(config-if)# description CONNECTED TO PHONE+PC
Small Enterprise Design Profile Reference Guide
2-62
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
cr24-3560r-1(config-if)# service-policy input Phone+PC-Policy
cr24-3560r-1#show policy-map interface f0/4 | inc Service|Class
Service-policy input: Phone+PC-Policy
Class-map: VVLAN-VOIP (match-all)
Class-map: VVLAN-SIGNALING (match-all)
Class-map: MULTIMEDIA-CONFERENCING (match-all)
Class-map: SIGNALING (match-all)
Class-map: TRANSACTIONAL-DATA (match-all)
Class-map: BULK-DATA (match-all)
Class-map: SCAVENGER (match-all)
Class-map: DEFAULT (match-all)
Class-map: class-default (match-any)
Applying Ingress Queueing
Fixed configuration Cisco Catalyst switches (2960 and 3xxx) not only offer differentiated services on
the network ports but also internally on the switching fabric. Note, Cisco Catalyst 2960-S Series
platform do not support ingress queueing and buffer allocation. After enabling QoS and attaching
inbound policies on the physical ports, all the packets that meet the specified policy are forwarded to the
switching fabric for egress switching. The aggregate bandwidth from all edge ports may exceed
switching fabric bandwidth and cause internal congestion.
Cisco Catalyst 2960 and 3xxx platforms support two internal ingress queues: normal queue and priority
queue. The ingress queue inspects the DSCP value on each incoming frame and assigns it to either the
normal or priority queue. High priority traffic, like DSCP EF marked packets, are placed in the priority
queue and switched before processing the normal queue.
The Catalyst 3750-X family of switches supports the weighted tail drop (WTD) congestion avoidance
mechanism. WTD is implemented on queues to manage the queue length. WTD drops packets from the
queue, based on dscp value, and the associated threshold. If the threshold is exceeded for a given internal
DSCP value, the switch drops the packet. Each queue has three threshold values. The internal DSCP
determines which of the three threshold values is applied to the frame. Two of the three thresholds are
configurable (explicit) and one is not (implicit). This last threshold corresponds to the tail of the queue
(100 percent limit).
Figure 2-34 depicts how different class-of-service applications are mapped to the Ingress Queue
structure (1P1Q3T) and how each queue is assigned a different WTD threshold.
Small Enterprise Design Profile Reference Guide
2-63
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
Catalyst 2960 and 3xxx Ingress Queueing Model
Ingress Queue
1P1Q3T
Application
PHB
Network Control
CS7
EF
Internetwork Control
CS6
CS5
VoIP
EF
CS4
Broadcast Video
CS5
CS7
Realtime Interactive
CS4
CS6
Multimedia Conferencing
AF4
CS3
Multimedia Streaming
AF3
AF4
Signaling
CS3
AF3
Transactional Data
AF2
AF2
OAM
CS2
CS2
Bulk Data
AF1
AF1
Best Effort
DF
DF
Scavenger
CS1
CS1
Queue 2
Priority-Queue
Q1T3
Q1T2
Q1T1
Queue 1
Normal Queue
227564
Figure 2-34
The DSCP marked packets in the policy-map must be assigned to the appropriate queue and each queue
must be configured with the recommended WTD threshold as defined in Figure 2-34. The following
ingress queue configuration must be enabled in global configuration mode on every access-layer switch.
cr25-3750-1(config)#mls qos srr-queue input priority-queue 2 bandwidth 30
! Q2 is enabled as a strict-priority ingress queue with 30% BW
cr25-3750-1(config)#mls qos srr-queue input bandwidth 70 30
! Q1 is assigned 70% BW via SRR shared weights
! Q1 SRR shared weight is ignored (as it has been configured as a PQ)
cr25-3750-1(config)#mls qos srr-queue input threshold 1 80 90
! Q1 thresholds are configured at 80% (Q1T1) and 90% (Q1T2)
! Q1T3 is implicitly set at 100% (the tail of the queue)
! Q2 thresholds are all set (by default) to 100% (the tail of Q2)
! This section configures ingress DSCP-to-Queue Mappings
cr25-3750-1(config)# mls qos srr-queue input dscp-map queue 1 threshold 1 0 8 10 12 14
! DSCP DF, CS1 and AF1 are mapped to ingress Q1T1
cr25-3750-1(config)# mls qos srr-queue input dscp-map queue 1 threshold 1 16 18 20 22
! DSCP CS2 and AF2 are mapped to ingress Q1T1
cr25-3750-1(config)# mls qos srr-queue input dscp-map queue 1 threshold 1 26 28 30 34 36 38
! DSCP AF3 and AF4 are mapped to ingress Q1T1
cr25-3750-1(config)#mls qos srr-queue input dscp-map queue 1 threshold 2 24
! DSCP CS3 is mapped to ingress Q1T2
cr25-3750-1(config)#mls qos srr-queue input dscp-map queue 1 threshold 3 48 56
! DSCP CS6 and CS7 are mapped to ingress Q1T3 (the tail of Q1)
cr25-3750-1(config)#mls qos srr-queue input dscp-map queue 2 threshold 3 32 40 46
! DSCP CS4, CS5 and EF are mapped to ingress Q2T3 (the tail of the PQ)
Small Enterprise Design Profile Reference Guide
2-64
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
cr25-3750-1#show mls qos input-queue
Queue:
12
----------------------------------buffers
:9010
bandwidth :7030
priority :030
threshold1:80100
threshold2:90100
cr25-3750-1#show mls qos maps dscp-input-q
Dscp-inputq-threshold map:
d1 :d2
0
1
2
3
8
9
4
5
6
7
---------------------------------------------------------------------------------------0 :
01-01 01-01 01-01 01-01 01-01 01-01 01-01 01-01 01-01 01-01
1 :
01-01 01-01 01-01 01-01 01-01 01-01 01-01 01-01 01-01 01-01
2 :
01-01 01-01 01-01 01-01 01-02 01-01 01-01 01-01 01-01 01-01
3 :
01-01 01-01 02-03 01-01 01-01 01-01 01-01 01-01 01-01 01-01
4 :
02-03 02-01 02-01 02-01 02-01 02-01 02-03 02-01 01-03 01-01
5 :
01-01 01-01 01-01 01-01 01-01 01-01 01-03 01-01 01-01 01-01
6 :
01-01 01-01 01-01 01-01
Deploying Egress QoS
The QoS implementation for egress traffic toward the network edge on access-layer switches is much
simpler than the ingress traffic QoS. The egress QoS implementation provides optimal queueing policies
for each class and sets the drop thresholds to prevent network congestion and application performance
impact. Cisco Catalyst switches support 4 hardware queues that are assigned the following policies:
•
Real-time queue (to support a RFC 3246 EF PHB service)
•
Guaranteed bandwidth queue (to support RFC 2597 AF PHB services)
•
Default queue (to support a RFC 2474 DF service)
•
Bandwidth constrained queue (to support a RFC 3662 scavenger service)
As a best practice each physical or logical link must diversify bandwidth assignment to map with
hardware queues:
•
Real-time queue should not exceed 33 percent of the link's bandwidth.
•
Default queue should be at least 25 percent of the link's bandwidth.
•
Bulk/scavenger queue should not exceed 5 percent of the link's bandwidth.
Figure 2-35 shows best practice egress queue bandwidth allocation for each class.
Small Enterprise Design Profile Reference Guide
2-65
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
Figure 2-35
Engress QoS
5%
33%
25%
Real-Time
Guaranteed
Best-Effort
37%
227565
Scavenger/Bulk
Given these minimum queuing requirements and bandwidth allocation recommendations, the following
application classes can be mapped to the respective queues:
•
Realtime Queue—Voice, broadcast video, and realtime interactive may be mapped to the realtime
queue (per RFC 4594).
•
Guaranteed Queue—Network/internetwork control, signaling, network management, multimedia
conferencing, multimedia streaming, and transactional data can be mapped to the guaranteed
bandwidth queue. Congestion avoidance mechanisms (i.e., selective dropping tools), such as
WRED, can be enabled on this class. If configurable drop thresholds are supported on the platform,
these may be enabled to provide intra-queue QoS to these application classes, in the respective order
they are listed (such that control plane protocols receive the highest level of QoS within a given
queue).
•
Scavenger/Bulk Queue—Bulk data and scavenger traffic can be mapped to the
bandwidth-constrained queue and congestion avoidance mechanisms can be enabled on this class.
If configurable drop thresholds are supported on the platform, these may be enabled to provide
inter-queue QoS to drop scavenger traffic ahead of bulk data.
•
Default Queue—Best effort traffic can be mapped to the default queue; congestion avoidance
mechanisms can be enabled on this class.
The egress queueing is designed to map traffic, based on DSCP value, to four egress queues. as shown
above. The egress QoS model for a platform that supports DSCP-to-queue mapping with a 1P3Q8T
queuing structure is depicted in Figure 2-36.
Small Enterprise Design Profile Reference Guide
2-66
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
Figure 2-36
Access-Layer 1P3Q3T Egress Queue Model
Egress Queue
1P3Q3T
Application
PHB
Network Control
CS7
Internetwork Control
CS6
VoIP
EF
Broadcast Video
CS5
CS7
Realtime Interactive
CS4
CS6
Queue
Multimedia Conferencing
AF4
CS3
2
Multimedia Streaming
AF3
CS2
Signaling
CS3
Transactional Data
AF2
OAM
CS2
Bulk Data
AF1
Best Effort
DF
Scavenger
CS1
CS1
AF1
DF
Queue 4
(5%)
Q4T2
Q4T1
Queue 3
(55%)
Q2T3
Q2T2
Q2T1
(30%)
AF4
AF3
AF2
EF
Queue 1
CS5 Priority-Queue
(30%)
CS4
227566
Chapter 2
DSCP marked packets are assigned to the appropriate queue and each queue is configured with
appropriate WTD threshold as defined in Figure 2-36. Egress queueing is the same on network edge port
as well as on uplink connected to internal network, and it is independent of trust mode. The following
egress queue configuration in global configuration mode, must be enabled on every access-layer switch
in the network.
cr25-3750-1(config)#mls qos queue-set output 1 buffers 15 30 35 20
! Queue buffers are allocated
cr25-3750-1(config)#mls qos queue-set output 1 threshold 1 100 100 100 100
! All Q1 (PQ) Thresholds are set to 100%
cr25-3750-1(config)#mls qos queue-set output 1 threshold 2 80 90 100 400
! Q2T1 is set to 80%; Q2T2 is set to 90%;
! Q2 Reserve Threshold is set to 100%;
! Q2 Maximum (Overflow) Threshold is set to 400%
cr25-3750-1(config)#mls qos queue-set output 1 threshold 3 100 100 100 400
! Q3T1 is set to 100%, as all packets are marked the same weight in Q3
! Q3 Reserve Threshold is set to 100%;
! Q3 Maximum (Overflow) Threshold is set to 400%
cr25-3750-1(config)#mls qos queue-set output 1 threshold 4 60 100 100 400
! Q4T1 is set to 60%; Q4T2 is set to 100%
! Q4 Reserve Threshold is set to 100%;
! Q4 Maximum (Overflow) Threshold is set to 400%
! This section configures egress DSCP-to-Queue mappings
cr25-3750-1(config)# mls qos srr-queue output dscp-map queue 1 threshold 3 32 40 46
! DSCP CS4, CS5 and EF are mapped to egress Q1T3 (tail of the PQ)
cr25-3750-1(config)# mls qos srr-queue output dscp-map queue 2 threshold 1 16 18 20 22
! DSCP CS2 and AF2 are mapped to egress Q2T1
Small Enterprise Design Profile Reference Guide
2-67
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
cr25-3750-1(config)# mls qos srr-queue output dscp-map queue 2 threshold 1 26 28 30 34 36 38
! DSCP AF3 and AF4 are mapped to egress Q2T1
cr25-3750-1(config)#mls qos srr-queue output dscp-map queue 2 threshold 2 24
! DSCP CS3 is mapped to egress Q2T2
cr25-3750-1(config)#mls qos srr-queue output dscp-map queue 2 threshold 3 48 56
! DSCP CS6 and CS7 are mapped to egress Q2T3
cr25-3750-1(config)#mls qos srr-queue output dscp-map queue 3 threshold 3 0
! DSCP DF is mapped to egress Q3T3 (tail of the best effort queue)
cr25-3750-1(config)#mls qos srr-queue output dscp-map queue 4 threshold 1 8
! DSCP CS1 is mapped to egress Q4T1
cr25-3750-1(config)# mls qos srr-queue output dscp-map queue 4 threshold 2 10 12 14
! DSCP AF1 is mapped to Q4T2 (tail of the less-than-best-effort queue)
! This section configures interface egress queuing parameters
cr25-3750-1(config)#interface range GigabitEthernet1/0/1-48
cr25-3750-1(config-if-range)# queue-set 1
! The interface(s) is assigned to queue-set 1
cr25-3750-1(config-if-range)# srr-queue bandwidth share 1 30 35 5
! The SRR sharing weights are set to allocate 30% BW to Q2
! 35% BW to Q3 and 5% BW to Q4
! Q1 SRR sharing weight is ignored, as it will be configured as a PQ
cr25-3750-1(config-if-range)# priority-queue out
! Q1 is enabled as a strict priority queue
cr25-3750-1#show mls qos interface GigabitEthernet1/0/27 queueing
GigabitEthernet1/0/27
Egress Priority Queue : enabled
Shaped queue weights (absolute) : 25 0 0 0
Shared queue weights : 1 30 35 5
The port bandwidth limit : 100 (Operational Bandwidth:100.0)
The port is mapped to qset : 1
Table 2-10 and Table 2-11 summarize the ingress and egress QoS policies at the access-layer for several
types of validated endpoints.
Table 2-10
Summarized Network Edge Ingress QoS Deployment Guidelines
Trust
Model
endpoint
DSCP Trust
Classification
Marking
Policing
Ingress
Queueing
Unmanaged devices,
printers etc
UnTrusted
Don’t Trust.
Default.
None
None
Yes
Yes
Managed secured
devices, Servers etc
Trusted
Trust
8 Class
Yes
Yes
Yes
Phone
Trusted
Trust
Yes
Yes
Yes
Yes
Phone + Mobile PC
Conditionally-Tru Trust
sted
Yes
Yes
Yes
Yes
IP Video surveillance
Camera
Trusted
Trust
No
No
No
Yes
Digital Media Player
Trusted
Trust
No
No
No
Yes
Core facing Uplinks
Trusted
Trust
No
No
No
Yes
Model
Small Enterprise Design Profile Reference Guide
2-68
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
Table 2-11
Summarized Network Edge Egress QoS Deployment Guidelines
Trust
Model
Classification / Marking /
Policing
Egress
Queueing
Bandwidth
Share
Unmanaged devices, printers
etc
UnTrusted
None
Yes
Yes
Managed secured devices,
Servers etc
Trusted
None
Yes
Yes
Phone
Trusted
None
Yes
Yes
Phone + Mobile PC
Conditionally-Truste None
d
Yes
Yes
IP Video surveillance Camera
Trusted
None
Yes
Yes
Digital Media Player
Trusted
None
Yes
Yes
Core facing Uplinks
Trusted
None
Yes
Yes
Endpoint
Deploying Network Core QoS
All connections between internal network devices that are deployed within the network domain
boundary are classified as trusted devices and follow the same QoS best practices recommended in the
previous section. Ingress and egress core QoS policies are simpler than those applied at the network
edge, See Figure 2-37.
Figure 2-37
Network Core QoS Boundary
Distribution/Core
Ingress QoS Ploicy
Egress QoS Ploicy
227567
Access
The core network devices are considered trusted and rely on the access-switch to properly mark DSCP
values. The core network is deployed to ensure consistent differentiated QoS service across the network.
This ensures there is no service quality degradation for high-priority traffic, such as IP telephony or
video.
Small Enterprise Design Profile Reference Guide
2-69
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
The QoS implementation at the main and remote large site differ from the remote small site, due to
different platforms used as the collapsed core router (Catalyst 4500-E vs Catalyst 3750-X StackWise
Plus).
Deploying Main or Remote Large Site Ingress QoS
The collapsed core at main site is deployed with Cisco Catalyst 450R-E with Supervisor-6E, whereas
the remote large site collapsed core is deployed with Cisco Catalyst 4507R-E with either Supervisor-6E,
Supervisor-6LE or Supervisor-V. The next-generation Sup-6E and Sup-6LE module has a redesigned
QoS implementation which matches Cisco IOS routers. No ingress QoS configuration is required, since
QoS is enabled by default, and all ports are considered trusted.
The Cisco Catalyst 4507R-E with Supervisor-V requires ingress QoS configuration similar to trusted
endpoints in the access-layer. Following is a sample configuration which enables QoS in the Catalyst
4507R-E with Supervisor-V:
cr35-4507-1(config)#qos
! Enables QoS function in the switch
cr35-4507-1#show qos
QoS is enabled globally
IP header DSCP rewrite is enabled
After QoS is globally enabled, all interfaces are in the untrusted mode by default. QoS trust settings must
be set on each Layer 2 or Layer 3 port that is physically connected to another device within the network
trust boundary. When Cisco Catalyst 4500 is deployed in EtherChannel mode, the QoS trust settings
must be applied to every physical member-link and logical port-channel interface. Best practice is to
enable trust DSCP settings on each physical and logical interface that connects to another internal trusted
device (e.g., access-layer switches in wiring closet or data-center, a router, wireless LAN controller
(WLC)).
cr35-4507-1(config)#interface range Po11 , Gi1/2 , Gi2/2
cr35-4507-1(config-if-range)#description Connected to cr35-2960-1
cr35-4507-1(config-if-range)#qos trust dscp
cr35-4507-1#show qos interface Port-channel 11
QoS is enabled globally
Port QoS is enabled
Administrative Port Trust State: 'dscp'
Operational Port Trust State: 'dscp'
Trust device: none
Default DSCP: 0 Default CoS: 0
Additional ingress QoS techniques (such as classification, marking, and policing) are not required at the
collapsed core layer since these functions are already performed by the access-layer switches. The
architecture of Catalyst 4500-E with classic or next-generation Supervisor do not need ingress queueing
since all of the forwarding decisions are made centrally on the supervisor. There are no additional QoS
configurations required at the collapsed core-layer system.
Small Enterprise Design Profile Reference Guide
2-70
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
Deploying Remote Small Site Ingress QoS
The remote small site is deployed using Cisco Catalyst 3750-X StackWise Plus as the collapsed core
switch. The QoS implementation remains the same whether deployed as 3750-X StackWise or as a
standalone switch. By default, QoS is disabled on the 3750-X switch. Following is a sample
configuration to enable QoS in global configuration mode:
cr36-3750s-1(config)#mls qos
! Enables QoS function in the switch
cr36-3750s-1#show mls qos
QoS is enabled
QoS ip packet dscp rewrite is enabled
After QoS is globally enabled, all interfaces are in the untrusted mode by default. QoS trust settings must
be set on each Layer 2 or Layer 3 port that is physically connected to another device within the network
trust boundary. When Cisco Catalyst 3750-E StackWise Plus is deployed in EtherChannel mode, the QoS
trust settings must be applied to every physical member-link. Best practice is to enable trust DSCP
settings on each physical and logical interface that connects to another internal trusted device (e.g.,
access-layer switches in wiring closet or data-center, a router, wireless LAN controller (WLC)).
cr36-3750s-1(config)#int range gi1/0/49 , gi3/0/49
cr36-3750s-1(config-if-range)# description Connected to cr36-2960-1
cr36-3750s-1(config-if-range)#mls qos trust dscp
cr36-3750s-1#show mls qos interface Gi1/0/49
GigabitEthernet1/0/49
trust state: trust dscp
trust mode: trust dscp
trust enabled flag: ena
COS override: dis
default COS: 0
DSCP Mutation Map: Default DSCP Mutation Map
Trust device: none
qos mode: port-based
Additional ingress QoS techniques (such as classification, marking, and policing) are not required at the
collapsed core layer since these functions are already performed by the access-layer switches. The
ingress queueing and DSCP-Ingress-Queue function in 3750-X StackWise Plus must be enabled to allow
differentiation between normal versus high-priority traffic. The ingress queuing configuration is
consistent with the implementation at the access-edge. Following is a sample configuration for the
ingress queues of the Catalyst 3750-X StackWise collapsed core switch:
cr36-3750-1(config)#mls qos srr-queue input priority-queue 2 bandwidth 30
! Q2 is enabled as a strict-priority ingress queue with 30% BW
cr36-3750-1(config)#mls qos srr-queue input bandwidth 70 30
! Q1 is assigned 70% BW via SRR shared weights
! Q1 SRR shared weight is ignored (as it has been configured as a PQ)
cr36-3750-1(config)#mls qos srr-queue input threshold 1 80 90
! Q1 thresholds are configured at 80% (Q1T1) and 90% (Q1T2)
! Q1T3 is implicitly set at 100% (the tail of the queue)
! Q2 thresholds are all set (by default) to 100% (the tail of Q2)
! This section configures ingress DSCP-to-Queue Mappings
cr36-3750-1(config)# mls qos srr-queue input dscp-map queue 1 threshold 1 0 8 10 12 14
! DSCP DF, CS1 and AF1 are mapped to ingress Q1T1
cr36-3750-1(config)# mls qos srr-queue input dscp-map queue 1 threshold 1 16 18 20 22
Small Enterprise Design Profile Reference Guide
2-71
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
! DSCP CS2 and AF2 are mapped to ingress Q1T1
cr36-3750-1(config)# mls qos srr-queue input dscp-map queue 1 threshold 1 26 28 30 34 36 38
! DSCP AF3 and AF4 are mapped to ingress Q1T1
cr36-3750-1(config)#mls qos srr-queue input dscp-map queue 1 threshold 2 24
! DSCP CS3 is mapped to ingress Q1T2
cr36-3750-1(config)# mls qos srr-queue input dscp-map queue 1 threshold 3 48 56
! DSCP CS6 and CS7 are mapped to ingress Q1T3 (the tail of Q1)
cr36-3750-1(config)# mls qos srr-queue input dscp-map queue 2 threshold 3 32 40 46
! DSCP CS4, CS5 and EF are mapped to ingress Q2T3 (the tail of the PQ)
cr36-3750s-1#show mls qos input-queue
Queue
:
1
2
------------------------------------buffers
:
90
10
bandwidth :
70
30
priority :
0
30
threshold1:
80
100
threshold2:
90
100
cr36-3750s-1#show mls qos maps dscp-input-q
Dscp-inputq-threshold map:
d1 :d2
0
1
2
3
8
9
4
5
6
7
---------------------------------------------------------------------------------------0
1
2
3
4
5
6
:
:
:
:
:
:
:
01-01
01-01
01-01
01-01
02-03
01-01
01-01
01-01
01-01
01-01
01-01
02-01
01-01
01-01
01-01
01-01
01-01
02-03
02-01
01-01
01-01
01-01
01-01
01-01
01-01
02-01
01-01
01-01
01-01
01-01
01-02
01-01
02-01
01-01
01-01
01-01
01-01
01-01
02-01
01-01
01-01
01-01
01-01
01-01
02-03
01-03
01-01
01-01
01-01
01-01
02-01
01-01
01-01
01-01
01-01
01-01
01-03
01-01
01-01
01-01
01-01
01-01
01-01
01-01
Deploying Main Site Egress QoS
The main site is deployed with Cisco Catalyst 4507R-E with Supervisor-6E as the collapsed core router.
Egress QoS from the collapsed core router provides optimized queueing and drop thresholds to drop
excess low-priority traffic and protect high-priority traffic.
The Supervisor-6E supports up to 8 traffic classes for QoS mapping. It also supports a platform-specific
congestion avoidance algorithm to provide Active Queue Management (AQM) with Dynamic Buffer
Limiting (DBL). DBL tracks the queue length for each traffic flow in the switch. When the queue length
of a flow exceeds its limit, DBL drops packets or sets the Explicit Congestion Notification (ECN) bit in
the TCP packet header. With 8 egress (1P7Q1T) queues and DBL capability in the Sup-6E, the
bandwidth distribution for each class changes, as shown in Figure 2-38.
Small Enterprise Design Profile Reference Guide
2-72
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
Figure 2-38
Small Enterprise Main Site Network
Egress Queue
1P7Q1T (+DBL)
Application
PHB
Network Control
CS7
EF
Internetwork Control
CS6
CS5
VoIP
EF
CS4
Broadcast Video
CS5
Realtime Interactive
CS4
Multimedia Conferencing
AF4
Multimedia Streaming
AF3
Signaling
CS3
Transactional Data
AF2
OAM
CS2
Bulk Data
AF1
Best Effort
Scavenger
Priority-Queue
(30%)
CS7 & CS6
Q7 (10%)
CS3 & CS2
AF4
Q6 (10%)
AF3
Q5 (10%)
AF2
Q4 (10%)
AF1
Q3 (4%)
DF
DF
Q2 (1%)
CS1
CS1
Q1 (25%)
227568
Chapter 2
Implementing QoS policies on Sup-6E-based Catalyst 4500 platform follows IOS (MQC)-model. The
egress QoS implementation bundles the queueing and policing functions on EtherChannel based
networks. To provide low-latency for high priority traffic, all lower priority traffic must wait until the
priority-queue is empty. Best practice includes implementing a policer along with the priority-queue to
provide more fair treatment for all traffic.
The following sample configuration shows how to create an 8-class egress queueing model and protect
from high-priority traffic consuming more bandwidth than global policies allow. The egress QoS
service-policy must be applied to all the physical EtherChannel member-links connected to different
service-blocks (i.e., WAN edge, serverfarm, access-layer switches, etc).
! Creating class-map for each classes using match dscp statement as marked by edge systems
cr24-4507-1(config)#class-map match-all PRIORITY-QUEUE
cr24-4507-1(config-cmap)# match dscp ef
cr24-4507-1(config-cmap)# match dscp cs5
cr24-4507-1(config-cmap)# match dscp cs4
cr24-4507-1(config-cmap)#class-map match-all CONTROL-MGMT-QUEUE
cr24-4507-1(config-cmap)# match dscp cs7
cr24-4507-1(config-cmap)# match dscp cs6
cr24-4507-1(config-cmap)# match dscp cs3
cr24-4507-1(config-cmap)# match dscp cs2
cr24-4507-1(config-cmap)#class-map match-all MULTIMEDIA-CONFERENCING-QUEUE
cr24-4507-1(config-cmap)# match dscp af41 af42 af43
cr24-4507-1(config-cmap)#class-map match-all MULTIMEDIA-STREAMING-QUEUE
cr24-4507-1(config-cmap)# match dscp af31 af32 af33
cr24-4507-1(config-cmap)#class-map match-all TRANSACTIONAL-DATA-QUEUE
cr24-4507-1(config-cmap)# match dscp af21 af22 af23
cr24-4507-1(config-cmap)#class-map match-all BULK-DATA-QUEUE
cr24-4507-1(config-cmap)# match dscp af11 af12 af13
cr24-4507-1(config-cmap)#class-map match-all SCAVENGER-QUEUE
Small Enterprise Design Profile Reference Guide
2-73
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
cr24-4507-1(config-cmap)# match
dscp cs1
! Creating policy-map and configure queueing for class-of-service
cr24-4507-1(config)#policy-map EGRESS-POLICY
cr24-4507-1(config-pmap)# class PRIORITY-QUEUE
cr24-4507-1(config-pmap-c)# priority
cr24-4507-1(config-pmap-c)# class CONTROL-MGMT-QUEUE
cr24-4507-1(config-pmap-c)# bandwidth remaining percent 10
cr24-4507-1(config-pmap-c)# class MULTIMEDIA-CONFERENCING-QUEUE
cr24-4507-1(config-pmap-c)# bandwidth remaining percent 10
cr24-4507-1(config-pmap-c)# class MULTIMEDIA-STREAMING-QUEUE
cr24-4507-1(config-pmap-c)# bandwidth remaining percent 10
cr24-4507-1(config-pmap-c)# class TRANSACTIONAL-DATA-QUEUE
cr24-4507-1(config-pmap-c)# bandwidth remaining percent 10
cr24-4507-1(config-pmap-c)# dbl
cr24-4507-1(config-pmap-c)# class BULK-DATA-QUEUE
cr24-4507-1(config-pmap-c)# bandwidth remaining percent 4
cr24-4507-1(config-pmap-c)# dbl
cr24-4507-1(config-pmap-c)# class SCAVENGER-QUEUE
cr24-4507-1(config-pmap-c)# bandwidth remaining percent 1
cr24-4507-1(config-pmap-c)# class class-default
cr24-4507-1(config-pmap-c)# bandwidth remaining percent 25
cr24-4507-1(config-pmap-c)# dbl
! Attaching egress service-policy on all physical member-link ports
cr24-4507-1(config)#int range Gi1/1 - 6 , Gi2/1 - 6
cr24-4507-1(config-if-range)# service-policy output EGRESS-POLICY
EtherChannel is an aggregated logical bundle interface that does not perform queueing and relies on
individual member-links to queue egress traffic. The policer to rate-limit priority class traffic must be
implemented on EtherChannel and not on individual member-links since it governs the aggregate egress
traffic limits. The following additional policy-map must be created to classify priority-queue class traffic
and rate-limit the traffic to 30 percent of egress link capacity:
cr24-4507-1(config)#class-map match-any PRIORITY-QUEUE
cr24-4507-1(config-cmap)# match dscp ef
cr24-4507-1(config-cmap)# match dscp cs5
cr24-4507-1(config-cmap)# match dscp cs4
cr24-4507-1(config)#policy-map PQ-POLICER
cr24-4507-1(config-pmap)# class PRIORITY-QUEUE
cr24-4507-1(config-pmap-c)# police cir 300 m conform-action transmit exceed-action drop
cr24-4507-1(config)#interface range Port-Channel 1 , Port-channel 11 - 17
cr24-4507-1(config-if-range)#service-policy output PQ-POLICER
Deploying Remote Large Site Egress QoS
The remote large site is deployed with Cisco Catalyst 450R-E and either Supervisor-6E, Supervisor-6LE
or Supervisor-V as the collapsed core router. If the remote large site network is deployed with Sup-6E
or Sup-6LE, then the configuration is the same as described in the previous section.
Small Enterprise Design Profile Reference Guide
2-74
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
The QoS deployment and implementation guidelines differ when the Cisco Catalyst 4500-E is deployed
with the classic Supervisor-V module. The SupV supervisor can have up to four egress queues like the
Cisco Catalyst 2960, 2960-S and 35xx/37xx-X Series switches. Before forwarding egress traffic, each
packet must be internally classified and placed in the appropriate egress-queue. Placing traffic into
different class-of-service queues, will offer traffic prioritization and guaranteed bandwidth to the
network. The following sample configuration shows how to implement egress QoS on the Catalyst
4500-E with Supervisor-V:
cr35-4507-1(config)#qos dbl
! DBL is globally enabled
cr35-4507-1(config)#no qos dbl dscp-based 32
cr35-4507-1(config)#no qos dbl dscp-based 40
cr35-4507-1(config)#no qos dbl dscp-based 46
! DBL is explicitly disabled on DSCP CS4, CS5 and EF
! as these DSCP values are assigned to the PQ
! and as such should never experience congestion avoidance drops
cr35-4507-1(config)#qos dbl exceed-action ecn
! DBL will mark IP ECN bits in the event of congestion
! This section configures the DBL policy-map
cr35-4507-1(config)#policy-map DBL
cr35-4507-1(config-pmap)# class class-default
cr35-4507-1(config-pmap-c)# dbl
! DBL is enabled on all flows
! (with the exception of DSCP CS4, CS5 and EF)
! This section configures the DSCP-to-Queue mappings
cr35-4507-1(config)#qos map dscp 8 10 12 14 to tx-queue 1
! DSCP CS1 and AF1 are mapped to Q1 (the less than best effort queue)
cr35-4507-1(config)#qos map dscp 0 to tx-queue 2
! DSCP DF is mapped to Q2 (the best effort/default queue)
cr35-4507-1(config)#qos map dscp 32 40 46 to tx-queue 3
! DSCP CS4, CS5 and EF are mapped to Q3 (the PQ)
cr35-4507-1(config)#qos map dscp 16 18 20 22 to tx-queue 4
! DSCP CS2 and AF2 are mapped to Q4 (guaranteed BW queue)
cr35-4507-1(config)#qos map dscp 24 26 28 30 to tx-queue 4
! DSCP CS3 and AF3 are mapped to Q4 (guaranteed BW queue)
cr35-4507-1(config)#qos map dscp 34 36 38 to tx-queue 4
! DSCP AF4 is mapped to Q4 (guaranteed BW queue)
cr35-4507-1(config)#qos map dscp 48 56 to tx-queue 4
! DSCP CS6 and CS7 are mapped to Q4 (guaranteed BW queue)
! This section configures all the EtherChannel member-link for egress queuing
cr35-4507-1(config)#interface range Gig1/1 - 6 , Gig2/1 - 6
cr35-4507-1(config-if-range)# tx-queue 1
cr35-4507-1(config-if-tx-queue)# bandwidth percent 5
! Q1 (less than best effort queue) is assigned 5% BW
cr35-4507-1(config-if-tx-queue)# tx-queue 2
cr35-4507-1(config-if-tx-queue)# bandwidth percent 35
! Q2 (default/best effort queue) is assigned 35% BW
cr35-4507-1(config-if-tx-queue)# tx-queue 3
cr35-4507-1(config-if-tx-queue)# priority high
cr35-4507-1(config-if-tx-queue)# bandwidth percent 30
! Q3 is enabled as a PQ and assigned 30% BW
cr35-4507-1(config-if-tx-queue)# tx-queue 4
cr35-4507-1(config-if-tx-queue)# bandwidth percent 30
! Q4 (guaranteed BW queue) is assigned 30% BW
cr35-4507-1(config-if-range)# service-policy output DBL
! DBL policy-map is attached to the interface(s)
cr35-4507-1#show qos dbl
QOS is enabled globally
Small Enterprise Design Profile Reference Guide
2-75
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
DBL is enabled globally on DSCP values:
0-31,33-39,41-45,47-63
DBL flow includes vlan
DBL flow includes layer4-ports
DBL uses ecn to indicate congestion
DBL exceed-action probability: 15%
DBL max credits: 15
DBL aggressive credit limit: 10
DBL aggressive buffer limit: 2 packets
cr35-4507-1#show qos maps dscp tx-queue
DSCP-TxQueue Mapping Table (dscp = d1d2)
d1 : d2 0
1
2
3
4
5
6
7
8
9
-----------------------------------------------0 :
02 01 01 01 01 01 01 01 01 01
1 :
01 01 01 01 01 01 04 02 04 02
2 :
04 02 04 02 04 02 04 02 04 02
3 :
04 02 03 03 04 03 04 03 04 03
4 :
03 03 03 03 03 03 03 03 04 04
5 :
04 04 04 04 04 04 04 04 04 04
6 :
04 04 04 04
cr35-4507-1#show qos interface Gig1/2
QoS is enabled globally
Port QoS is enabled
Administrative Port Trust State: 'dscp'
Operational Port Trust State: 'dscp'
Trust device: none
Default DSCP: 0 Default CoS: 0
Appliance trust: none
Tx-Queue
Bandwidth
ShapeRate
Priority
(bps)
(bps)
1
50000000
disabled
N/A
2
350000000
disabled
N/A2080
3
300000000disabled
high2080
4
300000000disabledN/A2080
QueueSize
(packets)
2080
Deploying Remote Small Site Egress QoS
Collapsed Core— Catalyst 3750-X StackWise Plus
The remote small site is deployed with Cisco Catalyst 3750-X StackWise Plus as the collapsed core
router. The Catalyst 3750-X can have up to four egress queues. Before forwarding egress traffic, each
packet is placed in the appropriate egress-queue as shown in Figure 2-36. The Catalyst 3750-E switch
supports Shaped Round Robin (SRR) packet schedule service which can be deployed in two different
modes:
•
Shaped—To provide guaranteed bandwidth, the shaped egress queue reserves some of the
bandwidth of the port for each queue. Traffic load exceeding the shape parameter gets dropped. The
queue cannot take advantage of excess bandwidth capacity when other queues are not using their
bandwidth allocations.
•
Shared—Shared mode also provides guaranteed bandwidth for each queue; however, it allows the
flexibility of using excess bandwidth when there is any available.
The following sample configuration shows how to implement egress QoS on the Catalyst 3750-X:
Small Enterprise Design Profile Reference Guide
2-76
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Deploying QoS in Network
cr36-3750s-1(config)#mls qos queue-set output 1 buffers 15 30 35 20
! Queue buffers are allocated
cr36-3750-1(config)#mls qos queue-set output 1 threshold 1 100 100 100 100
! All Q1 (PQ) Thresholds are set to 100%
cr36-3750s-1(config)#mls qos queue-set output 1 threshold 2 80 90 100 400
! Q2T1 is set to 80%; Q2T2 is set to 90%;
! Q2 Reserve Threshold is set to 100%;
! Q2 Maximum (Overflow) Threshold is set to 400%
cr36-3750s-1(config)#mls qos queue-set output 1 threshold 3 100 100 100 400
! Q3T1 is set to 100%, as all packets are marked the same weight in Q3
! Q3 Reserve Threshold is set to 100%;
! Q3 Maximum (Overflow) Threshold is set to 400%
cr36-3750s-1(config)#mls qos queue-set output 1 threshold 4 60 100 100 400
! Q4T1 is set to 60%; Q4T2 is set to 100%
! Q4 Reserve Threshold is set to 100%;
! Q4 Maximum (Overflow) Threshold is set to 400%
! This section configures egress DSCP-to-Queue mappings
cr36-3750s-1(config)# mls qos srr-queue output dscp-map queue 1 threshold 3 32 40 46
! DSCP CS4, CS5 and EF are mapped to egress Q1T3 (tail of the PQ)
cr36-3750s-1(config)# mls qos srr-queue output dscp-map queue 2 threshold 1 16 18 20 22
! DSCP CS2 and AF2 are mapped to egress Q2T1
cr36-3750s-1(config)#mls qos srr-queue output dscp-map queue 2 threshold 1 26 28 30 34 36 38
! DSCP AF3 and AF4 are mapped to egress Q2T1
cr36-3750s-1(config)#mls qos srr-queue output dscp-map queue 2 threshold 2 24
! DSCP CS3 is mapped to egress Q2T2
cr36-3750s-1(config)#mls qos srr-queue output dscp-map queue 2 threshold 3 48 56
! DSCP CS6 and CS7 are mapped to egress Q2T3
cr36-3750s-1(config)#mls qos srr-queue output dscp-map queue 3 threshold 3 0
! DSCP DF is mapped to egress Q3T3 (tail of the best effort queue)
cr36-3750s-1(config)#mls qos srr-queue output dscp-map queue 4 threshold 1 8
! DSCP CS1 is mapped to egress Q4T1
cr36-3750s-1(config)# mls qos srr-queue output dscp-map queue 4 threshold 2 10 12 14
! DSCP AF1 is mapped to Q4T2 (tail of the less-than-best-effort queue)
! This section configures interface egress queuing parameters
cr36-3750s-1(config)#interface range GigabitEthernet1/0/1-48
cr36-3750s-1(config-if-range)# queue-set 1
! The interface(s) is assigned to queue-set 1
cr36-3750s-1(config-if-range)# srr-queue bandwidth share 1 30 35 5
! The SRR sharing weights are set to allocate 30% BW to Q2
! 35% BW to Q3 and 5% BW to Q4
! Q1 SRR sharing weight is ignored, as it will be configured as a PQ
cr36-3750s-1(config-if-range)# priority-queue out
! Q1 is enabled as a strict priority queue
cr36-3750s-1#show mls qos interface GigabitEthernet1/0/49 queueing
GigabitEthernet1/0/49
Egress Priority Queue : enabled
Shaped queue weights (absolute) : 25 0 0 0
Shared queue weights : 1 30 35 5
The port bandwidth limit : 100 (Operational Bandwidth:100.0)
The port is mapped to qset : 1
Small Enterprise Design Profile Reference Guide
2-77
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
Building a Resilient Network
The Small Enterprise Design Profile is a high performance, resilient and scalable network design. A
network outage may be caused by the system, human error, or natural disaster. The small enterprise
network is designed to minimize the impact of a failure regardless of the cause. Network outages may
be either planned or unplanned.
•
Planned Outage—Planned network outage occurs when a portion of the network is taken out of
service as part of a scheduled event (e.g., a software upgrade).
•
Unplanned Outage—Any unscheduled network outage is considered an unplanned outage. Such
outages may be caused by internal faults in the network, or devices due to hardware or software
malfunctions.
The network is designed to recover from most un planned outages in less than a second (milliseconds).
In many situations, the user will not even notice the outage occurred. If the outage lasts longer (several
seconds), then the user will notice the lack of application responsiveness. The network is designed to
minimize the overall impact of a unplanned network outage, and gracefully adjust and recover from
many outage conditions. Figure 2-39 shows an example of a real-time VoIP application and user impact
depending on duration of outage event.
Figure 2-39
VoIP User Impact for Minor and Major Network Outage
50
45
40
35
30
25
20
Data Lose (Seconds)
15
10
5
No Impact
Minimal Impact
to Voice
User Hangs Up
Phone Reset*
227569
0
Several techniques are used to make the network design more resilient. Deploying redundant devices and
redundant connections between devices, enables the network to recover from fault conditions.
Identifying critical versus non critical applications, and network resources optimizes cost performance,
by focusing on the most important elements of the network design. The resiliency of a system design is
often categorized as follows:
•
Network Resiliency—Provides redundancy during physical link outages (e.g., fiber cut, bad
transceivers, incorrect cabling, etc).
•
Device Resiliency—Protects network during device outage triggered by hardware or software (e.g.
software crash, non-responsive supervisor, etc).
Small Enterprise Design Profile Reference Guide
2-78
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
•
Operational Resiliency—Capabilities which provide network availability even during planned
network outage conditions (e.g., ISSU features which enable software upgrades while device is
operating).
Figure 2-40
Resiliency Deployment Strategy
Network Resiliency
Device Resiliency
Distribution/Core
Network Resiliency
Device Resiliency
Operational Resiliency
Serverfarm
Network Resiliency
Device Resiliency
Network Resiliency
229432
Access
The high availability framework is based upon the three resiliency categories described in the previous
section. Figure 2-41 shows which technologies are implemented to achieve each category of resiliency.
Figure 2-41
High-Availability Categories and Technologies
Resilient
Goal
Resilient
Strategies
Network Service Availability
Network
Resiliency
Device
Resiliency
Operational
Resiliency
EtherChannel
NSF/SSO
UDLD
ISSU
Stack Wise
IP Event Dampening
227571
Resilient
Technologies
Redundant Hardware Components
Redundant hardware implementations vary between fixed configuration and modular Cisco Catalyst
switches. Selective deployment of redundant hardware is an important element of the Small Enterprise
Design Profile which delivers device resiliency.
Redundant hardware component for device resiliency varies between fixed configuration and modular
Cisco Catalyst switches. To protect against common network faults or resets, all critical main and remote
site campus network devices must be deployed with similar device resiliency configuration. This
subsection provides a basic redundant hardware deployment guideline at the access-layer and collapsed
core switching platforms in the campus network.
Small Enterprise Design Profile Reference Guide
2-79
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
Redundant Power System
Redundant power supplies protect the device from power outage or power supply failure. Protecting the
power is not only important for the network device, but also the endpoints that rely on power delivery
over the Ethernet network. Redundant power supplies are deployed differently depending on the switch
type:
•
Modular Switch—Dual power supplies can be deployed in the modular switching platforms like the
Cisco Catalyst 4500-E. By default, the Cisco Catalyst 4500 power supply operates in 1+1 redundant
mode (both power supplies are active).
•
Fixed Configuration Switch—Fixed configuration switches are deployed with internal power
supplies and they may also use Cisco RPS 2300 external power supply. A single Cisco RPS 2300
power supply has modular power supplies and fans to deliver power to multiple switches. Deploying
internal and external power-supplies provides a redundant power solution for fixed configuration
switches.
Redundant Network Connectivity
Redundant network connections protect the system from failure due to cable or transceiver faults.
Redundant network connections attached to a single fixed configuration switch or network module in the
Cisco Catalyst 4500 switch do not protect against internal device hardware or software fault.
Best practice design is to deploy redundant network modules within the Catalyst 4500 switch and the
Cisco 3750-X StackWise Plus solution in the small remote site collapsed core network. Deploying the
3750-X StackWise Plus in critical access-layer switches in the serverfarm network and in the main site
is also best practice. Connecting redundant paths to different hardware elements provides both network
and device resiliency.
Figure 2-42
Redundant Network Connectivity
Redundant StackWise Plus Switch
Stack Ring
Distribution/Core
Distribution/Core
Redundant
Line Cards
Diversed
Fiber Paths
Diversed
Fiber Paths
Redundant
Switch
Redundant
Switch
227572
Access
Access
Redundant Control-Plane
The processing software operation is different in standalone or StackWise fixed configuration switches,
and on a supervisor module of a modular switch. Network communication and forwarding operations
can be disrupted when the processing unit fails, causing a network outage. Network recovery techniques
vary based on the different platforms.
Small Enterprise Design Profile Reference Guide
2-80
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
The standalone and non-stackable fixed configuration switches like the Cisco Catalyst 2960 or 3560-E
feature power redundancy and network resiliency support; however they do not protect against a
processing unit failure. During a processing unit failure event, all endpoints attached to the switch are
impacted and network recovery time is undeterministic.
Device resiliency in Cisco StackWise and modular switching platforms provides 1+1 redundancy with
enterprise-class high availability and deterministic network recovery time.
Cisco StackWise Plus
Cisco Catalyst 3750-X switches can be deployed in StackWise mode using a special stack cable. Up to
nine switches can be integrated into a single stack that delivers distributed forwarding architecture and
unified single control and management plane. Device level redundancy in StackWise mode is achieved
via stacking multiple switches using the Cisco StackWise technology. One switch from the stack is
selected automatically to serve as the master, which manages the centralized control-plane process.
Cisco StackWise solution provides 1:N redundancy. In the event of a active master-switch outage, a new
master is selected automatically. See Figure 2-43.
Figure 2-43
Cisco Stack Wise Switching Architecture
Member
Switch – 1
Master
Switch – 2
Member
Switch – 3
Single Logical Switch
227573
Stack Ring
Since Cisco StackWise enables up to 9 switches to appear as one logical switch, it has centralized
management and control functions. Most Layer 2 and Layer 3 functions are centrally performed,
however Layer-2 topology development is distributed (i.e., each switch performs the function
independently). Table 2-12 lists network protocol functions and identifies which are centralized and
which are distributed.
Table 2-12
Cisco StackWise Centralized and Distributed Control-Plane
Protocols
Layer 2 Protocols
Layer 3 Protocols
Function
MAC Table
Distributed
Spanning-Tree Protocol
Distributed
CDP
Centralized
VLAN Database
Centralized
EtherChannel - LACP
Centralized
Layer 3 Management
Centralized
Layer 3 Routing
Centralized
Cisco StackWise Plus solution offers network and device resiliency with distributed forwarding. In the
event of a master switch outage, Non-Stop Forwarding (NSF) enables packet forwarding to continue
based on current state information, while a new master switch is selected. New master switch selection
Small Enterprise Design Profile Reference Guide
2-81
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
is accomplished in the range of 700 to 1000 milliseconds; the amount of time to reestablish the
control-plane and develop distributed forwarding will vary depending on the size and complexity of the
network.
Following is a best practice to reduce Layer-3 disruption in the event of a master switch outage:
Determine the master switch with the higher switch priority, and isolate the uplink Layer-3 EtherChannel
bundle path by using physical ports from member switches (i.e. don't use the master switches ports for
Etherchannel uplinks). With NSF capabilities enabled, this design decreases network downtime during
a master-switch outage.
An understanding of SSO and StackWise components and failover events associated with NSF provides
significant insight in designing a network that enables supervisor redundancy. The following subsection
uses the above concepts and principles to identify the design parameters, and applies them to develop a
best-practice hierarchical network with the highest availability.
Cisco FlexStack
The next-generation Catalyst 2960-S Series Layer 2 access-switch introduces high-speed, low-latency
stacking capability based on “pay as you grow” model. Following the Catalyst 3750-X StackWise Plus
success, the Catalyst 2960-S model offers high availability, increased port-density with unified single
control-plane and management to reduce the cost for small enterprise network. However, the architecture
of FlexStack on Catalyst 2960-S Series platform differs from StackWise Plus. The Cisco FlexStack is
comprised with hardware module and software capabilities. The FlexStack module must be installed on
each Catalyst 2960-S switches that are intended to be deployed in stack-group. Cisco FlexStack module
is hot-swappable module providing flexibility to deploy FlexStack without impacting business network
operation.
Cisco FlexStack allows up to four Catalyst 2960-S Series switches into a single stacking group; it is
recommended to deploy each switch member with dual FlexStack cable to provide increased 20G
bidirectional stack bandwidth capacity and FlexStack redundancy. The FlexStack protocol dynamically
detects switch member and allows it to join the stack group if all stacking criteria is met. The unique
data forwarding architecture and FlexStack QoS is on per-hop basis, the unknown unicast, broadcast,
and multicast traffic will be flooded between stack group switch members. The FlexStack protocol
detects and breaks the loop between the FlexStack group switches. Once the destination switch member
is determined, Catalyst 2960-S use shortest egress stack port path to forward traffic. Any packet traverses
across FlexStack is encapsulated with 32 bytes of FlexStack header carrying unique information to
provide centralized control-plane and distributed forwarding design.
Cisco Modular Switch
The Cisco Catalyst 4500-E modular switch supports redundant supervisors, and Stateful Switch Over
(SSO). When deployed along with NSF, the 4500-E provides a enterprise-class highly available system
with network and device resiliency.
SSO is a Cisco IOS service used to synchronize critical forwarding and protocol state information
between redundant supervisors configured in a single chassis. With SSO enabled, one supervisor in the
system assumes the role of active and the other supervisor becomes the hot-standby. Each is ready to
backup the other, thus providing 1:1 hot redundancy to protect from a control-plane outage. Since both
supervisors are active, the system benefits by using the physical ports from both supervisors during
normal operation. SSO synchronizes system services such as DHCP snooping, Switched Port Analyzer
(SPAN), security access control lists (ACLs), and QoS policies so ensure the switch provides the same
level of protection and service after a supervisor failover event.
Small Enterprise Design Profile Reference Guide
2-82
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
NSF enables packets to continue to be forwarded using existing routing table information, during
switchover. NSF also provides graceful restart to the routing protocol such that during the failover, the
routing protocol remains aware of the change and does not react by resetting its adjacency. If the routing
protocol were to react to the failure event, and alter routing path information, the effectiveness of stateful
switch over would be diminished.
Operational Resiliency Strategy
Designing the network to recover from unplanned outages is important. It is also important to consider
how to minimize the disruption caused by planned outages. These planned outages can be due to standard
operational processes, configuration changes, software and hardware upgrades, etc.
The same redundant components which mitigate the impact of unplanned outages can also be used to
minimize the disruption caused by planned outages. The ability to upgrade individual devices without
taking them out of service is enabled by having internal component redundancy (such as with power
supplies, and supervisors) complemented with the system software capabilities. Two primary
mechanisms exist to upgrade software in a live network:
•
Full-image In-Service Software Upgrade (ISSU) on the Cisco Catalyst 4500-E leverages dual
supervisors to allow for a full, in-place Cisco IOS upgrade. This leverages the NSF/SSO capabilities
of the switch and provides for less than 200 msec of traffic loss during a full Cisco IOS upgrade.
•
Network and device level redundancy, along with the necessary software control mechanisms,
guarantee controlled and fast recovery of all data flows following a fault condition, and provide the
ability to manage the fault tolerant infrastructure during planned outage events.
Validating operational resiliency is beyond the scope of this design guide, refer to CCO documentation
for deployment guidelines.
Deploying High Availability in Network
Many of the design features of the Small Enterprise Design Profile which were described in “Deploying
Network Foundation Services” section on page 2-16, contribute to the network high availability
capabilities. This section focuses on how to implement additional features which complete the high
availability design in small enterprise network design.
Network Resiliency
Etherchannel and UDLD are two design features which are included in the network foundation services,
which contribute to network resiliency.
Implementing IP Event Dampening
Poor signaling or a loose connection may cause continuous port-flap (port alternates between active state
and inactive state). A single interface flapping can impact the stability and availability of the network.
Route summarization is one technique which mitigates the impact of a flapping port. Summarization
isolates the fault domain with a new metric announcement by the aggregator and thereby hides the local
networks fault within the domain.
A best practice to mitigate local network domain instability due to port-flap, is implementing IP Event
Dampening on all layer 3 interfaces. Each time the Layer-3 interface flaps the IP dampening tracks and
records the flap event. Upon multiple flaps, a logical penalty is assigned to the port and suppresses link
status notification to IP routing until the port becomes stable. IP event dampening is a local function and
Small Enterprise Design Profile Reference Guide
2-83
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
does not have a signaling mechanism to communicate with a remote system. It can be implemented on
each individual physical or logical Layer-3 interface: physical ports, SVI or port-channels. Following is
a example configuration to implement IP Event Damenping:
Distribution/Core
cr24-4507-1(config)#int range Port1 , Gig5/6 , Gig6/6 , Vlan 101 - 110
cr24-4507-1(config-if-range)#dampening
cr24-4507-1#show interface dampening | be Port
Port-channel1 Connected to cr24-3750ME-1
Flaps Penalty
Supp ReuseTm
HalfL ReuseV
SuppV
0
0
FALSE
0
5
1000
16000
0
MaxSTm
2000
MaxP Restart
20
The following output illustrates how the IP event dampening keeps track of port flaps and makes a
decision to notify IP routing process based on interface suppression status:
cr24-4507-1#debug dampening interface
cr24-4507-1#show logging | inc EvD|IF-EvD
12:32:03.274: EvD(GigabitEthernet5/6): charge penalty 1000, new accum. penalty 1000, flap
count 2
12:32:03.274: EvD(GigabitEthernet5/6): accum. penalty 1000, not suppressed
12:32:03.274: IF-EvD(GigabitEthernet5/6): update IP Routing state to DOWN, interface is
not suppressed
cr24-4507-1#show interface dampening | be 5/6
Flaps Penalty
Supp ReuseTm
HalfL ReuseV
SuppV
2
0
FALSE
0
5
1000
0
MaxSTm
2000
MaxP Restart
20
16000
In a multilayer access-distribution design, the Layer-2 and Layer-3 demarcation is at the collapsed
core-distribution device. IP event dampening is enabled on per-logical VLAN (SVI) interface basis on
the collapsed core device. IP event dampening becomes more effective when each access-layer switch is
deployed with a unique set of Layer-2 VLANs.
Assigning unique VLANs on each access-layer switch also helps IP event dampening to isolate the
problem and prevent network faults triggered in a multilayer network. The following output illustrates
how IP event dampening keeps track of individual logical VLAN networks associated to same Layer-2
physical trunk ports. When a Layer-2 trunk port flaps, the state of SVI also flaps, and forces dampening
to track and penalize unstable interfaces:
12:58:41.332: EvD(Vlan101): charge penalty 1000, new accum. penalty 2627, flap count 3
12:58:41.332: EvD(Vlan101): accum. penalty 2627, now suppressed with a reuse intervals of
7
12:58:41.332: IF-EvD(Vlan101): update IP Routing state to DOWN, interface is suppressed
cr24-4507-1#show interface dampening
Vlan101 Connected to cr24_2960_Dept_1_VLAN
Flaps Penalty
Supp ReuseTm
HalfL ReuseV
SuppV MaxSTm
MaxP Restart
3
71
FALSE
0
5
1000
2000
20
16000
0
Small Enterprise Design Profile Reference Guide
2-84
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
Device Resiliency
As described earlier, redundant hardware is an important technique for achieving device resiliency. The
small enterprise network design applies hardware redundancy considering the cost/performance
tradeoffs.
Implementing Redundant Power Supply
Redundant power supplies can prevent a system outage due to power outage, power supply or fan
hardware failure. All Cisco Catalyst switching platforms supports robust 1+1 redundant power
capabilities that can be deployed with internal or external power source management. See Figure 2-44.
Cisco Catalyst Internal and External Power Redundancy Option
Redundant
Internal
Power
Supply
Redundant
External
Power
Supply
Redundant Dual
Power Supply
S W IT C H E D S H O U LD B E IN T H E O FF ‘O ’ P O S IT IO N T O IN S T A LL /
R E M O V E P O W E R S U P P LIE S . FA S T E N E R S M U S T B E FU LLY E N G A G E D
P R IO R T O O P E R A T IN G P O W E R S U P P LY
100-240V ~
12A
50/60H z
O U T P U T FA IL
O U T P U T FA IL
P O E E N A B LE D
P O E E N A B LE D
FA N O K
FA N O K
IN P U T 1
OK
IN P U T 1
OK
IN P U T 2
OK
IN P U T 2
OK
100-240V ~
12A
50/60H z
100-240V ~
12A
50/60H z
4200A C V
Master
Redundant
Power
Supply
Cisco 2300 RPS
StackPower
Cable
4200A C V
StackWise
Plus
1
Catalyst
4507R-E
1
SYSTEM
S W IT C H E D S H O U LD B E IN T H E O FF ‘O ’ P O S IT IO N T O IN S T A LL /
R E M O V E P O W E R S U P P LIE S . FA S T E N E R S M U S T B E FU LLY E N G A G E D
P R IO R T O O P E R A T IN G P O W E R S U P P LY
100-240V ~
12A
50/60H z
2
3
4
5
6
RPS
Cable
2
Redundant
Power
Supply
SUPERVISOR
4
SUPERVISOR
3
Member
Catalyst 3750-X
StackWise Plus
7
Master
Member
Catalyst 2960-S
Catalyst 4507R-E
229288
Figure 2-44
Catalyst 4500-E—Redundant Internal Power Supply
The Cisco Catalyst 4500-E provides power to internal hardware components and external devices like
IP phones. All the power is provided by the internal power supply. Dual-power supplies in the Catalyst
4500-E can operate in one of two different modes:
•
Redundant Mode—By default, Catalyst 4500-E power supply operates in redundant mode offering
1+1 redundant option. The system determines power capacity and number of power supplies
required based on the power required for all internal and external power components. Both power
supplies must have sufficient power to support all the installed modules and operate in 1+1
redundant mode.
cr24-4507-1(config)#power redundancy-mode redundant
cr24-4507-1#show power supplies
Power supplies needed by system
:1
Power supplies currently available :2
•
Combined Mode—If the system power requirement exceeds the capacity of a single power supply,
then both power supplies can be combined to increase the capacity. In this mode, the power system
does not provide 1+1 power redundancy. The following global configuration will enable power
supplies to operate in combined mode:
cr24-4507-1(config)#power redundancy-mode combined
Small Enterprise Design Profile Reference Guide
2-85
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
cr24-4507-1#show power supplies
Power supplies needed by system:2
Power supplies currently available:2
Catalyst 3750-X— Cisco StackPower Redundancy
The next-generation Catalyst 3750-X Series platform introduces innovative Cisco StackPower
technology to provide power redundancy solution for fixed configuration switches. Cisco StackPower
unifies the individual power supplies installed in the switches and creates a pool of power, directing that
power where it is needed. Up to four switches can be configured in a StackPower stack with the special
Cisco proprietary StackPower cable. The StackPower cable is different than the StackWise data cables
and is available on all Cisco Catalyst 3750-X models.
During individual power supply fault from the stack can regain power from global power pool to provide
seamless operation in the network. With the modular power supply design in Catalyst 3750-X series
platform the defective power supply can be swapped without disrupting network operation. The Cisco
StackPower can be deployed in following two modes:
•
Sharing Mode—All input power is available to be used for power loads. The total aggregated
available power in all switches in the power stack (up to four) is treated as a single large power
supply. All switches in stack can share power with available power to all powered devices connected
to PoE ports. In this mode, the total available power is used for power budgeting decisions and no
power is reserved to accommodate power-supply failures. If a power supply fails, powered devices
and switches could be shut down. Default mode.
•
Redundant Mode—The power from the largest power supply in the system is subtracted from the
power budget, which reduces the total available power, but provides backup power in case of a
power-supply failure. Although there is less available power in the pool for switches and powered
devices to draw from, the possibility of having to shut down switches or powered devices in case of
a power failure or extreme power load is reduced. It is recommended to budget the required power
and deploy each Catalyst 3750-X switch in stack with dual power supply to meet the need. Enabling
redundant mode will offer power redundancy as a backup during one of the power supply unit failure
event.
Since Cisco StackWise Plus can group up to nine 3750-X Series switches in the stack-ring, the Cisco
StackPower must be deployed with two power stack group to accommodate up to four switches.
Following sample configuration demonstrate deploying Cisco StackPower redundancy mode and
grouping the stack-member into power stack group, to make new power configuration effective, it is
important that network administrator must plan downtime as all the switches in the stack ring must be
reloaded:
cr36-3750X-xSB(config)# stack-power stack PowerStack
cr36-3750X-xSB(config-stackpower)#mode redundant
cr36-3750X-xSB(config)#stack-power switch 1
cr36-3750X-xSB(config-switch-stackpower)#stack-id PowerStack
%The change may not take effect until the entire data stack is reloaded
cr36-3750X-xSB(config)#stack-power switch 2
cr36-3750X-xSB(config-switch-stackpower)#stack-id PowerStack
%The change may not take effect until the entire data stack is reloaded
Small Enterprise Design Profile Reference Guide
2-86
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
Catalyst 2960 (External Power Redundancy)
The Cisco Redundant Power Supply (RPS) 2300 can support up to six RPS ports to provide seamless
power backup to critical access-layer switches in the campus network. Upto two devices can be protected
by Cisco RPS 2300 against power or power supply failure. Additional power resiliency on Cisco RPS
2300 can be added by deploying dual power supply to backup to two devices simultaneously. Note that
external power redundancy requires special RPS cable and specific 2960 models currently do not support
external power redundancy. Deploying external power redundancy on Cisco Catalyst 2960 and 2960-S
with Cisco RPS 2300 is performed automatically and do not require any extra configuration. Cisco
Catalyst 3560-X and 3750-X switches can be used if RPS 2300 configuration is required.
Implementing Redundant Control Plane System
The collapsed core device in the main and remote sites (Catalyst 4500-E or 3750-X StackWise Plus) is
deployed with redundant supervisor, or StackWise Plus to enable graceful recovery from switch
hardware outage. Any access-switch which is deemed critical may be deployed as StackWise Plus and
FlexStack to improve device resiliency. The implementation for each switch is different, and is discussed
separately in the sections which follow.
Resilient Cisco FlexStack and StackWise Plus
Starting in Cisco IOS Release 12.2(53)SE1, Cisco Catalyst 2960-S supports FlexStack and it can be used
when higher port-density, increased uplink bandwidth capacity, and the resiliency at Layer-2 access is
required. Starting in Cisco IOS Release 12.2(53)SE2, Cisco Catalyst 3750-X supports StackWise Plus,
and is used when a resilient Layer 2 or Layer 3 access-switch is required. Cisco 3750-X StackWise Plus
is deployed for the collapsed core in the small remote site network.
Cisco Catalyst 3750-X and 2960-S switches are provisioning dynamically in the stack group by the
StackWise or FlexStack protocols. Cisco IOS automatically adjusts the interface addressing and its
associated configuration based on the number of provisioned switches in the stack.
cr26-2960s-1#show run | inc provision
switch 1 provision ws-c2960s-48ts-s
switch 2 provision ws-c2960s-48ts-s
Master Switch Election
The centralized control-plane and management plane is managed by the master switch in the stack. By
default, the master switch selection within the ring is performed dynamically by negotiating several
parameters and capabilities between each switch within the stack. Each StackWise-capable switch is by
default configured with priority 1.
cr26-C2960S-1# show switch
Switch/Stack Mac Address : 0022.bdc4.1d80
H/W
Current
Switch# Role
Mac Address
Priority Version State
---------------------------------------------------------*1
Master 0022.bdc4.1d80
1
1
Ready
2
Member 0026.0ac1.3e00
1
1
Ready
As described in previous section, the Cisco StackWise Plus architecture is not SSO-capable. This means
that all the centralized Layer-3 functions must be reestablished with the neighbor switch during a
master-switch outage. To minimize the control-plane impact and improve network convergence, the
Layer-3 uplinks should be diverse, originating from member switches instead of the master switch. The
Small Enterprise Design Profile Reference Guide
2-87
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
default switch priority must be increased manually after identifying the master switch and switch
number. The new switch priority becomes effective after switch reset. It is recommended to modify
default switch priority on the Catalyst 2960-S FlexStack group as Catalyst 3750-X StackWise.
cr26-3750r-1(config)#switch 1 priority
Changing the Switch Priority of Switch
cr26-3750r-1(config)#switch 2 priority
Changing the Switch Priority of Switch
15
Number 1 to 15
14
Number 2 to 15
cr26-3750r-1#show switch
Switch/Stack Mac Address : 0023.eb7b.e580
H/W
Current
Switch# RoleMac AddressPriorityVersionState
---------------------------------------------------------------1
Member0023.eb7b.e580150
Ready
* 2
Master0026.5284.ec80140
Ready
StackWise Layer-3 MAC Management
To provide a single unified logical network view in the network, the MAC addresses of Layer-3
interfaces on the StackWise (physical, logical, SVIs, port channel) are derived from the Ethernet MAC
address pool of the master switch in the stack. All the Layer-3 communication from the StackWise
switch to the endpoints (like IP phone, PC, servers and core network system) is based on the MAC
address pool of the master switch.
cr26-3750r-1#show switch
Switch/Stack Mac Address : 0026.5284.ec80
H/W
Current
Switch# Role
Mac AddressPriority
Version
State
----------------------------------------------------------------------------------------1Member0023.eb7b.e580
1
0
Ready
* 2 Master 0026.5284.ec80
cr26-3750r-1#show version
. . .
Base ethernet MAC Address
. . .
1
0
Ready
: 00:26:52:84:EC:80
After a master-switch outage, the new master switch in the stack assigns new MAC addresses to all
Layer-3 interfaces, from the local MAC address pool. Once the new MAC address is assigned, it will
force the switch to generate a gratuitous ARP in the network to make sure no other system is using the
same MAC address. The default timer to retain the MAC address from the failed master switch is four
minutes. While the new MAC address is not assigned on Layer-3 interface and not being propagated and
updated in the network, the traffic will blackhole in the network.
cr26-3750r-1#reload slot 2
Proceed with reload? [confirm]
Switch 2 reloading...
cr26-3750r-1#show switch
Switch/Stack Mac Address : 0023.eb7b.e580
H/W
Current
Switch# Role
Mac AddressPriority
Version
State
----------------------------------------------------------------------------------------* 1 Master 0023.eb7b.e580
1
1
Ready
2Member000.0000.0000
0
1
Removed
Small Enterprise Design Profile Reference Guide
2-88
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
To prevent this network instability, the old MAC address assignments on Layer-3 interfaces can be
retained even after the master switch fails. The new active master switch can continue to use the MAC
addresses assigned by the old master switch, which prevents ARP and routing outages in the network.
The default stack-mac timer settings must be changed in Cisco Catalyst 2960-S FlexStack and 3750-E
StackWise Plus switch mode using the global configuration CLI mode as shown below:
cr26-3750r-1(config)#stack-mac persistent timer 0
cr26-3750r-1#show switch
Switch/Stack Mac Address : 0023.eb7b.e580
Mac persistency wait time: Indefinite
H/W
Current
Switch# Role
Mac AddressPriority
Version
State
----------------------------------------------------------------------------------------1Master0023.eb7b.e580
1
0
Ready
* 2 Master 0026.5284.ec80
1
0
Ready
Non-Stop Forwarding (NSF)
The Cisco Catalyst 3750-X switch in StackWise Plus mode is not SSO-capable. When the master switch
fails, the new master switch is required to reform the Layer-3 adjacencies with the neighbors in the
network. The forwarding architecture in StackWise switch is designed to provide non-stop forwarding
during the master switch outage using NSF technology. Each 3750-X switch in the stack maintains
distributed Layer-3 FIB from the old master switch and continues to forward upstream traffic, until they
are updated by the new master in the stack ring.
To enable NSF capability, explicit configuration must be enabled under the routing process. NSF-aware
feature is enabled by default on all Layer-3 Ethernet switches to function in helper mode to perform
graceful recovery during NSF-capable Cisco 3750-X master switch outage. NSF-capable system can
also operate in NSF aware role:
NSF Capable Layer-3 Switch
cr36-3750s-1(config)#router eigrp 100
cr36-3750s-1(config-router)#nsf
cr36-3750s-1#show ip protocols | inc NSF
*** IP Routing is NSF aware ***
EIGRP NSF-aware route hold timer is 240
EIGRP NSF enabled
NSF signal timer is 20s
NSF converge timer is 120s
NSF-Aware Layer-3 Switch
cr24-3560r-1#show ip protocols | inc NSF
*** IP Routing is NSF aware ***
EIGRP NSF-aware route hold timer is 240
Small Enterprise Design Profile Reference Guide
2-89
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
NSF Timers
As depicted in the above show commands, the default NSF-aware system hold timer is 240 seconds.
Lowering the timer value may abruptly terminate graceful recovery, causing network instability. Best
practice is to use the default NSF hold timer, unless it is observed that NSF recovery takes longer than
240 seconds.
600 seconds after a graceful-recovery starts on a NSF-aware system, NSF clears the route stale marking
and resumes using the synchronized routing database.
! NSF Aware received graceful-restart message from new master switch
11:56:15.365: %DUAL-5-NBRCHANGE: EIGRP-IPv4:(100) 100: Neighbor 10.125.32.3
(Port-channel15) is resync: peer graceful-restart
11:56:15.365: EIGRP: NSF: AS100, NSF or GR initiated by 10.125.32.3 at 00:00:00, flags 0x4
! NSF route hold timer expires and searches and removes all stale route entries
received graceful-restart message from new master switch
12:00:15.392: EIGRP: NSF: AS100. route hold timer expiry
12:00:15.392: DUAL: Search for outdated routes from 10.125.32.3
Resilient Cisco Catalyst 4500-E
A modular switching platform like the Cisco Catalyst 4507R-E is fully NSF/SSO-capable, providing 1+1
control plane redundancy. In the Catalyst 4507R-E, all the intelligent Layer-2 and Layer-3 functions
are performed centrally on the supervisor module. Deploying redundant supervisor in SSO mode in same
system will allow the primary supervisor to fully synchronize the adjacencies, forwarding,
configuration, counters, and more information on redundant hot-standby supervisor.
The Cisco Catalyst 4507R-E ports are independent of the supervisor state. Because of this hardware
design, during a supervisor switchover, the ports connected to the failed supervisor do not go down.
Because paths and ports are not down, hardware keeps forwarding the packet to a valid next-hop while
supervisor switchover is occurring.
The configuration and implementation guidelines for implementing NSF/SSO on the Cisco Catalyst
4507R-E are the same for main and remote site network designs.
Increasing Supervisor Uplink Port Availability
There are restrictions on which supervisor uplink ports can be actively configured. Multiple ports can be
simultaneously active on the supervisor. However Cisco IOS Release 12.2(25)SG or later is required for
concurrent use of both 10G and 1G. The Small Enterprise Design Profile uses the 1G interface to connect
to the Cisco 3750-MetroE WAN aggregation switch. To use 10G port in 1G mode with redundancy, the
following configuration must be applied on collapsed core Catalyst 4507R-E switch:
cr24-4507-1(config)#hw-module uplink mode shared-backplane
cr24-4507-1(config)#hw-module module 3 port-group 1 select gigabitethernet
cr24-4507-1(config)#hw-module module 4 port-group 1 select gigabitethernet
cr24-4507-1#show hw-module uplink
Active uplink mode configuration is Shared-backplane
Small Enterprise Design Profile Reference Guide
2-90
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Building a Resilient Network
Stateful Switchover (SSO)
SSO redundancy mode in the Cisco Catalyst 4507R-E supervisor is turned on by default starting with
Cisco IOS Release 12.2(20)EWA. To provide 1+1 redundancy, all the technical specifications between
active and standby supervisor must be identical. Also note that the Cisco Catalyst 4507R-E and 4510R-E
are the only models that support supervisor redundancy. SSO is supported on all supervisors running IOS
except Sup II-Plus-TS. The NSF-awareness feature is supported by all the supervisors supporting
EIGRP, OSPF, IS-IS, and BGP routing protocols, while the NSF-capable feature is supported only on
supervisors IV, V, and V-10G. NSF/SSO on Catalyst 4500-E requires a minimum boot ROM version and
must be the same on both supervisors. For additional details on hardware requirements, refer to the
Release Notes at the following URL:
http://www.cisco.com/en/US/docs/switches/lan/catalyst4500/12.2/53SG/configuration/NSFwSSO.html
#wp1135767
cr24-4507-1(config)#redundancy
cr24-4507-1(config-red)# mode sso
cr24-4507-1(config-red)# main-cpu
cr24-4507-1(config-r-mc)# auto-sync standard
cr24-4507-1#show module | inc Chassis | 6-E | SSO
Chassis Type : WS-C4507R-E
3
4
3
4
6
6
Sup 6-E 10GE (X2), 1000BaseX (SFP)
Sup 6-E 10GE (X2), 1000BaseX (SFP)
Active Supervisor
Standby Supervisor
WS-X45-SUP6-E
WS-X45-SUP6-E
JAE1132SXQ3
JAE1132SXRQ
SSOActive
SSOStandby hot
The active supervisor dynamically detects the secondary supervisor installed in the same chassis and
initiates several SSO dependency configuration checks. If the SSO dependency check fails, then the
standby supervisor falls back into RPR mode. For example, IOS release mismatch between two
supervisors may not allow SSO to synchronize.
If the SSO dependency configuration checks successfully pass, then SSO communication between both
supervisors goes through several synchronization states before it transitions to hot-standby state as
illustrated in the following output:
cr24-4507-1#show redundancy states
my state = 13 -ACTIVE
peer state = 8 -STANDBY HOT
. . .
Redundancy Mode (Operational) = Stateful Switchover
Redundancy Mode (Configured) = Stateful Switchover
Redundancy State
= Stateful Switchover
Maintenance Mode = Disabled
Manual Swact = enabled
Communications = Up
. . .
All the state-machines and dynamic information of SSO-capable protocols are automatically
synchronized to the standby supervisor module without any additional operational requirement. The
hot-standby supervisor takes over the ownership of control-plane process when the active supervisor
outage or removal from the chassis is detected.
Small Enterprise Design Profile Reference Guide
2-91
Chapter 2
Small Enterprise Design Profile(SEDP)—Network Foundation Design
Summary
Non-Stop Forwarding (NSF)
All the state-machines and dynamic information of SSO-capable protocols are automatically
synchronized to the standby supervisor module. The hot-standby supervisor takes over the ownership of
control-plane process if the active supervisor suffers an outage or is removed from the chassis.
NSF-Capable Layer 3 Switch
cr24-4507-1(config)#router eigrp 100
cr24-4507-1 (config-router)#nsf
cr24-4507-1#show ip protocols | inc NSF
*** IP Routing is NSF aware ***
EIGRP NSF-aware route hold timer is 240
EIGRP NSF enabled
NSF signal timer is 20s
NSF converge timer is 120s
NSF-Aware Layer 3 Switch
cr24-3560r-1#show ip protocols | inc NSF
*** IP Routing is NSF aware ***
EIGRP NSF-aware route hold timer is 240
Summary
Designing the LAN network aspects for the small enterprise network design establishes the foundation
for all other aspects within the design profile including WAN, Security, and Mobility.
This chapter reviews LAN design models recommended by Cisco, as well as where to apply these models
within the various locations of a small enterprise network. Each of the layers is discussed and design
guidance is provided on where to place and how to deploy these layers. Finally, key network foundation
services such as routing, switching, QoS, multicast, and high availability best practices are given for the
entire small enterprise design.
Small Enterprise Design Profile Reference Guide
2-92
CH A P T E R
3
Small Enterprise Design Profile(SEDP)—WAN
Design
This chapter discusses how to design and deploy WAN architecture for Small Enterprise Design Profile.
The primary components of the WAN architecture are as follows:
•
WAN technology
•
Bandwidth capacity planning
•
WAN IP addressing structure
•
Routing
•
QoS
WAN Design
The success of Small Enterprise Design depends on having an effective and better collaborative
environment between employees located at different geographic locations. Some of the technologies that
enhance this environment are interactive-video, on-demand video, voice and web collaboration, video to
mobile devices, and TelePresence. To enable these technologies that foster the collaborative
environment, WAN design is a critical consideration. There are several WAN technologies available
today to provide WAN services, such as MPLS/VPN, Internet, and Metro Ethernet. See Figure 3-1.
Figure 3-1
WAN Technology Representation
MPLS VPN
Ethernet over MPLS
Ethernet access to
MPLS
Internet
Leased Line
Ethernet Handoff
Broadband
Metro
Ethernet
227580
Next-Generation WAN/MAN
MPLS/VPN
The MPLS/VPN provides Layer-2 or Layer-3 VPN services. It provides the capability to an IP network
infrastructure that delivers private network services over a shared network infrastructure. For more
information about deploying MPLS/VPN, refer to the following URL: www.cisco.com/go/srnd
Small Enterprise Design Profile Reference Guide
3-1
Chapter 3
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Design
Internet
Out of the three WAN technologies, Internet is the cheapest and easiest to deploy. However, to deploy
a VPN service over Internet requires an overlay VPN network such as DMVPN to provide secure VPN
service.
Metro Ethernet
Metro Ethernet is one of the fastest growing transport technologies in the telecommunications industry. In
the current WAN design, it is recommended to use Metro Ethernet Service as WAN transport between
remote sites and main office. The following subsection describes the advantages of Metro service.
Why Metro Ethernet Service Is Needed
The deployment of carrier Ethernet services had raced to an astounding $7 billion by 2007* with
predictions of more than $30 billion by 2012. This is driven by the following benefits to end users (IT,
network, and applications departments) and service providers alike:
•
Scalability, ubiquity, and reachability
– Global availability of standardized services independent of physical access type dramatically
reduce complexity and cost
•
Performance, QoS, and suitability for convergence
– Inherently, Ethernet networks require less processing to operate and manage at higher
bandwidth than other technologies
– Low latency and delay variation make it the best solution for video, voice, and data
•
Cost savings
–
•
Carrier Ethernet brings the cost model of Ethernet to the wide area network WAN)
Control, simplicity, familiarity
– IT departments manage Ethernet networks every day and now are in control of their IP routed
networks, worldwide
•
Expediting and enabling new applications
– Accelerates implementations with reduced resources for overburdened IT departments
– Enables new applications requiring high bandwidth and low latency that were previously not
possible or prohibited by high cost
Types of Services Available in Metro Ethernet
The following are two popular methods of Metro service:
•
E-line, also known as Ethernet Virtual Private Line (EVPL), provides a point-to-point service.
•
E-LAN provides multipoint or any to any connectivity.
EVPL, like Frame Relay, provides multiplexing multiple point-to-point connections over a single
physical link. In the case of Frame Relay, the access link is a serial interface to a Frame Relay switch
with individual data-link connection identifiers (DLCIs) identifying the multiple virtual circuits or
connections. In the case of EVPL, the physical link is Ethernet, typically FastEthernet or Gigabit
Ethernet, and the multiple circuits are identified as VLANs by way of an 802.1q trunk.
Small Enterprise Design Profile Reference Guide
3-2
Chapter 3
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Service Deployed in the Small Enterprise Design
One of the major advantage E-LAN provides is any-to-any connectivity within the Metro area, which allows
flexibility. It passes 802.q trunks across the SP network known as Q-in-Q. Figure 3-2 shows the difference
between E-line and E-LAN.
Figure 3-2
Different Services Available
E-LAN
227581
E-Line
WAN Service Deployed in the Small Enterprise Design
In this version of the guide, E-line with point-to-point service is chosen as the model for connecting
remote sites to main site. All the remote sites have a point-to-point connection to the main site. Each
remote site is assumed to have 100Mpbs of Metro service to the service provider. As mentioned in the
previous section, each circuit is represented by a VLAN using dot1q trunk. Figure 3-3 shows how this
is implemented in this design.
EVPN Service Used in Remote WAN Architecture
Virtual Circuit
Point-To-Point
Main Site
Service
Provider
Remote Site 1
Remote Site 2
Remote Site 3
229310
Figure 3-3
Small Enterprise Design Profile Reference Guide
3-3
Chapter 3
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Service Deployed in the Small Enterprise Design
The following is the configuration of the WAN interface at the main site:
interface GigabitEthernet1/1/1
description Connected to SP-MPLS-Core-cr24-6500-1
switchport trunk native vlan 801
switchport trunk allowed vlan 501-550
switchport mode trunk
logging event trunk-status
load-interval 30
carrier-delay msec 0
priority-queue out
mls qos trust dscp
spanning-tree portfast trunk
spanning-tree bpdufilter enable
spanning-tree guard root
service-policy output Remote-1to50-Parent-Policy-Map
hold-queue 2000 in
hold-queue 2000 out
As shown in the above configuration, the link is carrying 50 VLANs, which are connected to 50 small
office sites. The solution shown above is one of the ways to deploy WAN service; there are other ways
to deploy the WAN service. The next section discusses when to look at alternative to Metro service.
When to Consider WAN Technologies
In this design, all the remote sites are connected to the main site using point-to-point links that provide
high bandwidth, symmetric traffic flows, and rich media services to all the sites. Since it is a
point-to-point circuit, when the number of circuits increases to 1K and beyond, then we need not only
to have 1K point-to-point circuits at the main site but also have 1K routing adjacency relationships,
which will increase the CPU utilization. Therefore, when the scalability numbers increase more than 1K,
an alternative Layer-3 WAN technology such as MPLS/VPN should be considered. MPLS/VPN Layer-3
technology deployments will not require different hardware, but only changing the WAN service.
Figure 3-4 shows the routing behavior difference between Layer 2 (Metro service) and Layer 3
(MPLS/VPN service) from the routing protocol perspective.
Small Enterprise Design Profile Reference Guide
3-4
Chapter 3
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Service Deployed in the Small Enterprise Design
Figure 3-4
Difference in Routing Adjacency Between Layer 2 and Layer 3 Service
Layer 2 (L2) Service
CE
PE
PE
CE
Direct Layer 2 Adjacency Between CE Routers
Layer 3 (L3) Service
PE
PE
Direct Layer 2 Adjacency Only
Between CE and PE Routers
CE
229311
CE
Bandwidth Capacity
This is critical component of the overall WAN design architecture because the application performance
largely depends upon guaranteed level of bandwidth at remote sites and the main site. This section
primarily discusses the general WAN bandwidth capacity planning and implementation steps.
Planning
To start with, each site must purchase the Metro Ethernet service from local service provider (SP) with
bandwidth capacity that can meet the current minimum network load and provides enough flexibility to
scale higher in the future, when additional applications need to be added. This bandwidth must be shared
based on 4 class QoS model as described in the “Deploying QoS in Network” section on page 2-51 for
optimal delivery. Logical bandwidth assignment to each circuit must be symmetric at the remote small
office sites and at the main site. A mismatch in bandwidth capacity will force the traffic drop in the SP
core due to inconsistent bandwidth service-level agreement.
This design requirement remains consistent for all the remote sites. Depending on large, medium, or
small size, the WAN bandwidth capacity may also vary; however, the bandwidth sharing, routing, QoS,
multicast etc. principles remains consistent.
Note
Main site WAN router is an aggregator that connects multiple small office sites logically over a single
media. This media type must be GigE, because the platform is 3750ME and the WAN egress port is a
non-negotiated fiber port.
Small Enterprise Design Profile Reference Guide
3-5
Chapter 3
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Service Deployed in the Small Enterprise Design
Calculating the optimum guaranteed bandwidth required at each remote site and the main site must be
done by considering the following factors:
•
Number of remote sites
•
Bandwidth required at each location
•
Platforms scalability limits, mainly, at main office
To explain how much bandwidth is required at each remote site, two terminologies (CIR and interface
bandwidth) are used:
•
CIR—The committed bandwidth that is guaranteed from the service provider, based on the logical
connection.
•
Interface speed—The interface speed is actual Ethernet handoff, which is 1Gbps for both the remote
sites and the main site.
For example, consider a scenario where there are three remote sites and one main site. The CIR required
at each remote site is 100Mbps and, because there are three remote sites, the main office needs three
virtual circuits each having a CIR of 100Mbs. This also means that the aggregate bandwidth at main
office is 300Mbps. Figure 3-5 illustrates this example.
Figure 3-5
Bandwidth Capacity Planning for Three Remote Sites
Interface Speed: 1000Mbps
CIR Bandwidth: 100Mbps
Interface Speed: 1000Mbps
CIR Bandwidth for each site: 100Mbps
Aggregate Bandwidth Capacity: 300Mbps
Remote Site 1
Service
Provider
Remote Site 2
Remote Site 3
229312
Main Site
Similarly, if the CIR required at each remote site is 100Mbps and if there are 100 remote sites, then the
bandwidth required at main office is 10Gbps. To support an aggregate CIR bandwidth of 10Gbps at main
office, we need to think about scalability limits on the WAN aggregation switch. If the aggregated logical
connection speed to the remote sites exceed the media capacity on 3750-ME, which is WAN aggregation
switch for this design, then there are two approaches to mitigate the problem:
•
Migrate to a Modular switching platform—Migrating to a modular switching platform for the WAN
aggregation tier will enable wan capability to scale higher, reduce operational and management
complexities and may offer better performance and resiliency depending on modular platform
capabilities. When planning to migrate to a modular switching platform, please make sure that it
supports the same QoS features of 3750ME, along with redundancy.
•
Deploy another 3750-ME in same tier—Alternatively, it is possible to deploy secondary 3750-ME
system to connect another set of remote networks. Deploying another set of 3750-ME does not
change any WAN principles, In fact, except VLAN and IP address, all the configurations can be
replicated to the secondary system.
Small Enterprise Design Profile Reference Guide
3-6
Chapter 3
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Service Deployed in the Small Enterprise Design
Implementation
This section discusses how WAN bandwidth was implemented and validated at remote sites and main
office of this design. The following are the considerations used to validate this model:
•
The main office has dual WAN connections, and each connection is 1Gbps.
•
On each WAN connection at the main office, there are 50 virtual circuits, each with CIR of 20Mbps.
•
The remote sites have 1Gbps connection, and CIR value of 20Mbs on each circuit.
Figure 3-6 shows how this design is validated with 100 remote sites.
Figure 3-6
Bandwidth Capacity Planning for 100 Remote Sites
Remote Site 1
Interface Speed: 1000Mbps
CIR Bandwidth: 20Mbps
Interface Speed: 1000Mbps
CIR Bandwidth : 20Mbps
Aggregate Bandwidth: 1000Mbps
Main Site
Remote Site 50
Service
Provider
Remote Site 51
Interface Speed: 1000Mbps
CIR Bandwidth: 20Mbps
Aggregate Bandwidth: 1000Mbps
Remote Site100
229313
Interface Speed: 1000Mbps
CIR Bandwidth: 20Mbps
The next section discusses how to implement an IP addressing structure for the WAN interfaces.
Aggregation Structure
This section describes how to aggregate IP address space at remote sites and the main site on the WAN
interface.
For designing IP address spaces for the LAN, refer to “Multicast IP Addressing” section on page 2-43.
As explained in the previous section, all the remote sites are connected to the main site using
point-to-point links, which means that every remote site should be on a different subnet. One common
method that some customers deploy is using /30 subnets at both ends. However, this would mean that
every subnet uses up to four addresses. The recommendation is to use /31 subnets, this way only two
addresses are used in every link. The following is a sample configuration for deploying this link at the
remote site:
Small Enterprise Design Profile Reference Guide
3-7
Chapter 3
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Service Deployed in the Small Enterprise Design
Remote Site
Main Office
interface Vlan501
description Connected to cr24-3750ME-DO
dampening
ip address 10.127.0.1 255.255.255.254
no ip redirects
no ip unreachables
ip authentication mode eigrp 100 md5
ip authentication key-chain eigrp 100 eigrp-key
ip pim sparse-mode
ip summary-address eigrp 100 10.127.0.0
255.255.248.0 5
load-interval 30
interface Vlan501
description Connected to cr35-4507-SS1
dampening
ip address 10.127.0.0 255.255.255.254
no ip redirects
no ip unreachables
ip authentication mode eigrp 100 md5
ip authentication key-chain eigrp 100 eigrp-key
ip pim sparse-mode
ip summary-address eigrp 100 10.124.0.0
255.252.0.0 5
load-interval 30
hold-queue 2000 in
hold-queue 2000 out
After planning an IP addressing scheme, the next part is to complete a routing design for the WAN
connections, which is discussed in the following section.
Routing for WAN Connections
This section discusses how to implement routing on WAN interfaces. The key consideration while
designing routing scheme is as follows:
•
Summarization on network boundaries—This is very important aspect of the design as it prevents
unnecessary routing updates to flow across the WAN interface when there is a link state change in
the network. For example, let us consider the main site where there are subnets in the following
range:
10.127.0.0/26
10.127.0.64/26
10.127.0.128/26
.
.
.
10.127.7.64/26
Since the above subnets belong to a particular remote site (the main site pair, they could be summarized
as 10.127.0.0/21. The following configuration shows how to perform summarization on the WAN
interface, which is Vlan501 in this example:
Remote Site
Main Office
interface Vlan501
description Connected to cr24-3750ME-DO
dampening
ip address 10.1270.1 255.255.255.254
no ip redirects
no ip unreachables
ip authentication mode eigrp 100 md5
ip authentication key-chain eigrp 100 eigrp-key
ip pim sparse-mode
ip summary-address eigrp 100 10.127.0.0 255.255.248.0 5
load-interval 30
interface Vlan501
description Connected to cr35-4507-SS1
dampening
ip address 10.1270.0 255.255.255.254
no ip redirects
no ip unreachables
ip authentication mode eigrp 100 md5
ip authentication key-chain eigrp 100 eigrp-key
ip pim sparse-mode
ip summary-address eigrp 100 10.124.0.0 255.252.0.0 5
load-interval 30
hold-queue 2000 in
hold-queue 2000 out
Small Enterprise Design Profile Reference Guide
3-8
Chapter 3
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Service Deployed in the Small Enterprise Design
The next most important component is QoS design, which the following section discusses.
QoS Design
This section discusses how QoS is implemented at the main office and remote sites. With Ethernet
defined as the WAN medium, QoS design becomes most important because the router or switch on the
WAN edge may believe that it has the complete line rate available for it to transmit. If so, the service
provider would start dropping packets due to policers defined at its end. Figure 3-7 shows what may
happen without a proper QoS design.
Figure 3-7
Policing at Service Provider Due to Lack of Proper QoS at Main Office and Remote
Sites
ow
Inbound Policing
c
ffi
a
Tr
Traffic Flow
Service
Provider
Main Site
Fl
Remote Site 1
Traffic Flow
Tr
Remote Site 2
af
fic
Fl
Remote Site 3
229314
ow
To prevent packets getting dropped at the service provider network, it is very important to have proper
and consistent QoS policies at the remote sites and main site. Having a proper QoS policy would ensure
that all the traffic is queued and serviced as per the bandwidth defined. This in turn ensures that packets
will not be dropped by the service provider. The following subsections discuss how QoS policies are
implemented at the remote and main sites.
QoS Policy at Main Site
As mentioned in the “When to Consider WAN Technologies” section on page 3-4, the main office has
several point-to-point circuits to the remote sites, and the number (point-to-point circuits) depends upon
the number of remote sites. Therefore, the QoS policy at the main site has the following objectives:
•
The aggregate traffic going out to the remote site does not exceed 20Mbs.
•
All the traffic going out is put in four classes.
To accomplish the above objectives, an hierarchical CBWFQ (HCBWFQ) is implemented. For more
information about HCBWFQ, refer to “Chapter 4, Medianet QoS Design Considerations” of the
Medianet Reference Guide at following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Video/Medianet_Ref_Gd/medianet_ref_gd.htm
l
Small Enterprise Design Profile Reference Guide
3-9
Chapter 3
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Service Deployed in the Small Enterprise Design
To implement HCWFQ, the following two policies are needed:
1.
Parent policy that defines the aggregate shape rate.
2.
Child policy that enables queuing within the shaped rate.
Figure 3-8 shows the representation of hierarchical policy.
Figure 3-8
Hierarchical Policy Implementation at Main Office
GE Speed Interface
to SP MetroE
1
Parent Policy: Per-Port aggregated Traffic Shape
from Main Site to all remote School Sites
VLAN 502
Remote Site 1
1
VLAN 501
Remote Site 2
Parent Policy:
Per-VLAN Sub-rate
Traffic Shaping to
Remote Site 1
Parent Policy:
Per-VLAN Sub-rate
Traffic Shaping to
Remote Site 2
1
2
229315
Child Policy: Per-VLAN Queueing
Queue-1 – Real-time Class (PQ) – 30% of Per-VLAN BW
Queue-2 – Gold Class (CBWFQ) – 5% of Per-VLAN BW
Queue-3 – Silver Class (CBWFQ) – 30% of Per-VLAN BW
Queue-4 – Best-Effort (WFQ/WRED) – 35% of Per-VLAN BW
Table 3-1 shows how these classes are defined.
Table 3-1
QoS Policies
Class
Queuing type
Bandwidth Allocation
REAL_TIME
LLQ
30%
GOLD
CBWFQ
5%
SILVER
CBWFQ
30%
DEFAULT
CBWFQ
35%
The following is the configuration of the QoS policy at the main office. For brevity, only configuration
for one VLAN is shown.
class-map match-all Remote_Site1  This class map would match the traffic going to a
Remote site
description cr2-4507-SS1
match vlan 501
policy-map Remote-Child-Policy-Map  This is child policy
class REAL_TIME
priority
police cir percent 30 conform-action set-cos-transmit 5 exceed-action drop
violate-action drop
class GOLD
bandwidth percent 5
set cos 3
Small Enterprise Design Profile Reference Guide
3-10
Chapter 3
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Service Deployed in the Small Enterprise Design
class SILVER
bandwidth percent 30
set cos 2
class class-default
bandwidth percent 35
set cos 0
!
The following configuration is of the parent policy, which in this design shapes to 20Mbps for each
remote site:
Policy Map Remote-1to50-Parent-Policy-Map  This is parent policy map
Class Remote_Site1
shape average 20000000 (bits/sec)
service-policy Remote-Child-Policy-Map  This is child policy map
After defining the policies, they are applied to WAN interfaces. The following example shows the
configuration of the Metro switch on its WAN interface:
interface GigabitEthernet1/1/1
description Connected to SP-MPLS-Core-cr24-6500-1
switchport trunk native vlan 801
switchport trunk allowed vlan 501-550
switchport mode trunk
logging event trunk-status
load-interval 30
carrier-delay msec 0
priority-queue out
mls qos trust dscp
spanning-tree portfast trunk
spanning-tree bpdufilter enable
spanning-tree guard root
max-reserved-bandwidth 100
service-policy output Remote-1to50-Parent-Policy-Map  The policy-map
hold-queue 2000 in
hold-queue 2000 out
After completing the QoS policy at main office, we need to define the QoS policy at remote sites. The
following subsection describes how to define QoS policy at the remote site.
QoS Policy at Remote Site
At the remote sites, the objectives are similar to the one at main site, which is to ensure that 20Mbps is
the aggregate traffic leaving out the switch and the ingress traffic is queued in 4 classes. However, the
implementation is different from the main office due to lack of hierarchical QoS support on the particular
switch deployed at the remote site. It is however possible to have a line card with the capability to do
HCBWFQ, if it is needed. The following configuration shows how to do QoS policy when the device
does not support HCBWFQ.
The first step is to queue the ingress traffic in the four queues. Table 3-2 shows the queues and the
bandwidth allocated.
Table 3-2
QoS Policies for Remote Site
Class
Queuing type
Bandwidth Allocation
REAL_TIME
Per-class
6mbps
GOLD
Per-class
1mbps
Small Enterprise Design Profile Reference Guide
3-11
Chapter 3
Small Enterprise Design Profile(SEDP)—WAN Design
WAN Service Deployed in the Small Enterprise Design
Table 3-2
QoS Policies for Remote Site
SILVER
Per-class
7mpbs
DEFAULT
Per-class
6mbps
The following is configuration of egress interface on the remote site:
interface GigabitEthernet1/1
description Connected to MetroE-Core-cr25-6500-1
switchport trunk encapsulation dot1q
switchport trunk native vlan 801
switchport trunk allowed vlan 501
switchport mode trunk
logging event link-status
load-interval 30
carrier-delay msec 0
qos trust dscp
udld port disable
tx-queue 1
bandwidth 1 mbps
tx-queue 2
bandwidth 7 mbps
tx-queue 3
bandwidth 6 mbps
priority high
tx-queue 4
bandwidth 6 mbps
no cdp enable
spanning-tree portfast trunk
spanning-tree bpdufilter enable
spanning-tree guard root
service-policy output WAN-EGRESS-PARENT
The above configuration ensures that each class of the traffic is queued as shown in Table 3-2. However,
the bandwidth only ensures the minimum amount of bandwidth available. It does not control the upper
threshold, which needs to be 20Mbs in our solution. Therefore, to ensure that the traffic on the egress
interface does not exceed 20Mps, WAN-EGRESS-PARENT policy is used to police the traffic to
20Mbps. The following is configuration of the WAN egress policy:
cr35-4507-SS1#show policy-map WAN-EGRESS-PARENT
Policy Map WAN-EGRESS-PARENT
Class class-default
police 20 mbps 1000 byte conform-action transmit exceed-action drop
service-policy WAN-EGRESS-CHILD
cr35-4507-SS1#
Small Enterprise Design Profile Reference Guide
3-12
CH A P T E R
4
Small Enterprise Design
Profile(SEDP)—Wireless LAN Design
Cisco Unified Wireless Network Architecture
WLANs have emerged as one of the most effective means for connecting to a network, given the mobility
of users. The Cisco Unified Wireless Network (CUWN) is a unified wired and wireless network solution
that addresses the wireless network security, deployment, management, and control aspects of deploying
a wireless network. It combines the best elements of wireless and wired networking to deliver secure,
scalable wireless networks with a low total cost of ownership.
Figure 4-1 shows a high-level topology of the CUWN architecture, which includes Lightweight Access
Point Protocol (LWAPP) access points (APs), mesh LWAPP APs (MAPs), the Wireless Control System
(WCS), and the Wireless LAN Controller (WLC). Alternate WLC platforms include the Wireless LAN
Controller Module (WLCM) or Wireless Services Module (WiSM). The Cisco Access Control Server
(ACS) and its Authentication, Authorization, and Accounting (AAA) features complete the solution by
providing RADIUS services in support of wireless user authentication and authorization.
Small Enterprise Design Profile Reference Guide
4-1
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Cisco Unified Wireless Network Architecture
Figure 4-1
Cisco Unified Wireless Network Architecture Overview
Cisco Catalyst
3750G Integrated
Cisco WCS
Wireless LAN
Navigator
Controller
Browser Based
W
Cisco Wireless
Control System
(WCS)
Cisco Wireless
LAN Controller
Module (WLCM)
Cisco
WCS
Cisco
WCS
Cisco Wireless
LAN Controller
N
S
E
Cisco
Mobile
Services
Engine
Third Party
Integrated
Applications:
E911, Asset
Tracking, ERP,
Workflow
Automation
Cisco Aironet
Wireless Bridge
Cisco Catalyst 6500
Series Wireless
Services Module
(WiSM)
Cisco Aironet
Lightweight
Access Points
(802.11a/b/g
and 802.11n)
Chokepoint
125 kHz
Cisco
Compatible
Client
Devices
Cisco Aironet
Wireless LAN
Client Adapters
Cisco Aironet
1500 Series
Lightweight
Outdoor Mesh
Access Points
225263
Cisco
Compatible
Wi-Fi Tags
The Cisco Unified Wireless Network is composed of two key elements: Wireless LAN Controllers and
Access Points (APs). These form the core of the Wireless LAN system, where the APs provide the radio
connection between wireless clients and the network, and the WLCs provide network.
Note
Figure 4-2 illustrates one of the primary features of the architecture: how LWAPP or Control and
Provisioning of Wireless Access Points (CAPWAP) access points use the LWAPP/CAPWAP protocol to
communicate with and tunnel traffic to a WLC.CUWN is migrating from the LWAPP protocol to
CAPWAP, and the WLC software version in the Small Enterprise Design Profile uses CAPWAP. The
Small Enterprise Design Profile Reference Guide
4-2
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Cisco Unified Wireless Network Architecture
fundamentals of the architecture and operation are the same. Documents discussing the LWAPP
architecture operation and behavior are still valid for CAPWAP, apart from the UDP port numbers. For
the purposes of this document and other documents referring to LWAPP, the Cisco CAPWAP
implementation can be considered as a superset of LWAPP features and behavior.
Figure 4-2
LAP and WLC Connection
AP
LWAPP
AP
LWAPP
Network
LWAP
P/CAP
WAP
WLC
PWAP
LWAPP/CA
AP
P
AP
LW
227453
LWAPP
P
WA
P
/CA
LWAPP/CAPWAP has three primary functions:
•
Control and management of the LAP
•
Tunneling of WLAN client traffic to the WLC
•
Collection of 802.11 data for the management of the Cisco Unified Wireless System
LWAPP Features
The easier a system is to deploy and manage, the easier it will be to manage the security associated with
that system. Early implementers of WLAN systems that used “fat” APs (autonomous or intelligent APs)
found that the implementation and configuration of such APs was the equivalent of deploying and
managing hundreds of individual firewalls, each requiring constant attention to ensure correct firmware,
configuration, and safeguarding. Even worse, APs are often deployed in physically unsecured areas
where theft of an AP could result in someone accessing its configuration to gain information to aid in
some other form of malicious activity.
LWAPP addresses deployment, configuration, and physical security issues by doing the following:
•
Removing direct user interaction and management of the AP. Instead, the AP is managed by the
WLC through its LWAPP connection. This moves the configuration and firmware functions to the
WLC, which can be further centralized through the use of the WCS.
•
Having the AP download its configuration from the WLC, and be automatically updated when
configuration changes occur on the WLC.
•
Having the AP synchronize its firmware with its WLC, ensuring that the AP is always running the
correct software version
•
Storing sensitive configuration data at the WLC, and storing only IP address information on the AP.
In this way, if the AP is physically compromised, there is no configuration information resident in
NVRAM that can be used to perform further malicious activity.
•
Mutually authenticating LAPs to WLCs, and AES encrypting the LWAPP control channel.
Small Enterprise Design Profile Reference Guide
4-3
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Cisco Unified Wireless Network Architecture
In addition to the improvements in physical security, firmware, and configuration management offered
by LWAPP, the tunneling of WLAN traffic in an LWAPP-based architecture improves the ease of
deployment without compromising the overall security of the solution. LAPs that support multiple
WLAN VLANs can be deployed on access-layer switches without requiring dot1q trunking or additional
client subnets at the access switches. All WLAN client traffic is tunneled to centralized locations (where
the WLC resides), making it simpler to implement enterprise-wide WLAN access and security policies.
Small Enterprise Design Profile
Figure 4-3 shows a simple schematic of the CUWN integration into the small enterprise design profile.
The key features of the CUWN integration is the use of a WLC at each location, with the management
function (WCS) located at the main site. If context-aware services are implemented, the Cisco Mobility
Services Engine (MSE) may be placed at the remote site; for smaller remote sites, an MSE at the main
site may provide a centralized service.
The standalone WLCs used in this design support AP capacities from 12 to 250 APs per WLC, and
multiple WLCs may be deployed at the same site if more than 250 APs are required or if a load sharing
or higher availability WLAN solution is required. An alternate higher availability solution is to use a
WLC at the main site as a backup WLC for the remote site’s WLCs. This is known as an N+1 solution,
where a main site WLC maintains sufficient capacity to support the APs of any individual remote site.
A similar principle to N+1 is used to provide high availability for the AAA service provided by the Cisco
ACS server. Each remote site will have a local ACS server to provide AAA services, and use the main
site ACS server as its secondary AAA server.
Small Enterprise Design Profile Reference Guide
4-4
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Figure 4-3
High level view of the CUWN Integration
ACS
MSE
W
N
S
WCS
E
WLC
MAN
Main Site
ACS
W
N
S
MSE
ACS
E
WLC
WLC
229316
Cisco Catalyst 3750 Remote Site
Cisco Catalyst 4500 Remote Site
Management
Each WLCs has both a CLI and web interface to provide WLAN configuration and management features,
but for a complete lifecycle management solution, the Cisco Wireless Control System (WCS) is needed.
The WCS supports the delivery of high-performance applications and mission-critical solutions that
simplify business operations and improve productivity. This comprehensive platform scales to meet the
needs of small-, mid-, and large-scale wireless LANs across local, remote, national, and international
locations. The WCS provides IT managers immediate access to the tools they need, when they need
them, to more efficiently implement and maintain new or expanding WLANs—all from a centralized
location requiring minimal IT staffing. Operational costs are significantly reduced through the Cisco
WCS’s intuitive GUI, simplified ease-of-use, and built-in tools that deliver improved IT efficiency,
Small Enterprise Design Profile Reference Guide
4-5
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
lowered IT training costs, and minimized IT staffing requirements, even as the network grows. Cisco
WCS lowers operational costs by incorporating the full breadth of management requirements, from radio
frequency to controllers services, into a single unified platform.
The Cisco WCS scales to manage hundreds of Cisco wireless LAN controllers, which in turn can manage
thousands of Cisco Aironet® access points, including the next-generation Cisco Aironet 1140 and 1250
Series 802.11n access points. For large-scale indoor and outdoor deployments, Cisco WCS Navigator
can be included to simultaneously support up to 20 Cisco WCS platforms and 30,000 Cisco access
points. Adding mobility services such as context-aware software and adaptive wireless intrusion
prevention systems (wIPS) is simplified through Cisco WCS integration with the Cisco MSE.
Designing a wireless LAN that effectively supports business-critical data, voice, and video services is
simplified with the Cisco WCS suite of built-in planning and design tools. Figure 4-4 shows an example
of the simplified Wireless LAN Planning and Design Cisco WCS planning and design tools, simplify the
process of defining access-point placement and determining access-point coverage areas for standard
and irregularly shaped buildings. These tools give IT administrators clear visibility into the radio
frequency (RF) environment. They make it easier to visualize the ideal RF environment, anticipate future
coverage needs, and assess wireless LAN behavior. They help IT administrators reduce, and in many
cases eliminate, improper RF designs and coverage problems that can lead to end-user trouble tickets.
Specialized Cisco WCS planning tools enable real-time assessment of the WLAN's readiness to support
voice-over-WLAN (VoWLAN) and context-aware (location) services. VoWLAN services support single
and dual-mode Wi-Fi-enabled phones. Context-aware services use Cisco's patent pending “RF
fingerprinting” technology to locate, track, and manage Wi-Fi-enabled devices and their contextual
information in conjunction with Cisco MSE.
Figure 4-4
WCS Planning Tools
Getting the WLAN up and running quickly and cost-effectively to meet end-user needs is streamlined
with the broad array of Cisco WCS integrated configuration templates. These easy-to-use templates and
deployment tools help IT managers provision and configure the wireless LAN to expressly deliver the
services that their business requires. Figure 4-5 shows an example of the Flexible Deployment Tools and
Configuration Templates available through an easy-to-use interface, make it simple to apply common
configurations across one or more wireless LAN controllers, regardless of their location in the
network—whether on the same LAN as Cisco WCS, on separate routed subnets, or across a wide-area
connection. At the click of a button, IT administrators can streamline even the most complex controller
configurations, updates, and scheduling across the entire wireless network. Auto-provisioning access
points is just as simple, with easy-to-use templates that support customized configuration of single or
multiple access points.
Small Enterprise Design Profile Reference Guide
4-6
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Figure 4-5
WCS Deployment Templates
Cisco WCS is the ideal management platform for monitoring the entire WLAN to maintain robust
performance and deliver an optimal wireless experience to mobile end users. Cisco WCS centralized
interface makes it easy to access information where it is needed, when it is needed, on-demand or as
scheduled. Figure 4-6 shows an example of the customizable dashboard and easy-to-use web-based
interface. The Cisco WCS easy-to-use graphical displays contained within Cisco WCS serve as a starting
point for maintenance, security, troubleshooting, and future capacity planning activities. Quick access
to actionable data about healthy and unhealthy events occurring on the network is available from a
variety of entry points, making Cisco WCS vital to ongoing network operations. The ever-present alarm
summary in the Cisco WCS simplifies access to critical information, faults, and alarms based on their
severity. Detecting, locating, and containing unauthorized (rogue) devices is fully supported when
location services are enabled. Figure 4-7 shows an example of the ever-present alarm summary and
simplified rogue device detection and location capabilities found within Cisco WCS.
Figure 4-6
WCS Monitoring Dashboard
Small Enterprise Design Profile Reference Guide
4-7
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Figure 4-7
WCS Alarm Panels
The integrated workflow and expansive array of troubleshooting tools in the Cisco WCS help IT
administrators quickly identify, isolate, and resolve problems across all components of the Cisco Unified
Wireless Network. Cisco WCS supports rapid troubleshooting of any size WLAN with minimal IT
staffing. Figure 4-8 shows an example of the integrated workflows and troubleshooting tools found in
Cisco WCS. Cisco WCS makes it easy to quickly assess service disruptions, receive notices about
performance degradation, research resolutions, and take action to remedy non-optimal situations.
Integrated workflows support seamless linkage between all tools, alarms, alerts, searches, and reports
for all infrastructure components and client devices. A variety of tools work together to help IT
administrators understand the operational nuances occurring on the WLAN and discover non-optimal
events occurring outside baseline parameters such as client connection or roaming problems. The
ever-present search tool in Cisco WCS facilitates cross-network access to real-time and historic
information about devices and assets located anywhere in the wireless network. A built-in client
troubleshooting tool provides a step-by-step method to analyze problems for all client devices. Cisco
CleanAir supports finding, classifying, and correlating sources of interference from Wi-Fi and
non-Wi-Fi sources such as Bluetooth devices and cordless phones.
Figure 4-8
WCS Troubleshooting Tools
Small Enterprise Design Profile Reference Guide
4-8
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Cisco WCS includes customizable reporting that assists IT teams in more effectively managing,
maintaining, and evolving the wireless LAN to meet ongoing business and operations requirements.
Flexible reports provide access to the right data, at the right time, in a format to meet any requirement,
as illustrated in Figure 4-9. An extensive variety of reports is available to help IT managers stay on top
of network trends, maintain network control, audit operations, and quickly address changing business
and end-user requirements. Reports are customizable based on user-defined parameters. Detailed
analysis of what is going on, where and when in the network, as well as capacity planning, is simplified
by collecting data from several reports and analyzing trends to understand how the WLAN has changed
over time. Understanding WLAN trends makes it easier to plan for future enhancements and growth.
Figure 4-9
WCS Customizable Reports
Connection to the Small Enterprise Design Profile Network
Figure 4-10 and Figure 4-11 show the remote site switch to WLC physical connection in more detail. A
key feature of the WLC interface is its direct connection to the core distribution switch via a port-channel
interface. This uses multiple Gigabit Ethernet connections from the WLC to the core/distribution switch.
These Gigabit Ethernet connections are different line cards on switches or line card to ensure that a
single switch or line card failure does not result in the loss of the WLC connection to the remote site
network. The switch feature to achieve this is the same switch feature used for the Ether Channel
connections between switches in the Small Enterprise Design Profile. The WLC feature is called link
aggregation (LAG). LAG is disabled by default on the WLC and requires a WLC reboot to be enabled.
This allows the WLC to use the same port channel configuration as the access switches when connecting
to the core/distribution switch.
Small Enterprise Design Profile Reference Guide
4-9
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Figure 4-10
4500 Site Switch WLC Physical Connection
ACS
W
N
S
E
MSE
227461
WLC
Figure 4-11
3750 Site Switch WLC Physical Connection
ACS
227462
WLC
The WLC connects to the switch via a 802.1Q trunk connection, as shown in Figure 4-12. Multiple SVIs
need to be configured on the switch to support the CUWN implementation. The key SVIs are an SVI for
the management and AP manager interface of the WLC, and the SVIs for each of the different WLANs
implemented on the WLC. There is not always a one-to-one relationship between SVIs and WLANs, but
in most simple WLAN deployments this is the case.
Switch WLC Layer-2 Connection
Trunk
227463
Figure 4-12
Figure 4-13 shows an example of the interface configuration summary on the remote site WLC. The key
interfaces of interest are ap-manager, manager, and wlan data1, wlan data2, and wlan voice1 interfaces.
Small Enterprise Design Profile Reference Guide
4-10
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
The server port is an out-of-band management interface not used in this design guide. The virtual
interface and its interface address are used to assist in the provisioning of seamless mobility. The virtual
interface is assigned an address during the initial configuration of the WLC and this address is typically
1.1.1.1 for all controllers.
Figure 4-13
WLC Interface Example
Figure 4-14 shows the mapping of a particular WLAN SSID to a defined interface. A WLAN can be
mapped to the management interface (this is normally not recommended), or any dynamic interface.
Figure 4-14
WLAN Example
Small Enterprise Design Profile Reference Guide
4-11
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
RF Groups and Mobility Groups
Part of a WLCs role is to manage the RF network in its area, and to provide mobility services to WLCs
in its network. To define the area of the RF network that you are interested in managing, use an RF
group name. To define the mobility services domain, use a mobility group. The details of RF groups
and mobility groups are beyond the scope of this design guide, but the key point for the design is that
the RF network area and the mobility services domain will typically be a single remote site, and only
WLCs that are at the same site should have the same RF group name or mobility group name.
Figure 4-15 shows an example of the RF and mobility group configuration on the controllers. Each
remote site can be given a different RF group and mobility groups, since the WLCs are in different
remote sites and are not expected to be in the same RF group or mobility group.
Figure 4-15
Mobility Groups and RF Groups Example
A remote site with only one WLC will have a mobility group with only its own details in the mobility
group. If there is more than one WLC at the remote site, then the mobility group configuration will
contain both WLCs.
Figure 4-16 shows a single WLAN example and Figure 4-17 shows a multiple WLC example. If there is
only one WLC, the mobility group information is automatically populated. Additional WLCs must have
the MAC address and management IP address added manually.
Small Enterprise Design Profile Reference Guide
4-12
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Figure 4-16
Mobility Groups for a Single WLC
Figure 4-17
Mobility Groups for a Multiple WLCs
Example WLAN Configurations
In a typical remote site WLAN environment, it is expected that there be multiple WLANs (SSIDs)
serving different purposes and different client groups. This section addresses the examples of what
would be considered typical WLAN examples.
•
A secured data WLAN network that uses 802.1X/EAP to provide AAA functionality and
dynamically generated per-user, per-session encryption key.
•
A secured VoWLAN network that also uses 802.1X/EAP to provide AAA functionality and
optimized for voice.
Small Enterprise Design Profile Reference Guide
4-13
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
•
An open unencrypted WLAN for access to a WLAN network for unmanaged clients such as laptops,
iPods, and iPhones.
For ease of administration and support for users who visit multiple sites, the WLAN SSIDs should be
the same for each site in the enterprise. In addition, the SSIDs should be broadcast and have meaningful
names.
Secured Employee WLAN
Figure 4-18 shows the general WLAN configuration tab for the secured data WLAN network. The key
point shown are the security policy that has been set under the security tab and the WLC interface that
the WLAN has been mapped to. The security configuration recommended is to use WPA2 with
802.1X+CCKM. Most WLANs should now support WPA2, and CCKM has been added to 802.1X as it
provides a faster roaming for WLAN clients. This is for clients that support CCKM, while using the
AAA features of 802.1X/AP to secure the WLAN connection.
Figure 4-18
General Configuration for Secured WLAN
Figure 4-18 shows the QoS configuration for the secured data WLAN; in this case, the QoS profile is set
to Silver, which is best effort setting. The WMM policy is set to disabled, as disabled WMM is the
equivalent of best effort. The primary role of WMM is to give higher priority to voice and video traffic
over the WLAN. Unless the site is planning to deliver interactive voice and video applications to their
WLAN data clients, WMM can remain disabled.
Note
802.11n standard requires WMM be enabled and, therefore, WMM must be enabled on all WLANs in
the 802.11n deployments. In this case, the WMM policy would be set to allowed.
Small Enterprise Design Profile Reference Guide
4-14
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Figure 4-19
Secured Employee WLAN QoS
Figure 4-20 shows the secured data WLAN advanced configuration. The only change from the default
settings on the tab is enabling the DHCP address assignment required feature. Typically, WLAN mobile
clients use DHCP, and any statically configured client runs the risk of introducing an address duplication
issue.
Figure 4-20
Secured Employee Advanced Configuration
Secured VoWLAN
Figure 4-21 shows the General Tab of the voice over WLAN (VoWLAN). The primary difference
between this WLAN and the secured data WLAN is that the security policy is WPA with CCKM, because
this is the optimum security configuration for the Cisco 7921G and 7925G. The other difference is that
the radio policy has been set for 802.11a only.
The use of 802.11a for the VoWLAN will depend on a number of factors, but the Cisco 7921G and
7925G are dual-band phones, and can use both bands but do not roam between bands. This means that
once the handset associates with a network in one band, it will not leave that band while call quality is
maintained. Keeping the VoWLAN handsets in the 802.11a band will ensure that the 2.4GHz band
remains available for other client devices. Whether this is a viable option for a site depends on the
required call capacity of the site’s WLAN and the type of AP network that has been deployed.
Small Enterprise Design Profile Reference Guide
4-15
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Figure 4-21
VoWLAN General Configuration
Figure 4-22 shows the QoS Tab for the VoWLAN. In this WLAN configuration, WMM is required.
Both the 7921G and 7921G support WMM, and WMM will give voice traffic priority over other
WLAN traffic on the network. The QoS profile is set to Platinum to ensure that the QoS classification
is appropriate for voice. The QoS profile controls the maximum classification value for both the
WLAN frames and LWAPP packets.
Figure 4-22
VoWLAN QoS Configuration
The Advanced Tab for the VoWLAN is the secured data WLAN. There is an option for VoIP snooping
and reporting, but this option pertains only to a particular type of SIP and is not applicable to the Cisco
7921G and 7925G handsets.
To protect VoIP call quality, the WLC can perform call admission control (CAC) to prevent VoWLAN
calls being added to an access point that cannot take any additional VoWLAN calls without
compromising call quality. An example of the CAC configuration page is shown in Figure 4-23.
Note
There is a separate CAC page for each RF band.
Small Enterprise Design Profile Reference Guide
4-16
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Figure 4-23
VoWLAN Call Admission Control
The CUWN prioritizes traffic based upon the QoS profiles applied to each WLAN, but it does not change
the IP QoS classification (DSCP) of the client traffic carried by the CUWN. This means that client traffic
that leaves the CUWN may need to be reclassified based upon the network policy. There are two ways
of achieving this.
1.
Applying policy at each of the network SVIs that connect the WLC to the network.
2.
Learning the QoS policy that was applied within the CUWN as this should be in alignment with the
network policy.
The second method is preferable as it requires less configuration and maintenance of the policy; the
policy only needs to be maintained upon WLCs, and not open the WLCs and the connected switch. To
achieve this, the Wired Protocol in the QoS profiles (Platinum, Gold, Sliver, and Bronze) must be set to
802.1p and all other settings may remain as default. This configures the WLC to set the 802.1p marking
of the frames sent from the WLC to reflect QoS policy on that WLAN. For example, the IP packet was
from a Platinum WLAN and had a DSCP value of EF, the WLC would use a CoS value of 5 in the frame
header. If the same packet had been on a Silver WLAN, the CoS value would be 0. Therefore, if the WLC
is connected to switch network that is configured to trust CoS and maintains a translation table between
CoS and DSCP for its network, the translation between CUWN policy and network policy will occur
automatically. See Figure 4-24.
For a further information on WLAN QoS, refer to the Voice over WLAN Design Guide at the following
URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Mobility/vowlan/41dg/vowlan41dg-book.html
Small Enterprise Design Profile Reference Guide
4-17
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Figure 4-24
Controller QoS Profiles
AP Deployments Considerations
As with any other WLAN deployment, the key design decisions are as follows: which areas require
coverage and what level of performance is required in those areas with WLAN coverage. The general
guidance for enterprise AP deployments is 15 to 20 active clients per AP. The number of APs required
depends on many factors, including the number of clients, the type of applications, and the expected
performance.
AP 1250
The Cisco 1250 Series is a rugged indoor access point designed for challenging RF environments that
require the versatility associated with external antennas, a rugged metal enclosure, and a broad operating
temperature range. The combined data rates of up to 600 Mbps to provide users with mobile access to
high-bandwidth data, voice, and video applications. 802.11n provides reliable and predictable WLAN
coverage to improve the end-user experience for both existing 802.11a/b/g clients and new 802.11n
clients.
AP 1140
The Cisco 1140 Series Access Point is a business-ready, 802.11n access point designed for simple
deployment and energy efficiency. The high-performance platform, which offers at least six times the
throughput of existing 802.11a/g networks, prepares the business for the next wave of mobile devices
and applications. Designed for sustainability, the Cisco 1140 Series delivers high performance from
standard 802.3af PoE while decreasing waste with multi-unit eco-packs and Energy Star certified power
supplies. As part of the CUWN, the Cisco 1140 Series provides the industry's lowest total cost of
ownership and investment protection by integrating seamlessly with the existing network.
Small Enterprise Design Profile Reference Guide
4-18
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Coverage and Site Surveys
The WLAN coverage requirements can be expected to vary from enterprise to enterprise depending upon
their goals and their budget. If the enterprise is simply to try to provide wireless network connectivity in
selected areas, then simple tactical placement of APs in the selected rooms is likely to be sufficient. If
the enterprise is planning to leverage the productivity associated with mobile applications and mobile
access, then a more strategic approach is required.
If the enterprise is planning to implement a mobility solution, they need to examine the expected
workflow and movement of the users of these applications to determine the range of coverage required
and perform a site survey based on these coverage requirements. If the customer is considering WLAN
location-based services as a possibility for future deployments, this should also be taken into account
during the site survey process as the density and placement of APs can be substantially different when
providing a suitable WLAN platform for location-based services.
Single-Band versus Dual-Band APs
There are both single-band and dual-band APs available for remote site solutions. The single-band APs
support the 2.4GHz band and the dual-band APs support both the 2.4GHz and 5GHz band. It is a general
recommendation that a dual-band solution be deployed.
Number of APs Per Room, Coverage in the Remote site
Single band APs vs Dual Band APs
If your goal is to simply provide WLAN coverage without trying to optimize capacity and performance,
then a single-band AP is an appropriate choice; however, in most cases, a dual-band AP is a better long
term choice.
The longevity of a WLAN deployment is fundamentally determined by its capacity. A quick look at the
dual-band deployment shows that it has twice the capacity of a single-band solution, but a deeper look
will reveal that the advantage of a dual-band solution is much greater than an additional radio.
The additional 5GHz radio, of a dual-band AP, is able to support a much higher capacity WLAN
network, as it has access to approximately 7 times the number of non-overlapping channels as does a the
2.4GHz AP radio. In almost all 2.4GHz deployments, APs reusing the three non-overlapping channels
interfere with each other and prevent the WLAN deployment from delivering a full WLAN capacity
increase when the number of APs is increased, realizing its full theoretical capacity. A 5GHz AP is 7
times more likely to be able to delivery additional capacity for the addition of an AP.
Another consideration in the single-band versus dual-band AP discussion is 802.11n performance.
802.11n uses two primary mechanism to provide data rate improvements over the existing 802.11g and
802.11a standards. The first mechanism changes in the modulation, and error correction that can provide
a data rate of up to 150Mbps, and the second mechanism is channel binding that combines
non-overlapping channels to deliver data rates that are multiples of what a single channel could achieve.
Channel binding is only available for the 5GHz band, as there is not sufficient channel capacity to
support it in an enterprise 2.4GHz deployment.
Deploying a dual-band WLAN system is not a matter of simply replacing the APs in place, the 5GHz
band has different power constraints, and has different propagation properties that need to be considered
when deciding on AP density and placement. If fiscally possible, a dual-band AP solution should be
planned and deployed initially. This will save an expensive rework layer.
For further discussion on 2.4GHz vs 5GHz capacity, refer to the Voice over WLAN Design Guide at the
following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Mobility/vowlan/41dg/vowlan41dg-book.html
Small Enterprise Design Profile Reference Guide
4-19
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Management
Client Considerations
One additional consideration in the single-band versus dual-band AP decision is the client devices that
the WLAN network is going to support. Many earlier laptops and mobile devices only supported the
2.4GHz band, and this is still true for many consumer WLAN clients. To take advantage of a dual-band
solution, a concerted effort needs to be made to ensure that as many clients as possible are also
dual-band. For cases where the remote site is purchasing WLAN clients, they should favor dual-band
devices. When recommending WLAN client devices, they should point out that the dual-band client
devices will have access to a higher performance network. Of course, the first step is to have the dual
band network in place, in order to for client devices to take advantage of their investment in a higher
performance clients.
WLC Discovery
CUWN provides auto-discovery functionality for its APs, where an AP upon connection to an
appropriately connected network can automatically find and connect to a WLC. The WLC will ensure
that the AP is running the appropriate software version, apply the appropriate configuration to that AP,
and adjust the radio settings to optimize the AP for its current environment.
Multiple auto-discovery options are available in the CUWN:
•
Over the air: The APs learns the IP address of WLCs from APs in the area which are currently
attached to those WLCs
•
DHCP: The APs learns the IP address(es) of the WLCs as part of its DHCP address assignment
•
DNS: The APs learn the IP address(es) of the WLCs by querying and well known DNS name
CISCO-LWAPP-CONTROLLER.<localdomain.com>
•
Staging: Have the AP join a WLC prior to them being deployed, and the APs will attempt to rejoin
this WLC when reconnected to the network
•
Static Configuration: The APs can be manually configured with the WLC IP address prior to being
connected to the networks
Given that the small enterprise design profile uses a local DNS server for sites to ensure survivability,
the use of the DNS discovery provides the simplest WLC discovery mechanism.
For details about how to configure DHCP discovery, refer to the DHCP OPTION 43 for Lightweight
Cisco Aironet Access Points Configuration Example at the following URL:
http://www.cisco.com/en/US/tech/tk722/tk809/technologies_configuration_example09186a00808714f
e.shtml
WLC Failover Options
CUWN provides multiple failover options, allowing APs to make a choice between WLCs based on
configured priorities. When an AP goes through its discovery process, it learns about all of the WLCs in
the mobility group, and can prioritize based on its high availability (HA) configuration or choose an
WLC based on current loads.
In network architectures, such as the small enterprise design profile, where there is a high-speed
WAN/MAN that makes AP failover to a remote WLC—such as the main site WLC—feasible, APs can
be configured to failover to a WLC outside their mobility group. In this scenario, the remote WLC would
not be in the Mobility Group that is learned during the AP discovery process, and the IP address of the
remote WLC needs to be provided in the HA configuration.
Small Enterprise Design Profile Reference Guide
4-20
Chapter 4
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
Appendix A—Devices and Software Used
This feature allows the main site to become a backup WLC for the remote sites in an event of an WLC
outage at the remote site. For this to be effective, a common WLAN SSID naming policy for key WLANs
needs to be implemented within the enterprise to ensure that WLAN clients do not have to be
reconfigured in the event of an AP failover to the main site WLC. This type of HA configuration is call
N+1 where a single main site WLC is able to provide HA at a much lower cost than a traditional 1+1
design which would require additional WLCs at each remote site. See Figure 4-25.
Figure 4-25
AP High Availability Configuration Example
Appendix A—Devices and Software Used
Table 4-1 lists the devices and software used for the CUWN in this design guide.
Table 4-1
WLAN Devices and Software
Name
Version
WCS
6.0.132
WLC 4402
6.0.182.0
WLC 4404
6.0.182.0
AP1252
AIR-LAP1252AG-A-K9
AP1142
AIR-LAP1142N-A-K9
Small Enterprise Design Profile Reference Guide
4-21
Chapter 4
Appendix A—Devices and Software Used
Small Enterprise Design Profile Reference Guide
4-22
Small Enterprise Design Profile(SEDP)—Wireless LAN Design
CH A P T E R
5
Small Enterprise Design
Profile(SEDP)—Security Design
This chapter describes how the Small Enterprise Design Profile sets the foundation for safe and secure
enterprise networks by leveraging the proven design and deployment guidelines of the Cisco SAFE
security architecture. The Small Enterprise Design Profile is a well-designed and validated network
architecture that enables the small enterprise to deliver all of the services required for an enhanced
business environment. Cisco SAFE is a security reference architecture that provides detailed design and
implementation guidelines for organizations looking to build highly secure and reliable networks.
Small Enterprise Network Security Design
The Small Enterprise Design Profile was designed with built-in security to protect the infrastructure and
to provide a safe environment for business. Following the proven guidelines of the Cisco SAFE security
architecture, a series of network security technologies and products are strategically deployed
throughout the network to protect employees and company assets, to guarantee the confidentiality of
sensitive data, and to ensure the availability and integrity of systems and data.
Small enterprises are not immune to violence, theft, vandalism, and other threats. The adoption of new
network collaboration and Internet-based technologies also opens the possibility for a number of cyber
threats. Understanding the nature and diversity of these threats, and how they may evolve over time, is
the first step towards a successful enterprise security strategy. The following are some of the common
threats to enterprise environments:
•
Service disruption—Disruption to the infrastructure, applications and other business resources
caused by botnets, worms, malware, adware, spyware, viruses, denial-of-service (DoS) attacks, and
Layer 2 attacks.
•
Network abuse—Use of non-approved applications by employees; peer-to-peer file sharing and
instant messaging abuse; and access to non-business related content.
•
Unauthorized access—Intrusions, unauthorized users, escalation of privileges, and unauthorized
access to restricted resources.
•
Data loss—Theft or leakage of private and confidential data from servers, endpoints, while in
transit, or as a result of spyware, malware, key-loggers, viruses, etc.
•
Identity theft and fraud—Theft of personnel identity or fraud on servers and end users through
phishing and E-mail spam.
Small Enterprise Design Profile Reference Guide
5-1
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
The Small Enterprise Design Profile is based on a validated network architecture designed around both
business operations and technical considerations. Recognizing the fact cost is a common limiting factor
to small enterprise network designs, the architecture topologies and platforms were carefully selected to
increase productivity while reducing overall costs.
The Small Enterprise Design Profile accommodates a main site and multiple remote sites of various
sizes, all interconnected over a Metro Ethernet core. The architecture design is illustrated in Figure 5-1.
At the heart of the architecture is a robust routing and switching network. Operating on top of this
network are the all the services used within the enterprise. This often includes the majority of the
business most critical applications such as databases, payroll, accounting and customer relationship
management (CRM) applications. The core of those services are deployed and managed at the main site,
allowing the enterprise to reduce the need for separate services to be operated and maintained at the
various remote locations. Centralized systems and applications are served by a main site serverfarm.
Figure 5-1
Small Enterprise Network Design
Internet Edge
Access
Serverfarm
IP
Video
Surveillance
Media Server
Secure
Mobility
NAC
Manager
Core
Distribution
Cisco ACS
Appliance
SRST/Video
Gateway
Cisco Unified
Communications
Manager
Cisco ASA
NAC Server
Internet
www
V
WLAN
Controller
Web
Security
M
Main Site
SensorBase
Cisco IPS
Email
Security
Cisco Security
Intelligence Operation
SP Managed
MetroE Core
NAC
Server
WLAN
Controller
WLAN
Controller
IP
229289
IP
NAC
Server
Remote Sites
As shown in Figure 5-1, the small enterprise network design follows a defense-in-depth approach, where
multiple layers of protection are built into the architecture. The different security tools are combined
together for enhanced visibility and control.
The security design of the architecture focuses on the following key areas:
•
Network Foundation Protection (NFP)
– Ensuring the availability and integrity of the network infrastructure, protecting the control and
management planes.
Small Enterprise Design Profile Reference Guide
5-2
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
•
Internet Perimeter Protection
– Ensuring safe Internet connectivity, and protecting internal resources and users from malware,
viruses, and other malicious software.
– Protecting personnel from harmful and inappropriate content.
– Enforcing E-mail and web browsing policies.
•
Serverfarm Protection
– Ensuring the availability and integrity of centralized applications and systems.
– Protecting the confidentiality and privacy of sensitive data.
•
Network Access Security and Control
– Securing the access edges. Enforcing authentication and role-based access for users residing at
the main site and remote offices.
– Ensuring systems are up-to-date and in compliance with the enterprise network security
policies.
•
Secure Mobility
– Providing secure, persistent connectivity to all mobile employees on laptops, smartphones and
other mobile platforms. Enforcing encryption, authentication and role-based access to all
mobile users.
– Delivering consistent protection to all mobile employees from viruses, malware, botnets, and
other malicious software.
– Ensuring a persistent enforcement of enterprise network security policies to all users. Making
sure systems comply with corporate policies and have up-to-date security.
The following sections discuss the key areas of the small enterprise network security design.
Network Foundation Protection
Small enterprise networks are built with routers, switches, and other network devices that keep the
applications and services running. Therefore, properly securing these network devices is critical for
continued operation.
The Small Enterprise Design Profile protects the network infrastructure by implementing the Cisco
SAFE best practices for the following areas:
•
Infrastructure device access
– Restrict management device access to authorized parties and for the authorized ports and
protocols.
– Enforce Authentication, Authorization and Accounting (AAA) with Terminal Access Controller
Access Control System (TACACS+) or Remote Authentication Dial-In User Service (RADIUS)
to authenticate access, authorize actions and log all administrative access.
– Display legal notification banners.
– Ensure confidentiality by using secure protocols like Secure Shell (SSH) and HTTPS.
– Enforce idle and session timeouts. Disable unused access lines.
•
Routing infrastructure
– Restrict routing protocol membership by enabling Message-Digest 5 (MD5) neighbor
authentication and disabling default interface membership.
Small Enterprise Design Profile Reference Guide
5-3
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
– Enforce route filters to ensure that only legitimate networks are advertised; and networks that
are not supposed to be propagated are never advertised.
– Log status changes of neighbor sessions to identify connectivity problems and DoS attempts on
routers.
•
Device resiliency and survivability
– Disable unnecessary services, implement control plane policing (CoPP).
– Enable traffic storm control.
– Implement topological, system and module redundancy for the resiliency and survivability of
routers and switches and to ensure network availability.
– Keep local device statistics.
•
Network telemetry
– Enable Network Time Protocol (NTP) time synchronization.
– Collect system status and event information with Simple Network Management Protocol
(SNMP), Syslog, TACACS+/RADIUS accounting.
– Monitor CPU and memory usage on critical systems.
– Enable NetFlow to monitor traffic patterns and flows.
•
Network policy enforcement
– Implement access edge filtering.
– Enforce IP spoofing protection with access control lists (ACLs), Unicast Reverse Path
Forwarding (uRPF), and IP Source Guard.
•
Switching infrastructure
– Implement a hierarchical design, segmenting the LAN into multiple IP subnets or virtual LANs
(VLANs) to reduce the size of broadcast domains.
– Protect the Spanning Tree Protocol (STP) domain with BPDU Guard and STP Root Guard.
– Use per-VLAN Spanning Tree to reduce the scope of possible damage.
– Disable VLAN dynamic trunk negotiation on user ports.
– Disable unused ports and put them into an unused VLAN.
– Implement Catalyst Infrastructure Security Features (CISF) including port security, dynamic
ARP inspection, DHCP snooping.
– Use a dedicated VLAN ID for all trunk ports.
– Explicitly configure trunking on infrastructure ports.
– Use all tagged mode for the native VLAN on trunks and drop untagged frames.
•
Network management
– Ensure the secure management of all devices and hosts within the enterprise network.
– Authenticate, authorize, and keep record of all administrative access.
– If possible, implement a separate out-of-band (OOB) management network (hardware or VLAN
based) to manage systems at the main site.
– Secure the OOB by enforcing access controls, using dedicated management interfaces or virtual
routing and forwarding (VRF) tables.
Small Enterprise Design Profile Reference Guide
5-4
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
– Provide secure in-band management access for systems residing at the remote sites by
deploying firewalls and ACLs to enforce access controls, using Network Address Translation
(NAT) to hide management addresses, and using secure protocols like SSH and HTTPS.
– Ensure time synchronization by using NTP.
– Secure servers and other endpoint with endpoint protection software and operating system (OS)
hardening best practices.
For more detailed information on the Network Foundation Protection best practices, refer the “Chapter
2, Network Foundation Protection” of the Cisco SAFE Reference Guide at the following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Security/SAFE_RG/chap2.html
Internet Perimeter Protection
The Small Enterprise Design Profile assumes the existence of a centralized Internet connection at the
headquarters or main site serving users at all locations. Common services include E-mail and web
browsing for employees, the hosting of a company’s website accessible to clients and partners over the
Internet, and secure remote access for mobile users and remote workers. Other services may also be
provided using the same infrastructure.
The network infrastructure that provides Internet connectivity is defined as the Internet perimeter,
illustrated in Figure 5-2.
Figure 5-2
Internet Perimeter
Access
Core
Distribution
Internet Edge
Internet
Border
Router
Cisco ASA
Internet
SensorBase
Cisco IPS
MetroE Core
www
Web
Security
Mail
Server
Email
Security
Web
Server
227629
SP Managed
MetroE Core
The primary functions of the Internet perimeter is to allow for safe and secure access for users residing
at all locations, and to provide public services without compromising the confidentiality, integrity and
availability of the company’s resources and data. To that end, the Internet perimeter incorporates the
following security functions:
Small Enterprise Design Profile Reference Guide
5-5
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
•
Internet Border Router—The Internet border router is the Internet gateway responsible for routing
traffic between the enterprise network and the Internet. The Internet border router may be
administered by the company’s personnel or may be managed by the Internet service provider (ISP).
The router provides the first line of protection against external threats and should be hardened
following the Network Foundation Protection (NFP) best practices.
•
Internet Firewall—A Cisco ASA provides stateful access control and deep packet inspection to
protect company resources and data from unauthorized access and disclosure. The security
appliance is configured to prevent incoming access from the Internet, to protect the enterprise
Internet public services, and to control user traffic bound to the Internet. In addition, the Cisco ASA
Botnet Traffic Filter feature can be enabled to defend the enterprise against botnet threats. Once
enabled, the Botnet Traffic Filter feature monitors network traffic across all ports and protocols for
rogue activity and to prevent infected internal endpoints from sending command and control traffic
back to external hosts on the Internet. The security appliance may also provide secure remote access
to employees with the Cisco AnyConnect Secure Mobility client.
•
Intrusion Prevention—An Advanced Inspection and Prevention Security Service Module (AIP
SSM) on the Cisco ASA or a separate IPS appliance can be implemented for enhanced threat
detection and mitigation. The IPS module or appliance is responsible for identifying and blocking
anomalous traffic and malicious packets recognized as well-known attacks. IPS can be deployed
either in inline or promiscuous mode. The module or appliance may be configured to participate in
Cisco IPS Global Correlation, allowing the IPS to gain visibility on global threats as they emerge in
the Internet, and to quickly react to contain them. IPS may also be configured to help block certain
Internet applications such as AOL Messenger, BitTorrent, Skype, and more.
•
Public Services DMZ— The company’s Internet website, mail server and other public-facing
services may be placed on a demilitarized zone (DMZ) for security and control purposes. The DMZ
acts as a middle stage between the Internet and company’s private resources, preventing external
users from directly accessing any internal servers and data. The Internet firewall is responsible for
restricting incoming access to the public services and by limiting outbound access from DMZ
resources out to the Internet. Systems residing on the DMZ are hardened with endpoint protection
software and operating system (OS) hardening best practices.
•
E-mail Security—A Cisco Ironport C Series E-mail Security Appliance (ESA) is deployed at the
DMZ to inspect incoming and outgoing E-mails and eliminate threats such as E-mail spam, viruses
and worms. The ESA appliance also offers E-mail encryption to ensure the confidentiality of
messages, and data loss prevention (DLP) to detect the inappropriate transport of sensitive
information.
•
Web Security—A Cisco IronPort S Series Web Security Appliance (WSA) is deployed at the
distribution switches to inspect HTTP and HTTPS traffic bound to the Internet. This system enforces
URL filtering policies to block access to websites containing non-business related content or that
are known to be the source of spyware, botnets or other type of malware. The WSA may also be
configured to block certain Internet applications such as AOL Messenger, BitTorrent, Skype, etc.
Following are the design guidelines for implementing the security functions.
Internet Border Router Security Guidelines
The Internet border router provides connectivity to the Internet through one or more Internet service
providers. The router act as the first line-of-defense against unauthorized access, DDoS, and other
external threats. Access control lists (ACLs), uRPF, and other filtering mechanisms may be implemented
for anti-spoofing and to block invalid packets. NetFlow, Syslog, SNMP may be used to gain visibility on
traffic flows, network activity and system status. In addition, the Internet border router should be secured
Small Enterprise Design Profile Reference Guide
5-6
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
following the practices explained in the “Network Foundation Protection” section on page 5-3. This
includes restricting and controlling administrative access, protecting the management and control
planes, and securing the dynamic exchange of routing information.
The “Internet Border Router Deployment” section on page 5-41provides an example of Internet edge
ACL. For more information on how to configure the Internet border router, refer to the “Chapter 6,
Enterprise Internet Edge” of the Cisco SAFE Reference Guide at the following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Security/SAFE_RG/chap6.html
Internet Firewall Guidelines
The Cisco ASA deployed at the Internet perimeter is responsible for protecting the enterprise internal
resources and data from external threats by preventing incoming access from the Internet; protecting
public resources served by the DMZ by restricting incoming access to the public services and by limiting
outbound access from DMZ resources out to the Internet; and controlling user's Internet-bound traffic.
To that end, the security appliance is configured to enforce access policies, keep track of connection
status, and inspect packet payloads following these guidelines:
•
Deny any connection attempts originating from the Internet to internal resources and subnets.
•
Allow outbound Internet access for users residing at any of the enterprise locations and for the
protocols permitted by the organization's policies (i.e., HTTP and HTTPS).
•
Allow outbound Internet SSL access for administrative updates, SensorBase, IPS signature updates,
etc.
•
Allow users access to DMZ services such as company’s website, E-mail, and domain name
resolution (HTTP, SMTP, POP, IMAP, and DNS).
•
Restrict inbound Internet access to the DMZ for the necessary protocols and servers (HTTP to web
server, SMTP to the mail transfer agent, DNS to DNS server, etc.).
•
Restrict connections initiated from DMZ to the only necessary protocols and sources (DNS from
DNS server, SMTP from the mail server, HTTP/SSL from Cisco IronPort ESA).
•
Enable stateful inspection for the used protocols to ensure returning traffic is dynamically allowed
by the firewall.
•
Implement Network Address Translation (NAT) and Port Address Translation (PAT) to shield the
internal address space from the Internet.
Figure 5-3 illustrates the protocols and ports explicitly allowed by the Cisco ASA.
Figure 5-3
Allowed Protocols and Ports
INSIDE
OUTSIDE
HTTP, HTTPS
DMZ
HTTP, HTTPS, DNS, SMTP
HTTP, HTTPS, DNS, SMTP
227422
HTTP, HTTPS, DNS, POP,
IMAP, SMTP
Small Enterprise Design Profile Reference Guide
5-7
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Note
Figure 5-3 does not include any management traffic destined to the firewall. Whenever available, a
dedicated management interface should be used. In case the firewall is managed in-band, identify the
protocols and ports required prior to configuring the firewall ACLs.
It is also important to remember that the Cisco ASA should be hardened following the Network
Foundation Protection best practices. This includes restricting and controlling administrative access,
securing the dynamic exchange of routing information with MD5 authentication, and enabling firewall
network telemetry with SNMP, Syslog, and NetFlow.
In the Small Enterprise Design Profile, high availability is achieved by using redundant physical
interfaces. This represents the most cost-effective solution for high availability. As an alternative, a pair
of firewall appliances could be deployed in stateful failover, as discussed in the “Internet Firewall
Deployment” section on page 5-42.
Cisco ASA Botnet Traffic Filter
The Cisco ASA Botnet Traffic Filter feature can be enabled to monitor network ports for rogue activity
and to prevent infected internal endpoints from sending command and control traffic back to an
external host on the Internet. The Botnet Traffic Filter on the ASA provides reputation-based control
for an IP address or domain name, similar to the control that Cisco IronPort SensorBase provides for
E-mail and web servers.
The Cisco Botnet Traffic Filter is integrated into all Cisco ASA appliances, and inspects traffic
traversing the appliance to detect rogue traffic in the network. When internal clients are infected with
malware and attempt to phone home to an external host on the Internet, the Botnet Traffic Filter alerts
the system administrator of this through the regular logging process and can be automatically blocked.
This is an effective way to combat botnets and other malware that share the same phone-home
communications pattern.
The Botnet Traffic Filter monitors all ports and performs a real-time lookup in its database of known
botnet IP addresses and domain names. Based on this investigation, the Botnet Traffic Filter determines
whether a connection attempt is benign and should be allowed, or is a risk and should be blocked.
The Cisco ASA Botnet Traffic Filter has three main components:
•
Dynamic and administrator blacklist data—The Botnet Traffic Filter uses a database of malicious
domain names and IP addresses that is provided by Cisco Security Intelligence Operations. This
database is maintained by Cisco Security Intelligence Operations and is downloaded dynamically
from an update server on the SensorBase network. Administrators can also configure their own local
blacklists and whitelists.
•
Traffic classification and reporting—Botnet Traffic Filter traffic classification is configured through
the dynamic-filter command on the ASA. The dynamic filter compares the source and destination
addresses of traffic against the IP addresses that have been discovered for the various lists available
(dynamic black, local white, local black), and logs and reports the hits against these lists
accordingly.
•
Domain Name System (DNS) snooping—To map IP addresses to domain names that are contained
in the dynamic database or local lists, the Botnet Traffic Filter uses DNS snooping in conjunction
with DNS inspection. Dynamic Filter DNS snooping looks at User Datagram Protocol (UDP) DNS
replies and builds a DNS reverse cache (DNSRC), which maps the IP addresses in those replies to
the domain names they match. DNS snooping is configured via the Modular Policy Framework
(MPF) policies
Small Enterprise Design Profile Reference Guide
5-8
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
The Botnet Traffic Filter uses two databases for known addresses. Both databases can be used together,
or the dynamic database can be disabled and the static database can be used alone. When using the
dynamic database, the Botnet Traffic Filter receives periodic updates from the Cisco update server on
the Cisco IronPort SensorBase network. This database lists thousands of known bad domain names and
IP addresses.
Note
The SensorBase network is an extensive network that monitors global E-mail and web traffic for
anomalies, viruses, malware, and other abnormal behavior. The network is composed of the Cisco
IronPort appliances, Cisco ASA, and Cisco IPS appliances and modules installed in more than 100,000
organizations worldwide, providing a large and diverse sample of Internet traffic patterns.
The Cisco ASA uses the dynamic database as follows:
1.
When the domain name in a DNS reply matches a name in the dynamic database, the Botnet Traffic
Filter adds the name and IP address to the DNS reverse lookup cache.
2.
When the infected host starts a connection to the IP address of the malware site, the Cisco ASA
sends a syslog message reporting the suspicious activity and optionally drops the traffic if the Cisco
ASA is configured to do so.
3.
In some cases, the IP address itself is supplied in the dynamic database, and the Botnet Traffic Filter
logs or drops any traffic to that IP address without having to inspect DNS requests.
The database files are stored in running memory rather than Flash memory. The database can be deleted
by disabling and purging the database through the configuration.
Note
To use the database, be sure to configure a domain name server for the Cisco ASA so that it can access
the URL of the update server. To use the domain names in the dynamic database, DNS packet inspection
with Botnet Traffic Filter snooping needs to be enabled; the Cisco ASA looks inside the DNS packets
for the domain name and associated IP address.
In addition to the dynamic database, a static database can be used by manually entering domain names
or IP addresses (host or subnet) that you want to tag as bad names in a blacklist. Static blacklist entries
are always designated with a Very High threat level. Domain names or IP addresses can also be entered
in a whitelist,
When a domain name is added to the static database, the Cisco ASA waits one minute, and then sends a
DNS request for that domain name and adds the domain name/IP address pairing to the DNS host cache.
This action is a background process, and does not affect your ability to continue configuring the Cisco
ASA. Cisco also recommends that DNS packet inspection be enabled with Botnet Traffic Filter
snooping. When enabled, the Cisco ASA uses Botnet Traffic Filter snooping instead of the regular DNS
lookup to resolve static blacklist domain names in the following cases:
•
The Cisco ASA DNS server is unavailable.
•
A connection is initiated during the one minute waiting period before the Cisco ASA sends the
regular DNS request.
If DNS snooping is used, when an infected host sends a DNS request for a name on the static database,
the Cisco ASA looks inside the DNS packets for the domain name and associated IP address and adds
the name and IP address to the DNS reverse lookup cache.
If Botnet Traffic Filter snooping is not enabled, and one of the above circumstances occurs, that traffic
is not monitored by the Botnet Traffic Filter.
Small Enterprise Design Profile Reference Guide
5-9
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Note
It is important to realize that a comprehensive security deployment should include Cisco Intrusion
Prevention Systems (IPS) with its reputation-based Global Correlation service and IPS signatures in
conjunction with the security services provided by the Cisco ASA security appliance such as Botnet
Traffic Filter.
For more information on the Cisco ASA Botnet Traffic Filter feature, see the following URL:
http://www.cisco.com/en/US/prod/vpndevc/ps6032/ps6094/ps6120/botnet_index.html
Intrusion Prevention Guidelines
IPS is responsible for identifying and blocking anomalous traffic and packets recognized as well-known
attacks. An AIP SSM module on the Cisco ASA Internet firewall or a separate IPS appliance can be
implemented in the Internet perimeter for enhanced threat detection and mitigation. IPS may also be
configured to help block certain Internet applications such as AOL Messenger, BitTorrent, Skype, and
so on.
Integrating IPS on a Cisco ASA appliance using an AIP SSM provides a cost-effective solution for small
enterprise networks. The AIP SSM is supported on Cisco ASA 5510 and higher platforms. The AIP SSM
runs advanced IPS software providing proactive, full-featured intrusion prevention services to stop
malicious traffic before it can affect the enterprise network.
The AIP SSM may be deployed in inline or promiscuous mode:
Inline mode—The AIP SSM is placed directly in the traffic flow (see the left side of Figure 5-4).
Traffic identified for IPS inspection cannot continue through the ASA without first passing through
and being inspected by the AIP SSM. This mode is the most secure because every packet that has
been identified for inspection is analyzed before being allowed through. Also, the AIP SSM can
implement a blocking policy on a packet-by-packet basis. This mode, however, can affect
throughput if not designed or sized appropriately.
•
Promiscuous mode—A duplicate stream of traffic is sent to the AIP SSM. This mode is less secure,
but has little impact on traffic throughput. Unlike inline mode, in promiscuous mode the AIP SSM
can block traffic only by instructing the ASA to shun the traffic or by resetting a connection on the
ASA. Also, while the AIP SSM is analyzing the traffic, a small amount of traffic might pass through
the ASA before the AIP SSM can shun it. The right side of Figure 5-4 shows the AIP SSM in
promiscuous mode.
IPS Inline and Promiscuous Modes
Inline Mode
Promiscuous Mode
IPS Inspection
IPS Inspection
VPN
Policy
Inside
Diverted
Traffic
Firewall
Policy
Outside
Small Enterprise Design Profile Reference Guide
5-10
VPN
Policy
Inside
Copied
Traffic
Firewall
Policy
Outside
227632
Figure 5-4
•
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
The recommended IPS deployment mode depends on the goals and policies of the enterprise. IPS inline
mode is more secure because of its ability to stop malicious traffic in real-time; however, it may impact
traffic throughput if not properly designed or sized. Conversely, IPS promiscuous mode has less impact
on traffic throughput but is less secure because there may be a delay in reacting to the malicious traffic.
Although the AIP SSM runs as a separate application within the Cisco ASA, it is integrated into the
traffic flow. The AIP SSM contains no external interfaces itself, except for the management interface on
the SSM itself. When traffic is identified for IPS inspection on the ASA, traffic flows through the ASA
and the AIP SSM in the following sequence:
Step 1
Traffic enters the adaptive security appliance.
Step 2
Firewall policies are applied.
Step 3
Traffic is sent to the AIP SSM over the backplane.
Step 4
The AIP SSM applies its security policy to the traffic and takes appropriate actions.
Step 5
Valid traffic (for inline mode only) is sent back to the adaptive security appliance over the backplane;
the AIP SSM might block some traffic according to its security policy and that traffic is not passed on.
Step 6
VPN policies are applied (if configured).
Step 7
Traffic exits the adaptive security appliance.
The AIP SSM card may be configured to fail open or close when the module becomes unavailable. When
configured to fail open, the ASA allows all traffic through, uninspected, if the AIP SSM becomes
unavailable. Conversely, when configured to fail close, the ASA blocks all traffic in case of an AIP SSM
failure.
Cisco IPS Global Correlation
If desired, the AIP SSM module on the Cisco ASA (or IPS appliance) may participate in Cisco IPS
Global Correlation for further threat visibility and control. Once enabled, the participating IPS sensor
receives threat updates from the Cisco SensorBase Network at regular intervals. The Cisco SensorBase
Network contains detailed information about known threats on the Internet, including serial attackers,
botnet harvesters, malware outbreaks, and dark nets. The IPS uses this information to filter out the worst
attackers before they have a chance to attack critical assets. It then incorporates the global threat data
into its system to detect and prevent malicious activity even earlier.
IPS Global Correlation is an important improvement in the basic functions of IPS because it enables the
system to understand the world in which it operates: an understanding of who the attacker is and whether
the attacker has a record of bad behavior. With Global Correlation, the sensor does not have to rely on
just the data in the packet or connection to make a decision about the intent of the activity and determine
whether the activity is malicious. Now, the sensor can look at a ping sweep and know that the source of
the ping sweep does not have a negative reputation, but later can look at another ping sweep and see that
the source is a known malicious site with a history of web attacks, and the sensor can block access to
and from that site. Global Correlation gives users greater confidence in the actions the sensor takes
because these actions are applied to attackers that have shown a predisposition for malicious behavior.
Global Correlation provides a process through which security data is collected for IP addresses and a
reputation score is developed for each IP address globally by Cisco. Cisco IPS 7.0 uses this reputation
data in two ways: for its reputation filters and for Global Correlation inspection.
Small Enterprise Design Profile Reference Guide
5-11
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
•
Reputation filters are used to block a subset of IP networks that are owned wholly by malicious
groups or were unused and have been hijacked. This first line of defense helps prevent malicious
contact ranging from spam to intelligence gathering in preparation for directed attacks. Reputation
filters also prevent attempts by botnets to phone home if the botnet controller machine resides in one
of these networks.
•
Global Correlation inspection uses reputation scores for normal IP addresses to increase the
percentage of attacks that the sensor can block. First, the sensor must detect some sort of malicious
activity and fire an event as a result. When an event is triggered, that event is processed to determine
whether the attacker’s IP address has a negative reputation and to what degree. If the event is sourced
from an attacker with a negative reputation, the sensor will add risk to the event, raising its risk
rating and making it more likely that the sensor will deny the event. This enables the sensor to deny
packets and attackers based on the fact that the event has a negative reputation in addition to a high
risk rating calculated on the sensor.
Once Global Correlation is configured, the IPS works in the following manner:
Step 1
When a packet enters the sensor, the first check is against the preprocessor, which performs Layer-2
packet normalization and atomic signature checks. Here, the packet header is processed to help ensure
that the packet is an IP packet, that the header is not incorrectly formed, and that the packet does not
match any atomic signatures.
Step 2
The packet is sent through the Cisco IPS reputation filters.
a.
Packets that match are discarded immediately, assuming that the reputation filters are enabled and
not in Audit mode.
b.
Packets that do not match go to the inspection engines, starting with the Layer 3 and 4 normalization
engine, then all the signature engines, and then anomaly detection.
Step 3
If any events are triggered, alerts are sent to the Global Correlation inspection processor, where the
source IP address for any alert is checked for negative reputation, and the risk rating is modified and
actions are added as appropriate.
Step 4
Any actions assigned to alerts are processed and acted upon, including event action overrides to add new
actions and event action filters to remove actions.
Reputation Filters
Cisco IPS reputation filters use a list of hundreds of networks that can be safely blocked because they
are not owned by any legitimate source. The reputation of the networks on this list can be considered to
be –10. This list includes only IP networks consisting entirely of stolen, “zombie” address blocks and
address blocks controlled entirely by malicious organizations. Individual IP addresses are never found
on this list. Because there is no way that a legitimate IP address can go from a positive or neutral
reputation and then, because of malicious activity, earn a place on the Cisco IPS reputation filter list,
users can confidently block all activity to and from networks on this list.
The primary purpose of the IPS reputation filters is to provide protection from direct scanning, botnet
harvesting, spamming, and distributed denial-of-service (DDoS) attacks originating from these
malicious address blocks, and from connections being attempted back to these networks from systems
already infected.
Note
There is currently no capability to view the networks on this list, but the networks that are being blocked
get logged by the sensor in the Statistics section of Cisco IPS Manager Express (IME).
Small Enterprise Design Profile Reference Guide
5-12
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
The only user configuration required for reputation filters is enabling or disabling them and specifying
whether Global Correlation is set to Audit mode (a global configuration setting for the entire sensor). In
Audit mode, the sensor will report potential deny actions due to reputation filters instead of actually
denying the activity.
Global Correlation Inspection
The primary activity of an IPS sensor is detection of malicious behavior. After the packet goes through
the IPS reputation filter process, the signature inspection occurs. This involves inspection of packets
flowing through the sensor by the various engines looking for the various types of malicious behavior.
Alerts that are created are passed to the Global Correlation inspection process for reputation lookups.
When an event occurs, the Global Correlation inspection process performs a local lookup of the source
(attacker) IP address of the event in its reputation database. This lookup process returns a value ranging
from –.1 to –10; the more negative the value, the more negative the reputation of the source IP address.
This reputation score is calculated for Cisco IPS sensors using the data in Cisco SensorBase and is sent
to the sensor as a reputation update. If an IP address returns no value for reputation, than it is considered
to be neutral. Cisco IPS, unlike E-mail and web security reputation applications, has no concept of
positive reputation. When an event is checked for reputation, this checking occurs entirely on the sensor
using data downloaded previously from Cisco SensorBase. Unlike other devices, the sensor will not send
a live request for information about an IP address that it has just seen. It will look in the data that it has,
and if it finds the address, it will use that data; otherwise, the sensor will assume that the address has a
neutral reputation.
Global Correlation inspection has three modes of primary operation: Permissive, Standard (default), and
Aggressive; you can also select Off:
•
Permissive mode tells the sensor to adjust the risk rating of an event, but not to assign separate
reputation only actions to the event.
•
Standard mode tells the sensor to adjust the risk rating and to add a Deny Packet action due to
reputation if the risk rating is greater than or equal to 86. It also adds a Deny Attacker action due to
reputation if the risk rating is greater than or equal to 100.
•
Aggressive mode also adjusts the risk rating due to reputation, adds a Deny Packet action due to
reputation if the risk rating is greater than or equal 83, and adds a Deny Attacker action due to
reputation if the risk rating is greater than or equal to 95.
•
Selecting Off in the Global Correlation Inspection window prevents the sensor from using updates
from Cisco SensorBase to adjust reputation.
If Global Correlation inspection is enabled and an event is generated by an attacker with a negative
reputation, the risk rating for the event will be elevated by a certain amount that is determined by a
statistical formula. The amount by which the risk rating is raised depends on the original risk rating of
the event and the reputation of the attacker.
Network Participation and Correlation Updates
The IPS sensor pulls reputation information for addresses on the global Internet from Cisco SensorBase.
When the sensor is configured initially, a DNS server needs to be configured for the sensor to use to
connect to Cisco SensorBase or an HTTP or HTTPS proxy (that has DNS configured) needs to be
configured. After the sensor has this information, the sensor will make an outbound connection to check
for the latest updates from Cisco SensorBase. It will initiate an HTTPS request to Cisco SensorBase
update servers and download a manifest that contains the latest versions of the files related to Global
Correlation. The sensor will check Cisco SensorBase every 5 minutes for updates. If changes are needed,
the sensor will perform a DNS lookup of the server name returned in the initial request. This lookup will
return the location of the server nearest to the sensor. The sensor will then initiates an HTTP connection
Small Enterprise Design Profile Reference Guide
5-13
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
that will actually transfer the data. The size of a full update is about 2 MB; incremental updates average
about 100 KB. If a sensor loses connection to Cisco SensorBase, Global Correlation information will
begin to timeout within days, and sensor health will change accordingly.
The other component of Global Correlation is network participation. This feature sends data from events
that the sensor fires back to Cisco SensorBase to adjust the reputation of IP addresses; this information
is then packaged in future reputation data downloads from Cisco SensorBase. The sensor passes this
information back to Cisco SensorBase according to the sensor configuration. The possible configuration
options are Off, Partial, and Full.
Note
•
With the Off (default) setting, the sensor will not send back any data. The sensor will still receive
reputation data, and this setting does not affect its use of that data except that the reputations of
addresses attacking the network being protected will not be influenced by their generation on the
sensor.
•
With the Partial setting, the sensor will send back alert information. This information consists of
protocol attributes such as the TCP maximum segment size and TCP options string, the signature ID
and risk rating of the event, the attacker IP address and port, and Cisco IPS performance and
deployment mode information.
•
The Full setting adds victim IP address and port information to the information reported with the
Partial setting.
No actual packet content information is sent to Cisco. In addition, events having RFC 1918 addresses,
because they are not unique, are not considered interesting. So all events reported to Cisco SensorBase
will have any such IP address information stripped from the reported data.
The mechanism used to update Cisco SensorBase with new attack information is fairly straightforward.
The sensor takes event information, parses out the important pieces of data, and buffers this data for
transmission back to Cisco SensorBase. It sends this data in the form of an HTTPS connection that it
initiates on average every 10 minutes. The average size of an update is 2 to 4 KB, with weekly averages
of about 0.5 to 1 MB. Some higher-volume sensors have average update sizes of about 50 KB, with
weekly totals in the 45-MB range. Sensors with very high alert volumes can have average update sizes
of about 850 KB, with weekly totals of up to 900 MB; these sensors, though, are at the extreme end of
the range.
For more information on IPS Global Correlation including configuration information, see the following
URL:
http://www.cisco.com/en/US/docs/security/ips/7.0/configuration/guide/cli/cli_collaboration.html.
E-mail Security Guidelines
The Small Enterprise Design Profile implements a Cisco Ironport C Series E-mail Security Appliance
(ESA) at the DMZ with the purpose of inspecting E-mails and eliminating threats such as E-mail spam,
viruses, and worms. The ESA can be described as a firewall and threat monitoring system for Simple
Mail Transfer Protocol (SMTP) traffic (TCP port 25). Logically speaking, the ESA acts as a Mail
Transfer Agent (MTA) within the E-mail delivery chain, as illustrated in Figure 5-5. Upon reception,
E-mails are evaluated using a reputation score mechanism based on the SensorBase network. The
SensorBase network is an extensive network that monitors global E-mail and web traffic for anomalies,
viruses, malware and other and abnormal behavior. The network is composed of Cisco IronPort
appliances, Cisco ASA and Cisco IPS appliances and modules installed in more than 100,000
organizations worldwide, providing a large and diverse sample of Internet traffic patterns. By leveraging
the SensorBase Network, messages originating from domain names or servers known to be the source of
spam or malware, and therefore with a low reputation score, are automatically dropped or quarantined
Small Enterprise Design Profile Reference Guide
5-14
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
by preconfigured reputation filters. Optionally, the enterprise may choose to implement some of the
other functions offered by the ESA appliance, including anti-virus protection with virus outbreak filters
and embedded anti-virus engines (Sophos and McAfee), encryption to ensure the confidentiality of
messages, and data loss prevention (DLP) for E-mail to detect the inappropriate transport of sensitive
information.
Note
Alternatively, Cisco offers managed hosted and hybrid hosted E-mail security services. These services
are provided through a dedicated E-mail infrastructure hosted in a network of Cisco data centers. For
more information, refer to http://www.cisco.com/go/designzone.
Figure 5-5
E-mail Delivery Chain
Internet
Firewall
Cisco IronPort ESA
Email Server
Firewall
Client
227423
Chapter 5
Note
Figure 5-5 shows a logical implementation of a DMZ hosting the E-mail server and ESA appliance. This
can be implemented physically by either using a single firewall or two firewalls in sandwich.
There are multiple deployment approaches for the security appliance depending on the number of
interfaces used (see Figure 5-6):
•
Dual-armed configuration—Two physical interfaces used to serve a public mail listener and a
private mail listener, each one configured with a separate logical IP address. The public listener
receives E-mail from the Internet and directs messages to the internal E-mail servers; while the
private listener receives E-mail from the internal servers and directs messages to the Internet. The
public listener interface may connect to the DMZ, while the private listener interface may connect
to the inside of the firewall.
Small Enterprise Design Profile Reference Guide
5-15
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
•
One-armed configuration—A single ESA interface configured with a single IP address and used for
both incoming and outgoing E-mail. A public mail listener is configured to receive and relay E-mail
on that interface. The best practice is to connect the ESA interface to the DMZ where the E-mail
server resides.
For simplicity, the Small Enterprise Design Profile implements the ESA appliance with a single
interface. In addition, using a single interface leaves other data interfaces available for redundancy.
Figure 5-6
Common ESA Deployments
Dual-Armed
Single-Armed
Internet
Outside
Internet
Outside
DMZ
DMZ
Inside
IP x.x.x.x
Public Listener
Inside
Public Listener
IP x.x.x.x
Private Listener
IP y.y.y.y
Email
Server
Figure 5-7 illustrates the logical location of the ESA within the E-mail flow chain.
Small Enterprise Design Profile Reference Guide
5-16
227424
Email
Server
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Figure 5-7
Typical Data Flow for Inbound E-mail Traffic
Mail
Server
DNS
Server
2
1
3
Internet
4
6
7
5
ESA
8
Mail
Server
9
227425
Chapter 5
The following steps explain what is taking place in Figure 7:
Step 1
Sender sends an E-mail to [email protected] X.
Step 2
What’s the IP address of domain X?
Step 3
It’s a.b.c.d (public IP address of ESA).
Step 4
E-mail server sends message to a.b.c.d using SMTP.
Step 5
Firewall permits incoming SMTP connection to the ESA, and translates its public IP address.
Step 6
ESA performs a DNS query on sender domain and checks the received IP address in its reputation
database, and drops, quarantines E-mail based on policy.
Step 7
ESA forwards E-mail to preconfigured inbound E-mail server.
Step 8
E-mail server stores E-mail for retrieval by receiver.
Step 9
Receiver retrieves E-mail from server using POP or IMAP.
The Cisco IronPort ESA appliance functions as a SMTP gateway, also known as a mail exchange (MX).
The following are the key deployment guidelines:
•
Ensure that the ESA appliance is both accessible via the public Internet and is the first hop in the
E-mail infrastructure. If you allow another MTA to sit at your network’s perimeter and handle all
external connections, then the ESA appliance will not be able to determine the sender’s IP address.
Small Enterprise Design Profile Reference Guide
5-17
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
The sender’s IP address is needed to identify and distinguish senders in the Mail Flow Monitor, to
query the SensorBase Reputation Service for the sender’s SensorBase Reputation Service Score
(SBRS), and to improve the efficacy of the anti-spam and virus outbreak filters features.
•
Features like Cisco IronPort Anti-Spam, Virus Outbreak Filters, McAfee Antivirus and Sophos
Anti-Virus require the ESA appliance to be registered in DNS. To that end, create an A record that
maps the appliance’s hostname to its public IP address, and an MX record that maps the public
domain to the appliance’s hostname. Specify a priority for the MX record to advertise the ESA
appliance as the primary (or backup during testing) MTA for the domain. A static address translation
entry needs to be defined for the ESA public IP address on the Internet firewall if NAT is configured.
•
Add to the Recipient Access Table (RAT) all the local domains for which the ESA appliance will
accept mail. Inbound E-mail destined to domains not listed in RAT will be rejected. External E-mail
servers connect directly to the ESA appliance to transmit E-mail for the local domains, and the ESA
appliance relays the mail to the appropriate groupware servers (for example, Exchange™,
Groupwise™, and Domino™) via SMTP routes.
•
For each private listener, configure the Host Access Table (HAT) to indicate the hosts that will be
allowed to send E-mails. The ESA appliance accepts outbound E-mail based on the settings of the
HAT table. Configuration includes the definition of Sender Groups associating groups or users, upon
which mail policies can be applied. Policies include Mail Flow Policies and Reputation Filtering.
Mail Flow Policies are a way of expressing a group of HAT parameters (access rule, followed by
rate limit parameters and custom SMTP codes and responses). Reputation Filtering allows the
classification of E-mail senders and to restrict E-mail access based on sender's trustworthiness as
determined by the IronPort SensorBase Reputation Service.
•
Define SMTP routes to direct E-mail to appropriate internal mail servers.
•
If an out-of-band (OOB) management network is available, use a separate interface for
administration.
A failure on the ESA appliance may cause service outage; therefore, a redundant design is
recommended. There are multiple ways to implement redundancy:
•
IronPort NIC Pairing—Redundancy at the network interface card level by teaming two of the
Ethernet interfaces on the ESA appliance. If the primary interface fails, the IP addresses and MAC
address are assumed by the secondary.
•
Multiple MTAs—Consists of adding a second ESA appliance or MTA with an equal cost secondary
MX record.
•
Load Balancer—A load balancer such as Cisco ACE Application Control Engine (ACE)
load-balances traffic across multiple ESA appliances.
IronPort NIC pairing is the most cost-effective solution (see Figure 5-8), because it does not require the
implementation of multiple ESA appliances and other hardware. It does not, however, provide
redundancy in case of chassis failure.
Figure 5-8
Cisco IronPort ESA NIC Pairing
Internet
Firewall
DMZ
IP address, Secondary NIC
Small Enterprise Design Profile Reference Guide
5-18
227426
IP address, Primary NIC
ESA
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Finally, the Internet firewall should be configured to accommodate traffic to and from the Cisco IronPort
ESA deployed at the DMZ. The protocols and ports to be allowed vary depending on the services
configured on the appliance. For details, refer to the Cisco IronPort User’s Guide at the following URL:
http://www.ironport.com/support/
The following are some of the most common services required:
•
Outbound SMTP (TCP/25) from ESA to any Internet destination
•
Inbound SMTP (TCP/25) to ESA from any Internet destination
•
Outbound HTTP (TCP/80) from ESA to downloads.ironport.com and updates.ironport.com
•
Outbound SSL (TCP/443) from ESA to updates-static.ironport.com and phonehome.senderbase.org
•
Inbound and Outbound DNS (TCP and UDP port 53)
•
Inbound IMAP (TCP/143), POP (TCP/110), SMTP (TCP/25) to E-mail server from any internal
client
Also remember that if the ESA is managed in-band, appropriate firewall rules need to be configured to
allow traffic such as SSH, NTP, and syslog.
For more information on how to configure the Cisco Ironport ESA, see the following guides:
•
Cisco SAFE Reference Guide—
http://www.cisco.com/en/US/docs/solutions/Enterprise/Security/SAFE_RG/SAFE_rg.html
•
Cisco IronPort ESA User Guide—http://www.ironport.com/support
Web Security Guidelines
The Small Enterprise Design Profile implements a Cisco IronPort S Series Web Security Appliance
(WSA) to block access to sites with non-business related content, and to protect the enterprise from
web-based malware and spyware.
Cisco IronPort WSA’s protection relies in two independent services:
Note
•
Web Proxy—This provides URL filtering, web reputation filters, and optionally anti-malware
services. The URL filtering capability defines the handling of each web transaction based on the
URL category of the HTTP requests. Leveraging the SensorBase network, the web reputation filters
analyze the web server behavior and characteristics to identify suspicious activity and protect
against URL-based malware. The anti-malware service leverages anti-malware scanning engines
such as Webroot and McAfee to monitor for malware activity.
•
Layer 4 Traffic Monitoring (L4TM)—Service configured to monitor all Layer-4 traffic for rogue
activity and to detect infected clients.
The SensorBase network is an extensive network that monitors global E-mail and web traffic for
anomalies, viruses, malware, and other abnormal behavior. The network is composed of the Cisco
IronPort appliances, Cisco ASA, and Cisco IPS appliances and modules installed in more than 100,000
organizations worldwide, providing a large and diverse sample of Internet traffic patterns.
As the small enterprise network design assumes a centralized Internet connection, the WSA is
implemented at the core/distribution layer of the main site network. This allows the inspection and
enforcement of web access policies to all users residing at any of the enterprise locations. Logically, the
WSA sits in the path between web users and the Internet, as shown in Figure 5-9.
Small Enterprise Design Profile Reference Guide
5-19
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Figure 5-9
Cisco IronPort WSA
Internet
Firewall
www
Cisco IronPort WSA
227427
Client
There are two deployment modes for the Web Proxy service:
•
Explicit Forward Proxy—Client applications, such as web browsers, are aware of the Web Proxy and
must be configured to point to the WSA. The web browsers can be either configured manually or by
using Proxy Auto Configuration (PAC) files. The manual configuration does not allow for
redundancy, while the use of PAC files allows the definition of multiple WSAs for redundancy and
load balancing. If supported by the browser, the Web Proxy Autodiscovery Protocol (WPAD) can be
used to automate the deployment of PAC files. WPAD allows the browser to determine the location
of the PAC file using DHCP and DNS lookups.
•
Transparent Proxy—Client applications are unaware of the Web Proxy and do not have to be
configured to connect to the proxy. This mode requires the implementation of a Web Cache
Communications Protocol (WCCP) enabled device or a Layer-4 load balancer in order to intercept
and redirect traffic to the WSA. Both deployment options provide for redundancy and load
balancing.
Explicit forward proxy mode requires the enterprise to have control over the configuration of the
endpoints, which may not be always possible. For example, the enterprise may allow the use of personal
laptops, smart-phones and other devices outside the company’s administration. Transparent proxy mode,
on the other hand, provides a transparent integration of WSA without requiring any configuration control
over the endpoints. It also eliminates the possibility of users reconfiguring their web browsers to bypass
the appliance without knowledge of the administrators. For these reasons, the Small Enterprise Design
Profile implements transparent proxy with WCCP. In this configuration, the Cisco ASA at the Internet
perimeter is leveraged as a WCCP server while the WSA act as a WCCP Traffic Processing Entity.
The Cisco ASA uses WCCP version 2, which has a built-in failover and load balancing mechanism. Per
WCCPv2 specification, multiple appliances (up to 32 entities) can be configured as part of the same
service group. HTTP and HTTPS traffic is load-balanced across the active appliances based on source
and destination IP addresses. The server (Cisco ASA) monitors the availability of each appliance in the
group, and can identify appliance failures within 30 seconds. After failure, traffic is redirected across the
remaining active appliances. In the case where no appliances are active, WCCP takes the entire service
group offline and subsequent requests bypass redirection. In addition, WCCPv2 supports MD5
authentication for the communications between WCCP server and WSA appliances.
Small Enterprise Design Profile Reference Guide
5-20
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Note
In the event the entire service group fails, WCCP automatically bypasses redirection, allowing users to
browse the Internet without the Web controls. In case it is desired to handle a group failure by blocking
all traffic, an outbound ACL may be configured on the Cisco ASA outside interface to permit
HTTP/HTTPS traffic originated from the WSA appliance itself and to block any direct requests from
clients. The ACL may also have to be configured to permit HTTP/HTTPS access from IPS and other
systems requiring such access.
WCCPv2 supports Generic Route Encapsulation (GRE) and Layer-2-based redirection; however, the
Cisco ASA only supports GRE. In addition, WCCP is supported only on the ingress of an interface. The
only topology supported is one where both clients and WSA are reachable from the same interface, and
where the WSA can directly communicate with the clients without going through the Cisco ASA. For
these reasons, the WSA appliance is deployed at the inside segment of the Cisco ASA.
Figure 5-10 illustrates the how WCCP redirection works in conjunction with Cisco ASA.
Figure 5-10
WCCP Redirection
Website.com
Internet
4
2
3
www
1
WSA
5
227428
Chapter 5
The following steps describe what takes place in Figure 5-10:
Step 1
Client’s browser requests connection to http://website.com.
Step 2
Cisco ASA intercepts and redirects HTTP requests over GRE.
Step 3
If content not present in local cache, WSA performs a DNS query on destination domain and checks the
received IP address against URL and reputation rules, and allows/denies request accordantly.
Step 4
WSA fetches content from destination web site.
Small Enterprise Design Profile Reference Guide
5-21
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Step 5
Content is inspected and then delivered directly to the requesting client.
The WSA appliance may also be configured to control and block peer-to-peer file-sharing and Internet
applications such as AOL Messenger, BitTorrent, Skype, Kazaa, etc. The way WSA handles these
applications depends on the TCP port used for transport:
Note
•
Port 80—Applications that use HTTP tunneling on port 80 can be handled by enforcing access
policies within the web proxy configuration. Application access may be restricted based on
applications, URL categories, and objects. Applications are recognized and blocked based on their
user agent pattern, and by the use of regular expressions. The user may also specify categories of
URL to block, including the predefined chat and peer-to-peer categories. Custom URL categories
may also be defined. Peer-to-peer access may also be filtered based on object and MIME
Multipurpose Internet Mail Extensions (MIME) types.
•
Ports other than 80—Applications using ports other than 80 can be handled with the L4TM feature.
L4TM blocks access to a specific application by preventing access to the server or block of IP
addresses to which the client application must connect.
The Cisco IPS appliances and modules, and the Cisco ASA (using the modular policy framework), may
also be used to block peer-to-peer file sharing and Internet applications.
The following are the guidelines for implementing a Cisco IronPort WSA appliance with WCCP on a
Cisco ASA:
•
Deploy WSA on the inside of the firewall so that the WSA can communicate with the clients without
going through the firewall.
•
Implement MD5 authentication to protect the communications between the Cisco ASA and the
WSA(s).
•
Configure a redirect-list on the firewall to indicate what traffic needs to be redirected. Make sure the
WSA is always excluded from redirection.
•
Ingress ACL on the firewall takes precedence over WCCP redirection, so make sure the ingress ACL
is configured to allow HTTP and HTTPS traffic from clients and the WSA itself.
•
In an existing proxy environment, deploy the WSA downstream from the existing proxy servers
(closer to the clients).
•
Cisco ASA does not support WCCP IP source address spoofing, therefore any upstream
authentication or access controls based on client IP addresses are not supported. Without IP address
spoofing, requests originating from a client are sourced with the IP address of the Web Proxy, and
not the one of the client.
•
TCP intercept, authorization, URL filtering, inspect engines, and IPS features do not apply to
redirected flows of traffic served by the WSA cache. Content requested by the WSA is still subject
to all the configured features on the firewall.
•
Configure WSA access policies to block access to applications (AOL Messenger, Yahoo Messenger,
BitTorrent, Kazaa, etc.) and URL categories not allowed by the enterprise Internet access policies.
•
If an out-of-band (OOB) management network is available, use a separate interface for
administration.
Small Enterprise Design Profile Reference Guide
5-22
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Note
WCCP, firewall, and other stateful features usually require traffic symmetry, whereby traffic in both
directions should flow through the same stateful device. The Small Enterprise Design Profile is designed
with a single Internet path ensuring traffic symmetry. Care should be taken when implementing
active-active firewall pairs as they may introduce asymmetric paths.
The Layer-4 Traffic Monitor (L4TM) service is deployed independently from the Web Proxy
functionality, and its mission is to monitor network traffic for rogue activity and for any attempts to
bypass port 80. L4TM works by listening to all UDP and TCP traffic and by matching domain names
and IP addresses against entries in its own database tables to determine whether to allow incoming and
outgoing traffic. The L4TM internal database is continuously updated with matched results for IP
addresses and domain names. Additionally, the database table receives periodic updates from the
IronPort update server (https://update-manifests.ironport.com).
For more information on how to configure the L4TM feature on Cisco Ironport WSA, see the following
guides:
•
Cisco SAFE Reference Guide—
http://www.cisco.com/en/US/docs/solutions/Enterprise/Security/SAFE_RG/SAFE_rg.html
•
Cisco IronPort WSA User Guide—http://www.ironport.com/support
Configuration steps and examples are included in the “Web Security Deployment” section on page 5-60.
Serverfarm Protection
Small enterprise networks typically include a serverfarm at the main site that hosts the systems that serve
business applications and store the data accessible to internal users. The infrastructure supporting it may
include application servers, the storage media, routers, switches, load balancers, off-loaders, application
acceleration devices and other systems. In addition, they may also host foundational services as part of
the enterprise network such as identity and security services, unified communication services, mobility
services, video services, partner applications, and other services.
Depending on the size of the enterprise network, the serverfarm may be constructed following different
design models. Figure 5-11 illustrates a collapsed design, and the less common for small enterprise
networks, multi-tier design. In the collapsed design all services are hosted in a shared physical
server-farm, and high availability is achieved by using redundant processors and interfaces. Large
enterprises may implement a more scalable multi-tier design serverfarm with chassis redundancy.
Small Enterprise Design Profile Reference Guide
5-23
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Figure 5-11
Serverfarm Designs
Collapsed Serverfarm
Serverfarm
Multi-tier Serverfarm
Core
Distribution
Serverfarm Tier
Application Tier
Cisco
IPS
Web Tier
Core
Application
Servers
Cisco ASA
with IPS
Module
229402
Cisco
ASA
Independently from the design model adopted by the enterprise, the following are the primary security
guidelines for the serverfarm design:
•
Network Foundation Protection— All infrastructure equipment should be protected following the
Network Foundation Protection best practices described earlier in this document. This includes
restricting and controlling administrative access, protecting the management and control planes, and
securing the switching and routing planes.
•
Firewall—A stateful firewall may be deployed to limit access to only the necessary applications and
services, and for the intended users. The firewall should be configured to control and inspect both
traffic entering and leaving the server farm segments. The firewall may also be leveraged to ensure
the appropriate segregation between application layers or groups. In addition, the firewall’s deep
packet inspection may be used to mitigate DoS attacks and enforce protocol compliance.
•
Intrusion Prevention— An IPS module on the Cisco ASA or a separate IPS appliance may be
implemented for enhanced threat detection and mitigation. The IPS is responsible for identifying
and blocking anomalous traffic and packets recognized as well-known attacks. The Cisco IPS may
be configured either in inline or promiscuous mode. When deployed in inline mode, the Cisco IPS
is placed in the traffic path and is capable of stopping malicious traffic before it reaches the intended
target.
•
Service Isolation— Services and applications serving different group of users or under different
security requirements should be properly isolated. Isolation helps prevent data leakage and contain
possible compromises from spreading across different server farm groups. Logical isolation may be
achieved by separating applications and services in different VLANs and by assigning them into
different firewall interfaces (physical or logical). This is illustrated in Figure 5-12.
•
Switch Security—Private VLANs, port security, storm control and other switch security features
may be leveraged to mitigate spoofing, man-in-the-middle, denial-of-service and other
network-based attacks directed to the serverfarm applications and the switching infrastructure.
•
Endpoint Protection— Servers residing at the different layers should be protected with host-based
IPS or other endpoint security software.
Small Enterprise Design Profile Reference Guide
5-24
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Figure 5-12
Service Isolation
Serverfarms
Core
Distribution
Group 3
Group 2
Cisco ASA
229403
Group 1
SSL termination and inspection, Web Application Firewall (WAF), Application Control Engine (ACE),
and other solutions may be leveraged to complement the guidelines described above. For a more detailed
discussion of serverfarm security, refer to “Chapter 4, Intranet Data Center” of the Cisco SAFE
Reference Guide at the following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Security/SAFE_RG/chap4.html
Network Access Security and Control
Some of the most vulnerable points of the network are the access edges where users connect to the
network. With the proliferation of wireless networks, increased use of laptops and smart mobile devices,
the enterprise cannot simply rely on physical controls hoping to prevent unauthorized systems from
being plugged into the ports of the access switches. Protection should rather be embedded into the
network infrastructure, leveraging the native security features available in switches, routers, and WLAN
systems. Furthermore, the network infrastructure should also provide dynamic identity or role-based
access controls for all systems attempting to gain access.
Implementing role-based access controls for users and devices helps reduce the potential loss of sensitive
information by enabling enterprises to verify a user or device identity, privilege level, and security policy
compliance before granting network access. Security policy compliance could consist of requiring
antivirus software, OS updates or patches. Unauthorized or noncompliant devices can be placed in a
quarantine area where remediation can occur prior to gaining access to the network.
The Small Enterprise Design Profile achieves access security and control by leveraging the following
technologies:
•
Catalyst Integrated Security Features (CISF), wired
•
Cisco Unified Wireless Network (CUWN) Integrated Security Features, wireless
•
Cisco NAC Appliance, wired and wireless
•
Cisco Identity-Based Network Networking Services (IBNS), wired and wireless
Small Enterprise Design Profile Reference Guide
5-25
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Catalyst Integrated Security Features
Catalyst Integrated Security Features (CISF) is a set of native security features available on Cisco
Catalyst Switches and designed to protect the access infrastructure and users from spoofing,
man-in-the-middle, DoS and other network-based attacks. CISF includes features such as private
VLANs, port security, DHCP snooping, IP Source Guard, secure Address Resolution Protocol (ARP)
detection, and Dynamic ARP Inspection (DAI). CISF features are considered to be part of a security
baseline and should be deployed on all access ports.
•
Port Security—Mitigates MAC flooding and other Layer 2 CAM overflow attacks by restricting the
MAC addresses that are allowed to send traffic on a particular port. After Port Security is enabled
on a port, only packets with a permitted source MAC address are allowed to pass through the port.
A permitted MAC address is referred to as a secure MAC address.
•
DHCP snooping—Inspects and filters DHCP messages on a port to ensure DHCP server messages
come only from a trusted interface. Additionally, it builds and maintains a DHCP snooping binding
table that contains the MAC address, IP address, lease time, binding type, VLAN number, and
interface information corresponding to the local untrusted interfaces of a switch. This binding table
is used by the other CISF features.
•
IP Source Guard—Restricts IP traffic on a port based on DHCP or static IP address MAC bindings
to prevent IP spoofing attacks. IP address bindings are validated using information in the DHCP
Snooping binding table.
•
Dynamic ARP inspection—Validates that the source MAC and IP address in an ARP packet received
on an untrusted interface matches the source MAC and IP address registered on that interface (using
the DHCP snooping binding table) to prevent ARP spoofing and MITM attacks.
•
ARP rate limiting—Where an excessive rate of ARP request (which must be processed by network
hosts CPUs), and the switch responds with access restriction if this rate is exceeded.
•
Storm Control—Prevents broadcast and multicast storms by monitoring packets passing from an
interface to the switching bus and determines whether the packet is unicast, multicast, or broadcast.
The switch counts the number of packets of a specified type received within the 1-second time
interval and compares the measurement with a predefined suppression-level threshold. When the
suppression-level threshold is reached, the port blocks traffic until the traffic falls below the
threshold level.
Cisco Unified Wireless Network (CUWN) Integrated Security Features
The Cisco Unified Wireless Network adds to the 802.11 security standards by providing additional
security features. Some of these are the WLAN equivalent of CiSF features such as, Dynamic Host
Configuration Protocol (DHCP) and Address Resolution Protocol (ARP) protection, peer-to-peer
blocking, and access control list and firewall features. Additionally other more WLAN specific features
are provided, including Enhanced WLAN security options, wireless intrusion detection system (IDS),
client exclusion, rogue AP detection, management frame protection, dynamic radio frequency
management, and network IDS integration.
The Cisco Unified Wireless Network solutions are discussed in the Wireless and Network Security
Integration Solution Design Guide at the following URL:
http://www.cisco.com/en/US/solutions/ns340/ns414/ns742/ns820/landing_sec_wireless.html
Small Enterprise Design Profile Reference Guide
5-26
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Cisco NAC Appliance
The Cisco Network Admission Control (NAC) Appliance (formerly known as Cisco Clean Access) uses
the network infrastructure to enforce security policy compliance on all devices seeking to access network
computing resources. With Cisco NAC Appliance, network administrators can authenticate, authorize,
evaluate, and remediate wired, wireless, and remote users and their machines before network access. The
NAC Appliance identifies whether networked devices such as laptops, IP phones, or game consoles are
compliant with your network security policies, and can repair any vulnerability before permitting access
to the network.
The Cisco NAC Appliance provides the following benefits:
•
Recognizes users, their devices, and their roles in the network. This first step occurs at the point of
authentication, before malicious code can cause damage.
•
Evaluates whether machines are compliant with security policies. Security policies can include
specific anti-virus or anti-spyware software, OS updates, or patches. Cisco NAC Appliance supports
policies that vary by user type, device type, or operating system.
•
Enforces security policies by blocking, isolating, and repairing non-compliant machines.
•
Non-compliant machines are redirected to a quarantine network, where remediation occurs at the
discretion of the administrator.
The NAC solution provides the following four functions, as shown in Figure 5-13:
•
Authenticates and authorizes
•
Scans and evaluates
•
Quarantines and enforces
•
Updates and remediates
Figure 5-13
The Four Functions of the NAC Framework
Authenticate and Authorize
Scan and Evaluate
• Enforces authorization
policies and privilges
• Agent scan for required versions
of hotfixes, AV, etc.
• Supports multiple user roles
• Network scan for virus
and worm infections and
port vulnerbilities
Quarantine and Enforce
• Isolate non-compliant devices
from rest of network
• Network-based tools for
vulnerbility and threat
remediation
• Help-desk integration
227497
• MAC and IP-based quarantine
effective at a per-user level
Update and Remediate
For more details of the NAC Appliance solution, see the following URL:
http://www.cisco.com/go/nacappliance.
Small Enterprise Design Profile Reference Guide
5-27
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Cisco NAC Appliance is a network-centric integrated solution administered from the Cisco Clean
Access Manager (NAC Manager) web console and enforced through the Clean Access Server (NAC
Server) and (optionally) the Clean Access Agent or Cisco NAC Web Agent. Cisco NAC Appliance
checks client systems, enforces network requirements, distributes patches and antivirus software, and
quarantines vulnerable or infected clients for remediation before clients access the network. Cisco NAC
Appliance consists of the components shown in Figure 5-14.
Figure 5-14
NAC Appliance Components
Internet
Switch
L2
Router
L3
eth1
Firewall
eth0
LAN/Intranet
Clean Access
Server (CAS)
PCs with
Clean Access
Agent (CAA)
Clean Access
Manager (CAM)
Authentication sources
(LDAP, RADIUS, Kerberos,
WindowsNT)
Admin laptop
DNS
server
228544
Clean Access Manager
Web admin console
Clean Access Manager (CAM)
The Cisco CAM (also known as NAC Manager) is the administration server for Clean Access
deployment. The secure web console of the Clean Access Manager is the single point of management
for up to 20 Clean Access Servers in a deployment (or 40 CASs if installing a SuperCAM). For
Out-of-Band (OOB) deployment, the web admin console allows you to control switches and VLAN
assignment of user ports through the use of SNMP. In the Small Enterprise Design Profile, the CAM
would be located at the main site.
Clean Access Server (CAS)
The Cisco CAS (a.k.a NAC Server) is the enforcement server between the untrusted (managed) network
and the trusted network. The CAS enforces the policies you have defined in the CAM web admin
console, including network access privileges, authentication requirements, bandwidth restrictions, and
Small Enterprise Design Profile Reference Guide
5-28
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Clean Access system requirements. You can install a CAS either as a standalone appliance (like the
Cisco NAC-3300 Series) or as a network module (Cisco NME-NAC-K9) in a Cisco ISR chassis and
deploy it in-band (always inline with user traffic) or OOB (inline with user traffic only during
authentication/posture assessment).
The CAS can also be deployed in Layer 2 mode (users are Layer-2-adjacent to the CAS) or Layer 3 mode
(users are multiple Layer-3 hops away from the CAS). You can also deploy several CASs of varying
size/capacity to fit the needs of varying network segments. You can install Cisco NAC-3300 Series
appliances in your company headquarters core, for example to handle thousands of users and
simultaneously install one or more Cisco NAC network modules in ISR platforms to accommodate
smaller groups of users at a satellite office, for example.
In the Small Enterprise Design Profile, the CAS would be located at the main site and the remote
locations, and it would be used to provide Layer-2 or Layer-3 OOB authentication/posture assessment.
Clean Access Agent (CAA)
CAA is optional read-only agent that resides on Windows clients. It checks applications, files, services,
or registry keys to ensure that clients meets your specified network and software requirements prior to
gaining access to the network.
Note
There is no client firewall restriction with CAA posture assessment. The agent can check the client
registry, services, and applications even if a personal firewall is installed and running.
If NAC is implemented as part of the Small Enterprise Design Profile it is recommended that the CAA
be used.
Cisco NAC Web Agent
The Cisco NAC Web Agent provides temporal posture assessment for client machines. Users launch
the Cisco NAC Web Agent executable, which installs the Web Agent files in a temporary directory on
the client machine via ActiveX control or Java applet. When the user terminates the Web Agent
session, the Web Agent logs the user off of the network and their user ID disappears from the Online
Users list.
Clean Access Policy Updates
Regular updates of prepackaged policies/rules that can be used to check the up-to-date status of
operating systems, antivirus (AV), antispyware (AS), and other client software. It provides built-in
support for 24 AV vendors and 17 AS vendors.
NAC Appliance Modes and Positioning
NAC Appliance allows multiple deployment options and may be placed at different points in the
network. The modes of operation can be generally defined as follows:
•
Out-of-band (OOB) virtual gateway
•
OOB IP real gateway
•
In-band (IB) virtual gateway
•
IB real IP gateway
Small Enterprise Design Profile Reference Guide
5-29
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Out-of-Band Modes
Out-of-Band (OOB) deployments require user traffic to traverse through the NAC Appliance only during
authentication, posture assessment, and remediation. When a user is authenticated and passes all policy
checks, their traffic is switched normally through the network and bypasses the appliance. See
Figure 5-15.
Figure 5-15
Layer-2 OOB Topology
Clean Access VLAN 10
Manager (CAM)
VLAN 10
Clean Access
Server (CAS)
VLAN 99
VLAN 110
802.1q Trunk
VLANs 110 and 10
VLAN 110
Posture Assessment
Authenticated Access
228545
VLAN 10
To deploy the NAC Appliance in this manner, the client device must be directly connected to the network
via a Catalyst switch port. After the user is authenticated and passes posture assessment, the Clean
Access Manager (CAM) instructs the switch to map the user port from an unauthenticated VLAN (which
switches or routes user traffic to the NAC) to an authenticated (authorized) VLAN that offers full access
privileges. For example, as shown in Figure 5-15, the client PC is connected through VLAN 110 to the
NAC Clean Access Server for the authentication and posture assessment, and is moved to VLAN 10 once
it successfully completes the authentication and authorization, scan, and evaluation phases of the NAC
solution.
In-Band Modes
When the NAC Appliance is deployed in-band, all user traffic, both unauthenticated and authenticated,
passes through the NAC Appliance, which may be positioned logically or physically between end users
and the network(s) being protected. See Figure 5-16 for a logical in-band topology example and
Figure 5-17 for a physical in-band topology example.
Small Enterprise Design Profile Reference Guide
5-30
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Figure 5-16
In-Band Virtual Gateway Topology
Clean Access VLAN 10
Manager (CAM)
VLAN 10
Clean Access
Server (CAS)
VLAN 99
VLAN 110
VLAN 110
Posture Assessment
Authenticated Access
Figure 5-17
228546
VLAN 110
Physical In-Band Topology
Clean Access
Manager (CAM)
VLAN 10
VLAN 99
VLAN 10
Clean Access
Server (CAS)
VLAN 110
Posture Assessment
Authenticated Access
228547
Chapter 5
Small Enterprise Design Profile Reference Guide
5-31
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
In-Band Virtual Gateway
When the NAC Appliance is configured as a virtual gateway, it acts as a bridge between end users and
the default gateway (router) for the client subnet being managed. The following two bridging options are
supported by the NAC Appliance:
•
Transparent—For a given client VLAN, the NAC Appliance bridges traffic from its untrusted
interface to its trusted interface. Because the appliance is aware of “upper layer protocols”, by
default it blocks all traffic except for Bridge Protocol Data Unit (BPDU) frames (spanning tree) and
those protocols explicitly permitted in the “unauthorized” role; for example, DNS and DHCP. In
other words, it permits those protocols that are necessary for a client to connect to the network,
authenticate, undergo posture assessment, and remediation. This option is viable when the NAC
Appliance is positioned physically in-band between end users and the upstream network(s) being
protected, as shown in Figure 5-17.
•
VLAN mapping—This is similar in behavior to the transparent method except that rather than
bridging the same VLAN from the untrusted side to the trusted side of the appliance, two VLANs
are used. For example, Client VLAN 131 is defined for the untrusted interface of the NAC
Appliance. There is no routed interface or switched virtual interface (SVI) associated with VLAN
131. VLAN 31 is configured between the trusted interface of the NAC Appliance and the next-hop
router interface/SVI for the client subnet. A mapping rule is made in the NAC Appliance that
forwards packets arriving on VLAN 131 and forwards them out VLAN 31 by swapping VLAN tag
information. The process is reversed for packets returning to the client. Note that in this mode,
BPDUs are not passed from the untrusted-side VLANs to their trusted-side counterparts.
The VLAN mapping option is usually selected when the NAC Appliance is positioned logically in-band
between clients and the networks being protected. This is the bridging option that should be used if the
NAC Appliance is going to be deployed in the virtual gateway mode.
In-Band Real IP Gateway
When the NAC Appliance is configured as a “real” IP gateway, it behaves like a router and forwards
packets between its interfaces. In this scenario, one or more client VLAN/subnets reside behind the
untrusted interface. The NAC Appliance acts as a default gateway for all clients residing on those
networks. Conversely, a single VLAN/subnet is defined on the trusted interface, which represents the
path to the protected upstream network(s).
After successful client authentication and posture assessment, the NAC Appliance by default routes
traffic from the untrusted networks to the trusted interface, where it is then forwarded based on the
routing topology of the network.
The NAC Appliance is not currently able to support dynamic routing protocols. As such, static routes
must be configured within the trusted side of the Layer 3 network for each client subnet terminating on
or residing behind the untrusted interface. These static routes should reference, as a next hop, the IP
address of the trusted interface of the NAC.
If one or more Layer-3 hops exist between the untrusted NAC interface and the end-client subnets, static
routes to the client networks must be configured in the NAC Appliance. Likewise, a static default route
(0/0) is required within the downstream Layer 3 network (referencing the IP address of the untrusted
NAC interface) to facilitate default routing behavior from the client networks to the NAC Appliance.
Depending on the topology, multiple options exist to facilitate routing to and from the NAC Appliance,
including static routes, VRF-Lite, MPLS VPN, and other segmentation techniques. It is beyond the
scope of this design guide to examine all possible methods.
Small Enterprise Design Profile Reference Guide
5-32
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
In-Band Versus Out-of-Band
Table 5-1 summarizes different characteristics of each type of deployment.
Table 5-1
In-Band Versus Out-of-Band Deployment Characteristics
In-Band Deployment Characteristics
Out-of-Band Deployment Characteristics
The CAS is always inline with user traffic (both before
and following authentication, posture assessment and
remediation). Enforcement is achieved through being
inline with traffic.
The CAS is inline with user traffic only during the process of
authentication, assessment and remediation. Following that, user
traffic does not come to the CAS. Enforcement is achieved through
the use of SNMP to control switches and VLAN assignments to
ports.
The CAS can be used to securely control authenticated The CAS can control user traffic during the authentication,
and unauthenticated user traffic by using traffic policies assessment and remediation phase, but cannot do so
post-remediation since the traffic is out-of-band.
(based on port, protocol, subnet), bandwidth policies,
and so on.
Does not provide switch port level control.
Provides port-level control by assigning ports to specific VLANs
as necessary using SNMP.
In-band deployment is supported for wired and wireless OOB deployments support wired and wireless clients. Wireless
clients.
OOB requires a specific network topology.1
Cisco NAC Appliance In-Band deployment with
supported Cisco switches is compatible with 802.1x
1.
Cisco does not recommend using 802.1x in an OOB deployment,
as conflicts will likely exist between Cisco NAC Appliance OOB
and 802.1x to set the VLAN on the switch interfaces/ports.
OOB NAC deployments for wireless require the NAC server to be deployed in Layer 2 OOB virtual gateway (bridge)
mode, and the NAC server must be placed Layer 2-adjacent to the wireless LAN controller (WLC).
Out-of-Band Requirements
OOB implementation of Cisco NAC Appliance requires the switches and Wireless LAN Controllers to
be supported by the Cisco NAC Appliance software. All the switches tested as part of the development
of the Small Enterprise Design Profile, apart from the Cisco Catalyst 2975, are supported by the Cisco
NAC OOB, and the Wireless LAN Controllers are also supported by the NAC Appliance software used
in this design guide. If the Catalyst 2975 is to be used as an access switch with the Cisco NAC Appliance,
the NAC solution must be an in-band solution.
Note
To obtain the latest list of supported devices, check the latest version of the Cisco NAC Appliance-Clean
Access Manager Installation and Administration Guide at the following URL:
http://www.cisco.com/en/US/docs/security/nac/appliance/configuration_guide/45/cam/45cam-book.
html
Out-Of-Band, Layer 2 and Layer 3
The proposed design for the Small Enterprise Design Profile is an OOB design, in order to get the highest
possible performance and scalability for traffic that has passed through the authentication, posture
assessment, and remediation stages of NAC. The Small Enterprise Design Profile offers two different
access layer options, a Layer-2 access layer for smaller sites and a hybrid Layer-2/Layer-3 access layer
for larger sites. This means that either a Layer-2 OOB solution or a Layer-3 OOB NAC solution may be
deployed.
Small Enterprise Design Profile Reference Guide
5-33
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
NAC Appliance Deployment in Small Enterprise Networks
Within the Small Enterprise Design Profile, a Cisco NAC Appliance is deployed at each of the site types,
the headquarters or main site, and remote sites. A centralized CAM is deployed at the main site, likely
within the serverfarm. A CAS is deployed at the main site and each remote site and is directly connected
to the core/distribution layer at each of the locations.
The simple topology used in each site type means that a VLAN from an access layer to the untrusted
interface of the NAC Appliance is always available as a standard component of the design, and untrusted
traffic should never need to be tunneled to the CAS. This allows a common network configuration, to
support NAC at any of the enterprise sites, regardless of whether the client devices are using a Layer-2
or Layer-3 access model. As the client can use a Layer-2 connection to the untrusted interface of the NAS
in either Layer 2 or Layer 3 access mode (this requires a trunk between the Layer-3 access switch and
the core/distribution. One VLAN of the trunk would carry the untrusted VLAN, and the other VLAN the
IP routing for all other traffic), and the VLAN used once the client is trusted will be either be a Layer-2
access VLAN from the core/distribution switch or a Layer-3 access switch VLAN depending upon the
site requirements.
This is illustrated in Figure 5-18 and Figure 5-19. In Figure 5-19, there is a simple Layer-2 NAC OOB
connection where a client device upon initial connection to the network is given VLAN 264, which
connects them directly to the untrusted interface of the NAS. VLAN 264 on the untrusted interface is
mapped to VLAN 64 on the trusted interface within the NAC appliance, which allows the client to obtain
an IP address that belongs on VLAN 64. Upon successful completion of the NAC authentication and
validation functions, the access switch is instructed, via SNMP from the CAM, to change the client
VLAN to VLAN 64. Even though the client has changed Layer-2 VLANs, its Layer-3 network
connections are unchanged and the traffic from the client no longer passes through the NAC appliance.
Figure 5-18
Layer 2 OOB Topology
VLAN 264
L2 Untrusted
227499
L2 Trusted
VLAN 64
In Figure 5-19, the same processes are followed when the client is untrusted, but once the client has
successfully completed its NAC functions the access switch is instructed via SNMP to change the client
VLAN to VLAN 67—a subnet local to the access switch. As the Layer-3 information for the client has
changed, the switch is also instructed to “bounce” the client switch port to initiate a new DHCP request
for an IP address appropriate to VLAN 67.
Small Enterprise Design Profile Reference Guide
5-34
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Figure 5-19
Layer 3 OOB Topology
VLAN 264
L2 Untrusted
L2 Trusted
VLAN 64
227500
L3 Trusted
VLAN 67
Cisco Identity-Based Network Networking Services (IBNS)
The best and most secure solution to vulnerability at the access edge is to leverage the intelligence of
the network. The Cisco Identity-Based Network Networking Services solution (IBNS) is a set of Cisco
IOS software services designed to enable secure user and host access to enterprise networks powered by
Cisco Catalyst switches and wireless LANs. It provides standards-based network access control at the
access layer by using the 802.1X protocol to secure the physical ports where end users connect. 802.1X
is an IEEE standard for media-level (Layer 2) access control, offering the capability to permit or deny
network connectivity based on the identity of the end user or device. 802.1X is well-known as a way to
secure wireless network access. It is equally essential in securing wired network access.
IEEE 802.1X Protocol
The IEEE 802.1X protocol allows Cisco Catalyst switches to offer network access control at the port
level. Every port on the switch is individually enabled or disabled based on the identity of the user or
device connecting to it. When 802.1X is first enabled on a port, the switch automatically drops all traffic
received on that port. There is one exception to this rule. The only traffic a switch will accept is a request
to start 802.1X authentication. Only after the 802.1X authentication has successfully completed will the
switch accept any other kind of traffic on the port.
The high-level message exchange in Figure 5-20 illustrates how port-based access control works within
an identity-based system.
Small Enterprise Design Profile Reference Guide
5-35
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Figure 5-20
Port-Based Access Control
7 Switch Enables Port
Authentication
Server
Authenticator
1 EAPOL Start
2 Login Request
3 Login Response
4 Check with Policy DB
5 Policy DB Confirms
ID and Grants Access
227515
6 Policy DB Informs Switch
The following steps describe the port-based access control flow shown in Figure 5-20:
Step 1
A client, such as a laptop with an 802.1X supplicant, connects to an IEEE 802.1X-enabled network and
sends a start message to the LAN switch (the authenticator).
Step 2
When the start message is received, the LAN switch sends a login request to the client.
Step 3
The client replies with a login response.
Step 4
The switch forwards the response to the policy database (authentication server).
Step 5
The authentication server authenticates the user.
Step 6
After the user identity is confirmed, the policy database authorizes network access for the user and
informs the LAN switch.
Step 7
The LAN switch then enables the port connected to the client.
The user and device credentials are processed by an AAA server. The AAA server is able to reference
user or device policy profile information either internally, using the integrated user database, or
externally, using database sources such as Microsoft Active Directory, LDAP, Novell NDS or Oracle
databases. This enables the integration of the system into exiting user management structures and
schemes, thereby simplifying overall deployment.
802.1X and EAP
When authenticating users for the purposes of network access control, the system must provide user
and/or device identification using strong authentication technologies known to be secure and reliable.
IEEE 802.1X does not by itself dictate how this is achieved. Rather, the 802.1X protocol defines an
encapsulation for the transport of the Extensible Authentication Protocol (EAP) from the client to the
switch. The 802.1X encapsulation is sometimes referred to as EAP over LAN (EAPoL). The switch in
turn relays the EAP information to the authentication server using the RADIUS protocol (EAP over
RADIUS).
Small Enterprise Design Profile Reference Guide
5-36
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
EAP, which is defined by RFC 3748, is itself a framework---not a specific authentication method. EAP
provides a way for the client and the authentication server to negotiate an authentication method that
they both support. There are many EAP methods but the ones used more frequently for 802.1X wired
authentication include EAP-TLS, EAP-PEAP, and EAP-FAST.
Impacts of 802.1X on the Network
Before enabling 802.1X in the network, it is essential to review the default security posture of a port
enabled for 802.1X authentication: all traffic is dropped except 802.1X EAPoL packets. This is a
fundamental change from the traditional model in which the port is enabled and all traffic is allowed
from the moment that a device plugs into the port. Ports that were traditionally open will now be closed
by default. This is one of the cornerstones of the strong security and network access control provided by
802.1X. However, this change in the default network access model can have a profound impact on
network devices and applications. Understanding and providing for the impacts of this change will make
for a smooth deployment of 802.1X network access control.
Non-802.1X-Enabled Devices
802.1X must be enabled on both the host device and on the switch to which the device connects. If a
device without an 802.1X supplicant attempts to connect to a port that is enabled for 802.1X, it will be
subjected to the default security policy. The default security policy says that 802.1X authentication must
succeed before access to the network is granted. Therefore, by default, non-802.1X-capable devices
cannot get access to an 802.1X-protected network.
Although many devices increasingly support 802.1X, there will always be devices that require network
connectivity but do not and/or cannot support 802.1X. Examples of such devices include network
printers, badge readers, legacy servers, and PXE boot machines. Some provision must be made for these
devices.
Cisco provides two features to accommodate non-802.1X devices. These are MAC Authentication
Bypass (MAB) and the Guest VLAN. These features provide fallback mechanisms when there is no
802.1X supplicant. After 802.1X times out on a port, the port can move to an open state if MAB succeeds
or if the Guest VLAN is configured. Judicious application of either or both of these features will be
required for a successful 802.1X deployment.
Note
Network-specific testing will be required to determine the optimal values for 802.1X timers to
accommodate the various non-802.1X-capable devices on your network.
802.1X in Small Enterprise Networks
As mentioned above, one of the requirements for 802.1X authentication is the requirement for a
supplicant. This has typically been a challenge in enterprise environments with a wide range of devices
and limited or no management of many of these devices. In many enterprises this is still the case, and
this makes a company-wide 802.1X very challenging. At the same time there are pockets of an enterprise
network where 802.1X may be a good choice.
For example, 802.1X protected ports may be a good choice for the network ports in the company’s
headquarters or main site, because these locations are more likely to have managed PCs.
Small Enterprise Design Profile Reference Guide
5-37
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Other locations in the enterprise network still need protection, but user network access may be better
served by a NAC Appliance solution. For networks requiring role-based access control using posture
assessments to ensure security compliance, Cisco NAC Appliance should be considered. In addition,
network access ports in open areas such as lobbies and meeting rooms may use 802.1X or Cisco Clean
Access NAC to protect these ports.
When considering the 802.1X deployment, the following four main 802.1X authentication options need
to be considered:
•
Basic 802.1X Authentication—An 802.1X controlled port with an 802.1X client directly connected
•
IP Phone Ports—An IP Phone and an 802.1X controlled port with an 802.1X client connected to the
phone
•
MAC Auth By-Pass—Using the MAC address of the client to provide authentication and bypass the
802.1X authentication process. Printer and legacy device support are typical applications
•
Web Auth—Allowing a user to authenticate by entering username and passwords in a web page.
Legacy device support and guest access are typical deployment applications
For more information on the Cisco IBNS 802.1X network access solution, see the following URL:
http://www.cisco.com/go/ibns.
NAC 802.1X and CISF in Combination
The three key access security features discussed above have been discussed in isolation, but can be
combined. In particular, the CISF features should be considered “baseline” features that are applied on
all access ports, and either NAC or 802.1X maybe overlaid on top of the CISF configuration.
The Cisco Clean Access and 802.1X configuration are also compatible (although they are not often
combined in wired networks), the key consideration in combining the two is how to give the appearance
of a SSO for the end user. Both 802.1X and NAC require authentication, as 802.1X authenticates the
client initially, a mechanism of communicating the 802.1X authentication result to the Cisco Clean
Access system is required.
If the authenticating clients join an Windows Active Directory network, the Cisco Clean Access Active
Directory SSO feature allows the clients to authenticate to active directory once they have performed
their 802.1X authentication. The CAM, when a client is detected, checks Active Directory to see if the
client has authenticated; this allows a SSO experience for client devices that are using 802.1X and NAC.
Secure Mobility
Today workers use laptops, smartphones and other smart mobile devices to access information and
applications at anytime and from anywhere there is an Internet connection. While embracing a mobile
workforce clearly boosts productivity and makes the small enterprise more competitive, there are a
number of challenges that arise from the use of mobile technologies. To start with, workers often use the
same devices to access both business and personal information. Devices used outside the enterprise
onsite controls may potentially introduce viruses, worms, spyware and other type of malware as mobile
workers connect back to the corporate network. Confidential and proprietary information may also be
lost or stolen while mobile users connect outside the company premises. At the other hand, the great
variety in hardware types, operating systems, and applications represents a clear challenge to the
enforcement of security controls and policies. In order to continue to foster innovation, enable
productivity, and meet the needs of the mobile workforce, companies must adapt to the changing trends
in mobility. A viable solution is one that enables access for mobile workers while ensuring that the
Small Enterprise Design Profile Reference Guide
5-38
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
corporate data, assets and network remain secure. At the other hand, users want the flexibility of
choosing how, when, and where to access both personal and professional information to be productive
without being inconvenienced by security checks.
The Small Enterprise Design Profile provides persistent and secure mobile access by implementing the
Cisco AnyConnect Secure Mobility solution (see Figure 5-21). This solution delivers secure, persistent
connectivity to all mobile employees independently from the type of mobile device used. The Cisco
AnyConnect Secure Mobility solution also ensures a consistent enforcement of the network security
policies to all users, no matter when, where and how they connect to the network.
The Cisco AnyConnect Secure Mobility is a collection of features across multiple Cisco products that
extends control and security into borderless networks. The products that work together to provide
AnyConnect Secure Mobility are as follows:
•
Cisco IronPort Web Security appliance (WSA)
•
Cisco ASA 5500 series adaptive security appliance (ASA)
•
Cisco AnyConnect client
Cisco AnyConnect Secure Mobility addresses the challenges of a mobile workforce by offering the
following features:
•
Secure, persistent connectivity—Cisco AnyConnect (with the Cisco ASA at the headend) provides
the remote access connectivity portion of AnyConnect Secure Mobility. The connection is secure
because both the user and device must be authenticated and validated prior to being provided access
to the network. The connection is persistent because Cisco AnyConnect is typically configured to
be always-on even when roaming between networks. Although Cisco AnyConnect is always-on, it
is also flexible enough to apply different policies based on location, allowing users access to the
Internet in a “captive portal” situation, when users must accept terms of agreement before accessing
the Internet.
•
Persistent security and policy enforcement—The Web Security appliance applies context-aware
policies, including enforcing acceptable use policies and protection from malware for all users,
including mobile (remote) users. The WSA also accepts user authentication information from the
AnyConnect client, providing an automatic authentication step for the user to access web content.
Figure 5-21
Cisco AnyConnect Secure Mobility Solution
Combined Solution
End-to-End Seamless Security
Information Sharing
Between Cisco ASA and
Cisco WSA
News
Email
Social
Networking
Enterprise
SaaS
www
AnyConnect
ASA
Cisco Web
Security
Appliance
Cisco AD
229290
Chapter 5
Small Enterprise Design Profile Reference Guide
5-39
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Small Enterprise Network Security Design
Figure 5-21 illustrates the relationship between the various elements of the Cisco AnyConnect Secure
Mobility solution. Remote and mobile users use the Cisco AnyConnect Secure VPN client to establish
VPN sessions with the Cisco ASA appliance. The Cisco ASA sends web traffic to the WSA appliance
along with information identifying the user by IP address and user name. The WSA scans the traffic,
enforces acceptable use policies, and protects the user from security threats. The Cisco ASA returns all
traffic deemed safe and acceptable to the user.
All Internet traffic scanning is done by the WSA, not the client on the mobile device. This improves
overall performance by not burdening the mobile device, some of which have limited processing power.
In addition, by scanning Internet traffic on the network, the enterprise can more easily and quickly
update security updates and acceptable use policies since the enterprise does not have to wait days,
weeks, or even months to push the updates to the client. The WSA tracks the requests it receives and
applies policies configured for remote users to traffic received from remote users.
For complete details about the Cisco AnyConnect Secure Mobility solution, refer to the documentation
available at the following URL:
http://www.cisco.com/en/US/netsol/ns1049/index.html
Threats Mitigated
The success of the security tools and measures in place ultimately depends on the degree they enhance
visibility and control. Simply put, security can be defined as a function of visibility and control. Without
any visibility, it is difficult to enforce any control, and without any control it is hard to achieve an
adequate level of security. Therefore, the security tools selected in the enterprise network design were
carefully chosen not only to mitigate certain threats but also to increase the overall visibility and control.
Table 5-2 summarizes how the security tools and measures used in the Small Enterprise Design Profile
help mitigate certain threats, and how they contribute to increasing visibility and control. Please note the
table is provided for illustration purposes and it is not intended to include all possible security tools, and
threats.
Table 5-2
Security Measures of the Small Enterprise Design Profile
Service
Disruption
Network Foundation
Protection
Yes
Stateful Firewall
Yes
IPS
Yes
Mobile Security
Yes
Harmful
Content
Network
Abuse
Unauthorized
Access
Data Loss
Visibility
Control
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Web Security
Yes
Yes
Yes
Yes
Yes
Yes
E-mail Security
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Access Security and
Control
Small Enterprise Design Profile Reference Guide
5-40
Yes
Yes
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Network Security Deployment
This section describes the deployment best practices for the key security platforms and features used in
the Small Enterprise Design Profile. This includes deployment and setting guidelines, and configuration
examples.
Internet Border Router Deployment
Whether the Internet border router is managed by the enterprise or the ISP, it must be hardened following
the best practices listed in the “Network Foundation Protection” section on page 5-3. This includes
restricting and controlling administrative access, protecting the management and control planes, and
securing the dynamic exchange of routing information. In addition, the Internet border router may be
leveraged as the first layer of protection against outside threats. To that end, edge ACLs, uRPF, and other
filtering mechanisms may be implemented for anti-spoofing and to block invalid packets.
The following configuration snippet illustrates the structure of an edge ACL applied to the upstream
interface of the Internet border router. The ACL is designed to block invalid packets and to protect the
infrastructure IP addresses from the Internet. The configuration assumes the enterprise is assigned the
198.133.219.0/24 address block for its public-facing services, and that the upstream link is configured
in the 64.104.10.0/24 subnet.
! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!--- Module 1: Anti-spoofing Denies
!--- These ACEs deny fragments, RFC 1918 space,
!--- invalid source addresses, and spoofs of
!--- internal space (space as an external source).
!
!--- Deny fragments.
access-list 110 deny tcp any 198.133.219.0 0.0.0.255 fragments
access-list 110 deny udp any 198.133.219.0 0.0.0.255 fragments
access-list 110 deny icmp any 198.133.219.0 0.0.0.255 fragments
!--- Deny special-use address sources.
!--- See RFC 3330 for additional special-use addresses.
access-list 110 deny ip host 0.0.0.0 any
access-list 110 deny ip 127.0.0.0 0.255.255.255 any
access-list 110 deny ip 192.0.2.0 0.0.0.255 any
access-list 110 deny ip 224.0.0.0 31.255.255.255 any
!--- Filter RFC 1918 space.
access-list 110 deny ip 10.0.0.0 0.255.255.255 any
access-list 110 deny ip 172.16.0.0 0.15.255.255 any
access-list 110 deny ip 192.168.0.0 0.0.255.255 any
!--- Deny packets spoofing the enterprise public addresses
access-list 110 deny ip 198.133.219.0 0.0.0.255 any
!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!--- Module 2: Explicit Permit
!--- Permit only applications/protocols whose destination
!--- address is part of the infrastructure IP block.
!--- The source of the traffic should be known and authorized.
!
!--- Permit external BGP to peer 64.104.10.113
access-list 110 permit tcp host 64.104.10.114 host 64.104.10.113 eq bgp
access-list 110 permit tcp host 64.104.10.114 eq bgp host 64.104.10.113
!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!--- Module 3: Explicit Deny to Protect Infrastructure
access-list 110 deny ip 64.104.10.0 0.0.0.255 any
!
Small Enterprise Design Profile Reference Guide
5-41
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!--- Module 4: Explicit Permit for Traffic to Company’s Public
!--- Subnet.
access-list 110 permit ip any 198.133.219.0 0.0.0.255
!
Note
The 64.104.0.0/16 and 198.133.219.0/24 address blocks used in the examples in this chapter are reserved
for the exclusive use of Cisco Systems, Inc.
Internet Firewall Deployment
The mission of the Internet firewall is to protect the company’s internal resources and data from external
threats, secure the public services provided by the DMZ, and to control user’s traffic to the Internet. The
Small Enterprise Design Profile uses a Cisco ASA appliance as illustrated in Figure 5-22.
Figure 5-22
Core
Distribution
Internet Edge Firewall
Internet Edge
Cisco
Catalyst
4507
Cisco ASA
5520
Inside (100)
Outside (0)
Internet
Border
Router
Internet
Mail
Server
Email
Security
Web
Server
229291
DMZ (50)
The Cisco ASA is implemented with three interface groups, each one representing a distinct security
domain:
•
Inside—The interface connecting to the core/distribution switch that faces the interior of the
network where internal users and resources reside. The inside interface connects to the internal
trusted networks, therefore it is given the highest security level, 100.
•
Outside—Interface connecting to the Internet border router. The router may be managed either by
the enterprise or a service provider. The outside interface connects to the Internet, hence it’s given
the lowest security level, 0.
•
Demilitarized Zone (DMZ)—The DMZ hosts services that are accessible over the Internet. These
services may include the company’s website and E-mail services. The DMZ serves a medium level
security segment, therefore should be given any security value between the ones defined for the
inside and the outside interfaces, for example, 50.
Small Enterprise Design Profile Reference Guide
5-42
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
The Internet firewall acts as the primary gateway to the Internet; therefore, its deployment should be
carefully planned. The following are key aspects to be considered when implementing the firewall:
•
Firewall hardening and monitoring
•
Network Address Translation (NAT)
•
Firewall access policies
•
Botnet Traffic Filter
•
Firewall redundancy
•
Routing
Firewall Hardening and Monitoring
The Cisco ASA should be hardened in a similar fashion as the infrastructure routers and switches.
According to the Cisco SAFE security best practices, the following is a summary of the measures to be
taken:
•
Implement dedicated management interfaces to the OOB management network.
•
Present legal notification for all access attempts.
•
Use HTTPS and SSH for device access. Limit access to known IP addresses used for administrative
access.
•
Configure AAA for role-based access control and logging. Use a local fallback account in case AAA
server is unreachable.
•
Use NTP to synchronize the time.
•
Use syslog or SNMP to keep track of system status, traffic statistics, and device access information.
•
Authenticate routing neighbors and log neighbor changes.
•
Implement firewall access policies (explained in Firewall Access Policies).
The Cisco ASA 5510 and higher appliance models come with a dedicated management interface that
should be used whenever possible. Using a dedicated management interface keeps the management plane
of the firewall isolated from threats originating from the data plane. The management interface should
connect to the OOB management network, if one is available.
The following is an example of the configuration of a dedicated management interface.
interface Management0/0
nameif management
security-level 100
ip address 172.26.160.225 255.255.252.0
management-only
!
Note
Any physical interface or logical sub-interface can be configured as a management-only interface using
the management-only command.
Small Enterprise Design Profile Reference Guide
5-43
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
It is recommended that a legal notification banner is presented on all interactive sessions to ensure that
users are notified of the security policy being enforced and to which they are subject. The notification
banner should be written in consultation with your legal advisors.
The following example displays the banner after the user logs in:
banner motd UNAUTHORIZED ACCESS TO THIS DEVICE IS PROHIBITED.
banner motd You must have explicit, authorized permission to access or configure this
device.
banner motd Unauthorized attempts and actions to access or use this system may result in
civil and/or criminal penalties.
banner motd All activities performed on this device are logged and monitored.
Management access to the firewall should be restricted to SSH and HTTPS. SSH is needed for CLI
access and HTTPS is needed for the firewall GUI-based management tools such as CSM and ADSM.
Additionally, this access should only be permitted for users authorized to access the firewalls for
management purposes.
The following ASA configuration fragment illustrates the configuration needed to generate a 768 RSA
key pair and enabling SSH and HTTPS access for devices located in the management subnet.
! Generate RSA key pair with a key modulus of 768 bits
crypto key generate rsa modulus 768
! Save the RSA keys to persistent flash memory
write memory
! enable HTTPS
http server enable
! restrict HTTPS access to the firewall to permitted management stations
http <CSM/ADSM-IP-address> 255.255.255.255 management
! restrict SSH access to the firewall to well-known administrative systems
ssh <admin-host-IP-address> 255.255.255.255 management
! Configure a timeout value for SSH access to 5 minutes
ssh timeout 5
Administrative users accessing the firewalls for management must be authenticated, authorized, and
access should be logged using AAA. The following ASA configuration fragment illustrates the AAA
configurations needed to authenticate, authorize, and log user access to the firewall:
aaa-server tacacs-servers protocol tacacs+
reactivation-mode timed
aaa-server tacacs-servers host <ACS-Server>
key <secure-key>
aaa authentication ssh console tacacs-servers LOCAL
aaa authentication serial console tacacs-servers LOCAL
aaa authentication enable console tacacs-servers LOCAL
aaa authentication http console tacacs-servers LOCAL
aaa authorization command tacacs-servers LOCAL
aaa accounting ssh console tacacs-servers
aaa accounting serial console tacacs-servers
aaa accounting command tacacs-servers
aaa accounting enable console tacacs-servers
aaa authorization exec authentication-server
! define local username and password for local authentication fallback
username admin password <secure-password> encrypted privilege 15
As with the other infrastructure devices in the network, it is important to synchronize the time on the
firewall protecting the management module using NTP.
The following configuration fragment illustrates the NTP configuration needed on an ASA to enable
NTP to an NTP server located in the management network:
ntp authentication-key 10 md5 *
ntp authenticate
ntp trusted-key 10
Small Enterprise Design Profile Reference Guide
5-44
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
ntp server <NTP-Server-address> source management
Syslog and SNMP can be used to keep track of system status, device access and session activity. NetFlow
Security Event Logging (NSEL), now supported on all Cisco ASA models, may also be used for the
monitoring and reporting of session activity. The following configuration fragment illustrates the
configuration of Syslog.
logging trap informational
logging host management <Syslog-Server-address>
logging enable
The routing protocol running between the Internet firewall and the distribution/core should be secured.
The following ASA configuration fragment illustrates the use of EIGRP MD5 authentication to
authenticate the peering session between the inside firewall interface and the core/distribution switch:
interface Redundant1
nameif inside
security-level 100
ip address 10.125.33.10 255.255.255.0
authentication key eigrp 100 <removed> key-id 1
authentication mode eigrp 100 md5
Network Address Translation (NAT)
NAT is required because the enterprise typically gets a limited number of public IP addresses. In
addition, NAT helps shield the company’s internal address space from reconnaissance and another
malicious activity.
The following illustrates the NAT configuration:
! Static translation for servers residing at DMZ
static (dmz,outside) 198.133.219.10 10.25.34.10 netmask 255.255.255.255
static (dmz,outside) 198.133.219.11 10.25.34.11 netmask 255.255.255.255
static (dmz,outside) 198.133.219.12 10.25.34.12 netmask 255.255.255.255
static (dmz,outside) 198.133.219.13 10.25.34.13 netmask 255.255.255.255
!
! Dynamic Port Address Translation (PAT) for inside hosts going to the Internet
global (outside) 10 interface
nat (inside) 10 10.0.0.0 255.0.0.0
!
! Static translation for inside hosts going to the DMZ and vice-versa. The inside IP
addresses are visible to the DMZ.
static (inside,dmz) 10.0.0.0 10.0.0.0 netmask 255.0.0.0
Firewall Access Policies
As previously explained, the Internet firewall should be configured to:
•
Protect the company’s internal resources and data from external threats by preventing incoming
access from the Internet.
•
Protect public resources served by the DMZ by restricting incoming access to the public services
and by limiting outbound access from DMZ resources out to the Internet.
•
Control user's Internet-bound traffic.
Enforcing such policies requires the configuration of the appropriate interface security levels and the
deployment of ACLs governing what traffic is allowed or prevented from transiting between interfaces.
Small Enterprise Design Profile Reference Guide
5-45
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
By default, the Cisco ASA appliance allows traffic from higher to lower security level interfaces (i.e.,
from inside to outside). However, due to the sensitivity of enterprise environments, the security
administrators are recommended to override the default rules with more stringent rules indicating
exactly what ports and protocols are permitted.
In our configuration example the inside, DMZ and outside interfaces were configured with the security
levels of 100, 50 and 0 respectively. With this, by default any traffic originating from the inside to the
DMZ and outside, and from the DMZ to the outside interface will be allowed freely. At the same time,
any traffic originating from the outside to the DMZ and inside, and from the DMZ to the inside interface
will be blocked. While this may satisfy the basic access control requirements of the organization, it is
always a good idea to reinforce the policies by enforcing granular ACLs.
Before designing the ACLs, it should also be noted that, as the Cisco ASA inspects traffic it is able to
recognize packets belonging to already established sessions. The stateful inspection engine of the
firewall dynamically allows the returning traffic. Therefore, the firewall ACLs should be constructed to
match traffic in the direction in which it is being initiated. In our sample configurations ACLs are applied
in the ingress direction.
The following are the guidelines and configuration examples of ACLs controlling access and traffic
flows:
•
Ingress Inside
Allow Internet access to users residing at all enterprise sites for the allowed ports and protocols. This
typically includes HTTP and HTTPS access.
access-list Outbound extended permit tcp 10.0.0.0 255.0.0.0 any eq http
access-list Outbound extended permit tcp 10.0.0.0 255.0.0.0 any eq https
Allow users access to DMZ services such as the company’s website, E-mail, and domain name resolution
(HTTP, HTTPS, SMTP, POP, IMAP, and DNS). Note that the previous entries in the ACL already permit
HTTP and HTTPS traffic.
! Allow DNS queries to DNS server
access-list Outbound extended permit udp 10.0.0.0 255.0.0.0
! Allow SMTP, POP3 and IMAP access to DMZ mail server
access-list Outbound extended permit tcp 10.0.0.0 255.0.0.0
access-list Outbound extended permit tcp 10.0.0.0 255.0.0.0
access-list Outbound extended permit tcp 10.0.0.0 255.0.0.0
! Apply ACL to inside interface
access-group Outbound in interface inside
•
host 10.25.34.13 eq domain
host 10.25.34.12 eq smtp
host 10.25.34.12 eq pop3
host 10.25.34.12 eq imap4
Ingress DMZ
Restrict connections initiated from DMZ only to the necessary protocols and sources. This typically
includes DNS queries and zone transfer from DNS server, SMTP from E-mail server, HTTP/SSL
access from the Cisco IronPort ESA for updates, Sensorbase, etc.
! Allow DNS queries and zone transfer from DNS server
access-list DMZ extended permit udp host 10.25.34.13 any eq
access-list DMZ extended permit tcp host 10.25.34.13 any eq
!
! Allow SMTP from Cisco IronPort ESA
access-list DMZ extended permit tcp host 10.25.34.11 any eq
!
! Allow update and SensorBase access to Cisco IronPort ESA
access-list DMZ extended permit tcp host 10.25.34.11 any eq
access-list DMZ extended permit tcp host 10.25.34.11 any eq
!
! Apply ACL to DMZ interface
access-group DMZ in interface dmz
Small Enterprise Design Profile Reference Guide
5-46
domain
domain
smtp
http
https
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
•
Ingress Outside
Inbound Internet access should be restricted to the public services provided at the DMZ such as
SMTP, web, and DNS. Any connection attempts to internal resources and subnets from the Internet
should be blocked. ACLs should be constructed using the servers’ global IP addresses.
! Allow DNS queries and zone transfer to DNS server
access-list Inbound extended permit udp any host 198.133.219.13
access-list Inbound extended permit tcp any host 198.133.219.13
!
! Allow SMTP to Cisco IronPort ESA
access-list Inbound extended permit tcp any host 198.133.219.11
!
! Allow HTTP/HTTPS access to the company’s public web portal
access-list Inbound extended permit tcp any host 198.133.219.10
access-list Inbound extended permit tcp any host 198.133.219.10
!
! Apply ACL to outside interface
access-group Inbound in interface outside
eq domain
eq domain
eq smtp
eq http
eq https
Botnet Traffic Filter
The Small Enterprise Design Profile uses the Cisco ASA Botnet Traffic Filter on the Internet Firewall
to detect malware that attempts network activity such as sending private data (passwords, credit card
numbers, key strokes, or other proprietary data) when the malware starts a connection to a known bad
IP address. The Botnet Traffic Filter checks incoming and outgoing connections against a dynamic
database of known bad domain names and IP addresses (the blacklist), and then logs or blocks any
suspicious activity.
Configuring the Botnet Traffic Filter requires the following steps:
Step 1
Configure DNS Server
Step 2
Enable Use of the Dynamic Database
Step 3
Enable DNS Snooping
Step 4
Enable Traffic Classification and Actions for the Botnet Traffic Filter
Step 5
Verify and Monitor Botnet Traffic Filter Operation
The following sections provides configuration examples for each of these steps.
Configure DNS Server
The Botnet Traffic Filter requires a DNS Server to access Cisco’s dynamic database update server and
to resolve entries in the static database. The following configuration illustrates this configuration.
! Enable DNS requests to a DNS Server out the outside interface
dns domain-lookup outside
! Specify the DNS Server Group and the DNS Servers
dns server-group DefaultDNS
name-server 68.238.112.12
name-server 68.238.96.12
domain-name cisco.com
Small Enterprise Design Profile Reference Guide
5-47
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Enable Use of the Dynamic Database
The Botnet Traffic Filter can receive periodic updates for the dynamic database from the Cisco update
server. This database lists thousands of known bad domain names and IP addresses. The following
configuration enables database updates, and also enables use of the downloaded dynamic database by
the adaptive security appliance.
! enable downloading of the dynamic database from the Cisco Update server
dynamic-filter updater-client enable
! enable use of the dynamic database
dynamic-filter use-database
Enable DNS Snooping
DNS Snooping enables inspection of DNS packets and enables Botnet Traffic Filter Snooping which
compares the domain name with those in the dynamic or static databases, and adds the name and IP
address to the DNS reverse lookup cache. This cache is then used by the Botnet Traffic Filter when
connections are made to the suspicious address.
It is recommended that DNS snooping is only enabled on interfaces where external DNS requests are
going. Enabling DNS snooping on all UDP DNS traffic, including that going to an internal DNS server,
creates unnecessary load on the Cisco ASA. For example, if the DNS server is on the outside interface,
you should enable DNS inspection with snooping for all UDP DNS traffic on the outside interface.
The following configuration example illustrates enabling DNS Snooping on the outside interface:
! create a class map to identify the traffic you want to inspect DNS
class-map dynamic-filter-snoop-class
match port udp eq domain
! create a policy map to enable DNS inspection with Botnet Traffic Filtering snooping for
the class map
policy-map dynamic-filter-snoop-policy
class dynamic-filter-snoop-class
inspect dns preset_dns_map dynamic-filter-snoop
! activate the policy map on the outside interface
service-policy dynamic-filter-snoop-policy interface outside
Enable Traffic Classification and Actions for the Botnet Traffic Filter
The Botnet Traffic Filter compares the source and destination IP address in each initial connection
packet to the IP addresses in the dynamic database, static database, DNS reverse lookup cache, and DNS
host cache, and sends a syslog message or drops any matching traffic. When an address matches, the
Cisco ASA sends a syslog message and can optionally be configured to drop the connection. You can
enable Botnet Traffic filter on a subset of traffic or for all traffic by enabling an access list to classify
traffic.
The following configuration example enables the Botnet Traffic Filter feature on all traffic and
additionally enables dropping of connections going to IP addresses with a severity of moderate and
higher.
! identify the traffic that you want to monitor or drop.
access-list btf-filter-acl extended permit ip any any
! enable Botnet Traffic Filter on the outside interface for traffic classified by the
btf-filter-acl access list
dynamic-filter enable interface outside classify-list btf-filter-acl
! enable automatic dropping of traffic with threat level moderate or higher
dynamic-filter drop blacklist interface outside action-classify-list btf-filter-acl
threat-level range moderate very-high
Small Enterprise Design Profile Reference Guide
5-48
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Botnet Traffic Filter Verification
To monitor and verify the operation of the Botnet Traffic Filter feature, the following commands can be
used:
•
show dynamic-filter updater-client—Shows information about the updater server, including the
server IP address, the next time the adaptive security appliance will connect with the server, and the
database version last installed.
cr12-asa-1-ie# show dynamic-filter updater-client
Dynamic Filter updater client is enabled
Updater server URL is https://update-manifests.ironport.com
Application name: threatcast, version: 1.0
Encrypted UDI:
0bb93985f42d941e50dc8f022350d1a8a8c5097dc1d252b9cff62d26d4ec58e202883d704fc62b85bf8629
fa757fe36b
Last update attempted at 15:14:11 UTC Apr 7 2010,
with result: Downloaded file successfully
Next update is in 00:52:14
Database file version is '1270651144' fetched at 15:14:11 UTC Apr 7 2010, size:
2097152
cr12-asa-1-ie#
•
show dynamic-filter data—Shows information about the updater server, including the server IP
address, the next time the adaptive security appliance will connect with the server, and the database
version last installed.
cr12-asa-1-ie# show dynamic-filter data
Dynamic Filter is using downloaded database version '1270651144'
Fetched at 15:14:11 UTC Apr 7 2010, size: 2097152
Sample contents from downloaded database:
win-antimalware2.com firstlook.com red-devil-sport-club.gymdb.com
mswindowsupdate.info
zardoz.wizardz.com exchange.bg bisexual-photo.com lookmbox.com
Sample meta data from downloaded database:
threat-level: very-high,
category: Malware,
description: "These are sources that use various exploits to deliver adware, spyware
and other malware to victim computers. Some of these are associated with rogue online
vendors and distributors of dialers which deceptively call premium-rate phone
numbers."
threat-level: high,
category: Bot and Threat Networks,
description: "These are rogue systems that control infected computers. They are
either systems hosted on threat networks or systems that are part of the botnet
itself."
threat-level: moderate,
category: Spyware,
description: "These are sources that distribute spyware, adware, greyware, and other
potentially unwanted advertising software. Some of these also run exploits to install
such software."
threat-level: low,
category: Ads,
description: "These are advertising networks that deliver banner ads, interstitials,
rich media ads, pop-ups, and pop-unders for websites, spyware and adware. Some of
these networks send ad-oriented HTML emails and email verification services."
Total entries in Dynamic Filter database:
Dynamic data: 82119 domain names , 2565 IPv4 addresses
Local data: 0 domain names , 0 IPv4 addresses
Active rules in Dynamic Filter asp table:
Dynamic data: 0 domain names , 2565 IPv4 addresses
Local data: 0 domain names , 0 IPv4 addresses
cr12-asa-1-ie#
Small Enterprise Design Profile Reference Guide
5-49
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
•
show dynamic-filter statistics detail—Shows how many connections were monitored and dropped
with the Botnet Traffic Filter, and how many of those connections match the whitelist, blacklist, and
greylist. (The greylist includes addresses that are associated with multiple domain names, but not
all of these domain names are on the blacklist.) The detail keyword shows how many packets at each
threat level were classified or dropped.
cr12-asa-1-ie# show dynamic-filter statistics detail
Enabled on interface outside using classify-list btf-filter-acl
Total conns classified 35, ingress 0, egress 35
Total whitelist classified 0, ingress 0, egress 0
Total greylist classified 16, dropped 0, ingress 0, egress 16
Threat-level very-high: classified
0, dropped
0,
egress
0
Threat-level
high: classified
0, dropped
0,
egress
0
Threat-level
moderate: classified
0, dropped
0,
egress
0
Threat-level
low: classified
16, dropped
0,
egress
16
Threat-level
very-low: classified
0, dropped
0,
egress
0
Total blacklist classified 19, dropped 0, ingress 0, egress 19
Threat-level very-high: classified
9, dropped
0,
egress
9
Threat-level
high: classified
0, dropped
0,
egress
0
Threat-level
moderate: classified
0, dropped
0,
egress
0
Threat-level
low: classified
10, dropped
0,
egress
10
Threat-level
very-low: classified
0, dropped
0,
egress
0
cr12-asa-1-ie#
Note
ingress
0,
ingress
0,
ingress
0,
ingress
0,
ingress
0,
ingress
0,
ingress
0,
ingress
0,
ingress
0,
ingress
0,
To clear the statistics, enter the clear dynamic-filter statistics [interface name] command.
Other commands that are useful for monitoring the Botnet Traffic filter include the following:
•
show dynamic-filter reports top [malware-sites | malware-ports | infected-hosts]—Generates
reports of the top 10 malware sites, ports, and infected hosts monitored. The top 10 malware-sites
report includes the number of connections dropped, and the threat level and category of each site.
This report is a snapshot of the data, and may not match the top 10 items since the statistics started
to be collected.
•
show dynamic-filter reports infected-hosts {max-connections | latest-active | highest-threat |
subnet ip_address netmask | all}—Generates reports about infected hosts. These reports contain
detailed history about infected hosts, showing the correlation between infected hosts, visited
malware sites, and malware ports. The max-connections keyword shows the 20 infected hosts with
the most number of connections. The latest-active keyword shows the 20 hosts with the most recent
activity. The highest-threat keyword shows the 20 hosts that connected to the malware sites with the
highest threat level. The subnet keyword shows up to 20 hosts within the specified subnet. The all
keyword shows all buffered infected-hosts information. This display might include thousands of
entries. You might want to use ASDM to generate a PDF file instead of using the CLI.
•
show dynamic-filter dns-snoop [detail]—Shows the Botnet Traffic Filter DNS snooping summary,
or with the detail keyword, the actual IP addresses and names. All inspected DNS data is included
in this output, and not just matching names in the blacklist. DNS data from static entries are not
included.
Small Enterprise Design Profile Reference Guide
5-50
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
•
show asp table dynamic-filter [hits]—Shows the Botnet Traffic Filter rules that are installed in the
accelerated security path.
Firewall Redundancy
The Internet perimeter of the Small Enterprise Design Profile uses a single Cisco ASA appliance
configured with redundant interfaces. The use of redundant interfaces makes the design resilient to link
level failures, representing an affordable option for high availability. In cases where chassis redundancy
is desirable, the enterprise may consider deploying a pair of Cisco ASA appliances configured for
stateful failover. Both active/active and active/standby failover modes are supported. While stateful
failover protects against chassis failures, it requires the deployment of two identical Cisco ASA
appliances and the adjustment of the topologies around the firewalls, so its deployment should be
carefully planned.
This guide explains the use of redundant interfaces. For information on how to configure stateful
failover, refer to the Cisco ASA 5500 Series Adaptive Security Appliances Configuration Guides at the
following URL:
http://www.cisco.com/en/US/products/ps6120/products_installation_and_configuration_guides_list.ht
m
A Cisco ASA redundant interface is a logical interface that pairs two physical interfaces, called active
and standby interfaces. Under normal operation the active interface is the only one passing traffic. The
active interface uses the IP address defined at the redundant interface, and the MAC address of the first
physical interface associated with the redundant interface. When the active interface fails, the standby
interface becomes active and starts passing traffic. The same IP address and MAC address are maintained
so that traffic is not disrupted. Figure 5-23 illustrates the concept of redundant interface.
Figure 5-23
Cisco
Catalyst
4507
Cisco ASA Redundant Interface
Cisco
Inside ASA
10.125.33.10 5520
Outside
Gi5/3
Gi0/0
Internet
Gi4/4
Gi0/1
Redundant
Interface
227431
VLAN 200
10.125.33.9
The configuration of a redundant interface consists in the configuration of the physical interface
parameters and the logical redundant interface. Physical parameters such as media type, duplex, and
speed are still configured within the physical interface. IP address, interface name, routing protocols,
security level are configured as part of the redundant interface. The following configuration example
corresponds to Figure 5-23.
! Physical interface and Ethernet parameters
interface GigabitEthernet0/0
description Connected to cr24-4507-DO
no nameif
no security-level
no ip address
!
interface GigabitEthernet0/1
description backup to cr24-4507-DO
no nameif
Small Enterprise Design Profile Reference Guide
5-51
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
no security-level
no ip address
!
! Defines logical redundant interface associated with physical
! interfaces. Configures IP and logical interface parameters.
interface Redundant1
description Connected to cr24-4507-DO
member-interface GigabitEthernet0/0
member-interface GigabitEthernet0/1
nameif inside
security-level 100
ip address 10.125.33.10 255.255.255.0
authentication key eigrp 100 <removed> key-id 1
authentication mode eigrp 100 md5
!
Routing
An interior gateway protocol, EIGRP in our configuration examples, is used for dynamic routing. The
Internet firewall may participate in routing by learning the internal routes and by injecting a default route
pointing to the Internet. The default route should be removed dynamically if the Internet connection
becomes unavailable.
As part of the Small Enterprise Design Profile, two different approaches were validated for the injection
of the default route:
•
OSPF—The Cisco ASA appliance learns the default route from the Internet border router using
OSPF. The default route is then redistributed into EIGRP, and from there propagated into the rest of
the internal network.
•
Static Route—The Cisco ASA appliance is configured with a static default route pointing to the
Internet gateway. Object tracking is configured to dynamically remove the default route when the
Internet connection becomes unavailable. The default route is redistributed into EIGRP, and from
there propagated into the rest of the internal network.
Injecting a default route with OSPF requires the configuration of an OSPF process between the Cisco
ASA and the Internet border router, as illustrated in Figure 5-24. If the router is managed by the ISP, the
configuration will require coordination with the service provider. This scenario also requires the default
route to be propagated over OSPF. The actual default route may originate from the Internet border router
itself or somewhere in the ISP network
Figure 5-24
OSPF Default Route Injection
EIGRP 100
OSPF 200
Cisco
Catalyst
4507
default
Internet
default
227432
VLAN 200
10.125.33.9/24
Cisco
Internet
Inside ASA
Border
10.125.33.10/24 5520
Outside
Router
Gi0/0
Gi5/3
Gi0/2
.1
198.133.219.5/24
Gi0/1
Gi4/4
Small Enterprise Design Profile Reference Guide
5-52
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
The following are the guidelines for using OSPF for the injection of a default route:
•
Whenever possible, use MD5 authentication to secure the routing session between the Cisco ASA
and the Internet border router.
•
Since NAT is configured on the Cisco ASA and the inside address space is not visible outside the
firewall, there is no need to redistribute routes from the internal EIGRP into OSPF.
•
Route redistribution from OSPF into the internal EIGRP should be limited to the default route only.
No other routes should be propagated into EIGRP.
The following configuration snippet illustrates the routing configuration of the Cisco ASA appliance.
The configuration includes the route redistribution from OSPF into EIGRP with the enforcement of a
route-map allowing only the injection of the default route. MD5 authentication is used for OSPF, and the
logging of neighbor status changes is enabled.
! Permit default only
access-list Inbound-Routes standard permit host 0.0.0.0
!
interface GigabitEthernet0/2
ospf message-digest-key 1 md5 <removed>
ospf authentication message-digest
!
route-map Inbound-EIGRP permit 10
match ip address Inbound-Routes
!
router eigrp 100
no auto-summary
network 10.125.33.0 255.255.255.0
passive-interface default
no passive-interface inside
redistribute ospf 200 metric 1000000 2000 255 1 1500 route-map Inbound-EIGRP
!
router ospf 200
network 198.133.219.0 255.255.255.0 area 100
area 100 authentication message-digest
log-adj-changes
!
Note
The hello-interval and dead-interval OSPF timers can be adjusted to detect topological changes faster.
The other validated alternative for the default route injection is the definition of a static default route,
which then can be redistributed into the internal EIGRP process. This is shown in Figure 5-25. This
option does not require the configuration of the Internet border router.
Small Enterprise Design Profile Reference Guide
5-53
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-25
Static Default Route with Object Tracking
EIGRP 100
VLAN 200
10.125.33.9
Cisco
Internet
Inside ASA
Border
10.125.33.10/24 5520
Outside
Router
Gi5/3
Gi0/0
Gi0/2
.1
198.133.219.5/24
Gi0/1
Gi4/4
Internet
default
Object Tracking
227433
Cisco
Catalyst
4507
It is highly recommended to use object tracking so the default route is removed when the Internet
connection becomes unavailable. Without object tracking, the default route will be removed only if the
outside interface of the appliance goes down. So there is a possibility that the default route may remain
in the routing table even if the Internet border router becomes unavailable. To avoid that problem, the
static default route can be configured with object tracking. This consists in associating the default route
with a monitoring target. The Cisco ASA appliance monitors the target using ICMP echo requests. If an
echo reply is not received within a specified time period, the object is considered down and the
associated default route is removed from the routing table.
The monitoring target needs to be carefully selected. First, pick one that can receive and respond to
ICMP echo requests sent by the Cisco ASA. Second, it is better to use a persistent network object. In the
configuration example below the Cisco ASA monitors the IP address of the next hop gateway, which
helps identifying if the Internet gateway goes down, but it will not help if the connection is lost upstream.
If available, you may want to monitor a persistent network object located somewhere in the ISP network.
Static route tracking can also be configured for default routes obtained through DHCP or PPPoE.
In the following configuration the IP address of the next hop gateway (198.133.219.1) is used as the
monitoring target. The static default route is then redistributed into EIGRP.
router eigrp 100
no auto-summary
network 10.125.33.0 255.255.255.0
passive-interface default
no passive-interface inside
redistribute static metric 1000000 2000 255 1 1500
!
route outside 0.0.0.0 0.0.0.0 198.133.219.1 1 track 10
!
sla monitor 1
type echo protocol ipIcmpEcho 198.133.219.1 interface outside
sla monitor schedule 1 life forever start-time now
!
track 10 rtr 1 reachability
Note
The frequency and timeout parameters of object tracking can be adjusted to detect topological changes
faster.
Small Enterprise Design Profile Reference Guide
5-54
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Intrusion Prevention Deployment
The Small Enterprise Design Profile implements Intrusion Prevention using an Advanced Inspection and
Prevention Security Services Module (AIP SSM) on the Cisco ASA appliance deployed at the Internet
perimeter. This section describes the best practices for integrating and configuring the IPS module for
maximum threat control and visibility, as well as the deployment of the IPS Global Correlation feature.
Deploying IPS with the Cisco ASA
The Advanced Inspection and Prevention Security Services Module (AIP SSM) is supported on Cisco
ASA 5510 and higher platforms. The AIP SSM runs advanced IPS software that provides proactive,
full-featured intrusion prevention services to stop malicious traffic, including worms and network
viruses, before they can affect your network.
As described in the “Intrusion Prevention Guidelines” section on page 5-10, the AIP SSM may be
deployed in inline or promiscuous mode. In inline mode the AIP SSM is placed directly in the traffic
flow, while in promiscuous mode the Cisco ASA sends a duplicate stream of traffic to the AIP SSM.
When deploying the AIP SSM in inline mode it is particularly important to determine how traffic will
be treated in case of a module failure. The AIP SSM card may be configured to fail open or close when
the module becomes unavailable. When configured to fail open, the Cisco ASA appliance allows all
traffic through, uninspected, if the AIP SSM becomes unavailable. Per contrary, when configured to fail
close, the adaptive security appliance blocks all traffic in case of an AIP SSM failure.
The following example illustrates how a Cisco ASA can be configured to divert all IP traffic to the AIP
SSM in inline mode, and to block all IP traffic if the AIP SSM card fails for any reason:
access-list IPS permit ip any any
class-map my-ips-class
match access-list IPS
policy-map my-ips-policy
class my-ips-class
ips inline fail-close
service-policy my-ips-policy global
IPS Global Correlation Deployment
There are a number of aspects that need to be understood prior to the deployment of the IPS Global
Correlation feature. To start with, before configuration be sure that you are using Cisco IPS Version 7.0
with the latest patch and signature updates and that Cisco IPS is configured for network connectivity in
either IDS or IPS mode.
The configuration of Global Correlation can be performed using the command-line interface (CLI),
Cisco IDS Device Manager (IDM), Cisco IME, or the Cisco Security Manager. The figures here provided
are screenshots from Cisco IME that illustrate the basic steps in the configuration of IPS Global
Correlation.
The first step in configuring the IPS sensor (module or appliance) to use Global Correlation is to add
either a DNS address and/or the proxy server setup. This step, illustrated in Figure 5-26, enables
connection to Cisco SensorBase. After you configure the DNS and proxy settings, these settings will go
into effect as soon as the sensor has downloaded the latest Global Correlation updates.
Small Enterprise Design Profile Reference Guide
5-55
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-26
DNS and HTTP Proxy Within the Network Setting Configuration Screen
By default, a sensor runs Global Correlation Inspection in Standard mode, and enables the reputation
filters, both illustrated in Figure 5-27. A good practice is to configure Global Correlation Inspection
initially in permissive mode while monitoring the effects, and later change the configuration into
Standard or Aggressive mode.
Small Enterprise Design Profile Reference Guide
5-56
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-27
Global Correlation Inspection Settings
By default network participation is disabled, which means the sensor does not share any event data back to
the Cisco SensorBase network. The event data provided by all devices participating in the Cisco SensorBase
network is a key element that provides real-time and worldwide visibility into the threat activity,
accelerating the identification and mitigation of threats propagating throughout the Internet. For this reason,
Cisco recommends to configure the IPS sensors with partial or full network participation. See Figure 5-28.
Small Enterprise Design Profile Reference Guide
5-57
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-28
Network Participation Settings (Off by Default)
Event Monitoring with Global Correlation
Event monitoring with Global Correlation is similar to event monitoring with signature-only IPS. The
primary difference is the potential addition of reputation scores representing the Global Correlation data.
Figure 5-29 shows Cisco IPS events with reputation scores in Cisco IME.
Small Enterprise Design Profile Reference Guide
5-58
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-29
Event Monitoring with Global Correlation in Cisco IME
Figure 5-29 shows several TCP SYN Port Sweep and ICMP Network Sweep attacks that were seen by
the sensor. The first three events had no reputation, and the event’s risk ratings were 52 and 60 which
did not meet the threshold for the packets to be dropped. The next three events were identical except that
the attacker had a negative reputation of -1.8 which elevated the risk ratings to 70 and 75 which still did
not meet the thresholds to be dropped in Standard Mode. The last events were also identical except this
time the attacker has a negative reputation if -4.3 which elevated the risk ratings to 86 and 89. This time
the risk rating was high enough for the packets to be dropped.
Figure 5-30 illustrates the detail view of the TCP SYN Port Sweep event coming from the attacker with
a negative reputation of -4.3.
Small Enterprise Design Profile Reference Guide
5-59
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-30
Detailed view of a TCP SYN Port Sweep from an Attacker with a Negative Reputation
Score
Web Security Deployment
The Small Enterprise Design Profile implements a Cisco IronPort WSA at the core/distribution layer of
the main site, as illustrated in Figure 5-31. The WSA is located at the inside of the Cisco ASA acting as
the Internet firewall. That ensures that clients and WSA are reachable over the same inside interface of
the firewall, and that the WSA can communicate with them without going through the firewall. At the
same time, deploying the WSA at the core/distribution layer gives complete visibility to the WSA on the
traffic before getting out to the Internet through the firewall.
Small Enterprise Design Profile Reference Guide
5-60
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-31
Cisco IronPort WSA deployment
Core
Distribution
Cisco
Catalyst
4507
Internet Edge
Cisco ASA
5520
Inside
Internet
Border
Router
Outside
Internet
SensorBase
Cisco IronPort
WSA S160
227434
www
Following are the guidelines for the WSA configuration and deployment.
Initial System Setup Wizard
The WSA provides a browser-based system setup wizard that must be executed the first time the
appliance is installed. The System Setup Wizard guides the user through initial system configuration
such as network and security settings. It should be noted that running the initial System Setup Wizard
completely reconfigures the WSA appliance and resets the administrator password. Only use the System
Setup Wizard the first time you install the appliance, or if you want to completely overwrite the existing
configuration.
The following are some of the default settings when running the System Setup Wizard:
•
Web Proxy is deployed in transparent mode.
•
The L4 Traffic Monitor is active and set to monitor traffic on all ports.
All these settings can be changed any time after the initial configuration by running the WSA web-based
configuration GUI.
Interface and Network Configuration
The following need to be completed as part of the initial setup of the WSA appliance:
Step 1
Configuring network interfaces
Step 2
Adding routes
Step 3
Configuring DNS
Step 4
Setting time
Step 5
Working with upstream proxy (if present)
Configuring Network Interfaces
Independently from the model, all Cisco IronPort WSA appliances are equipped with six Ethernet
interfaces as shown in Figure 5-32.
Small Enterprise Design Profile Reference Guide
5-61
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-32
WSA Interfaces
Use the M1 port for administering
the appliance.
Use the “P” ports for the Web Proxy.
227435
Use the “T” ports for the L4 Traffic Monitor.
The WSA interfaces are grouped for the following functions:
•
Management—Interfaces M1 and M2 are out-of-band (OOB) management interfaces. However,
only M1 is enabled. In the Small Enterprise Design Profile, interface M1 connects to the out-of-band
management network. Interface M1 can optionally be used to handle data traffic in case the
enterprise does not have an out-band management network.
•
Web Proxy—Interfaces P1 and P2 are Web Proxy interfaces used for data traffic. Only the P1
interface is used in the Enterprise Design Profile for Small Enterprise Networks. P1 connects to the
inside subnet of the firewall.
•
L4 Traffic Monitor (L4TM)—T1 and T2 are the L4TM interfaces. L4TM is not used in the Small
Enterprise Design Profile, Cisco IPS Global Correlation and the Cisco ASA Botnet Filter features
are used instead.
Figure 5-33 illustrates the network topology around the WSA used in the Cisco validation lab.
Figure 5-33
WSA Network Topology
Cisco
Catalyst
4507
Cisco
IronPort
WSA S160
www
Gi7/3
P1
10.125.33.8
VLAN 200
10.125.33.9
Cisco
Inside ASA
10.125.33.10 5520
Outside
Gi5/3
Gi0/0
Internet
Gi4/4
Gi0/1
M1
172.26.191.105
SP Managed
MetroE Core
229297
OOB
Management
Figure 5-34 illustrates the IP address and hostname configurations for the interfaces used. In this case,
an out-of-band management network is used; therefore the M1 port is configured with an IP address in
the management subnet. In addition, the WSA is configured to maintain a separate routing instance for
the M1 management interface. This allows the definition of a default route for management traffic
separate from the default route used for data traffic.
Small Enterprise Design Profile Reference Guide
5-62
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-34
WSA Interface Configuration
Adding Routes
A default route is defined for management traffic pointing to the OOB management default gateway
(172.26.191.1). A separate default route is defined for the data traffic pointing to the inside IP address
of the firewall (10.125.33.10). As all internal networks are reachable throughout the core/distribution
switch, a route to 10.0.0.0/8 is defined pointing to the switch IP address (10.125.33.9) to allow the WSA
to communicate with the clients directly without having to go to the firewall first. These settings are
illustrated in Figure 5-35.
Figure 5-35
WSA Route Configuration
Configuring DNS
The initial setup requires the configuration of a hostname for the WSA appliance, and listing the DNS
servers. Figure 5-36 shows the DNS configuration.
Small Enterprise Design Profile Reference Guide
5-63
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-36
WSA DNS Configuration
Setting Time
Time synchronization is critical for forensic analysis and troubleshooting, therefore enabling NTP is
highly recommended. Figure 5-37 shows how the WSA is configured to synchronize its clock with an
NTP server located on the OOB management network.
Figure 5-37
WSA NTP Configuration
Working with Upstream Proxies
If Internet access is provided by an upstream proxy, then the WSA must be configured to use the proxy
for component updates and system upgrades. This is illustrated in Figure 5-38 and Figure 5-39.
Figure 5-38
WSA Upgrade Settings
Small Enterprise Design Profile Reference Guide
5-64
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-39
WSA Component Updates
WCCP Transparent Web Proxy
The configuration of the WCCP Transparent Web Proxy includes the following:
Step 1
Defining WSA WCCP Service Group
Step 2
Enabling WSA Transparent Redirection
Step 3
Enabling WCCP redirection on the Cisco ASA
Step 4
Enabling WSA HTTPS scanning
Step 5
Working with upstream proxy (if present)
Defining WSA WCCP Service Group
Web Proxy settings are configured as part of an initial setup using the System Setup Wizard and can be
later modified with the WSA Web-based GUI. The Web Proxy setting include the following:
•
HTTP Ports to Proxy—List the ports to be proxied. Default is 80 and 3128.
•
Caching—Defines whether or not the WSA should cache response and requests. Caching helps
reduce latency and the load on the Internet links. Default is enabled.
•
IP Spoofing — Defines whether or not the Web Proxy should spoof IP addresses when forwarding
requests to upstream proxies and servers. The Cisco ASA does not support source address spoofing.
Figure 5-40 illustrates the Web Proxy settings.
Small Enterprise Design Profile Reference Guide
5-65
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-40
WSA Proxy Settings
Enabling WSA Transparent Redirection
Configuring WCCP Transparent Redirection requires the definition of a WCCP service profile in the
WSA. If redirecting HTTP and HTTPS, define a dynamic service ID to be used with the Cisco ASA. Use
MD5 authentication to protect the WCCP communication between the WSA and Cisco ASA.
Figure 5-41 shows an example.
Figure 5-41
WSA Transparent Proxy
Small Enterprise Design Profile Reference Guide
5-66
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Enabling WCCP Redirection on Cisco ASA
The configuration of WCCP on the Cisco ASA appliance requires:
•
A group-list indicating the IP addresses of the appliances member of the service group. In the
example provided below the group-list is called wsa-farm.
•
A redirect-list indicating the ports and subnets of traffic to be redirected. In the example, the ACL
named proxylist is configured to redirect any HTTP and HTTPS traffic coming from the 10.0.0.0/8
subnet. It is critical to ensure traffic from the WSA(s) bypasses redirection. To that end, add an entry
to the redirect-list explicitly denying traffic sourced from the WSA(s).
•
WCCP service indicating the service ID. Make sure you use the same ID as defined on the WSAs.
Use a password for MD5 authentication.
•
Enabling WCCP redirection on an interface. Apply the WCCP service on the inside interface of the
Cisco ASA.
The following is a Cisco ASA WCCP configuration example:
! Group-list defining the IP addresses of all WSAs
access-list wsa-farm extended permit ip host 10.125.33.8 any
!
! Redirect-list defining what ports and hosts/subnets should
access-list proxylist extended deny ip host 10.125.33.8 any
access-list proxylist extended permit tcp 10.0.0.0 255.0.0.0
access-list proxylist extended permit tcp 10.0.0.0 255.0.0.0
!
! WCCP service
wccp 10 redirect-list proxylist group-list wsa-farm password
!
! Applies WCCP on an interface
wccp interface inside 10 redirect in
be redirected
any eq www
any eq https
cisco
The WCCP connection status and configuration can be monitored on the Cisco ASA with the show wccp
command, as shown below:
cr26-asa5520-do# show wccp
Global WCCP information:
Router information:
Router Identifier:
Protocol Version:
Service Identifier: 10
Number of Cache Engines:
Number of routers:
Total Packets Redirected:
Redirect access-list:
Total Connections Denied Redirect:
Total Packets Unassigned:
Group access-list:
Total Messages Denied to Group:
Total Authentication failures:
Total Bypassed Packets Received:
cr26-asa5520-do#
198.133.219.5
2.0
1
1
428617
proxylist
0
4
wsa-farm
0
0
0
Enabling WSA HTTPS Scanning
To monitor and decrypt HTTPS traffic, you must enable HTTPS scanning on the WSA. The HTTPS
Proxy configuration is illustrated in Figure 5-42.
Small Enterprise Design Profile Reference Guide
5-67
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-42
WSA HTTPS Proxy
Working with Upstream Proxies
In case Internet traffic is handled by one or more upstream proxies, follow these guidelines:
•
Add an Upstream Proxy Group
•
Define a routing policy to direct traffic to the upstream proxies
The Upstream Proxy Group lists the IP addresses or domain names of the proxies to be used for traffic
sent to the Internet. When multiple proxies are available, the WSA can be configured for failover or load
balancing.
The following are the options available:
•
None (failover)—The first proxy in the list is used. If one proxy cannot be reached, the Web Proxy
attempts to connect to the next one in the list.
•
Fewest connections—Transactions are directed to the proxy servicing the fewest number of
connections.
•
Hash-based—Requests are distributed using a hash function. The function uses the proxy ID and
URL as inputs so that requests for the same URL are always directed to the same upstream proxy.
•
Least recently used—Transactions are directed to the proxy that least recently received a transaction
if all proxies are currently active.
•
Round robin—The Web Proxy cycles transactions equally among all proxies in the group in the
listed order.
Figure 5-43 illustrates the upstream proxy group configuration. Two upstream proxies are used, and
transactions are forwarded to the proxy servicing the fewest number of connections.
Small Enterprise Design Profile Reference Guide
5-68
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-43
WSA Upstream Proxy Group
Next, a routing rule needs to be defined to indicate when and how to direct transactions to the upstream
proxy group. Use the Global Routing Policy if all traffic is to be handled by the upstream proxies. If no
proxies are present, then leave the routing destination of the Global Routing Policy configured as Direct
Connection. Figure 5-44 presents an example where all traffic is directed to the proxies in the
Upstream-Lab_proxy group.
Figure 5-44
WSA Routing Policies
Web Access Policies
The access policies define how the Web Proxy handles HTTP requests and decrypted HTTPS
connections for network users. By configuring access policies the enterprise can control what Internet
applications (instant messaging clients, peer-to-peer file-sharing, web browsers, Internet phone services,
etc.) and URL categories users may access. In addition, access policies can be used to block file
downloads based on file characteristics, such as file size and file type.
The WSA comes with a default Global Policy that applies to all users. However, multiple policies can
be defined when different policies need to be applied to different group of users. Figure 5-45 shows the
global policy.
Figure 5-45
Global Access Policy
Small Enterprise Design Profile Reference Guide
5-69
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
URL categories corresponding to non-business related content should be blocked in compliance with the
company’s Internet access policies. Figure 5-46 provides an example on how the “Adult/Sexually
Explicit” category is blocked.
Figure 5-46
URL Categories
CISF Protected Ports
Catalyst Integrated Security Features (CISF) is a set of native security features available on Cisco
Catalyst Switches that protect the network against attacks such as man-in-the-middle, spoofing, and
infrastructure denial-of-services (DoS) attacks. CISF includes the following:
•
Port Security
•
DHCP snooping
•
IP Source Guard
•
Dynamic ARP inspection
•
ARP rate limiting
•
Storm Control
The following is a sample CISF port configuration:
! configure port security parameters
switchport port-security maximum 2
switchport port-security
switchport port-security aging time 2
switchport port-security violation restrict
switchport port-security aging type inactivity
! configure arp inspection rate limiting
ip arp inspection limit rate 100
! configure dhcp snooping rate limiting
ip dhcp snooping limit rate 100
! configure storm control parameters
storm-control broadcast level 20.00 10.00
storm-control multicast level 50.00 30.00
NAC Appliance Deployment
In the Small Enterprise Design Profile, a NAC Appliance solution is deployed at all locations, the main
site and each of the remote offices. To that end, a centralized CAM is deployed at the main site, likely
within the serverfarm. And a CAS is deployed at the main site and each remote site, and each of which
directly connects to the core/distribution layer at each of the locations.
Small Enterprise Design Profile Reference Guide
5-70
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Deploying the NAC appliance solution requires the configuration of the CAM and CAS components. The
initial CAM and CAS configuration is performed via console access, and this is described in the
installation guide, Cisco NAC Appliance Hardware Installation. During the configuration stage the
multiple steps must be followed in configuring the NAC Appliance with IP addresses, VLANs.
passwords, etc. The installation guide contains worksheets that assist in the gathering and preparation of
this information for both CAM and CAS.
After performing the initial setup through the console, the rest of the configuration of the CAM and CAS
is performed using the CAM web-based GUI. The first task on the CAM before beginning configuration
is the installation the licenses for the solution. A license must be installed for the CAM, and the CAS
servers controlled by the CAM. The Cisco NAC Appliance Ordering Guide, provides information on the
ordering options.
Licenses can be entered via the CAM web-based GUI, as shown in Figure 5-47.
Figure 5-47
NAC Appliance Licensing
Adding a CAS to the CAM
For a CAS server to be managed by the CAM, it must be first added to the list of managed servers on the
CAM. To do this the CAM needs to know the IP address of the CAS, and the Server Type (its role in the
network) of the CAS. In addition to this, the CAS and the CAM must have the same shared secret. The
shared secret is configured during the server installation. An example of adding a CAS to the CAM is
shown in Figure 5-48.
Small Enterprise Design Profile Reference Guide
5-71
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-48
Adding a new CAS to the CAM
Once the CAS has been added to the CAM, it appears in the list of servers on the CAM. From this point,
it can be managed directly from the CAM for almost all tasks. An example of a list of servers is shown
in Figure 5-49.
Figure 5-49
List of CAS Servers
Managing the CAS
Once the CAS is in the list of servers managed by the CAM, it can be configured further for its role in
the network. To manage the server click the icon under the Manage heading in the server list, this will
connect you to the CAS server and present you with the summary menu shown in Figure 5-50.
Small Enterprise Design Profile Reference Guide
5-72
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-50
AS Management Menu
Under the CAS Network setting tab, shown in Figure 5-51, the basic network settings for the CAS can
be seen and altered, if needed. In this example, we are keeping the network configuration performed
during the server installation. The primary dialog under the Network Tab is the IP dialog, shown in
Figure 5-51, the other dialogs allow the DHCP options to the configured—our example uses the default
of DHCP passthrough— and the DNS options where host name, domain name, and DNS server
information is added, as shown in Figure 5-52.
Figure 5-51
CAS Network Settings
Small Enterprise Design Profile Reference Guide
5-73
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-52
CAS DNS Settings
The next tab in the CAS configuration is the Filter tab (see Figure 5-53), for the purposes of our example
the important dialog is the Roles where network traffic filters may be applied to different user Roles. The
Role of interest at this moment is the default Unauthenticated Role. By default the Unauthenticated Role
blocks all traffic. In this example we are allowing the Unauthenticated Role to pass Active Directory
client authentication traffic to pass to the Active Directory Sever. This will allow a windows client to
join the active Directory Domain, and windows users to authenticate to the domain although they have
not been through the NAC process. This is often important to allow printer and drive mapping
information to be sent to the winders users. As the user has already authenticated to the Active Directory
Domain the user authentication information maybe learned from Active Directory, and the user does not
have to reauthenticate for the NAC server.
Note
The creation of Roles and their associated filters is performed in the CAM User Management -> User
Roles menu.
Figure 5-53
CAS Filter Settings
The next tab that requires configuration is the Advanced tab. This has multiple dialogs that require
configuration. The first of these is the Managed Subnet dialog, where each of the trusted VLAN subnets
is added to the CAS for management. An example of this shown in Figure 5-54.
Small Enterprise Design Profile Reference Guide
5-74
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-54
CAS Managed Subnet
The next dialog of the Advanced Tab is the VLAN Mapping dialog, which tells the CAS which trusted
VLAN to be mapped to an untrusted VLAN, an example of this is shown in Figure 5-55. In our example,
VLAN Prunning and VLAN Mapping is also enabled.
Figure 5-55
VLAN Mapping
The next tab of interest is the Authentication Tab (see Figure 5-56), this tab has multiple dialogs for
configuring different authentication options. The first dialog is the Login Page Dialog. This allows the
configuration of different web login pages depending upon the untrusted subnet being used for
authenticating client.
Small Enterprise Design Profile Reference Guide
5-75
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-56
Authentication Login Page
The other Authentication dialog of interest in this example is the Windows Auth dialog, as Windows
Single Sign On (SSO) is used in this example. To perform Windows SSO the CAS needs to be able to
communicate with Active Directory to determine the authentication state of the windows user. If Active
Directory confirms that the user has authenticated to Active Directory the user doesn't need to perform
additional authentication to the CAS. An example of this configuration is shown in Figure 5-57. There
are a number of steps required to configure Active Directory SSO, as these are described in the Cisco
NAC Appliance —Clean Access Server Installation and Configuration Guide. The key components in
this configuration are:
•
The creation of a Active Directory client account for the CAS
•
Using the KTPass Application on Active Directory to convert the account encryption to DES
encryption
Small Enterprise Design Profile Reference Guide
5-76
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-57
CAS Windows Authentication
Clean Access Roles
The unauthenticated role is common to all clients, but once the client has been authenticated a different
role may be applied based upon the identity of the client, different roles may be assigned for admin staff,
users, etc.
User roles allow you to aggregate various policies into a user role. These policies include:
•
Traffic policies
•
Bandwidth policies
If bandwidth policies are to be enforced by the Clean Access Server it must be operating in band.
Note
•
VLAN ID retagging
•
Clean Access network port scanning plugins
•
Clean Access Agent/Cisco NAC Web Agent client system requirements
For example, the Admin and User roles could each have different traffic polices and VLANs, in addition
the User role may enforce bandwidth policies by keeping the user traffic in band.
For more information on roles, refer to the Cisco NAC Appliance - Clean Access Manager Installation
and Configuration Guide at the following URL:
http://www.cisco.com/en/US/docs/security/nac/appliance/configuration_guide/45/cam/45cam-book.ht
ml
Layer 2 OOB Example
Figure 5-58 shows an example of a Layer-2 OOB deployment. Where a wired client connected to an
access switch is originally on the untrusted VLAN 264 and is switched to a trusted VLAN 64 once it has
completed the NAC functions. The first NAC function is the authentication and authorization function,
and this is the first design decision in implementing the NAC solution. That is, how will authentication
and authorization be achieved, and what will the user experience be.
Small Enterprise Design Profile Reference Guide
5-77
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
This example is focused upon the virtual gateway example, as virtual gateway provides the simplest
deployment. In the virtual gateway example the original IP addressing, interfaces, and VLANs are
maintained, and normal traffic flows are maintained. The only changes are the addition of the untrusted
VLANs that carry client traffic during the NAC Authentication and Authorization, Scanning and
evaluation, remediation, and quarantine modes.
Figure 5-58
CAM
10.40.94.15
Layer 2 OOB Example
AD
MetroE
School District Network
VLAN 264
L2 Untrusted
CAS
L2 Trusted
VLAN 64
VLAN 64
227512
VLAN 264
NAC Authentication Options
The authentication option in the NAC solution can be broadly categorized as NAC Authentication or
NAC Single Sign On
•
NAC Authentication—NAC authentication gives the NAC system the role of authenticating users, a
user database, either local to the NAC system or a separate system such as RADIUS, or LDAP
•
NAC Single Sign On—NAC SSO, addresses systems that already perform authentication as part of
their normal operation. For example 802.1X, VPN access, or Active Directory. NAC SSO learns the
authentication state of clients through RADIUS accounting, or Active Director and therefore doesn't
require the user to reenter authentication.
Topology Considerations
The Layer-2 OOB solution relies upon there being a Layer-2 network connection available between the
client devices and the Cisco CAS. In Figure 5-58 a trunk connects the access switch to the
core/distribution switch. The Cisco CAS is connected to the core/distribution switch through two
Small Enterprise Design Profile Reference Guide
5-78
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
interfaces—trusted and untrusted. In such a simple network it is relatively easy to provide a Layer-2
connection between the client and the Cisco CAS, for larger networks Layer-3 OOB may be a better
choice.
The roles of the untrusted and trusted interfaces:
•
Untrusted Interface—The untrusted interface connects the client to the to the Cisco CAS during the
NAC Authentication and Authorization, Scanning and Evaluation, Remediation, and Quarantine
modes
•
Trusted Interface—The trusted interface connects the NAC CAS to the “normal” network interface.
This makes a network connection available to the CAS while it is sitting between the client and the
network, thus allowing client access to services such as DHCP and DNS -and user defined services.
Once a client has successfully completed its authentication and scanning phases, the CAM uses
SNMP to change the client VLAN, on the access switch, from the untrusted VLAN to the trusted
VLAN. Thus providing a direct connection to the network that was on the other side of the CAS (the
trusted network).
Availability Considerations
Both the CAS and CAM are both highly involved in client network access, and consideration must be
given to the impact on clients if either a CAS or CAM should fail or need to be taken out of service for
a period of time.
The CAS is inline with client devices during the authentication, authorization, and posture assessment
phases of NAC, and if “In Band NAC” is being used it may be inline at all times. A CAS outage in an
OOB deployment would not impact already connected clients but would prevent network access for new
clients. A CAS outage for “In Line” clients prevents access for all clients.
In situations where availability of a CAS is critical an HA CAS solution may be implemented where a
pair of CAS servers are installed using a primary CAS, and a secondary in hot standby. For more
information refer to the Cisco NAC Appliance - Clean Access Server Installation and Configuration
Guide.
The CAM is also a critical part of the authentication, authorization, and posture assessment phases of
NAC. Although the CAM doesn't pass client traffic, the impact of its availability needs to be considered
in the network design as well. Like the CAS the CAM has a HA solution that provides for a primary
server and a hot standby secondary server. In addition, each CAS may be configured with a “fallback”
option (as shown in Figure 5-59) that defines how it will manage client traffic in a situation where the
CAM is unavailable.
In both HA CAM, and HA CAS, HA licenses are use that address the HA role of the server.
The use of the HA features will be dependent upon the company unique requirements, but CAS fallback
should always configured to ensure that critical network services are available in even of a network
outage.
Small Enterprise Design Profile Reference Guide
5-79
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
Figure 5-59
CAS Fallback
Basic Clean Access Switch Configuration
For OOB-based Clean Access some simple configuration must be performed on the switches
implementing NAC This configuration is primarily to enable SNMP communication between the
switches and the CAM. The table below shows a simple SNMP v1 configuration (SNMPv2c and
SNMPv3 are supported).
In addition to the switch SNMP configuration, the required trusted and untrusted VLANs must exist and
be operational on the switch, as illustrated in the configuration shown in Table 4. If a switch has more
than one IP address the snmp-server source interface must be specified, as the CAM must be configured
with the source IP address that OOB SNMP messages will originate from, alternatively all IP addresses
of interfaces on the switch can be added to the CAM. If SNMP access filtering is applied on the switch
(as recommended as a best practice) the CAM must be added as a trusted address.
Basic Clean Access Out-of-Band Switch Configuration
Switch Port Configuration
Global Switch Configuration
snmp trap mac-notification
change added
snmp-server enable traps mac-notification
snmp-server enable traps snmp linkup linkdown
mac-address-table aging-time 3600
snmp-server host 172.16.1.61 traps version 1 cam_v1
udp-port 162 mac-notification snmp
NAC CAS Connection
The NAC CAS connects to the core/distribution switch using two switch ports (which are not configured
in EtherChannel). One is the untrusted port that serves the VLANs used by the clients prior to their
authentication and authorization. The second is the trusted port that connects to the VLANs used once
Small Enterprise Design Profile Reference Guide
5-80
Chapter 5
Small Enterprise Design Profile(SEDP)—Security Design
Network Security Deployment
clients have successfully completed the NAC process. Both trusted and untrusted ports are required, even
if OOB NAC is used, as the CAS requires access to the trusted VLANs during the NAC process. The
following is an example NAC CAS connection configuration.
interface GigabitEthernet1/0/4
description NAC Trusted Eth0
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 48,57,62
switchport mode trunk
spanning-tree portfast trunk
!
interface GigabitEthernet1/0/8
description NAC Untrusted Eth1
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 61,248,257
switchport mode trunk
spanning-tree portfast trunk
Basic 802.1X Switch Configuration
The basic 802.1X configuration controls access to an access VLAN depending upon the success or
failure of the 802.1X authentication. If the 802.1X authentication is successful, there are three basic
options:
•
Access to the VLAN configured on the switch port
•
Access to the VLAN configured on the switch port an controlled by a access list downloaded from
the AAA server
•
Access to a VLAN passed to the switch by the AAA server
The following is an example 802.1X configurations.
Example 3750 802.1X PC Port Configuration
Example 3750 Global Configuration
authentication port-control auto
authentication periodic
dot1x pae authenticator
aaa new-model
aaa authentication dot1x default group radius
dot1x system-auth-control
ip radius source-interface Vlan300
radius-server host 10.40.62.9 auth-port 1812 acct-port 1813
key cisco
radius-server host 10.40.94.9 auth-port 1812 acct-port 1813
key cisco
For more information about the 3750 802.1X configuration, refer to the following documents:
•
Catalyst 3750-E and 3560-E Switch Software Configuration Guide, 12.2(50)SE ->Configuring IEEE
802.1x Port-Based Authentication
http://www.cisco.com/en/US/docs/switches/lan/catalyst3750e_3560e/software/release/12.2_50_se/
configuration/guide/sw8021x.html
•
Catalyst 2960 Switch Software Configuration Guide, Rel. 12.2(50)SE Configuring IEEE 802.1x
Port-Based Authentication
http://www.cisco.com/en/US/docs/switches/lan/catalyst2960/software/release/12.2_50_se/configur
ation/guide/sw8021x.htm
Small Enterprise Design Profile Reference Guide
5-81
Chapter 5
Network Security Deployment
Small Enterprise Design Profile Reference Guide
5-82
Small Enterprise Design Profile(SEDP)—Security Design
CH A P T E R
6
Small Enterprise Design Profile
(SEDP)—Context-Aware Services
This chapter focuses on the application of general design best practices for Cisco Context-Aware
Services (CAS) and the Cisco Mobility Services Engine (MSE) as they relate to small enterprise designs.
Note that while this chapter attempts to be comprehensive, it is not intended to be a standalone technical
guide on Cisco CAS, RFID technology, or the Cisco MSE. For comprehensive configuration and
deployment information, the reader should refer to the in-depth configuration and deployment guides
mentioned throughout this chapter.
Introduction
What Are Context-Aware Services?
Context-Aware Services provides the ability to dynamically capture and use contextual information
about assets to optimize existing communications flows and organizational processes or facilitate the
establishment of new ones. Contextual information can be collected for assets involved in almost any
activity or process and this includes not just network endpoint devices (such as laptops and VoIP Phones)
and products (such as video projectors), but in some cases also the users that are associated with such
devices.
In environments where the Cisco Unified Wireless Network has been deployed, Context-Aware Services
makes use of embedded 802.11 wireless network interface adapters (radios) in wireless client devices to
accumulate contextual information about those assets or the user associated with the asset. For example,
the location of a wireless laptop can be calculated via several different approaches or the user name
associated with the laptop's user may be collected.
For assets that do not possess embedded wireless interfaces, external active Radio Frequency
Identification (RFID) tags and sensors can be used to provide location input and monitor ambient
environmental characteristics. Sensor capabilities can be directly embedded into active RFID tags in
order to link the data captured (for instance, whether the asset is currently in motion) with the location
of the asset. The algorithms used to determine location vary depending on the Radio Frequency (RF)
environment and the accuracy required for a specific application.
In some cases, it may be necessary to track an asset with a high degree of accuracy throughout a small
enterprise, such as when it is necessary to determine where a missing valuable asset is currently located).
On the other hand, some applications using context-aware services may only require general indication
of whether an asset is in or out of a permissible zone (such as the confines of an equipment storage area,
for example).
Small Enterprise Design Profile Reference Guide
6-1
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Introduction
Context-Aware Services can also provide location and other contextual information for wired devices
attached to Cisco Catalyst LAN switches, such as the 2960G, 3560E, 3750E, 3750G, 4500, and 4900
series.
Note
For the most current information regarding whether any specific Cisco Catalyst switch supports
context-aware services, please refer to your Cisco Catalyst switch documentation.
Catalyst switches that are context-aware can provide civic location details for wired devices to the Cisco
Mobility Services Engine based on pre-configured information specified for each switch port. This
information can then be presented to users in a tabular format combined with other contextual
information such as user name, device serial number and emergency location identifier numbers
(ELINs).
Note
A civic location specifies the civic address and postal information for a physical location using fields
such as the number, street or road name, community, and county assigned to residential, commercial,
institutional, and industrial buildings (e.g., 31 Main Street, AnyTown, AnyState 90004). An emergency
location identifier number (ELIN) is a number that can be used by the local public safety answering point
(PSAP) to look up the geographic location of the caller in a master database known as the automatic
location information (ALI) database. The ELIN also allows the PSAP to call back the emergency caller
directly in the event the phone call is disconnected.
For a more detailed overview of Context-Aware Services, refer to the following URL:
http://www.cisco.com/en/US/solutions/collateral/ns340/ns394/ns348/ns788/solution_overview_c22-47
5173.html.
Why Use Context-Aware Services?
The information that can be provided by Context-Aware Services across its application API can
generally be classified into five functional categories, as shown in Figure 6-1.
Is It Here?
Five Functional Categories of Context-Aware Services
Where Is It?
What Is Its
Condition?
Condition
Asset Tracking
Tracking
Zone/
Inventory
Management
Small Enterprise Design Profile Reference Guide
6-2
What Is Its
Status?
Where in
My Network
Is It?
Status
Network
Location
Services
227588
Figure 6-1
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Introduction
Is It Here?—Zone or Inventory Management
Zone or inventory management applications that utilize Cisco Context-Aware Services can define
specific zones in which they monitor mobile assets that possess embedded wireless interfaces or have
been outfitted with Cisco Compatible Extensions compliant RFID tags. These devices can be tracked and
monitored when they enter and exit both permissible and non-permissible areas. Notifications can be
generated when monitored assets stray into areas where they are not allowed. Examples of how zone or
inventory management may be used in the small enterprise environment can include the following:
•
Triggering the issuance of notifications to administrators or security personnel when monitored
assets are moved out of authorized areas
•
Alerting appropriate parties when persons equipped with RFID-enabled ID badges enter
unauthorized areas
•
Providing indication of proper staffing levels in critical environments. The number of doctors or
nurses currently located within a hospital emergency room or the confines of an intensive care unit
(ICU) can be carefully monitored. If the level of staff drops unacceptably, an alert can be issued to
the appropriate recipients so that minimum staffing requirements can be maintained.
Further details about zone or inventory management can be found at the following URL:
http://www.cisco.com/en/US/solutions/collateral/ns340/ns394/ns348/ns788/solution_overview_c22-47
5178.html.
Where Is It?—Asset Tracking
Asset tracking applications that incorporate Context-Aware Services can help locate assets within the
enterprise, whether they are connected to wired or wireless infrastructure. In this way, Cisco
Context-Aware Services can provide great value to administrators, staff, security personnel, or anyone
who must quickly and effectively locate and recover missing assets. Examples of how asset tracking may
be used in the small enterprise environment include:
•
Locating wired and wireless portable assets (such as a portable video projector or flat panel display)
for meetings, customer presentations or other activities
•
Quickly locating personnel possessing wireless VoWLAN phones, RFID-enabled badges or other
trackable devices in both emergency and non-emergency situations
•
Identifying the past pattern of movement associated with an asset by reviewing its past location
history, in both tabular and graphical formats. Such audit trails can be especially useful when
incorporated into security applications that can combine this information with other information
sources, such as video surveillance.
Further detail regarding asset tracking can be found at the following URL:
http://www.cisco.com/en/US/solutions/collateral/ns340/ns394/ns348/ns788/solution_overview_c22-47
5177.html.
What Is Its Condition?—Condition Tracking
Condition tracking applications utilizing Context-Aware Services can monitor select characteristics of
an asset's internal or external environment, such as variations in temperature, humidity, pressure,
quantity, fluid volume, etc. Any change in these characteristics beyond set thresholds can trigger alerts,
notifications, or other application-dependent actions. Examples of how condition tracking may be used
in the small enterprise environment include:
Small Enterprise Design Profile Reference Guide
6-3
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Introduction
•
Temperature and humidity telemetry passed by RFID tag external sensors can be utilized in food
service applications to monitor the temperature of food refrigeration units, guarding against costly
spoilage and premature replacement.
•
External fluid level sensors placed in combination with compatible RFID tags can be used to
monitor critical fluid levels in building maintenance applications, such as fuel oil levels for
emergency power generators and remote fuel storage for building heating.
•
Indication of excessive or insufficient pressure in heating and cooling and other critical systems,
which can be passed as telemetry data using properly equipped RFID tag sensors.
What Is Its Status?—Status Monitoring
Applications that monitor changes in user and asset status can use Context-Aware Services to detect
status transitions, such as a change from a normal state to one indicating that an extraordinary event has
occurred. Examples of how this may apply to the small enterprise environment include:
•
RFID tags with user signalling buttons could be used to covertly pass indication of situations where
assistance is needed, along with the location of the tag at the time of activation.
•
Attempted asset tampering, such as the removal of RFID asset tags themselves or jostling of any
type, can trigger status monitoring applications to generate alerts.
•
The introduction of new assets into any small enterprise site, or any changes in the motion status of
existing assets above a certain threshold, can serve as preliminary indication to building energy
management systems regarding potential building environmental modifications. Such modifications
may include changes to lighting, zone heating, or zone cooling settings for optimal efficiency.
Where Is It in My Network?—Network Location Services
Network location applications interfacing with the Context-Aware Services can help optimize the
distribution of both wired and wireless network resources, reduce troubleshooting time, help eliminate
waste due to use of network resources by unauthorized “rogue” devices, and help lower the overall total
cost of network operation. Examples of how this may apply to the small enterprise environment include:
•
Location and removal of unauthorized 802.11 wireless devices operating within a site, which can
help reduce the enterprise’s exposure to various risks, such as maliciously introduced software and
illegal file sharing.
•
Ongoing tuning of the wireless network by identifying areas where wireless users routinely
congregate. This may prompt adjustments in wireless network parameters, the number of access
points deployed, or the placement of access points.
Further detail regarding network location services can be found at the following URL:
http://www.cisco.com/en/US/solutions/collateral/ns340/ns394/ns348/ns788/solution_overview_c02-47
4514.html.
When combined with other applications constructed to take advantage of the Cisco Context-Aware
Services API, Context-Aware Services can serve as an enabler for entirely new application functionality.
For example, a context-aware application might provide the ability for an employee to locate other
employee team members with specialized skills (such as an employee who is also a certified emergency
medical technician or emergency response team member). Once such specialists are located, contact can
be initiated with the nearest such specialist that is available to deliver assistance. Context-aware services
enhance the experience of users while at the same time improving their overall efficiency and
productivity.
Small Enterprise Design Profile Reference Guide
6-4
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Introduction
Cisco Context-Aware Components
The components of the Cisco Context-Aware Services are shown in Figure 6-2.
Cisco Context-Aware Services Components
Application and
Management
Figure 6-2
W
Mobility Services Engine (MSE)
with Context-Aware Software
N
S
Cisco Wireless
Control System (WCS)
E
Open API
Other Context-Aware
Applications
Si
Network
WLAN Controller
WiFi TDoA
Receiver
CAPWAP
Access
Points
CAPWAP
LWAPP
Catalyst
Ethernet
Switch
TDoA
RSSI
Triggers
Devices
Wired Clients
Cisco Compatible Extensions RFID Tags
WiFi Clients
227589
Chokepoint
RSSI
Wired and Wireless Client Devices
•
Wired or wireless (Wi-Fi) devices—Mobile wireless devices (asset tags, WiFi equipped computers,
mobile stations, etc.) that interact with the network and whose location and other contextual
parameters can be monitored by Context-Aware Services. Wired devices are generally equipped
with an Ethernet interface which is attached to a Cisco Catalyst switch that supports context-aware
services. In addition, some devices may possess both wired and wireless interfaces. Without
context-aware services deployed, a wired Cisco IP phone originally deployed in a conference room
that is subsequently moved to an adjacent conference room across the hall, would require any
location information associated with it to be manually updated. With Context-Aware Services and a
context-aware Cisco Catalyst switch, the location of the Cisco IP can be dynamically updated to
Small Enterprise Design Profile Reference Guide
6-5
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Introduction
reflect its new location within seconds after it has been plugged into its new location. Via the
Mobility Services Engine's API, this information can be provided to various context-aware
applications, including the Cisco Wireless Control System (WCS).
•
Cisco Compatible Extensions RFID Tags—These RFID tags can be physically attached to assets
(regardless of whether the asset itself contains a wired or wireless network interface adapter) and
can pass some contextual information on behalf of the asset to Cisco Context-Aware Services.
Compliance with the Cisco Compatible Extensions program helps ensure that RFID tags adhere to
predefined transmission formats such that the contextual information they capture can be made
readily available to other Context-Aware Services components, including safety and security
applications from Cisco partners. Sensor capabilities can be externally mounted or directly
embedded into tags in order to link the data captured (for instance, motion, or temperature data) with
the location of the mobile asset. Externally mounted sensors can also be placed in fixed locations,
like a refrigerator or a storage room.
For more information about the Cisco Compatible Extensions for RFID Tags program, refer to the
following URL: http://www.cisco.com/web/partners/pr46/pr147/ccx_wifi_tags.html.
•
Chokepoint triggers—Chokepoint triggers (sometimes referred to as Exciters) are an optional
component that can greatly enhance asset tag functionality, providing for finer tag location
granularity and improved accuracy by localizing tagged assets within multi-floor structures, in
presence detection scenarios, or when passing through chokepoint areas, such as entrances and exits.
RFID asset tags that enter the proximity of a chokepoint trigger can change their normal behavior
based on a set of pre-programmed instructions. RFID tags that have been stimulated by chokepoint
triggers in this fashion send notifications and contextual information via the wireless infrastructure
to the Cisco Mobility Services Engine.
Cisco Unified Network
This multipurpose network contains the wired and wireless infrastructure required to address converged
data, voice, and video requirements, as well as providing the foundation for use of context-aware
services.
•
Note
•
Context-Aware Cisco Catalyst switches—Such as the 2960G, 3560E, 3750E, 3750G, and 4500
series switches that support the specification of civic and emergency location identification number
(ELIN) location information and the transmission of this information to Cisco Context-Aware
Services. This functionality allows for contextual information associated with wired devices to be
tracked using Context-Aware software on the Mobility Services Engine. Switches transmit relevant
contextual information to the MSE for all of the devices attached to them. This information may
include the physical mailing or street address location associated with the attached device (the civic
address) as well as other information such as the IP address, MAC address, port, VLAN, and user
name. Typically, this information is obtained using switch features such as IEEE 802.1x, Dynamic
Host Configuration Protocol (DHCP) snooping, Dynamic Address Resolution Protocol (ARP)
Inspection (DAI), and IP Source Guard. Additionally, if the end device runs the Cisco Discovery
Protocol or Link Layer Discovery Protocol for Media Endpoint Devices (LLDP-MED), additional
information, such as the version number and serial number, can also be sent to the MSE.
At this time, serial numbers of attached devices are reported to the MSE Context-Aware Service
only if the device supports LLDP-MED.
WLAN controllers—WLAN controllers (and the embedded software residing within them) provide
for the aggregation and transfer of device tracking and statistics information for RFID tags, mobile
wireless clients, and any rogue devices detected.
Small Enterprise Design Profile Reference Guide
6-6
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Introduction
– Access points—In addition to their fundamental role of providing access for wireless clients,
Cisco Aironet access points provide measurements of received signal strength from both
wireless client devices and RFID tags and subsequently forward this information to the Mobility
Services Engine via their registered WLAN controller.
– Received Signal Strength Indication (RSSI)—This is a mechanism used to determine device
location by carefully considering the measured strength of a radio signal at several points in an
indoor environment. Used by the Cisco Mobility Services Engine for WLAN clients, RFID tags,
and rogue devices, this algorithm is based on the signal sent from the mobile asset to different
access points deployed within the small enterprise site. RSSI is usually preferred for indoor or
low ceiling environments, both of which can result in high degrees of signal reflection.
– Wi-Fi Time Difference of Arrival (TDoA) Receiver—Wi-Fi TDoA receivers are optional
components used in very large, open environments to locate assets equipped with RFID tags
with greater accuracy and precision than is possible using other techniques.
Note
Although useful in extending Context-Aware Services for RFID tagged assets in outdoor venues,
the use of TDoA receivers were not included in the SEDP Mobility design.
For a much more detailed explanation of both RSSI and other wireless location algorithms, refer to the
Wi-Fi Location-Based Services Design Guide 4.1 at the following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Mobility/wifich2.html#wp1049520.
Management and Applications
•
Cisco Mobility Services Engine (MSE) with Context Aware Services Software—The Cisco Mobility
Services platform can host multiple independent services possessing high-level capabilities that can
enhance both wireless and wired network infrastructures. One of these is Cisco Context-Aware
Services which can capture, store, and analyze contextual information from multiple wired and
wireless networks simultaneously. When Context-Aware Services is deployed in accordance with
generally accepted best practices, both wired and wireless network infrastructure devices
(controllers and switches) may send raw location measurement data, device attachment, and other
contextual information to the MSE regarding the presence of any wired clients, wireless clients,
RFID tags, or rogue devices. Both wired and wireless network infrastructures communicate with the
MSE using the Cisco Network Management Services Protocol (NMSP), which is a Cisco-defined
protocol used for secure communication between the MSE and other context-aware network
infrastructure components. The MSE sits out of the data path of the wireless LAN and receives data
from WLAN controllers and context-aware switches via the use of NMSP.
The location of WLAN clients and RFID tags on the Cisco Mobility Service Engine is calculated by
one of two software service modules:
– Cisco Context-Aware Engine for Clients, which handles all context-aware operations involving
RSSI location of Wi-Fi clients, rogue clients, and rogue access points. This engine also handles
context-aware operations for wired clients.
– Cisco Context-Aware Engine for Tags, which handles all context-aware operations involving
TDoA and RSSI location of Cisco Compatible Extensions compliant RFID tags.
Context-Aware Services software, when operating alone on the Cisco MSE, is capable of servicing
up to a maximum of 18,000 simultaneously tracked devices per single MSE-3350 appliance and
2,000 simultaneously tracked devices per single MSE-3310 appliance.
Small Enterprise Design Profile Reference Guide
6-7
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Introduction
Refer to the following data sheet for more information regarding the Cisco 3300 Series Mobility
Services Engines
http://www.cisco.com/en/US/prod/collateral/wireless/ps9733/ps9742/data_sheet_c78-475378.html
•
Wireless Control System (WCS)—The Cisco Wireless Control System is a management platform
that also contains a context-aware client application that interacts with the Mobility Services
Engine. The primary role of the context-aware client application is to provide access to the
contextual information contained on the MSE using the MSE's application programming interface
(API). The application can then either present this information to the user directly (such as is seen
in a graphical location map or a table of location values) or enable other processes to accomplish
relevant tasks using this information that would otherwise be difficult to achieve. The Cisco WCS
can also serve in a special secondary role as a control client that possesses the ability to configure
MSE operational parameters.
For more detailed information on the Cisco Wireless Control System management server and its
capabilities, including its ability to serve as a context-aware application client, refer to the
documentation at the following URL:
http://www.cisco.com/en/US/prod/collateral/wireless/ps5755/ps6301/ps6305/product_data_sheet0
900aecd802570d0.html.
•
Other Context-Aware Applications—Other context-aware client applications from third party Cisco
Technology Development Partners may also access the MSE via its open API, which is based on
Simple Object Access Protocol (SOAP) and XML protocol. Access to this API is available to any
Cisco technology partner. Context-aware applications developed by Cisco Partners often deliver
specifically targeted application functionality that is often not available from other sources.
For more information on the Cisco Context-Aware Services API, refer to the following URL:
http://developer.cisco.com/web/contextaware/home.
Context-Aware Component Interaction
Figure 6-3 provides a more detailed illustration of the protocol interaction between the various
Context-Aware Service components.
Small Enterprise Design Profile Reference Guide
6-8
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Introduction
Protocol Interaction Between Context-Aware Components
WCS
Client
Browser
Cisco Partner
Context-Aware
Applicatons
HT
TP
SOAP/XML
S
WCS Server
SN
MP
W
Wireless LAN
Controller
NMSP
CA
AP
PW
PW
AP
CA
CAPWAP
L
Access
Points
CAPWAP
WiFi Handsets, clients, rogues,
WFi Tags and chokepoint triggers
N
S
E
ns
atio
ic
otif
/XM
N
Mobility Services
Engine (MSE)
Catalyst
Switch
Wired Ethernet Clients
227590
MP
SN
NSMP
Figure 6-3
For wireless clients, tags, and rogues, Cisco Aironet access points use the Control and Provisioning of
Wireless Access Points (CAPWAP) protocol to forward the RSSI of detected clients, tags, and rogues to
the WLAN controller to which they are currently registered. The wireless LAN controller aggregates this
information on a per device basis from all registered access points detecting the wireless device’s signal.
This information is then forwarded to the MSE using the NMSP protocol via an authenticated and
encrypted session. The appropriate Context Aware software engine on the MSE then uses the RSSI data
received from one or more WLAN controllers to determine the location of the wireless device
Note
A rogue access point is any access point that is determined not to be a member of the same mobility
group as the WLAN controller to which the detecting access points belong. A rogue client is any client
that is currently associated to a rogue access point.
For wireless clients, rogue access points, and rogue clients, the Context-Aware Engine for Clients is used
to process the received RSSI information. For operations involving RFID tags, however, the
Context-Aware Engine for Tags is used instead. Using Context-Aware Services and the Mobility
Services Engine, the Cisco Unified Network can readily detect 802.11 Wi-Fi active RFID tags that are
compliant with the Cisco Compatible Extensions for Wi-Fi Tags specification (such as those from
AeroScout, WhereNet, G2 Microsystems, and others). Through the MSE and WCS, the location of these
RFID tags can then be displayed on WCS floor maps using a yellow tag icon.
Small Enterprise Design Profile Reference Guide
6-9
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Introduction
Note
The Context-Aware Engine for Tags in Cisco Context-Aware Services Release 6.0 only tracks RFID tags
that are compliant with the Cisco Compatible Extensions for Wi-Fi Tags specification.
Cisco Compatible Extensions compliant active RFID tags are detected on a Wi-Fi network based on
periodic frames1 that are sent by the tag using a Layer-2 multicast. The delay between these periodic
frames can be programmed based on the specific application use case. In most cases, tags are configured
to transmit periodic frames every three to five minutes in order to strike an equitable balance between
location accuracy and good tag battery life.
RFID tags can also pass tag telemetry information upstream to the MSE as part of the tag message
payload. This contextual information (battery status, motion, temperature, pressure, humidity, etc.) is
received by access points and collected by WLAN controllers in a similar fashion as that described
earlier in this section. WLAN controllers will aggregate telemetry traffic from multiple tags and
eliminate any duplicate tag telemetry values that might be received. After the telemetry has been distilled
and cleansed of any duplicate information, the WLAN controller passes it to the MSE. The MSE updates
its internal databases with this information and in turn makes this information available to application
programs.
Properly equipped RFID tags can also indicate the occurrence of a priority event, such as one that might
result from the triggering of a tag tamper sensor or the depression of a tag call button. RFID tags indicate
that these types of events have occurred via additional information embedded in the tag messages that
are sent to the WLAN controller. This information is in turn passed northbound from the WLAN
controller to the MSE and the MSE can make this information available to application programs.
Several manufacturer's RFID tags include a secondary on-board magnetic signaling receiver, typically
set up to respond to the magnetic field component of a 125 kHz RF carrier. This secondary receiver
provides for additional tag functionality when tags enter into areas that are within close proximity to a
magnetic signaling transmitter or chokepoint trigger. Chokepoint triggers are proximity communication
devices that trigger asset tags to alter their configuration or behavior when the tag enters the chokepoint
trigger's area of operation or stimulation zone. This alteration could be as simple as causing the asset tag
to transmit its unique MAC address identifier. It could be significantly more complex, including causing
the tag to change its internal configuration and status. One of the prime functions of a chokepoint trigger
is to stimulate the asset tag such that it provides indication to the system that tag has entered (or exited)
the confines of a constricted physical area known as a chokepoint. Typical chokepoints include
entrances, exits or other types of physical constrictions that provide passage between connected regions
of a facility (such as a corridor or hallway).
Note
While chokepoint triggers are electronic devices that are typically deployed within physical chokepoints,
it is not unusual to hear the term “chokepoint” used rather loosely to refer to both the physically
constricted area and the associated electronic device.
Chokepoint triggers (including a very popular model manufactured by AeroScout Ltd. known as an
Exciter) may be connected to the wired infrastructure and are configured using each vendor's
configuration software. Once they have been configured, they can either remain connected to the
infrastructure full-time for management purposes or can be disconnected and operate in a standalone
mode, requiring a source of electrical power but no actual connectivity to the network.
1. These periodic frames are also sometimes referred to as “beacons”. They should not be confused with the
802.11 beacons that are sent by access points.
Small Enterprise Design Profile Reference Guide
6-10
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Introduction
The chokepoint trigger address information contained in the tag packet provides the MSE with enough
information to temporarily override any RSSI or TDoA localization currently in place for the tag and set
the current location of the RFID tag to the location of the chokepoint trigger. The size of a chokepoint
trigger's stimulation zone, or range, can extend from a radius one foot or less to over twenty feet,
dependent upon the vendor and the capabilities of the particular model.
Catalyst switches supporting context-aware services also make use of NMSP to interact with the MSE
similar to the manner described earlier for WLAN controllers. A major difference between how WLAN
controllers and Catalyst switches interact with context-aware services lies with the method used to
determine the location of switch attached wired devices. As explained earlier, localization of wireless
devices is performed by the MSE using a signal or time based technique, whereas the location of wired
devices is based on information sent to the MSE that originates in the switch configuration. The
information recorded by the MSE for the wired device includes the device MAC address, switch MAC
address, slot or port, IP address, and user name (if available). This information is sent whenever a device
link changes state. Context-aware Cisco Catalyst switches provide the MSE with the latest relevant civic
location and emergency location identification number (ELIN) information for all attached IP endpoints.
These endpoints may include IP phones, PCs, access points and other devices.
In Release 6.0, all civic and ELIN location information is configured locally at the switch, and shortly
after location changes are made using the CLI, they are propagated to the MSE. NMSP is used between
the switches and the MSE to maintain synchronization, and alert the switch as to the connection or
disconnection of devices.
Additional information regarding civic address location is available from the IETF in the following
RFCs:
Note
•
DHCP Option for Civic Addresses Configuration Information
http://www.rfc-editor.org/rfc/rfc4776.txt
•
Revised Civic Location Format for Presence Information Data Format Location Object
http://www.rfc-editor.org/rfc/rfc5139.txt
Proper validation of certificates between context-aware service components requires the participants to
possess sane clocks (clocks whose configured time does not differ from one another by large amounts).
In order to facilitate this, it is highly recommended that the clocks in the MSE, WCS, WLAN
Controllers, and any context-aware Cisco Catalyst switches be synchronized to a common time base
using the Network Time Protocol (NTP). The lack of clock sanity amongst context-aware components
in the network can cause NMSP sessions to fail if the configured date and time fall outside of the
certificate validity period and cause certificate validation to fail.
Network Mobility Services Protocol (NMSP)
The Network Management Service Protocol (NMSP) was designed to define intercommunication
between Mobility Service Engines and network access controllers over a switched or routed IP network.
An access controller can provide network access for either wired or wireless endpoints. Within the scope
of the small enterprise design discussed in this document, access controllers are represented by WLAN
controllers and context-aware Cisco Catalyst switches.
NMSP is a two-way protocol that can be run over a connection-oriented or a connectionless transport.
WLAN controllers and context-aware switches can use NMSP to communicate with one or more MSEs.
NMSP is based upon a bidirectional system of requests and responses between the MSE and the access
controllers.
Small Enterprise Design Profile Reference Guide
6-11
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
MSP also provides for a keep alive feature that allows either partner in a NMSP session to determine if
the adjacent partner is still active and responsive. Should an MSE fail, a WLAN controller or a
context-aware Cisco Catalyst switch will try to contact another MSE with which to communicate. If the
WLAN controller or context-aware Cisco Catalyst switch fails, all context-aware services being
provided to that WLAN controller or context-aware Cisco Catalyst switch are disabled until that WLAN
controller or context-aware Cisco Catalyst switch once again becomes active and re-establishes its
NMSP session.
Note
It is important to understand that the failure of an NMSP session has no direct impact on the ability of a
WLAN controller or context-aware capable Cisco Catalyst switch to pass normal client session traffic to
applications on the network. In other words, a failed NMSP session to a WLAN controller may affect
the ability of the MSE to provide updated contextual information on that controller and its resources.
But it does not affect the ability of the WLAN clients using that WLAN controller to logon to
applications residing on the network. This also applies to wired clients and context-aware Cisco Catalyst
switches.
NMSP uses Transport Layer Security (TLS) and TCP port 16113 on the WLAN controller or
context-aware Cisco Catalyst switch. The MSE will initiate the connection to the WLAN controller or
context-aware Cisco Catalyst switch, although once a secure session has been established between MSE
and its session partner, messages may be initiated in either direction. The TCP port (16113) that the
controller and mobility services engine communicate over must be open on any firewall that exists
between the controller and mobility services engine.
The MSE and the WLAN controller or context-aware Cisco Catalyst switch use Echo Request and Echo
Response control messages to maintain an active channel of communication so that the data messages
can be sent. The Echo Request message is a keep alive mechanism that allows either NMSP session
partner to determine if the other partner remains active and responsive. Echo Requests are sent
periodically (upon expiration of a heartbeat timer) by the MSE or its session partner to determine the
state of the NMSP session. When the Echo Request is sent, a NeighborDeadInterval timer is started. The
NeighborDeadInterval timer specifies the minimum time a session partner must wait without having
received Echo Responses to its Echo Requests, before the other session partner can be considered
non-responsive and the NMSP session is placed in an idle state.
Context-Aware Services in Small Enterprise Designs
Figure 6-4 provides a high level illustration of the integration of Cisco Context-Aware Services into the
Small Enterprise Design Profile. The key points of this integration into a Metropolitan Area Network
deployment are:
•
The presence of a centralized management entity (the wireless control system (WCS) at the main
site. In the case of Context-Aware Services, WCS also serves as a context-aware application client.
Main and remote small enterprise site context-aware users can log into WCS and query the
contextual characteristics (such as location) associated with wireless and wired client devices,
rogues and asset tags. In some cases, third-party context-aware application servers may also contain
context-aware applications located in the main site.
•
The presence of a per-site local Mobility Services Engine may be deployed to provide
Context-Aware Services to larger sites, where the anticipated number of total tracked devices is
significantly higher (e.g., greater than 500 and most likely 1000 or more). A locally deployed MSE
may also be justified if context-aware services are being utilized for applications and tasks that are
considered mission-critical to the function of the small enterprise or the safety and security of
employees, guests and visitors.
Small Enterprise Design Profile Reference Guide
6-12
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
•
Figure 6-4
The option of a centralized Mobility Services Engine with Context Aware software at the main site
used to provide the Context-Aware Services to smaller remote sites where the anticipated number of
total tracked devices per small enterprise is relatively low (e.g., less than 500).
High-Level View of the SEDP With Hybrid Context-Aware Services
Optional Backup
WLAN Controller(s)
Access
Control
System
Wireless
Control System
WLAN
Controller
MSE
Main Site
Services Block
Cisco 2960G
Mobility Services
Engine
Cisco 3750G
IP
IP
Metropolitan
Area
Network
Access
Control
System
Access
Control
System
MSE
Remote Site
Cisco 4500
Distribution
Mobility Services
Engine
WLAN
Controller
Cisco 2960G
Cisco 3560G
Cisco 2960G
IP
Cisco 3750G
IP
IP
229323
IP
Remote Site
Cisco 3750 Stack
Distribution
WLAN
Controller
It is our understanding that the enterprises addressed by the Small Enterprise Design Profile will contain
a mix of site sizes, with the main or headquarters site typically being larger than the remote sites. We
anticipate that many of the smaller remote sites may be well served at the distribution layer by the 3750
switch stack, whereas the larger main site may be outfitted instead with the 4500E distribution switch.
In situations such as this, the deployment model shown in Figure 6-4 can be used to provide
context-aware services to both types of locations.
Historically, context-aware services and the Cisco Mobility Services Engine (as well as its predecessor,
the Cisco Wireless Location Appliance) were designed to be deployed within modern switched LAN
environments. In such deployments, FastEthernet speeds and capacity (or better) are typically assumed
to be present throughout the local area network. Due to the lower speeds associated with traditional
wide-area networking technologies (such as frame relay, T-1, and so on), context-aware components
have not been recommended for deployment across traditional WANs. In fact, if context-aware services
are to be deployed in remote sites possessing only traditional WAN connectivity, Cisco Systems has
typically always recommended that the MSE be deployed locally along with WLAN controllers and any
Small Enterprise Design Profile Reference Guide
6-13
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
other components establishing NMSP sessions to the MSE. In addition, designers may wish to break up
large WCS network designs into smaller designs to avoid time outs that can occur between WCS and the
MSE during synchronization of very large network designs when using low-speed links2. While the cost
of local MSE deployment can be very applicable to very large remote sites whose device population can
justify it, this is typically not true in the case of smaller sites with lower device populations.
A key advantage of the SEDP design is that the use of a modern high-speed metropolitan area network
(MAN) to interconnect the main and remote sites offers far more bandwidth to each site than would be
seen with a traditional WAN deployment. In this case, with modern LAN-like speeds available across
the metropolitan area network, our approach begins to take on the look and feel of a local LAN
deployment, and thus the idea of deploying an MSE remotely from the other context-aware components
becomes much more feasible. Note however, this assumes sufficient bandwidth and infrastructure is in
place to assure that FastEthernet-like speeds are available to each site and that proper network protocol
identification, classification and QoS are all applied properly to manage congestion in the network.
In the context-aware services model shown in Figure 6-4, we take advantage of high-speed connectivity
across the MAN allow for a centralized MSE to provide context-aware services for smaller remote sites
in the enterprise. In this case, we assume that a small site might have a maximum of 250 to 500
simultaneously tracked devices. An exception clearly would be made for sites where context-aware
services are used for mission critical applications, such as enterprise safety and security applications. An
example of such a safety and security application might one that is used to ascertain the location of all
employees and staff during an emergency event, such as a fire, flood or crime evacuation. This type of
application obviously must be available at all times, including during any potential network outages,
hence the use of a centralized MSE would not be an option. Any other supporting applications that are
required for such mission-critical deployments of context-aware services should also be deployed locally
in this case.
Note
Based on our analysis of MAN capacity, traffic flows, classification and QoS, we believe the centralized
deployment of an MSE across a modern high speed metropolitan network is a viable concept. Although
a great deal of intensive functional testing was performed during the preparation of this chapter, time
constraints did not allow us to complete validation of centralized MSE deployments across metropolitan
area networks.
Larger enterprise sites using a 4500 series Catalyst switch for distribution are assumed to possess at least
500 or, more likely, 1000 or more simultaneously tracked wired or wireless devices. While it may be
possible to service these larger sites using a centralized MSE, the larger number of clients and the
increased amount of traffic placed onto the MAN between controllers, switches, and the MSE in this case
can pose more of a challenge, especially when there are large device populations that move frequently
and generate location updates on a regular basis. Careful analysis of data traffic and the judicious
application of QoS in the network becomes especially important.
At the current time, the most reliable and robust solution in the case of sites with large tracked device
populations is to deploy an MSE locally. Once again, in cases where context-aware services are regarded
as mission critical for the small enterprise site, other context-aware components (such as third-party
context aware application servers or in some cases the WCS as well) should also be deployed locally in
order to ensure that the context-aware solution is functional even in the rare case of a prolonged MAN
failure.
In small enterprises where there are many sites with either large tracked device populations or
mission-critical context-aware applications, it is important to keep in mind that at this time Cisco
officially supports the management of up to five (5) MSE platforms from a single WCS system. While
this is not a “hard” limitation on the number of MSE platforms that can be assigned to a single WCS
2. Or use a locally deployed WCS at each remote site that could optionally be managed by WCS-Navigator at a
central site.
Small Enterprise Design Profile Reference Guide
6-14
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
system, it is the limit to which testing has been performed. Therefore, in designs where there may be
greater than five sites equipped with locally deployed MSE platforms, you may wish to consider using
additional WCS systems as necessary. In this case, the use of WCS-Navigator (not shown in Figure 6-4)
should be considered at the main site to provide a single interface portal to as many as twenty (20) WCS
management systems and their associated Mobility Services Engines. WCS-Navigator is a management
aggregation platform that delivers enhanced scalability, manageability, and visibility of large-scale
implementations of Cisco WCS and the Cisco Unified Network. WCS-Navigator provides
straightforward access to information from multiple Cisco WCS management platforms. A single
WCS-Navigator management aggregator can support up to twenty (20) WCS management systems and
30,000 access points.
Note
Due to time constraints, scalability testing of WCS-Navigator in the Small Enterprise Design Profile
beyond four WCS systems was not able to be completed. Further information on WCS-Navigator can be
found at http://www.cisco.com/en/US/products/ps7305/index.html.
Component Capacities
Mobility Services Engine
Each Cisco Mobility Services Engine has a maximum device tracking capacity, defined as the maximum
number of active wired and wireless (clients, rogues, and RFID tags) that can be tracked by a single
Mobility Services Engine. This is a “hard” limit that is dictated by the licensing purchased for the
Context-Aware software as well as the presence of any other applications on the MSE. Once a Mobility
Services Engine has reached its maximum tracking capacity, any new devices that the MSE becomes
aware of beyond that limit are simply not tracked. It is important to note that while this section discusses
the maximum device tracking limits for the MSE, MSE licenses can be purchased supporting device
limits significantly lower than those shown here. Refer to the MSE Licensing and Ordering Guide
(http://www.cisco.com/en/US/prod/collateral/wireless/ps9733/ps9742/data_sheet_c07-473865.html)
for more information regarding the various combination of client and RFID tag tracking capacities
available for the MSE.
For Release 6.0, the maximum device tracking limits when using only the Context-Aware Services
software on the MSE are shown in the following table:
Mobility Service Engine Maximum Tracked Device Capacity
MSE-3350
18,000
MSE-3310
2,000
If you intend to use the MSE-3350 to deliver other services in addition to Context-Aware, the maximum
capacity shown above in Table 2 will likely be reduced. See the MSE Licensing and Ordering Guide
(http://www.cisco.com/en/US/prod/collateral/wireless/ps9733/ps9742/data_sheet_c07-473865.html)
for information on Context-Aware maximum tracked device limits with other co-resident MSE services.
When working within these maximum capacities, it is important to note that further category-specific
limits can be initiated at the designer's discretion via the MSE configuration. This allows, for example,
a maximum capacity of 2,000 tracked devices on a MSE-3310 to be further limited as 1,000 wired and
wireless clients, 500 RFID tags, and 500 rogue access points and clients. Partitioning the maximum
tracking capacity of the context-aware software in this manner prevents any single device category from
consuming more than its authorized share of the maximum tracking capacity of the system.
Small Enterprise Design Profile Reference Guide
6-15
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
In the high-level diagram shown in Figure 6-4, the MSE-3310 might be a good design choice for a locally
deployed MSE at our 4500-based main site. Its maximum tracked device capacity of 2000 devices should
scale well in this type of application, where we are assuming an estimated 1000-1250 total tracked
devices. This would leave ample MSE capacity in reserve for future growth at this location. Of course,
you could deploy with less capacity in reserve should you choose to, and elect to address future tracked
device growth at a later date via a hardware addition or upgrade. 4500-based sites that possess or
anticipate near-term tracked device needs exceeding 2000 devices should consider the MSE-3350
instead.
Except for very small enterprises, the MSE-3350 would typically be the best overall choice when
considering a centralized deployment using a high-speed metropolitan area network. In this way, it can
provide context-aware services for several smaller sites, each of which might possess an estimated 500
or fewer tracked devices. In very small enterprises (e.g., enterprises containing up to a total of four small
sites, for example) a centralized MSE-3310 may prove to be even more cost-effective.
WLAN Controllers
WLAN controllers also possess limitations on the maximum number of devices for which the controller
will track and aggregate contextual information. In Release 6.0, these limits are shown in Table 6-1.
Table 6-1
Maximum Device Limitation
WLAN Controller Clients Tags
Rogue Access Points Rogue Clients
4404
5000
2500
625
500
4402
2500
1250
625
500
2106
500
256
125
100
Note that these are indeed “hard” limits. In other words, once these limits have been achieved on a
WLAN controller, contextual tracking information for any new clients, RFID tags, rogue access points,
or rogue clients beyond these limits will be dropped until such time that older entries are pruned from
the controller's internal database. The client and tag limits are quite high and should not be easily
exceeded within most small enterprise sites.
The Context-Aware System Performance chapter of the Mobility Services Engine Context Aware
Deployment Guide
(http://www.cisco.com/en/US/products/ps9742/products_tech_note09186a00809d1529.shtml#casysper
f) also points out that a single MSE can support up to 500 total NMSP connections. Keep in mind this
includes not only the NMSP sessions to WLAN controllers, but NMSP sessions to any context-aware
Cisco Catalyst switches conducted from that MSE.
Note
Although a single MSE can technically support up to 500 NMSP sessions, scalability testing constraints
have only allowed for testing of 100 simulated NMSP connections to a single MSE at this time.
Wireless Control System (WCS)
With regard to the MSE, WCS interacts as a context-aware client and does not track devices itself when
used in conjunction with the MSE. Thus, there are no direct constraints on the maximum number of
tracked devices imposed by WCS itself.
Small Enterprise Design Profile Reference Guide
6-16
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
In addition to established WCS sizing and capacity guidelines for the number of supported controllers
and access points (listed in the System Requirements section of the Wireless Control System
Configuration Guide, Release 6.0,
http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/6_0wst.html#wp1061082),
there are a few indirect constraints relating to Context-Aware services that you should be aware of:
•
To maintain clarity and the speed of its graphical user interface, WCS only displays the first 250
wireless clients, RFID tags, rogue clients, or access points on a single floor map. To view graphical
location displays for any of these device categories beyond this limit, filtering (based on MAC
address, asset name, asset group, asset category, or controller) should be used to limit the number
of devices displayed at once (see the Floor Settings section of the Wireless Control System
Configuration Guide, Release 6.0,
http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/6_0maps.html#wp121096
9).
•
In Release 6.0, a single WCS can manage Context-Aware Services on up to five (5) Mobility Service
Engines. While defining more than five MSEs to a single WCS is possible, Cisco Systems has not
validated this level of operation.
•
In Release 6.0, Context-Aware Services on the Mobility Services Engine can be managed by only
one Wireless Control System.
•
In Release 6.0, WCS supports the creation of up to 124 WCS virtual domains.
•
Although there is no “hard” limit on the number of simultaneous WCS users supported per WCS
server in Release 6.0, Cisco has not validated more than five (5) simultaneous users per WCS server
at this time.
Context-Aware Engine for Tags (AeroScout)
The Context-Aware Engine for Tags used in version 6.0.85.0 of the MSE Context-Aware Services
software supports network designs containing up to 255 floor maps. In small enterprise designs, a
network design might be used to describe the overall small enterprise from the point of view of
context-aware services. An MSE network design typically consists of campus, buildings, and floor maps.
The limitation on the number of floor maps then could be interpreted as saying that a small enterprise
should not contain more than 255 floor maps (regardless of the number of buildings these floor maps are
spread amongst). If more than 255 floor maps are required, it is necessary to divide the small enterprise
up into two or more network designs, with each network design containing a subset of the total number
of floor maps.
Obviously, the impact of this restriction will vary with the number of floors in each building, multiplied
by the number of buildings contained within the entire small enterprise. For example, if all buildings in
the enterprise contain only a single floor, then up to 255 buildings could be defined in a single network
design before the conditions of this restriction are encountered. If each building each contained three
floors however, then only 85 site buildings could be defined before this limitation would take effect.
Small Enterprise Design Profile Reference Guide
6-17
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Integration within Small Enterprise Designs
MSE Connection to the Network
Figure 6-5 is an illustration of the rear panel of both the MSE-3350 and the MSE-3310. The Cisco
Mobility Services Engine is equipped with two 10/100/1000BASE-T Gigabit Ethernet ports (shown in
Figure 6-5 by the solid arrows) that can be used to directly connect the MSE to two different IP networks
(dual-homed). This makes it a simple affair, for example, to configure a MSE for service on network A
while affording it the capability to be managed out-of-band on network B if the need arises.
In the Small Enterprise Design Profile, we attach the MSE to the network via a single connection to the
NIC0 interface.
Figure 6-5
Note
Rear Panel Illustration of MSE-3350
The dual on-board Ethernet controllers on the MSE are not intended for redundant or simultaneous
connections to the same IP network. Any attempt to manually configure the MSE in order to try and
establish parallel, load balancing, or redundant Ethernet connections to the same IP network is not
recommended or supported at this time.
Clock Synchronization
The Mobility Services Engine, WCS, WLAN controllers, and any switches that support context-aware
services use Coordinated Universal Time (UTC)3 in their interaction. Because proper certificate
authentication relies on time base consistency between participating components, it is important to
ensure that these components are synchronized to a common time base throughout our design. In
addition, having components synchronized to a common time source makes troubleshooting much
easier, especially when having to look at events occurring within the logs of different network
components. And the output of coordinated information by a central source, such as WCS, makes much
more sense when the time stamps of all information displayed follow a logical flow and make sense to
the user.
Once time and date in each network component has been set initially, time synchronization should be
maintained using the Network Time Protocol (NTP). In the Small Enterprise Design Profile, these
components should be synchronized to the ISR router found at the site.
3. For applications such those anticipated in the small enterprise designs discussed in this document, Universal
Coordinated Time may be considered as equivalent to Greenwich Mean Time (GMT).
Small Enterprise Design Profile Reference Guide
6-18
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
NTP Configuration of the Mobility Services Engine
Configuration of the NTP server addresses used by the MSE is handled during installation and the
execution of the MSE automatic configuration script. An excerpt of that script is shown below. Detailed
information regarding the automatic configuration script can be found in the Automatic Installation
Script section
(http://www.cisco.com/en/US/docs/wireless/mse/3350/quick/guide/mse_qsgmain.html#wp1057105) of
the Mobility Services Engine Getting Started Guide,
http://www.cisco.com/en/US/docs/wireless/mse/3350/quick/guide/mse_qsgmain.html#wp1057105,
and in Appendix A
(http://www.cisco.com/en/US/products/ps9742/products_tech_note09186a00809d1529.shtml#appena)
of the Mobility Services Engine Context Aware Deployment Guide,
http://www.cisco.com/en/US/products/ps9742/products_tech_note09186a00809d1529.shtml#appena.
If you choose to enable NTP, the system time will be configured from NTP servers that you
select. Otherwise, you will be prompted to enter the current date and time.
NTP is currently disabled.
Configure NTP related parameters? (Y)es/(S)kip/(U)se default (S)kip: Y
Enter whether or not you would like to set up the Network Time Protocol (NTP) for this
machine.
If you choose to enable NTP, the system time will be configured from NTP servers that you
select. Otherwise, you will be prompted to enter the current date and time.
Enable NTP (yes/no) no : yes
Enter NTP server name or address: <IP address or DNS name of NTP server>
Enter another NTP server IP address (or none) none: none
NTP Configuration of WLAN Controllers
Configuration of the internal clock and the specification of which NTP servers to use for periodic time
synchronization can be performed on the WLAN controller using either the web GUI interface or the
command line interface.
If you did not configure the system date and time through the configuration wizard when the controller
was initially configured, or if you want to change your configuration, you can follow the instructions
located in the section entitled Managing the System Date and Time
(http://www.cisco.com/en/US/docs/wireless/controller/6.0/configuration/guide/c60intf.html) in the
WLAN Controller Configuration Guide 6.0
(http://www.cisco.com/en/US/docs/wireless/controller/6.0/configuration/guide/c60intf.html#wp11443
40) in order to configure the controller to obtain the date and time from a Network Time Protocol (NTP)
server.
NTP Configuration of the Wireless Control System (WCS) Server
Configuration of the internal clock and the specification of which NTP servers to use for periodic time
synchronization must be performed on the WCS server using the time and date capabilities of the WCS
host operating system in use (either Windows or Linux).
Small Enterprise Design Profile Reference Guide
6-19
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
RHEL-Based WCS Server
For a Redhat Linux-based WCS server, login to the host OS as root and use the following procedure to
synchronize the internal software clock to the NTP server, synchronize the software clock to the server's
hardware clock, and then maintain synchronization by starting the ntpd client daemon:
1.
clock—Displays the current setting of the software clock.
2.
/etc/init.d stop—Stops the ntpd client if it is already running.
3.
ntpdate <ntp server name or address>—Synchronizes the system software clock with the NTP
server.
4.
setup—Brings up a setup utility that allows you to choose to set the time zone (shown in Figure 6-6).
5.
hwclock-systohc—Writes the software clock settings to the hardware clock.
6.
/etc/init.d/ntpd start—Starts the ntpd daemon to keep the clock synchronized going forward. 4
Figure 6-6
Note
RHEL Setup Utility
There are various other approaches that can be used to set the time zone on a Linux system. The reader
is encouraged to consult the Redhat documentation for methods involving the use of the TZ variable or
symbolic links to the localtime file or a particular time zone file in the system’s time zone directory.
Windows 2003-Based WCS Server
For a WCS server based on the Microsoft Windows 2003 Server OS, use the following procedure to
synchronize and maintain the correct system time via the Windows Time service (see Figure 6-7):
1.
Check Settings>Control Panel >Administrative Tools>Services for the Windows Time service
and ensure that it has been started.
2.
Right click on the Task Bar clock and select Adjust Date/Time.
3.
Under the Date & Time tab, set the current date and clock time to the approximate time of your NTP
server.
4.
Set the Time Zone and Daylight Savings time selections appropriately.
4. If ntpd does not start as part of your system boot script, you might want to add it using the command chkconfig
--add ntpd.
Small Enterprise Design Profile Reference Guide
6-20
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
5.
Select the Internet Time tab, check the box to Automatically Synchronize With An Internet Time
Server, type in the DNS name or address of your NTP server, and then Apply.
Figure 6-7
Setting Time and NTP Server on Windows 2003
NTP Configuration of Context-Aware Catalyst Switches
In order to prevent any issues with authentication and NMSP session initiation, context-aware Cisco
Catalyst switches should be configured to utilize NTP in order to keep their clocks in synchronization
with other context-aware components. NTP is configured similarly amongst the various switch models
discussed in this chapter, and the most comprehensive information on how to configure a Catalyst switch
as an NTP client can usually be found in the configuration guide for the particular switch model. For
example, for the Catalyst 2960G NTP configuration is documented in Configuring NTP section of the
Catalyst 2960 Switch Software Configuration Guide
(http://www.cisco.com/en/US/docs/switches/lan/catalyst2960/software/release/12.2_50_se/configurati
on/guide/swadmin.html#wp1053923).
It is good general best practice to ensure time synchronization of all network components when possible.
However, from the perspective of context-aware services in the small enterprise designs discussed here,
only those switches that are actually participating in an NMSP session with the MSE require clock
synchronization.
Small Enterprise Design Profile Reference Guide
6-21
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
SEDP Wireless Control System (WCS) Context-Aware Considerations
The Wireless Control System is used for several important configuration tasks relating to context-aware
services in small enterprise designs:
•
Creation of a network design at either the enterprise or site level, which may include campus,
building, floor, and outdoor level maps.
•
Definition of WCS User Groups, which are used to define what management actions context-aware
users are authorized to perform.
•
Definition of Virtual Domains, which can be used to restrict which network resources users will
have the ability to manage via WCS.
•
Configuration of Mobility Service Engine operating parameters. This represents the next level of
MSE setup beyond that performed by the MSE automatic configuration script discussed in section
“NTP Configuration of the Mobility Services Engine” section on page 6-19.
•
Definition of Context-Aware Conditional Notifications, which defines how applications and parties
external to the site may receive notification of specific events pertaining to changes in contextual
characteristics associated with clients, tags or rogue devices.
In this section, we discuss only those areas where, in our testing we made use of significant WCS
features relevant to the integration of context-aware services, or where important configuration changes
were made that significantly differ from the defaults. This is not meant to serve as a comprehensive
configuration guide to all aspects of the WCS and MSE. Readers should refer to the WCS and MSE
configuration documents already cited throughout this document (including Context-Aware Services
General Best Practice References) for additional information regarding configuration parameters and
procedures that, while not discussed in detail here, must still be configured or performed properly.
Creation of a Network Design
Once access points have been installed and have registered with a controller, WCS can be configured to
manage the controllers and a network design can be set up. A network design is a representation within
WCS of the physical placement of access points and other context-aware components throughout a
facility or group of facilities. A hierarchy consisting of a single campus, the buildings that compose that
campus, the floors of each building, and any outdoor areas constitutes a single network design.
In small enterprise designs, the choice of whether to configure the entire enterprise or each individual
site at the campus layer depends to a large part on the on whether the sites each contain a single building,
or multiple buildings. If each site is comprised of one building and one building only, then the campus
layer of the network design can be the entire enterprise. On the campus map, each site would be
represented by a single building with one or more floors per building.
In other cases, each enterprise site might be composed of multiple buildings. In these scenarios, it makes
more sense to define each site as a campus unto itself.
A step-by-step set of configuration instructions regarding how to configure network designs consisting
of campus, building, and floor maps can be found in Chapter 5 of the Cisco Wireless Control System
Configuration Guide, Release 6.0,
http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/6_0maps.html#wp1203275.
Figure 6-8, Figure 6-9, and Figure 6-10 give an example of what a campus, building, and floor level
network design might look like for a small enterprise where all small sites are assumed to be comprised
of single buildings. In Figure 6-8, we use satellite imagery of the small enterprise’s physical location as
the backdrop for the campus map. Clicking on any of the icons takes us to the building map for the small
enterprise site. In this case, we select site number 1278, which then brings us to Figure 6-9.
Small Enterprise Design Profile Reference Guide
6-22
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Figure 6-8
Campus Level View of the Small Enterprise
The building map shown in Figure 6-9 indicates that this building contains a single floor. Clicking
directly on the building map takes us to the floor definition shown in Figure 6-10. This is where we
would actually see the location of wireless clients, active RFID tags and rogues displayed. Wired devices
that are attached to context-aware Cisco Catalyst switches are not displayed on floor maps in Release
6.0 of Context-Aware Services.
Figure 6-9
Building Level View for Site #1278
Small Enterprise Design Profile Reference Guide
6-23
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Figure 6-10
Note
Floor Level for Site #1278
Network designs are created in WCS, but they are not actually used for device tracking until they are
transmitted to the MSE via a process known as network design synchronization. Only after network
design synchronization has successfully occurred between WCS and its associated MSE will the network
design actually be used by the Context-Aware Engine for Clients and the Context-Aware Engine for
Tags. Synchronization of network designs and other components with the MSE is discussed in detail in
the chapter entitled “Synchronizing Mobility Services Engines” in the Context-Aware Service
Configuration Guide 6.0,
http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/msecg_ch3_CAS.h
tml.
WCS Users, User Groups, and Virtual Domains
When installed, WCS provides for a single root user, which will have access to all WCS functions. The
password for this root user should be protected and only known by those personnel at the main site data
center with a true need to know (e.g. those personnel responsible for the installation, maintenance, and
detailed administration of WCS). Instead of using the root user password for routine access to WCS, you
should create other users and grant them administrative access with privileges assigned as necessary via
the use of WCS user groups. Chapter 7 of the Cisco Wireless Control System Configuration Guide,
Release 6.0 (http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/6_0manag.html)
provides comprehensive instructions with regard to the proper procedure for configuring users and group
privileges on your WCS server. This chapter also contains a complete listing of the user groups available
in WCS as well as the privileges contained in each group.
Common sense should be applied in the assignment of user privileges in small enterprise designs. For
example, while only a very small set of key technical personnel should have access to the actual WCS
root user ID and password, you may wish to assign the ability to make WCS configuration changes to a
Small Enterprise Design Profile Reference Guide
6-24
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
somewhat larger audience. This larger group can be assigned as WCS “admin” users or assigned to the
“superuser” group. Most small enterprise users will likely require only the capability to view or monitor
network services that pertain to them or their workgroups, as opposed to administering the network in
WCS. For these users, the privileges accorded them by the WCS System Monitoring or Monitor Lite user
groups may be all that is required, depending upon the specific WCS monitoring functions you wish to
grant those users.
In our small enterprise design validation, the custom user-defined group shown in Figure 6-11 was found
to be very useful in limiting users to only monitoring context-aware information, as well as some basic
WCS alerts and events. For example, we may wish to allow a research specialist in a small enterprise
document center access to the context-aware functions listed under the WCS “Maps” function or search
for a device by name. But this same research specialist probably has no need to monitor network security
compliance reports, thus we have not enabled access to those reports for the research specialist’s
account. Keep in mind that the requirements of small enterprise users in your environment may be
different, so it may make sense for you to develop a custom WCS user group that closely fits your needs.
Small Enterprise Design Profile Reference Guide
6-25
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Figure 6-11
Custom User Group to Allow Context-Aware Monitoring
While WCS user groups define the WCS functionality users have been granted, WCS virtual domains
allow the network administrator logically partition the WCS management domain and limit management
access. In this way, the group of resources that the WCS functionality assigned to a user group may be
exercised against is restricted. A WCS virtual domain consists of a set of assigned devices and maps,
and restricts a user's scope to only information that is relevant to those devices and maps. Through a
assigned virtual domain, users are only able to utilize WCS functionality against a pre-defined subset of
the devices managed by WCS.
Small Enterprise Design Profile Reference Guide
6-26
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Users can be assigned one or more virtual domains, however only one assigned virtual domain may be
active for a user at WCS login. The user can change the current virtual domain in use by selecting a
different permitted virtual domain using the WCS Virtual Domain drop-down menu.
The WCS virtual domain can be used to limit the user's ability to even view certain resources inside WCS
that are not contained in their active assigned virtual domain. For example, the site manager of small
enterprise site A have the ability to view the location and other context-aware characteristics of wireless
assets due to his WCS user account being assigned to an appropriate user group permitting this level of
WCS functionality. But the virtual domain that this site manager is assigned may only allow such
functionality to be exercised against these assets if they are located within the small enterprise he is
assigned to manage. Thus, if our site manager attempted to discover the quantity and location of
RFID-tagged equipment in small enterprise site “B”, his assigned virtual domain prevent him from being
able to view site B resources.
Administrative personnel with enterprise-wide responsibilities, on the other hand, would be assigned a
virtual domain that includes all resources in the enterprise, across all sites, and could exercise the
functionality assigned to them by their user group against any of these resources. The virtual domain
assignment also helps prevent unnecessary inter-site WCS traffic, especially traffic whose nature might
be based more upon curiosity rather than actual need.
Note
WCS user groups assign what actions a user can take against a resource, whereas WCS virtual domains
determine what resources those user group actions can be applied towards.
There are two basic steps necessary to enable the use of virtual domains within WCS:
Note
1.
A virtual domain must be created, and the resources that we wish to include assigned to the virtual
domain. The process for creating and assigning network resources to the virtual domain is detailed
in Chapter 20 Virtual Domains of the WCS Configuration Guide 6.0
(http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/6_0virtual.html#wp1040
002).
2.
The virtual domain must be assigned to the user. The process for assigning a virtual domain to a user
is detailed in Chapter 7 Managing WCS User Accounts
(http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/6_0manag.html#wp1097
733).
It is important to note that in release 6.0, non-root WCS virtual domain users cannot access WCS
functions listed under the Services > Mobility Services main menu heading. This includes wired switch
and device location. Refer to Understanding Virtual Domains as a User, WCS Configuration Guide 6.0
(http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/6_0virtual.html#wp1120787)
for a complete list of WCS functions that are not available in non-root virtual domains.
In Release 6.0, since wired devices attached to context-aware Cisco Catalyst switches are displayed
using Services > Mobility Services > Context Aware Service > Wired > Wired Clients, only users
that are assigned to the root virtual domain are able to display context-aware information for these
devices.
Figure 6-12, Figure 6-13, and Figure 6-14 demonstrate the effectiveness of WCS virtual domains (note
that the current virtual domain in use by the logged-in WCS user is highlighted in each figure by the red
oval). In Figure 6-12, we can see that the “Small Enterprise” virtual domain user can see the entire set
of sites comprising the small enterprise, and is capable of applying any of the WCS functionality
accorded to them by their WCS user group assignment. This virtual domain setting might be appropriate,
for example, in the case of a person requiring the ability to view and potentially take action upon all
resources in the network. An example of personnel in a small enterprise that might require such
Small Enterprise Design Profile Reference Guide
6-27
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
capability would be network administration staff. However, keep in mind the immediately preceding
note regarding the viewing or administration of WCS resources appearing under the Services > Mobility
Services main menu heading.
Figure 6-12
Virtual Domain for Entire Small Enterprise
In contrast, Figure 6-13 and Figure 6-14 each illustrate how the user's view of the small enterprise can
be severely curtailed when a WCS virtual domain is applied. A virtual domain setting such as this might
be appropriate for most of the personnel within a site that might only need to work with resources located
in their site only. Figure 6-13 illustrates what a user in site #1278 would see if they were assigned a WCS
virtual domain that limited their resource visibility to only those resources associated with site #1278.
Figure 6-13
Virtual Domain Limited to Site #1278
In Figure 6-14, we see the results of a WCS virtual domain for a user in site #1266. Note that a user that
is assigned a virtual domain for site #1266 does not have visibility to any resources associated with other
sites within the enterprise. All that is visible to the site #1266 user in this case are the buildings and floor
maps that are associated with site #1266 and site #1266 only. We can see that plainly in Figure 6-14, as
the resources associated with site #1278 or any other site are simply not displayed to the site 1266 virtual
domain user.
Small Enterprise Design Profile Reference Guide
6-28
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Figure 6-14
Virtual Domain Limited to Site #1266
Figure 6-15 illustrates the typical result that occurs when a user attempts to view information for
resources outside the scope of their assigned virtual domain. In this case, we see the result of a user in
site #1266 attempting to access resources that are located in site #1278.
Figure 6-15
Virtual Domain Permission Error
Mobility Services NMSP Parameters
WCS 6.0 provides us with several NMSP parameters available that affect various NMSP protocol timing
characteristics between the MSE and its session partners. These parameters can be found at Services >
Mobility Services > System > NMSP Parameters, and apply globally to all NMSP sessions between
the selected MSE and any of its WLAN controller or Cisco Catalyst switch session partners. Complete
configuration information for configuring NMSP session timing parameters, as well as the default values
for these parameters, can be found at Configuring Mobility Services Engine Properties section of the
Context-Aware Service Configuration Guide
(http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/msecg_ch4_CAS.
html#wp1014368).
When deploying an MSE locally in the small enterprise, it is unlikely that these parameters would
require modification. This would be more the norm in large enterprise sites, where there might be several
thousand tracked devices, and a large quantity of wireless devices moving about on a regular basis.
However, in a centralized deployment, there is more of a chance that network congestion or other factors
may cause delays that could cause NMSP session time outs. While this should be minimized by the
appropriate identification and classification of NMSP data flows in the network along with properly
defined network QoS, there may be instances where adjustments to NMSP timing is required. In these
cases, the NMSP echo interval, neighbor dead interval, and response time out values can be increased to
limit the number of failed echo acknowledgments that may occur, especially in a centralized MSE
deployment.
Note
Readers are advised that a tremendous amount of functional validation was performed in association
with the content contained in this chapter. However, time constraints limited the degree of performance
validation that could be completed for Context-Aware Services across a simulated Metropolitan Area
Network (MAN). The deployment of Mobility Services Engines in a centralized fashion across a MAN
was not able to be fully validated due to these time constraints.
Small Enterprise Design Profile Reference Guide
6-29
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
To aid in determining whether echo packets are being dropped, you can use the WCS function Services
> Mobility Services > System > Status > NMSP Connection Status as shown in Figure 6-16. This
WCS menu panel displays all the NMSP session partners for this MSE, along with echo request and
response counts. For example, in the figure for the Mobility Services Engine with the hostname MSE1,
we see that there are currently two NMSP sessions to two different WLAN controllers.
Figure 6-16
NMSP Connection Status
We can see in Figure 6-16 that both of the NMSP WLAN controller sessions appear to be functioning
properly, as the number of Echo Requests issued is seen to be exactly equal to the number of Echo
Responses received. This might not always be the case and small static differences over a long period of
time do not necessarily indicate a serious problem. However, sluggish performance combined with a
regularly increasing discrepancy in the delta between the number of requests issued and responses
received could be indicative of the NMSP session timing out. In a centralized deployment, this may be
due to unforeseen levels of congestion or other resource constraint. In this case, raising the response
time-out may assist in alleviate the time-outs. Keep in mind, however, that a successful centralized
deployment assumes that there is sufficient MAN capacity available to each site and that QoS has been
applied appropriately.
Figure 6-17 gives an example of the NMSP Parameter screen that is located at Services > Mobility
Services > System > NMSP Parameters, which we can use to change the system defaults. In
Figure 6-17, the network administrator or network technician has instituted the following changes from
the defaults: the echo interval has been raised to 30 seconds, the neighbor dead interval has been raised
to 60 seconds, and the response time-out has been raised to 5 seconds.
Small Enterprise Design Profile Reference Guide
6-30
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Figure 6-17
Note
Example of NMSP Parameter Modification
Although the NMSP configuration worked well for us in our lab testing, we cannot predict and simulate
each and every condition that might occur in a production deployment. Therefore, it is important that
you take the time to understand the function of these parameters and especially the fact that they can be
adjusted beyond the values illustrated here in order to promote improved NMSP session stability and
network performance.
Further information on these NMSP parameters can be found in the Configuring Mobility Services
Engine Properties section of the Context-Aware Service Configuration Guide
(http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/msecg_ch4_CAS.
html#wp1014368).
Context-Aware Service Parameters—Tracking
As mentioned earlier, Context-Aware Services can track up to a maximum of 18,000 licensed devices
when using the MSE-3350 hardware platform, and up to a maximum of 2,000 licensed devices when
using the MSE-3310 platform. The absolute limit on the number of clients or tags that can be tracked is
determined by the hardware platform used, the presence of any other applications co-residing on the
MSE, and the level of licensing purchased. The WCS tracking parameters configuration panel (located
at Services > Mobility Services > Context Aware Service > Administration> Tracking Parameters)
allows the administrator to pre-determine just how much of the MSE's maximum licensed tracking
capacity will be allocated towards the tracking of specific device categories. This is useful in the small
enterprise environment in order to allow the tracking of device categories such as nearby rogue access
points and rogue clients, but also limit these categories such that an uncontrolled introduction of rogues
is not allowed to consume all of the remaining context-aware tracking capacity on the MSE.
We can use the Context-Aware Service Tracking configuration to:
•
Entirely enable or disable the tracking of wired and wireless client stations, asset tags, rogue access
points, and rogue clients.
Small Enterprise Design Profile Reference Guide
6-31
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
•
Set limits on how much MSE tracked device capacity will be allocated to certain device categories.
Figure 6-18 provides us with an example of how this can be achieved, where the maximum number
of tracked clients and rogue clients/APs are capped at 4,000 devices each. No limit is placed on the
number of RFID tags tracked, which in effect means that the maximum number of tags tracked will
be allowed to rise until the tag licensing limit is reached (3,000 tags).
Note that any devices that are detected but excluded from tracking due to the enforcement of a tracking
limit will be reflected in the “Not Tracked” device count column shown on the right side of the display.
Figure 6-18
Note
Context-Aware Service Tracking Parameters
In Release 6.0, wired client tracking can be enabled or disabled, but imposing a limit on the number of
wired clients that are tracked simultaneously is not supported at this time. As a work around, use switch
CLI commands such as the global nmsp disable or the interface nmsp suppress attachment commands to
limit the number of tracked wired clients that are presented to the MSE.
Context-Aware Service Parameters—History
The MSE records and maintains historical location and statistics information for wireless clients, tags,
rogue access points, and rogue clients. This information is available for viewing through WCS or via
third-party context-aware application clients, and can be very valuable in helping establish patterns of
movement for tracked assets and rogues. Historical information can be used for location trending, asset
loss investigation, RF capacity management, and facilitation of network problem resolution. Contextual
information such as whether an emergency button was depressed or whether an asset tag has moved into
close proximity to a chokepoint trigger is also tracked in the history data.
The collection of historical information must be explicitly enabled for each desired category of device
(as shown in Figure 6-19). By default, 30 days of historical data are stored in the MSE.
Small Enterprise Design Profile Reference Guide
6-32
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Figure 6-19
MSE History Parameters
There are several variables that can affect how much historical information can be stored by the MSE for
tracked assets. Among these variables are the average number of elements that move, average distance
covered every time there is a movement, information transitions, telemetry information from tags, and
so on. Depending on these variables as well as the number of items for which you are tracking history
information, you may wish to decrease the number of days of historical data to a value below 30 days.
Changes to the default history archive period should be done with careful consideration, since longer
history periods typically increase the amount of space consumed by the history database. Users
unfamiliar with the way in which Context-Aware Services on the Cisco MSE archives historical
information for tracked devices may wish to consult with your Cisco field technical representative or the
Cisco Technical Assistance Center
Figure 6-20 illustrates what you can expect to see when recalling the history for a tracked device. Here,
we recall the history for a specific WLAN client. As shown in Figure 6-20, the focus of the display is
the list of past locations recorded for the client. Setting the “change selection time” parameter on the
screen and then clicking play displays the various client locations on the small floor map at the right side
of the image (this can be enlarged for easier viewing). In this way, you can step back through all of the
stored locations for the client within the history database. Obviously, this information could be very
useful to administrators, WLAN engineers, as well as enterprise security officers and other law
enforcement officials that may be looking for information useful in recreating a pattern of potentially
suspicious past movement.
Figure 6-20
Display of Location History for a WLAN Client
Further information on the procedure to follow when making adjustments to history parameters can be
found in the following documents:
•
Context Aware Solution Deployment Guide
http://www.cisco.com/en/US/products/ps9742/products_tech_note09186a00809d1529.shtml#rfid
Small Enterprise Design Profile Reference Guide
6-33
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
•
Context Aware Service Configuration Guide 6.0
http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/msecg_ch7_C
AS.html#wp1128896
Context Aware Service Parameters—Notifications
Cisco WCS allows you to define certain conditions for tags, WLAN clients, and rogues that cause the
MSE to send notifications to application programs that are monitoring specific ports. You can use Cisco
WCS to define and enable both conditional notifications and northbound notifications.
Conditional notifications are those notifications that the mobility services engine sends to Cisco WCS
and other applications that can receive short, relatively simple messages via SOAP/XML (either HTTP
or HTTPS), SMTP, UDP Syslog, or as an SNMP trap. Conditional notifications can be triggered by
WLAN clients, tags, or rogue devices. The conditions available include:
•
Missing—The MSE generates a Missing Asset conditional notification if it has not located the asset
for more than a specified number of minutes.
•
In/Out—The MSE generates an In/Out conditional notification if the asset is found to be inside of,
or outside of, a selected area.
•
Distance From Marker—The MSE generates a Distance From Marker conditional notification if the
asset is found to be beyond a specified distance from a designated marker.
•
Battery Level—The MSE generates a Battery Level conditional notification if the battery level
reported by an asset tag is equal to a selected value.
•
Location Change—The MSE generates a Location Change conditional notification if the asset
experiences a change in location.
•
Emergency—The MSE generates an Emergency conditional notification if a tag button, tamper or
detached event is detected.
•
Chokepoint—The MSE generates a Chokepoint Conditional notification if a tag enters into the
proximity of a chokepoint trigger.
Northbound notifications are a special category of notification that is specific to RFID tags only. They
define which tag notifications the MSE will send to third-party applications using SOAP/XML.
Northbound notifications can include chokepoint, telemetry, emergency, battery, and tag vendor data.
Optionally, the tag's location coordinates can be included within the northbound notifications as well.
An important difference between conditional notifications and northbound notifications is that
northbound notifications can contain considerably more information. The information sent in the
northbound notification is sent in a pre-defined data format. Details regarding the data format for
northbound notifications are available on the Cisco developers support portal at
http://developer.cisco.com/web/contextaware.
Note
In Release 6.0, neither conditional nor northbound notifications can be applied to tracked wired devices.
Complete and detailed information regarding how to configure Context-Aware Notifications can be
found in the following documents:
•
“Configuring Event Notifications” chapter of the Context-Aware Configuration Guide 6.0
(http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/msecg_ch6_C
AS.html).
Small Enterprise Design Profile Reference Guide
6-34
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
•
“Enabling Notifications and Configuring Notification Parameters” section of the Context-Aware
Configuration Guide 6.0
(http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/msecg_ch7_
CAS.html#wp1129909).
•
“Context-Aware System Performance” section of the Context-Aware Solution Deployment Guide
(http://www.cisco.com/en/US/products/ps9742/products_tech_note09186a00809d1529.shtml#casy
sperf).
Via the menu panel located at Services > Mobility Services > Context Aware Service > Advanced>
Notification Parameters (Figure 6-21),WCS allows the user to modify several advanced timing
parameters pertaining to conditional notifications and northbound notifications. For example, you can
limit the rate at which the MSE generates notifications, set a maximum queue size for notifications, and
set a retry limit for notifications with in a certain period.
Figure 6-21
Note
Context Aware Advanced Notification Parameters
Modify advanced notification parameters only when you can reasonably expect the Mobility Services
Engine to transmit a large number of notifications or when it is noticed that notifications are being
dropped.
•
Rate Limit—(Pertains to northbound notifications only.) This is the rate in milliseconds at which the
MSE generates northbound notifications. A value of 0 (default) means that the Mobility Services
Engine generates event notifications as fast as possible. If you are using the northbound notifications
system to communicate to a third-party application (such as AeroScout MobileView, for example)
in a small enterprise design, it is recommended that you consider how often notifications are
expected to be generated and how many total notifications will be requested of the MSE per minute.
For example, an exception based notification (such as an emergency notification) is not a normal
event and thus it would be extremely unusual to see a great deal of emergency events generated from
many sites at the same time. In contrast, more routine events may be seen to generate substantially
greater traffic. In cases where a large number of notifications might be a normal and routine event,
it is recommended to increase the rate limit so as to allow the MSE to pace the transmission of
notifications (for example, a rate limit value of 200 would slow the rate of notification transmission
down to one every 200 msec). The correct notification delay will vary from deployment to
deployment depending on the amount of traffic generated.
•
Queue Limit—Specifies the size of the output notification queue of the MSE. Default queue limit
value for the MSE-3350 is 18000 and for the MSE-3310 it is 5,000. The MSE drops any outbound
notifications above this limit if the output notification queue size is exceeded. In small enterprise
designs, if you are using context-aware notifications, it is recommended that you use the Queue
Limit parameter in conjunction with the Notifications Dropped counter to avoid any notifications
from being dropped. If you notice that the Notifications Dropped counter is greater than zero, you
should consider increasing the Queue Limit parameter to avoid any future increases. Given the
relatively large default sizes for this parameter however, it is unlikely that adjustment will be
required except in rare cases where very many notifications are generated.
Small Enterprise Design Profile Reference Guide
6-35
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
•
Note
Retry Count—For each matching condition, the retry count specifies the number of times to generate
an event notification before the refresh timer expires. The default value is 1. The total number of
event notifications transmitted between Refresh Time periods is equal to one plus the value specified
for Retry Count. After the value of one plus the Retry Count has been reached, the location appliance
skips firing any further northbound notifications for this condition and device for the time period
specified by the Refresh Time. Once the Refresh Time has expired, this cycle repeats unless the
event has been cleared. Retry count is intended to help ensure (to a limited extent) that notifications
reach their intended destination. If notifications are not indicated as being dropped by the
Notifications Dropped counter, but are not reliably reaching their destination application, you may
wish to increase the number of notifications sent between Refresh Time periods by raising the Retry
Count judiciously.
The Mobility Service Engine transmits notifications using a “Fire and Forget” technique. Notifications
are not retained in any database within the MSE after they are transmitted.
•
Refresh Time—The wait time in minutes that must pass before an event notification is resent. The
default is 60 minutes. Refresh Time and Retry Count are used cooperatively to help limit the number
of notifications repeatedly generated for events that have not been cleared. Retry Count limits the
number of notifications that are sent by the MSE, while Refresh Time imposes a “waiting period”
during which time no further notifications are sent for this event condition and device. If you are
noticing that repeated notifications are being generated for the same event, it may be due to the
condition not clearing within the refresh time interval. In this case, you may wish to investigate why
the triggered condition does not clear and possibly extend the refresh time.
•
Notifications Dropped—The number of event notifications dropped from the queue since startup.
The Notifications Dropped counter should be used in conjunction with the Queue Limit parameter
to reduce the number of total dropped notifications.
WLAN Controller and Cisco Catalyst Switch Definition and Synchronization
To allow for proper tracking of the devices that may be registered or attached to them, WLAN controllers
and context-aware Cisco Catalyst switches must be defined to WCS and then synchronized with the
Mobility Services Engine.
Detailed information regarding how to add WLAN controller definitions to WCS using the WCS
Configure > Controllers menu panel can be found in the section titled “Adding Controllers” in the WCS
Configuration Guide 6.0:
(http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/6_0ctrlcfg.html#wp1041451)
.
Detailed information regarding how to add Cisco Catalyst switch definitions to WCS using the WCS
Configure > Add Ethernet Switches menu panel can be found in the WCS Configuration Guide 6.0:
(http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/6_0ctrlcfg.html#wp1089752)
.
Detailed information regarding how to synchronize the MSE with WLAN controllers and context-aware
Cisco Catalyst switches using the WCS Services >Mobility Services > Synchronize WCS and MSE(s)
menu panel can be found in the “Synchronizing Mobility Services Engines” chapter of the
Context-Aware Service Configuration Guide 6.0:
(http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/msecg_ch3_CAS.
html#wp998995).
Small Enterprise Design Profile Reference Guide
6-36
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Note
Always ensure that the MSE is synchronized with the primary WLAN controllers it is providing
Context-Aware Services for, as well as any backup WLAN controllers. If the MSE is not kept in
synchronization with your backup WLAN controllers, Context-Aware Services may not function
properly or may not be available at all for WLAN clients and RFID tags in the event of a primary WLAN
controller failure.
Wireless Client Context-Aware Considerations
The ability to track WLAN client location using Cisco Context-Aware Services can be useful in a small
enterprise to locate WLAN clients residing on the network (such as authorized Cisco 7921G and 7925
VoWLAN phones, laptops, wireless desktops, etc.). Generally speaking, provided that a 802.11 wireless
client device has its 802.11 wireless network interface adapter powered on and is sending periodic
transmissions (probe requests), these WLAN client devices can be located by Context-Aware Services.
Since devices such as portable VoWLAN phones are very often powered on for the entire day and carried
on the belt, purse, or pocket of the user, these devices become useful in helping locate not only the device
itself but the user to which the device is assigned. As the device becomes larger in size and heavier in
weight, we find that the chances of the device being with the assigned user all the time diminishes and
thus its usefulness as a way to determine the location of the user diminishes as well.
Tracking WLAN client devices by their wireless network interface adapter works well if the goal is to
track the location of the device when it is in operation and use this information to enhance the operation
of the device on the network, or its interaction with an application. For example, using WLAN client
tracking in the small enterprise to perform one or more of the following tasks are just some examples of
how this functionality can be put to use:
•
Determining where wireless users and their portable devices may congregate, allowing for the
placement of access points to be further optimized to enhance overall coverage and performance.
•
Investigate where a missing device is currently located within the main or remote site. A good
example might be when equipment is “borrowed” from one office or conference room to another (or
even from one site to another) without permission.
•
Determining the location of users that are known to be visitors to one or more enterprise sites, and
helping ensure that they do not stray into areas that are off-limits to them.
•
Using the location history of a wireless device to establish a pattern of usage and location in order
to clarify past actions, such as those that might relate to safety and security concerns.
•
Use the location of wireless LAN clients as a parameter for network troubleshooting and security
audits.
•
Use the location of wireless printers (or other output devices) as input to an application designed to
determine which device is available and most convenient for a particular wireless user depending
on the user's location.
Figure 6-22 illustrates how Cisco Context-Aware Services can be used with Cisco WCS to display the
current location of employees equipped with Cisco 7925G IP phones, laptops, and PDAs. Note that we
have chosen to assign and display the user name associated with each device, rather than the device
MAC address. Clicking on any of the blue WLAN client icons shown in Figure 6-22 displays a plethora
of information about the client device, including its client properties, association history, connection
troubleshooting information, as well as its event history.
Small Enterprise Design Profile Reference Guide
6-37
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Figure 6-22
WLAN Client Tracking
If the intention is to try and track a wireless device for inventory or loss prevention purposes, it is
important to ascertain whether the embedded network interface remains active for any devices that may
be in a powered-down or standby state. Even so, if the device remains active but accesses the network
only very infrequently in order to conserve power, the degree of location fidelity obtained will likely be
less than optimal, especially if the device can be expected to move frequently while in this state. A higher
accuracy solution for this type of full-time location tracking application would be to attach or embed a
RFID asset tag (see the “RFID Asset Tag Context-Aware Considerations” section on page 6-41) to the
device. RFID tags are independently powered and operate without concern for the power state of the
WLAN client device, and can be stimulated to transmit immediately when passing in proximity of
properly equipped doorways and exits.
In general, best results are obtained in context-aware designs when all WLAN clients are compliant with
the Cisco Compatible Extensions for WLAN Clients specification v2 at minimum, and preferably with
the Cisco Compatible Extensions for WLAN Clients specification v4 or v55.
Additional information on the Cisco Compatible Extensions program for Wi-Fi Client Devices is
available at http://www.cisco.com/web/partners/pr46/pr147/partners_pgm_concept_home.html.
Note that this document does not detail the steps involved with important procedures such as calibration
of the Context-Aware Engine for WLAN Clients, or the definition of location inclusion and exclusion
regions and other deployment procedures. For information on these and other procedures that should be
understood prior to deployment, refer to the MSE Context Aware Service Deployment Guide
(http://www.cisco.com/en/US/products/ps9742/products_tech_note09186a00809d1529.shtml).
5. Readers interested in the technical details concerning compatibility with the Cisco Compatible Extensions for
WLAN Clients specification v2 are referred to the section titled “Tracking Clients, Assets and Rogue Devices”
in the Wi-Fi Location-Based Services Design Guide 4.1,
http://www.cisco.com/en/US/docs/solutions/Enterprise/Mobility/wifich3.html#wp1049277
Small Enterprise Design Profile Reference Guide
6-38
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Intel® Wi-Fi Clients and ProSet Client Supplicant Context-Aware Considerations
When using clients equipped with the Intel® Wireless WiFi Link 4965AGN, Intel® PRO/Wireless
3945ABG Network Connection, or the Intel® PRO/Wireless 2915ABG Network Connection adapter, it
is important to note that the default “Personal Security” settings of the Intel® ProSet Configuration
Utility does not include compatibility with the Cisco Compatible Extensions specification. In order to
enable compatibility with the Cisco Compatible Extensions specification, the Intel ProSet client
supplicant must be used to reconfigure the client for “Enterprise Security” and to enable Cisco
Compatible Extensions using the “Cisco Options” (Figure 6-23).
Figure 6-23
Enabling Cisco Compatible Extensions on Intel® ProSet Clients
Small Enterprise Design Profile Reference Guide
6-39
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Cisco 7921 and 7925G Context-Aware Considerations
7921G and 7925G Unified IP Wireless Phone users should note that phones that are idle and not
currently participating in an active call may not transmit 802.11 Probe Requests with sufficient
frequency to ensure that changes in the actual location of the phone user are promptly reflected in the
calculated location coordinates provided to the contest-aware services software in the MSE. In some
cases this can be of concern, especially if the 7921G or 7925G user remains within the primary coverage
area of the same access point for long periods of time. In cases where an access point might have a large
coverage footprint, a roaming event may not occur very often despite a significant change in user
location.
If you experience situations where the location fidelity of a 7921G or 7925G Unified IP Wireless Phone
appears to be much better for those users actively participating on a call versus those that are on hook
and idle, you may wish to consider changing the scan mode parameter associated with the 7921G or
7925G device in the Cisco Unified Communications Manager. In some cases, improved location fidelity
can be achieved by enabling the “continuous” scan mode on the Device=>Phone configuration page of
Cisco Unified Communications Manager Administration (shown in Figure 6-24). Note that scan mode
options listed are “auto”, “continuous”, and “single AP”, where auto is the default. “Continuous” scan
mode causes the wireless IP phone to issue a probe request approximately every two seconds, whereas
“auto” scan mode causes the device to issue probe requests primarily only when the device is engaged
on an active call, when roaming, or when preparing to roam. “Single AP” is used in installations where
the wireless IP phone is only used in the vicinity of a single access point at all times, as probe requests
are issued only when the wireless IP phone is first powered on. “Single AP” mode is not applicable and
should not be used in the small enterprise designs discussed here.
Small Enterprise Design Profile Reference Guide
6-40
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Figure 6-24
Note
Scan Mode Option in Cisco Unified Communications Manager
It is recommended that continuous scan mode be used only in situations where the anomaly described
here is actually witnessed. This is because a trade-off associated with any increase in the frequency of
probe requests transmitted is the potential for reduced 7921G or 7925G battery life. If you do not notice
the anomaly described in this section, it is recommended that you leave the CUCM scan mode setting at
the “auto” default.
RFID Asset Tag Context-Aware Considerations
Radio Frequency Identification (RFID) has many potential safety and security applications in the
enterprise arena. Already used to improve efficiency and productivity across many business sectors,
there are a wide variety of applicable use cases that embrace RFID in the enterprise in order to address
a plethora of challenges.
Figure 6-25
Floor Map Showing Various Uses for RFID Tags and Cisco Context-Aware Services
Small Enterprise Design Profile Reference Guide
6-41
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
The illustration in Figure 6-25 provides us with a visual representation of just some of the ways RFID
can be used in an enterprise in conjunction with Context-Aware Services:
•
High value assets (such as projectors, audio/video equipment, etc.) can be kept safe and secure, since
the application of RFID tags to these assets provides small enterprise administrators and security
staff with the ability to locate the assets quickly and efficiently. Figure 6-25 illustrates how the
present location of assets equipped in this fashion can be quickly ascertained by a quick check of
the small enterprise floor map. Here, we can see the last known location of XVGA portable
projectors and other equipment belonging to different departments at this enterprise site, such as
sales, marketing, R&D, engineering, customer service, and technical support. For buildings
containing more than a single floor, database search techniques can be used to search for the desired
asset by name or attached RFID tag MAC address.
Figure 6-26
ID Badge with Embedded Active RFID
•
Employees, visitors, guests and administrators can wear specially manufactured badges (see
Figure 6-26) that combine a traditional identification card with active RFID technology, allowing
them to be located quickly in the event of an emergency. These same devices can also transmit
special notifications using a push button sequence, which can be interpreted in various ways by
Context-Aware Services, including as a signal that an emergency event is in progress. Figure 6-25
illustrates how a person can be located if they are carrying an RFID-enabled ID card, as seen in the
displayed location of the site’s security officer, whose location is currently indicated as being inside
of the site manager’s office. Being able to physically locate the site security officer using a tool such
as this could help in saving precious seconds in the event of an emergency.
•
The whereabouts of enterprise site visitors and guests (such as maintenance or repair contractors)
can be monitored and alerts triggered to security officers or others if these visitors stray into areas
that they do not have authorization for. For example, in Figure 6-25, we see that we have an HVAC
repair contractor as well as a guest visitor that has come in for an employment interview. In addition
to making us aware as to the presence of these visitors, our system can alert us if these personnel
access areas that we do not want them to by comparing their current locations to location boundaries
we have defined as notification rules. Third-party applications that can access context-aware
information from the MSE may perform more advanced tasks as well.
•
RFID tag technology can be combined with chokepoint triggers (Exciters) to enable the use of
proximity applications. In this fashion, chokepoint triggers can serve many purposes, including
stimulating asset tags affixed to assets that might be in the process of being removed from the site
without authorization. This makes it possible to notify officials or security officers of such action,
so as to act quickly and determine whether such movement is legitimate.
Small Enterprise Design Profile Reference Guide
6-42
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
In addition to simply displaying the location of assets, Cisco Context-Aware Services makes other
contextual characteristics of the asset and its environment available to applications accessing the
MSE via its SOAP/XML API. For instance, if the asset tags used contain on-board temperature
sensors, we can also read the current temperature surrounding the asset tag via the MSE. This can
be seen in Figure 6-27, where we see from WCS that the ambient temperature surrounding our
HVAC repairman is a comfortable 20 degrees Celsius (68 degrees Fahrenheit). From this
information, it certainly appears that the site air conditioning system is doing its job! It is important
to note that while WCS is used to view this information in Figure 6-27, any authorized
context-aware application that is written to use the MSE SOAP/XML API could have accessed this
data as well.
Figure 6-27
Rogue AP Details
Simply put, RFID technology and Cisco Context-Aware Services can help small enterprises improve
their overall safety, security, and efficiency.
Cisco Context-Aware Software is designed to function with active RFID asset tags from vendors
compliant with the Cisco Compatible Extensions for Wi-Fi Tags specification. A list of current Cisco
Compatible Extensions for Wi-Fi Tags compliant vendors can be found at
http://www.cisco.com/web/partners/pr46/pr147/ccx_wifi_tags.html. Although all vendor RFID tags
compliant with the Cisco Compatible Extensions for Wi-Fi Tags specification share a great deal of
functionality in common, the parameter names and means used to configure each brand of tags can differ
from vendor to vendor, therefore no set prescribed configuration parameter list would apply to all. That
being said, there are several general configuration functions that tag vendors share and although the
exact parameter names may differ, it is important that you understand them and use this knowledge
accordingly when configuring the particular tags of choice for your installation. More information
regarding the configuration procedure for AeroScout tags can be found in the section entitled RFID Tag
Small Enterprise Design Profile Reference Guide
6-43
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
and WLC Configuration/Tuning in the Cisco MSE Context-Aware Deployment Guide
(http://www.cisco.com/en/US/products/ps9742/products_tech_note09186a00809d1529.shtml#rfid-wlc)
6
.
•
RF Channel Configuration—It is recommended that tags be configured for the standard set of 2.4
Ghz non-overlapping channels, which is typically channels 1, 6, and 11 (this may vary depending
on your international regulatory domain).
•
Stationary Transmission Interval—This is the time between periodic tag transmissions that are
normally generated when the tag is stationary and not in motion. It is recommended that this be
configured for values between 3 and 5 minutes.
•
Motion Transmission Interval—(Applies only to tags with motion sensors.) This is the time between
periodic tag transmissions generated when the tag and asset is in motion. A recommended initial
value is 15 seconds.
•
Number of Tag Message Repetitions—Some popular RFID tags default to transmitting a single
transmission on all defined channels. Tag parameters that control the number of tag message
repetitions specify the number of times each transmitted message is repeated, per channel. It is
generally recommended that this parameter be set to a value of three. Doing this helps protect
against lost tag transmissions due to congestion or interference, which is a primary cause of poor
tag location accuracy.
•
Message Repetitions Interval—The delay between subsequent message repetitions on the same
channel. This is often defaulted to 512msec, although in lab testing we have seen some evidence of
improved location accuracy when a message repetition interval of 256 msec is used with the default
controller value of 2 seconds for NMSP notification interval. This parameter is discussed in more
detail the section entitled RFID Tag and WLC Configuration/Tuning of the Cisco MSE
Context-Aware Deployment Guide
(http://www.cisco.com/en/US/products/ps9742/products_tech_note09186a00809d1529.shtml#rfidwlc).
This chapter does not detail the steps involved with procedures such as calibration of the Context-Aware
Engine for Tags and other deployment procedures. For information on these and other procedures that
should be understood prior to deployment, refer to the MSE Context Aware Service Deployment Guide
(http://www.cisco.com/en/US/products/ps9742/products_tech_note09186a00809d1529.shtml). In
addition, the AeroScout Context-Aware Engine for Tags for the Cisco MSE Users Guide, version 3.2,
available from your AeroScout representative or https://support.aeroscout.com, is highly recommended.
RFID Tag Chokepoint Trigger Considerations
Chokepoint triggers are proximity communication devices that trigger RFID asset tags to alter their
behavior when the tag enters into close range (otherwise known as the stimulation zone) of the
chokepoint trigger. This behavioral modification may, for example, cause the RFID tag to immediately
transmit its unique identifier (MAC address) or cause the tag to change its internal configuration,
depending on how the tag is programmed. A very popular use of the chokepoint trigger is to stimulate
the asset tag such that it provides indication to the MSE that the tag has entered or exited a given area,
known as a chokepoint. Chokepoints are entry or exit points that provide passage between connected
regions. Common chokepoints are entrances and exits such as doorways, hallways, and stairwells.
Note
An Exciter is a registered trademark of AeroScout Ltd., and represent a popular example of a chokepoint
trigger.
6. Additional information can also be found in the Wi-Fi Location-Based Services Design Guide 4.1 in the section
titled “Configuring Asset Tags”, located at the following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Mobility/wifich6.html#wp1077248.
Small Enterprise Design Profile Reference Guide
6-44
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Chokepoint triggers are useful in causing RFID tags to react quickly when assets are moved past certain
points. This could be a server, for example, with an affixed RFID tag moving past a chokepoint trigger
located near a facility exit, an employee with an RFID badge entering the site through the front entrance
at 7:00 AM, or it could be a delivery vehicle that has a RFID tag on the front visor coming past a
chokepoint trigger located at the site’s property entrance. In all these cases, the chokepoint trigger causes
the tag to change its behavior, most likely to immediately transmit its MAC address (as well as the MAC
address of the chokepoint trigger that stimulated it) to the MSE via one or more access points that can
receive the tag's transmissions. The net result is to cause the MSE to indicate that the current location of
the asset and its asset tag is within a known, pre-determined proximity of the chokepoint trigger.
In order to use chokepoint triggers with Cisco Context-Aware Services, they must be properly configured
using the appropriate vendor-supplied software utility, defined to WCS, placed on floor maps, and
synchronized as part of an updated network design to the MSE. After all of this is complete, the MSE is
able to recognize the transmissions generated by asset tags that have been stimulated by specific
chokepoint trigger MAC addresses. Based on this information, the MSE can attempt to localize the tag
to the proximity of the chokepoint trigger. Applications such as WCS (or third party context-aware
applications) may then display the asset tag's location at the chokepoint icon associated with the
chokepoint trigger's MAC address.
Various chokepoint trigger specific parameters such as transmission range, IP address, transmission
interval, transmission repetitions, and so on are set using vendor-specific utilities. Note that each vendor
maintains their set of software tools necessary for configuration of their chokepoint triggers. These
software configuration tools are not interoperable between vendors (for example, AeroScout software
configuration tools cannot be used to configure WhereNet chokepoint triggers or vice-versa).
The individual configuration of each vendor's chokepoint trigger is beyond the scope of this chapter.
Complete and detailed configuration information relating to the specific configuration of each vendor's
chokepoint trigger can be found in the appropriate vendor's documentation, which can be obtained from
your tag vendor representative.
AeroScout EX-3200 User Guide https://support.aeroscout.com
AeroScout Exciter EX-2000 User Guide https://support.aeroscout.com
AeroScout Context-Aware Engine for Tags for the Cisco MSE Users Guide, version 3.2
https://support.aeroscout.com
Technical documentation for WhereNet WherePort chokepoint triggers and the necessary software and
hardware for configuration of WherePorts is available from WhereNet Corporation
(http://www.wherenet.com) or via your WhereNet account representative.
Rogue Device Context-Aware Considerations
The use of context-aware services for locating clients and RFID tags that we have defined and authorized
in the small enterprise environment is often what first comes to mind for many of us when we consider
this solution. However, another equally important and useful function of context aware services is the
ability to detect the location associated with those wireless clients and access points that we have not
authorized to operate within our domain. In other words, context-aware services in our small enterprise
can help us in locating rogue access points or rogue clients that may have been installed by employees,
contractors, visitors, vendors, or even administration members without authorization. Even if an
unauthorized access point is innocently installed by an enterprise user that is otherwise authorized to use
the network, the unauthorized and potentially insecure portal provided by such an access point can
unnecessarily expose our otherwise secure small enterprise network to outside intruders.
The Cisco Wireless Control System can use the location capabilities provided by the Cisco Mobility
Services Engine and the Cisco Context-Aware Engine for Clients to define the location of unauthorized
access points and the wireless clients that may be using these access points, as shown in Figure 6-28. In
Small Enterprise Design Profile Reference Guide
6-45
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Figure 6-28, we can see icons for both rogue access points as well as rogue clients displayed over their
predicted positions on a floor map of a site located within our small enterprise. This capability is very
useful both to the local site administrator, as well as the main site administrator and their technical teams.
It allows them to determine the location of wireless equipment that may have been brought into the
enterprise site and used in an attempt to gain unauthorized access to enterprise site computing resources.
In Figure 6-28, the round icon with the “skull and crossbones” logo represents the rogue access point
and the rectangular icon with the same logo represents a rogue client that is associated to this rogue
access point.
Figure 6-28
Note
Using Context-Aware Services to Determine Location of Rogue Access Points and
Clients
In comparison to WLAN clients and RFID tags, it is normal to experience a reduced accuracy when
localizing rogue access points and rogue clients. The very nature of “rogue” devices establishes that
these devices are not under the control of site administration. Therefore, the configuration of such these
rogue devices may not facilitate optimal context-aware location accuracy (for example, they may not be
Cisco Compatible Extensions devices, etc.).
As shown in Figure 6-29, clicking on either the rogue access point icon or the rogue client icon reveals
additional information and capabilities that can help in further identifying (and even isolating) these
devices. This includes the ability to perform switch port tracing with the latest versions of WCS, which
Small Enterprise Design Profile Reference Guide
6-46
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
allows tracing of rogue access points that have been connected to the switch infrastructure (for more
information on switch port tracing, refer to the WCS Configuration Guide 6.0 at the following URL:
(http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/6_0ctrlcfg.html#wp1089752)
Figure 6-29
Rogue AP Details
The collection of rogue AP and rogue client information is enabled by default on WLAN controllers. In
order to enable the tracking of rogue access points and rogue clients by the context-aware service, it is
important to ensure that the collection of rogue information has also been enabled for context-aware
services on the MSE. Refer back to Figure 6-18 in this chapter and ensure that the “Rogue Client and
Access Points” tracking parameter check box has been enabled under Mobility Services >
Context-Aware Services > Administration >Tracking Parameters.
Note
Be advised that depending on the environment in which your enterprise site is located, as well as the
number of rogue devices present, the number of rogue devices detected can rise very quickly. Since the
size of the rogue device population is typically not under the direct control of enterprise IT staff or local
site administration, it is highly advisable that you enable limiting for rogue clients and access points and
set a limiting value. This is so that any unforeseen increase in the number of rogue devices detected does
not consume all the remaining tracked device capacity on the MSE, thereby depriving the MSE of
capacity that might be of more significance to enterprise site administration and employees. This is
especially important if the MSE is used to service multiple sites within your small enterprise.
In addition to enabling the collection of rogue access point and rogue client information by the
context-aware services on the MSE, you must also enable the display rogue device location when
displaying floor maps using WCS. This is accomplished via the WCS “Floor Settings” submenu that is
displayed on every WCS floor map, as shown in Figure 6-30. Information on how to use the Floor
Settings sub-menu for all the categories of device shown here can be found in Chapter 5 of the WCS
Configuration Guide 6.0,
http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/6_0maps.html#wp1210969.
Small Enterprise Design Profile Reference Guide
6-47
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Figure 6-30
WCS Floor Settings Sub-Menu
For further information regarding rogue access points and clients, refer to the following documents:
•
Context Aware Service Configuration Guide, Rlease 6.0
http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/CAS_60.html
•
Cisco Wireless Control System Configuration Guide, Release 6.0
http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/WCS60cg.html
•
Cisco MSE Context-Aware Deployment Guide
http://www.cisco.com/en/US/products/ps9742/products_tech_note09186a00809d1529.shtml
Context-Aware Considerations for Wired Device Tracking
As described previously in this chapter, beginning with release 6.0 Cisco Context-Aware Services
provides the capability to determine the civic location and emergency line identifiers of devices
connected to Cisco Catalyst switches, such as the 2960G, 3560E, 3750E, 3750G, 4500, and 4900 series.
As participants in context-aware services, switches are configured with and provide the relevant
contextual information for all the IP endpoints attached to them. These endpoints may include IP phones,
PCs, host servers, access points, etc. The NMSP protocol is used between the switches and MSE to
deliver this contextual information to the MSE. Location information may include the physical location
address (also known as the civic address) as well as other information about endpoints such as the IP
address, MAC address, port, VLAN, and username. If the end device makes use of the Cisco Discovery
Protocol or Link Layer Discovery Protocol (LLDP), additional information, such as the version number
and serial number, can also be sent to the MSE.
Note
The use of Context-Aware Services for wired device location is entirely optional. In the Small Enterprise
Design Profile, Context-Aware Services may be deployed for wireless devices, wired devices, or for
both.
The Small Enterprise Design Profile provides that in the event of a failure of a switch line card in a 4500
series context-aware Catalyst switch, or a stack member in the 3750 switch stack, NMSP sessions
recover from the failure without intervention from the user. NMSP session recovery time from isolated
switch stack member or line card failure was observed to be very quick during lab testing, and in most
cases the recovery time was almost unnoticeable from the perspective of the Mobility Services Engine.
Careful examination of the NMSP session status during simulated failures indicated that sessions
remained up, intact, and passing NMSP data while stack member or line card hand-off occurred.
Small Enterprise Design Profile Reference Guide
6-48
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
While still a relatively new context-aware capability, the inclusion of wired device tracking provides new
visibility into the location of wired IP endpoints in your small enterprise network. For example, some of
the ways that this new exciting capability can be used in small enterprise designs include:
•
Determining the whereabouts of missing IP phones, PCs, and peripheral devices such as printers and
network scanners that have disconnected and moved from their originally installed locations to other
locations within a site (or moved to another site altogether). Once reconnected to the network, the
MSE would be updated with the wired device's new attachment information and any location and/or
ELIN information defined for that switch port.
•
Keeping track of the location of host computers that are physically present at the main site data
center or in any other site. The wired location capabilities contained in release 6.0 of Context-Aware
Services make it possible to specify the location of a device down to the room, cubicle, seat, or even
rack/slot location (see Figure 6-31). This can be important in verifying that equipment to be
de-commissioned is actually removed from service and sanitized of all corporate and employee
information.
Figure 6-31
•
Displaying the Location of a Wired Host
Determining the civic location of users based on their IP address or the user name that was specified
during 802.1x / EAP login (refer to Figure 6-32). Context-Aware Services for wired devices makes
it possible to quickly determine the civic or emergency line identifier information associated with
the switch port. Figure 6-32 illustrates how we can search for the wired device by the known
username.
Small Enterprise Design Profile Reference Guide
6-49
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Figure 6-32
•
Searching for Wired Device by Username
For devices that support it, impromptu inventory checks of devices across the main site by
examining wired device listings by serial number (see Figure 6-33) and comparing this information
to deployment records. This can help enterprise and site administrators better determine whether
assets have been moved between enterprise sites without authorization.
Figure 6-33
Examining Device Serial Numbers Via Wired Device Tracking
In contrast to the manner in which searches for wireless clients and tags is handled via the WCS Monitor
> Maps, in Release 6.0 of Context-Aware Services, all wired client searches are handled via the Services
> Mobility Services >Context Aware Services > Wired > Wired Clients menu panel. Searches using
the wired client WCS menu panel are specific to the particular MSE that you have selected.
Hardware and Software Requirements for Wired Device Tracking
As mentioned previously, at the current time wired device tracking is only performed on Catalyst switch
hardware supporting context-aware services.
Small Enterprise Design Profile Reference Guide
6-50
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Note
The images used for testing of wired device tracking during the production of this “Hardware/Software
Releases” section on page 6-60. Readers should note that cryptography-enabled (k9) switch images are
required in order to enable wired device tracking in Catalyst switches.
•
You should be aware that including wired device tracking in your design will require additional
tracked device licenses on the MSE over and above that required for wireless device tracking alone.
This is because wired tracked devices are included in the maximum number of simultaneous devices
that can are licensed for tracking by an MSE. For example, a small enterprise that would otherwise
possess a maximum licensed tracked device requirement of 500 for wireless LAN clients, RFID
tags, and rogues might require 750 or more when wired device tracking is considered, depending on
the number of switches deployed and whether wired device tracking is enabled for all of them. Be
sure to plan appropriately for the number of wired devices that you intend to track when purchasing
MSE client tracking licenses for the Cisco Context-Aware Engine for Clients.
In addition to planning appropriately for an increase in MSE tracked device licensing due to the use of
wired device tracking, keep in mind that although a single MSE can technically support up to 500 NMSP
sessions, scalability testing limits have only allowed Cisco to test up to 100 simulated NMSP
connections to a single MSE at this time. Each context-aware switch that is enabled for wired device
tracking in your network establishes one NMSP session to the MSE and counts against this limit.
Therefore, when enabling many switches for context-aware wired device tracking, it is recommended
that you plan for the total number of MSEs that may be required to support the total number of NMSP
sessions in your network.
Enabling Context-Aware Wired Device Tracking
In order to track wired devices on Catalyst switch ports, each switch whose devices we wish to track
must be configured to enable NMSP and other important parameters, and to contain the appropriate civic
address and ELIN location information. In addition, WCS must be configured to be aware of the
context-aware switches in the network and to be able to communicate with them. WCS is also used to
transmit information about the switches to the Mobility Services Engine via the synchronization process.
A complete, step by step guide to configuring Catalyst switches and WCS for wired device tracking can
be found in the Context-Aware Service Configuration Guide 6.0
http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/msecg_ch7_CAS.h
tml#wp1224011.
In addition, the following chapters and documents provide valuable and detailed background
information concerning the wired device tracking capability of Catalyst switches:
•
“Configuring LLDP, LLDP-MED, and Wired Location Service” in the Catalyst 3750 Switch
Software Configuration Guide, 12.2(50)SE
http://www.cisco.com/en/US/docs/switches/lan/catalyst3750/software/release/12.2_50_se/configur
ation/guide/swlldp.html
•
“Configuring LLDP, LLDP-MED, and Wired Location Service” in the Catalyst 2960 Switch
Software Configuration Guide, 12.2(50)SE
http://www.cisco.com/en/US/docs/switches/lan/catalyst2960/software/release/12.2_50_se/configur
ation/guide/swlldp.html
•
“Configuring LLDP and LLDP-MED” in the Catalyst 4500 Series Switch Software Configuration
Guide, 12.2(53)SG
http://www.cisco.com/en/US/docs/switches/lan/catalyst4500/12.2/53SG/configuration/swlldp.html
#wp1097119
Small Enterprise Design Profile Reference Guide
6-51
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
Readers are reminded about the following:
•
NMSP is disabled on switches by default and must be explicitly enabled via the nmsp enable global
configuration command.
•
IP device tracking must be enabled on the switch in order for Context-Aware wired device tracking
to function properly. It can be enabled by issuing the ip device tracking command in global
configuration mode on the context-aware switch.
•
The civic location identifier in the LLDP-MED TLV is limited to 250 bytes or less. To avoid
receiving error messages regarding available buffer space during switch configuration, the total
length of all civic location information specified for each civic-location identifier must not exceed
250 bytes.
•
In Release 6.0, all wired device client and switch tracking is available only to the root WCS virtual
domain user. Because of this, you may wish to limit the use of context-aware wired device tracking
to only those users with whom you are comfortable assigning WCS root virtual domain privileges.
NMSP Attachment Notification Interval
After an NMSP session is established between the MSE and a context-aware Catalyst switch, the MSE
transmits an Echo Response packet to the switch every echo interval period, which is specified on the
MSE (for all NMSP session partners) using the WCS menu entitled Services > Mobility Services >
System > NMSP Parameters. In addition to Echo Responses, the switch will periodically send
attachment notifications to the MSE via the NMSP session. Any link-up or link-down events that are
detected by the switch are aggregated during a configurable time interval and sent to the MSE via an
attachment notification at the conclusion of that time interval.
This interval is known as the nmsp attachment notification interval and is configurable on the switch via
the nmsp notification interval attachment interval-seconds command. The range of values for
interval-seconds is from 1 to 30 seconds, with 30 seconds being the default. In large networks where
there are many NMSP sessions active across the MAN to an MSE and the number of users connecting
and disconnecting from the switch is high, configuring nmsp attachment notification interval to a very
short interval can increase the amount of NMSP traffic between switches and the MSE and is not
recommended7 without carefully understanding the nature of the traffic present in your network.
Civic Address Configuration
The information contained in the Context-Aware Service Configuration Guide 6.0
(http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/msecg_ch7_CAS.
html#wp1224011) provides the information necessary to configure context-aware switches and the WCS
for wired device tracking.
As can be seen in the Context Aware Service Configuration Guide, in release 6.0 of Context-Aware
Services all configuration of civic and ELIN location information is performed on each context-aware
switch using the switch CLI. Once a switch is configured with the desired civic and ELIN location
information, the switch will share all of the configured port information with the MSE when the NMSP
session is initially established and will periodically update the MSE if any location updates are
performed.
Note
Readers should find the following IETF RFC documents helpful in better understanding the types of
values that should be specified for the various civic location fields: RFC 4776
(http://www.ietf.org/rfc/rfc4776.txt), RFC 4589 (http://www.ietf.org/rfc/rfc4589.txt), and RFC 5139
(http://www.ietf.org/rfc/rfc5139.txt).
7. Except for switches that are local to the Mobility Services Engine and need not traverse the MAN.
Small Enterprise Design Profile Reference Guide
6-52
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
During the course of our lab testing, we discovered other useful facets of information regarding civic
location and ELIN configuration in Catalyst switch IOS releases 12.2-52(SE) and 12.2-53(SG):
•
Global Scope of Civic and ELIN location—Civic address or ELIN information must be configured
at a global level and then assigned to each switch interface using the appropriate civic or ELIN
location identifier. Globally defined civic address parameters (such as the building, county, postal
code, floor, and so on) cannot be individually over-ridden at the interface level in release 6.0. If more
than one switch port shares the same civic location or ELIN, then the same globally specified civic
and ELIN location identifiers can be used on each switch port interface. However, if all ports possess
unique civic address characteristics, then uniquely specified global civic address parameters for
each port must be used. This can be seen in the following example where three unique civic location
identifiers are applied to three different ports (Gi1/0/3 - 1/0/5). The test application server that is
being tested on port Gi 1/0/6 is in the same physical location as the device connected to port Gi1/0/5,
hence they share civic-location identifier 3.
location civic-location identifier 1
additional-code “Small Enterprise”
building “Hart Building”
city Downey
country US
county “Los Angeles”
floor “Floor 3"
name “Main Site”
postal-code 90241
state California
street-group “Brookshire Avenue”
number 11627
room 300
seat 3H103
type-of-place “Main Site”
!
location civic-location identifier 2
additional-code “Small Enterprise”
building “Hart Building”
city Downey
country US
county “Los Angeles”
floor “Floor 3"
name “Main Site”
postal-code 90241
state California
street-group “Brookshire Avenue”
number 11627
room 300
seat 3H104
type-of-place “Main Site”
!
location civic-location identifier 3
additional-code “Small Enterprise”
building “Hart Building”
city Downey
country US
county “Los Angeles”
floor “Floor 3"
name “Main Site”
postal-code 90241
state California
street-group “Brookshire Avenue”
number 11627
room 300
seat 3H105
type-of-place “Main Site”
Small Enterprise Design Profile Reference Guide
6-53
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
!
interface GigabitEthernet0/3
description 802.1x data access only
location civic-location-id 1
switchport access vlan 88
switchport mode access
authentication port-control auto
dot1x pae authenticator
!
interface GigabitEthernet0/4
description 802.1x data access only
location civic-location-id 2
switchport access vlan 88
switchport mode access
authentication port-control auto
dot1x pae authenticator
!
interface GigabitEthernet0/5
description 802.1x data access only
location civic-location-id 3
switchport access vlan 88
switchport mode access
authentication port-control auto
dot1x pae authenticator
!
interface GigabitEthernet0/6
description local RFID gate application server
location civic-location-id 3
switchport access vlan 56
switchport mode access
!
•
Civic location additional-location-information subcommand—During the course of our testing we
found the additional-location-information subcommand useful in adding miscellaneous
information about devices that can be displayed on the civic address tab of the WCS wired clients
display under “Address Line 2". In our testing, we made use of this facility to label the rack and
position location of various servers and other devices that are deployed in a common location.
Figure 6-34 illustrates how the information entered using additional-location-information is then
displayed on the civic address tab of the WCS wired clients display for the device (indicated by the
red arrow). Below the configuration for the switch port, we can see the results of displaying the civic
location for the switch interface using the location civic-location interface interface command.
Figure 6-34
Additional-Location-Information
Small Enterprise Design Profile Reference Guide
6-54
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
•
Civic Location street-group subcommand—In order to display a value under the street name
component on the civic address tab of the WCS wired clients display, we found it was necessary to
use the civic location street-group subcommand in the switch configuration. If we look again at the
left hand portion of Figure 6-34, we can see the street-group is specified as “Brookshire Avenue”.
On the right hand of Figure 6-34, we see that this value appears under the civic address tab of the
WCS wired clients display under the label “Street”.
Excluding Device Tracking on Select Switch Ports
In the majority of cases, when using context-aware wired device tracking, it is usually acceptable to
simply enable NMSP on the switch and allow the attachment status of all ports to be reported to the MSE.
In this way, the attachment status and any location or ELIN information specified for each port in the
switch is reported and can be accessed from the MSE. However, in some cases, it might be desirable to
enable wired device tracking on a switch, but exclude selected switch ports from reporting device
attachments to the MSE. A reason for doing this might be to help reduce the number of MSE tracked
device licenses required by eliminating the NMSP reporting of device attachments on select ports where
not much chance of device migration is expected. Recall that in release 6.0, while it is possible to enable
or disable wired device tracking entirely on a per MSE basis, it is not possible to limit the number of
wired devices that are tracked in each MSE at this time (i.e., wired device tracking is either on or off).
Therefore, any such limiting must be done manually by either disabling NMSP sessions with selected
switches entirely (no nmsp enable) or disabling only the tracking of select switch ports on a
context-aware switch that is otherwise reported device attachments normally.
To disable device tracking for select switch ports on a switch where NMSP has been enabled, the switch
interface configuration nmsp attachment suppress command should be specified on each switch
interface where device tracking is not desired. The nmsp attachment suppress interface command is used
to configure the interface to not send any attachment notifications to a Cisco Mobility Services Engine
(MSE).
If you are using Location MAC Filtering (Services > Mobility Services > Context Aware Service >
Administration> Filtering Parameters) to specifically limit or block (by MAC address) tracked
wireless clients and tags, be advised that these address filters also apply to wired device clients as well.
Make sure that any filtering specifications that you set using Location MAC Filtering are flexible enough
to allow tracking of not only your wireless clients and tags, but wired devices as well. Any devices that
have been blocked from location tracking as a result of a defined filter will be viewable under the
“Blocked MACs” listing on the Filtering Parameters page. Detailed information regarding how to
configure Location MAC filtering can be found in Modifying Filtering Parameters section of the Cisco
Context-Aware Service Configuration Guide 6.0
(http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/msecg_ch7_CAS.
html#wp1100062).
Classification and Marking of NMSP Sessions
A vital component in assuring NMSP session stability and acceptable CAS performance during periods
of network congestion is the application of QoS to NMSP data flows between the MSE and any WLAN
controllers or context-aware Cisco Catalyst switches in the Small Enterprise Design Profile.
Classification, marking and QoS should be applied ideally in both cases of locally deployed as well as
centralized MSE implementations, however, it is especially important when using a centralized MSE at
the main site and remote WLAN controllers and context-aware Cisco Catalyst switches in the remote
sites.
Small Enterprise Design Profile Reference Guide
6-55
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
In order to ensure that QoS prioritization can occur properly for NMSP sessions the NMSP data flows
must be properly identified, classified and marked as close as possible to their points of origin. This
section explains where such marking of NMSP data flows should occur in the network, and how it should
be performed.
Figure 6-35 provides an example of where identification, classification and marking of NMSP data flows
should occur in the case of the small enterprise design containing:
•
A main site with adjoining data center
•
A remote site that is based on the Catalyst 4500 for distribution
•
A remote site that is based on the Catalyst 3750 switch stack for distribution.
In Figure 6-35, we have identified several points where classification and remarking needs to occur in
order to properly mark NMSP traffic inbound from WLAN controllers and context aware switches, and
outbound from Mobility Services Engines. These points are indicated by the yellow and red numbered
circles.
Figure 6-35
Points of NMSP Classification and Marking
Optional Backup
WLAN Controller(s)
Access
Control
System
Wireless
Control System
2
WLAN
Controller
2
1
3
Main Site
Services Block
4
MSE
Mobility Services
Engine
3
Cisco 2960G
Cisco 3750G
IP
IP
Metropolitan
Area
Network
Access
Control
System
3
1
2
Remote Site
Cisco 4500
Distribution
WLAN
Controller
Access
Control
System
4
3
2
MSE
Mobility Services
Engine
WLAN
Controller
4
3
Cisco 2960G
3
Cisco 3560G
IP
Cisco 3750G
IP
IP
229340
IP
Cisco 2960G
Remote Site
Cisco 3750 Stack
Distribution
Small Enterprise Design Profile Reference Guide
6-56
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
NMSP Traffic Flows Originating At The MSE
NMSP sessions between the MSE, WLAN controllers and context-aware switches are bi-directional in
nature. Here, we are referring simply to that portion of any flow whose source address is that belonging
to the Mobility Services Engine. Since the Mobility Services Engine does not mark the DSCP for the
NMSP traffic it transmits into the network, all such traffic originating at the MSE will contain the default
DSCP marking of 0x00. Left unchanged, when congestion is encountered, all NMSP traffic marked in
this fashion will be treated with the lowest priority. This increases the probability that NMSP session
data will be dropped in the network during periods of congestion. Left unchecked, this can result in
NMSP session stability issues, and poor context-aware performance. Clearly, this is not desirable. We
can avoid this by classifying and remarking NMSP data appropriately, as described in this section.
Since the MSE currently does not support classification and marking the DSCP value assigned to its
traffic, we must make use of the classification and marking capabilities available to us in the Cisco
Catalyst switch to which the MSE is attached. Thus, where 1 appears in Figure 6-35, we will:
•
Define the criteria to select our NMSP traffic
•
Define a class-map that will filter NMSP traffic using this criteria
•
Define a service-policy to assign our desired DSCP value to the filtered traffic using the class-map
•
Apply the service-policy to the port to which the MSE is attached.
Since we know that NMSP traffic will involve TCP port 16113, we can make use of this to identify
NMSP traffic and proceed to mark it appropriately.
The following example defines a policy map that will mark NMSP traffic inbound to the network from
the MSE as DSCP 0x12 (also referred to as DSCP 18, Assured Forwarding 21), police down to 10 Mbps
per 8k burst, and mark down any NMSP traffic exceeding this accordingly using the QoS map.
mls qos
!
class-map match-all CAS
match access-group name NMSP
!
policy-map MSE-Policy
class CAS
set dscp af21
police 10000000 8000 exceed-action policed-dscp-transmit
!
ip access-list extended NMSP
remark Identify NMSP traffic
permit tcp any any eq 16113
permit tcp any eq 16113 any
On the switch interface where the MSE is attached, it is imperative that service-policy statement appears
to assign the policy map to the interface. For example:
interface GigabitEthernet1/0/2
description Mobility Services Engine MSE1
.
.
service-policy input MSE-Policy
Small Enterprise Design Profile Reference Guide
6-57
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
NMSP Traffic Generated By WLAN Controllers
As was the case with the MSE, NMSP traffic entering the network originating at the WLC is also marked
with the default DSCP value of 0x00. Left unchanged, when congestion is encountered all NMSP traffic
marked in this fashion will be treated with the lowest priority. Once again, this is not desirable and we
shall address it in this section.
Like the MSE, the WLAN controller does not provide us with the ability to mark the NMSP traffic as
we see fit, thus we must instead classify and mark this traffic inside the network. In accordance with
general best practices, this operation is performed as close in the network as possible to the WLAN
controller. Therefore, in Figure 6-35, the points at which this should occur are shown by 2. Note that
traffic coming from both normally active as well as any backup WLAN controllers located in the data
center services switch block must be classified and marked.
The method used to accomplish this classification and marking for WLAN controllers is very similar to
that presented in the previous section for the MSE. However, since the WLAN controller is attached to
the Cisco Catalyst switch using an Etherchannel port-channel group, there will be some relevant
differences relating to whether the port-channel attachment is to a Catalyst 3750 switch stack or Catalyst
4500 distribution switch.
For example, when using the Catalyst 3750 switch stack in distribution, the following configuration
would apply:
mls qos
!
class-map match-all CAS
match access-group name NMSP
!
policy-map CAS-Policy
class CAS
set dscp af21
police 10000000 8000 exceed-action policed-dscp-transmit
!
ip access-list extended NMSP
remark Identify NMSP traffic
permit tcp any any eq 16113
permit tcp any eq 16113 any
The 3750 switch stack would also have a port-channel definition that would refer to the two physical
Ethernet interfaces comprising the port-channel group.
interface Port-channel4
description EC trunk to site 1266, WLC, interfaces gig1/0/28, 2/0/28
It is important to note that the 3750 switch stack does not support the use of a policy-map statement on
a port-channel definition. In order to assure that NMSP traffic originating at the WLAN Controller
in-bound to the network is properly classified and marked, ensure that a service-policy statement appears
on each of the two physical interfaces that comprise the WLAN controller port-channel group in the
3750 switch stack. For example:
interface GigabitEthernet1/0/28
description trunk to site 1266 WLC, port-channel4
.
.
mls qos trust dscp
channel-group 4 mode on
service-policy input CAS-Policy
!
interface GigabitEthernet2/0/28
description trunk to site 1266 WLC, port-channel4
Small Enterprise Design Profile Reference Guide
6-58
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services in Small Enterprise Designs
.
.
mls qos trust dscp
channel-group 4 mode on
service-policy input CAS-Policy
When using the Catalyst 4500 as the distribution switch, the scenario is a bit different. The Catalyst 4500
requires the service-policy to be assigned to the port-channel group used for the WLAN controller, and
not the physical interfaces. Thus, our recommended configuration when using a Catalyst 4500 in
distribution would be as follows:
!
class-map match-all CAS
match access-group name NMSP
!
policy-map CAS-Policy
class CAS
set dscp af21
police cir 10000000
conform-action transmit
exceed-action set-dscp-transmit default
!
ip access-list extended NMSP
remark Identify NMSP traffic
permit tcp any any eq 16113
permit tcp any eq 16113 any
!
interface Port-channel6
description trunk to main site, interfaces gig2/11, gig3/11
.
.
service-policy input CAS-Policy
!
!
interface GigabitEthernet2/11
description trunk to main site WLC, port-channel6
.
.
channel-group 6 mode on
!
interface GigabitEthernet3/11
description trunk to main site WLC, port-channel6
.
.
channel-group 6 mode on
!
NMSP Traffic Generated By Context-Aware Switches
As you will recall, specific models of Catalyst switches described earlier in this document (such as the
2960G, 3560, 3750, 4500, 4900 and others) can provide context-aware information to the MSE relating
to IP-based attached devices. When this capability is enabled in a context-aware Cisco Catalyst switch,
the switch itself participates in an NMSP session with the MSE. This NMSP session is separate and
independent of any NMSP sessions that may pass through the switch to or from attached devices (such
as the WLAN controllers described earlier, or other downstream context-aware Cisco Catalyst switches).
In order to ensure that the NMSP traffic originating at these switches is treated appropriately in the
network during times of congestion, it is important that we properly classify the NMSP data originating
at the context-aware switch and destined in-bound to the network for the MSE. Depending on the type
of switch used, the exact method we shall use to apply this classification and marking will vary.
Small Enterprise Design Profile Reference Guide
6-59
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Hardware/Software Releases
Layer-Three Context-Aware Switches
In the case of context-aware Layer-3 switches such as the 3750, 3560 and 4500, we can make use of local
policy routing to assign an IP precedence 2 (DSCP 0x10) to NMSP traffic originating from the switch
itself. Local policy routing in this fashion is performed internal to the context-aware L3 switch
originating the NMSP traffic, and the traffic is marked prior to being introduced into the network. At the
locations in Figure 6-35 marked with 3, we can perform the required classification using a local policy
route-map as follows:
ip local policy route-map switch-NMSP
!
route-map switch-NMSP permit 10
match ip address NMSP
set ip precedence 2
set ip tos max-throughput
!
ip access-list extended NMSP
permit tcp any any eq 16113
permit tcp any eq 16113 any
!
Layer Two Context-Aware Switches
However, the use of the Cisco 2960G context-aware Layer 2 (L2) switch in the access layer presents a
challenge. Local policy routing is not supported by the L2-only 2960G. Therefore, as a work around, we
must classify and mark the NMSP traffic originating from the 2960G at the next switch upstream to it in
the network. In the small enterprise designs discussed in this document, this upstream switch would be
the 3750 switch stack or 4500 distribution switch. In Figure 6-35, the points at where this must be
performed are indicated by 4.
Note that in the Small Enterprise Design Profile, the 2960G access layer switches are attached to the
distribution switches using an Etherchannel port-group, in a fashion very similar to that of the WLAN
controllers. Therefore, similar techniques can be used to classify and mark the NMSP traffic originating
at the 2960G and coming across the Etherchannel link.
Thus, for a 3750 switch stack used in distribution, a class-map and policy-map can be applied as
described previously, and the service-policy input CAS-Policy statement applied to the physical
interfaces in the 3750 switch stack that comprise the port-channel group to the 2960G switch. Just as for
an Etherchannel-attached WLAN controller, a service-policy cannot be applied to the port-channel
group definition used for the 2960G.
When using a Catalyst 4500 as the distribution switch with a context-aware 2960G at the access layer,
apply the service-policy input CAS-Policy statement to the port-channel group definition for the
2960G. This would be done in a similar fashion to that described earlier for a Etherchannel-attached
WLAN controller.
Hardware/Software Releases
Table 6-2
Hardware/Software Releases
Component
Version
Comments
Wireless Control System
6.0.132.0
For licensing and part number information, see
http://www.cisco.com/en/US/prod/collateral/wireless/ps5755/ps630
1/ps6305/product_data_sheet0900aecd804b4646.html
Small Enterprise Design Profile Reference Guide
6-60
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services—General Best Practice References
Table 6-2
Hardware/Software Releases
Mobility Services Engine 3350
6.0.85.0
For client and tag licensing information, see
http://www.cisco.com/en/US/prod/collateral/wireless/ps9733/ps974
2/data_sheet_c07-473865.html
Mobility Services Engine 3310
6.0.85.0
For client and tag licensing information, see
http://www.cisco.com/en/US/prod/collateral/wireless/ps9733/ps974
2/data_sheet_c07-473865.html
WLAN Controller 4404
6.0.182.0
Standalone Cisco Wireless LAN Controller
WLAN Controller 4402
6.0.182.0
Standalone Cisco Wireless LAN Controller
Catalyst 4500
12.2.53-SG
Large site distribution switch; must use crypto (K9) image if CAS for
wired devices is desired
Catalyst 3750E Switch Stack
12.2(52)SE
Small site distribution switch; must use crypto (K9) image if CAS for
wired devices is desired
Catalyst 2960G
12.2(52)SE
Access switch; must use crypto (K9) image if CAS for wired devices
is desired
Catalyst 3750E
12.2(52)SE
Access switch; must use crypto (K9) image if CAS for wired devices
is desired
LAP1252 Access Point
6.0.182.0
Cisco Aironet 1250 Series Wireless Access Point
LAP1142 Access Point
6.0.182.0
Cisco Aironet 1140 Series Wireless Access Point
WCS Navigator
1.5.132.0
Optional component; for licensing information, see
http://www.cisco.com/en/US/prod/collateral/wireless/ps5755/ps630
1/ps7305/product_data_sheet0900aecd8065bd19.html
AeroScout T2 RFID Asset Tag
4.33
Available from http://www.aeroscout.com/; RFID tags are required
only if RFID tracking is desired
AeroScout T3 RFID Asset Tag
6.05
Available from http://www.aeroscout.com/; RFID tags are required
only if RFID tracking is desired
AeroScout EX-3200 Exciter
33007/60007
Chokepoint trigger, optional RFID enhancement; EX-2000 model
recommended if outdoor placement is necessary
Context-Aware Services—General Best Practice References
The following are recommended references with regard to general best practice deployment
recommendations for Cisco Unified Networks making use of Context-Aware Services release 6.0:
•
A cornerstone of a successful design is knowledge of established best practices. Thus, it is highly
recommended that you become familiar with the material presented in the following documents:
– Context-Aware Solution Deployment Guide
http://www.cisco.com/en/US/products/ps9742/products_tech_note09186a00809d1529.shtml
– VoWLAN Design Guide 4.1
http://www.cisco.com/en/US/docs/solutions/Enterprise/Mobility/vowlan/41dg/vowlan41dg-bo
ok.html
– Context-Aware Services Configuration Guide, Release 6.0
http://www.cisco.com/en/US/docs/wireless/mse/3350/6.0/CAS/configuration/guide/CAS_60.h
tml
Small Enterprise Design Profile Reference Guide
6-61
Chapter 6
Small Enterprise Design Profile (SEDP)—Context-Aware Services
Context-Aware Services—General Best Practice References
– Wireless LAN Controller Configuration Guide, Release 6.0
http://www.cisco.com/en/US/docs/wireless/controller/6.0/configuration/guide/Controller60C
G.html
– Wireless Control System Configuration Guide, Release 6.0
http://www.cisco.com/en/US/docs/wireless/wcs/6.0/configuration/guide/WCS60cg.html
•
If you intend to make use of RFID tags in your Context-Aware solution, it is also recommended that
you become familiar with the following document which explains the operation of the Cisco
Context-Aware Engine for Tags:
– AeroScout Context-Aware Engine for Tags for Cisco MSE User Guide, version 3.2
http://support.aeroscout.com
•
If you intend to track the location and status of wired devices attached to Cisco Catalyst switches,
it is recommended that you familiarize yourself with the appropriate configuration guide for this
feature in the switches you will be using. For example:
– Catalyst 2960 Switch Software Configuration Guide, 12.2(50)SE
http://www.cisco.com/en/US/docs/switches/lan/catalyst2960/software/release/12.2_50_se/con
figuration/guide/scg.html
– Catalyst 3750 Switch Software Configuration Guide, 12.2(50)SE
http://www.cisco.com/en/US/docs/switches/lan/catalyst3750/software/release/12.2_50_se/con
figuration/guide/scg.html
– Catalyst 4500 Series Switch Software Configuration Guide, 12.2(53)SG
http://www.cisco.com/en/US/docs/switches/lan/catalyst4500/12.2/53SG/configuration/config.
html
During any deployment of Context-Aware Services and Cisco Unified Wireless Networks, detailed site
surveys should be performed by a Cisco Wireless LAN Specialized Partner with expertise in voice, high
speed data, and Context-Aware wireless network deployment. Cisco Systems also offers a complete
package of bundled design and deployment services via the Cisco Advanced Services team. Cisco and
our Wireless LAN Specialized Partners offer Context-Aware Design and Implementation Services to
help you successfully deploy enterprise-class wireless connectivity. These services include the
installation and configuration of crucial components such as the Mobility Services Engine (MSE),
helping you to take full advantage of the strong security, management, and investment protection
features that are built into Cisco Context-Aware components. In addition to planning, design, and
implementation, we also offer services based on proven methodologies for operating and optimizing the
performance of a Context-Aware Mobility solution, along with its associated technologies and strategies.
Note
The importance of a properly performed wireless site survey of your facility cannot be over-emphasized.
For more information on Cisco bundled planning, design, and deployment services, refer to
http://www.cisco.com/en/US/services/ps2961/ps6899/ps8306/services_overview_context_aware.pdf.
To locate a Cisco Wireless LAN Specialized Partner, refer to
http://tools.cisco.com/WWChannels/LOCATR/openAdvanceSearch.do.
Small Enterprise Design Profile Reference Guide
6-62
CH A P T E R
7
Small Enterprise Design Profile(SEDP)—Digital
Media and Video Surveillance Design
Digital Media System Overview
The Small Enterprise Design network architecture in combination with the Cisco Digital Media System
(DMS) creates an environment that streamlines and automates information flow and process throughout
enterprise network.
The evolution of digital communication in the 21st century enterprise network environment is
transforming processes and empowering educators to develop and deploy “eye-catching”, compelling,
and integrated video content.
•
Digital media communication in small enterprise systems is an extremely effective tool to deliver
dynamic and cost effective subject-matter to staff and customers.
•
Using digital media systems enables enterprise organization to keep their customers connected,
increases awareness, and integrates with safety and security system notifications.
•
Integrating digital media technologies assists corporate information, business events and internal
professional development.
•
Comprehensive digital video systems transform video delivery processes in individual business
function; network administrator can create customized, on-demand educational video content or
even relay live video feed during national events or local emergencies.
Video Surveillance Overview
Video surveillance has been a key component of safety and security groups for many organizations. As
an application, video surveillance has demonstrated its value and benefits countless times by providing
real-time monitoring of a facility’s environment, people, and assets as well as by recording events for
subsequent investigation, proof-of-compliance, and audit.
For small enterprise network systems that need to visually monitor or record events video, surveillance
has become more important as the number of security risks increases. In addition to video analytics, the
value of video surveillance has grown significantly with the introduction of motion, heat, and
environmental sensors.
In a typical enterprise network environment, several systems are deployed for disparate applications,
such as physical access control, fire and smoke detection, and video surveillance. These applications
typically do not communicate with each other and require different management and support personnel.
Small Enterprise Design Profile Reference Guide
7-1
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Video Surveillance Overview
As a result, owners and operators suffer from a lack of operational consistency, interoperability, and
capabilities that translate into higher capital and operational costs and limit the return on their
investment.
Cisco’s solution offers software and hardware to support video transmission, monitoring, recording, and
management. Cisco video surveillance solutions work in unison with the advanced features and
functions of the IP network infrastructure—switches, routers, and other network security devices—to
enable secure, policy-based access to live and recorded video.
Cisco Digital Media System Architecture
Cisco has developed a comprehensive, scalable, and network-centric DMS architecture that is built on
three major digital application components. Each application is specifically designed to address key
challenges, regardless of how or where the digital content is designed and developed. The multifunction
management and adaptive media system is common to these three applications:
•
Desktop video—Interactive education training application that allow employees to watch
instructional and corporate event videos on-demand or via live streaming. With the right privilege,
employees can navigate a Cisco Video Portal database to securely access relevant training video
content.
•
Enterprise TV—While the targeted users for desktop video applications are individuals or group of
employees, enterprise TV expands the same video capability to a larger audience. Live or
on-demand pre-recorded training video can be broadcast in a enterprise network independent of
geographical location. In addition to internally developed video material, main and remote site
network administration can also integrate live TV programming feed from the local content
provider.
•
Digital signage—Enables innovative ways to publish content and information that improves the
user experience, allows dynamic updates, and increases campus safety and security. Some of the
common digital signage use cases in enterprise networks include announcing main and remote site
news, major events, corporate initiatives, board meetings, etc.
Figure 7-1 shows a Cisco DMS solution suite that is a set of product and technologies developed to create
an end-to-end digital media network. The products are divided into three major functions, create,
manage, and access.
Small Enterprise Design Profile Reference Guide
7-2
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Video Surveillance Overview
Figure 7-1
Digital
Application
Cisco DMS Solution Suite
Create
Manage
Access
Cisco LCD
Display
Digital
Signage
Flash,
Shockwave,
HTML
Digital Media Manager
Digital Media
Player
Desktop
Video
Digital Media
Encoders
Cisco Video Portal
Cisco MCS
Cisco LCS
Display
Enterprise
TV
227853
Scientific Atlanta
Encoders
The digital media components in the Cisco DMS solution suite assist small enterprise network in
deploying digital media solutions at their own pace; e.g., initially deploy digital signage with simple
development applications and deploy interactive video solutions in subsequent phases.
The Cisco Digital Media Manager (DMM) is a Web-based application that simplifies deploying all three
digital media applications in small enterprise network.
DMS Solution for Small Enterprise
DMS solutions in small enterprise have proven valuable in overcoming key challenges in the
development of next-generation education delivery processes. DMS provides the flexibility to develop
training content that can be accessed by employees anytime, and anywhere. Cisco DMS relies on a
resilient, scalable, and reliable network infrastructure for seamless end-to-end content delivery.
Figure 7-2 shows an end-to-end digital media reference model with all three applications enabling a
unified digital network service for the main site of small enterprise network.
The following are some of the key benefits of this DMS design model:
•
Centralized management at the main site providing consistent publishing policies, security,
scalability and reduced operational and maintenance cost.
•
Distributed storage and media access points enabling main and remote site to use centrally
developed content with reduced bandwidth capacity and increased availability during network
instability
•
In large-scale video networks, the Cisco Application and Content Network Services (ACNS) or
WAN optimization appliances can be deployed to increase media performance and reduce expensive
WAN bandwidth requirements.
Small Enterprise Design Profile Reference Guide
7-3
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Video Surveillance Overview
Figure 7-2
End-to-End Reference DMS Solution in Small Enterprise Network Architecture
Main Site
Admin
Lobby
Cafe
HDTV
HDTV
HDTV
HDTV
HDTV
HDTV
Serverfarm
Studio
Conference
Room
DMM
Internet
DMS
Admin
Web Server
Streaming
Server
Video
HDTV
MXE
CVP
Scientific Atlantic
Encoders
ADS/LDAP
Digital Media
Encoder
WAN
Remote Site
Admin
Lobby
Cafe
HDTV
HDTV
HDTV
HDTV
HDTV
HDTV
Studio
Server Room
Streaming
Server
Digital Media Encoder
Recreation
Center
Video
Scientific Atlantic
Encoders
HDTV
HDTV
HDTV
HDTV
Web
Admin
Digital Signage
Enterprise TV
229299
Web Server
Conference
Room
Desktop Video
with Video Portal
Digital Media
Player
The following section provides an overview of various deployment scenarios, device components,
communication, and network requirements for digital media applications in small enterprise network.
For a detailed digital media design and implementation guide, refer to the following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Video/DMS_DG/DMS_dg.html
Small Enterprise Design Profile Reference Guide
7-4
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Desktop Video Application Overview
Desktop Video Application Overview
Employees and administrative staff can watch live or pre-recorded video events from their personal
computers at any location and at any time. The Cisco DMS Digital Video application empowers
employees to extend the audience reach ability to remote site locations by broadcasting live or recording
training sessions available as video on-demand (VoD).
Cisco Desktop Video applications offer several benefits:
•
Customizable interface with program guide and search window.
•
Employees can create a personalized video playlist.
•
Questions and comments can be made during live video broadcast events.
•
Restrict video content access based on Active Directory or LDAP authentication and privilege.
•
Wide format support—Adobe Flash, Windows Media, H.264, QuickTime, etc.
•
Player Controls—Synchronized slides, advanced video and controls, etc.
Desktop Video Components
As one of the integrated components of the Cisco DMS solution suite, the digital video application uses
common video development, management, and publishing components. In combination, external
authentication servers can provide secure, on-demand video content and live video broadcast services to
the desktop or on Cisco LCD TVs deployed in different physical locations.
The Cisco DMS solution suite consists of:
•
Digital Media Manager (DMM)—Centralized management appliance in main site governs the
content and communicates with local or remotely deployed critical desktop video components, e.g.,
CVP, HTTP server etc.
•
Digital Media Encoder (DME)—Single or multi-channel media encoder receives live analog/digital
feed from cameras or television service providers and transports over the IP network to the
streaming server.
•
Streaming Server—Provides stream splitting capabilities, allowing many clients to view a single
live stream from DME or pre-recorded source (replay)
•
Cisco Video Portal (CVP)—Web-based video navigation engine provides access to users after
successful authentication with Active Directory or LDAP server.
•
Web Server/Content Repository—Stores all VoDs referenced by Video Portal server. User triggers
VoD request to access video content through CVP and the request gets redirected to pull video file
from the content repository to the requested user.
Publishing Live and Video On-Demand Content
A critical feature of the Cisco Digital Media System for Cisco Desktop Video is its ability to simplify
the publishing of live and on-demand digital media files to the Cisco Video Portal (CVP). On-demand
video content can be uploaded from the developer’s computer directly to the DMM server for staging
and previewing prior to deployment. This staging capability includes the addition of an approval process
within the content workflow to help ensure that school branding, publishing policies, and messaging are
properly incorporated in the content. Post approval process, the content can be moved or deployed to the
Cisco Video Portal using secure file transmission.
Small Enterprise Design Profile Reference Guide
7-5
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Enterprise TV Application Overview
The Cisco DMM works in conjunction with Cisco Digital Media Encoders (DME) to create and deploy
live content to the Cisco Video Portal. The Cisco DMM first manages the Cisco DME to set up their
encoding profiles, defining the bit rate, format, and media type. The Cisco DMM also defines the port
that the Cisco Digital Media Encoders will stream from, so that the streaming servers can pull the stream
to their live publishing points. These publishing points are then deployed to the Cisco Video Portal
through the Cisco Digital Media Manager deployment process. The same workflow defined for the
on-demand digital media content is applied to live events, providing a consistent, easy-to-use process
for all types of deployments. Figure 7-3 shows a schematic of Cisco DMM video management.
Figure 7-3
VoD and Live Video Broadcast Using Digital Video Application
Video On-Demand (VoD)
Live Video Broadcast
Studio
Live
Video
Feed
Content
Sever
DME
DMS
Admin
DMM
Scientific Atlantic
Encoders
CVP
Video
ADS/LDAP
Portal Access
Multicast/HTTP
227855
HDTV
Enterprise TV Application Overview
The Enterprise TV (ETV) application brings standard or high-definition television network channels
into IP-based networks. Deploying Scientific Atlanta encoders in a video head-end role performs the
interworking function that transforms video source from television service provider to an IP based video
delivery within the campus network. When the ETV module is enabled in Cisco DMM appliance, the
network administrator can program the channel guide information to be broadcast in different physical
locations, e.g., channel number, name, port number, etc. To improve the user experience in navigating
video channels, ETV Electronic Program Guide (EPG) can be programmed to provide information on
channel lineup, and current and future programming information, which is similar to television service
Small Enterprise Design Profile Reference Guide
7-6
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Enterprise TV Application Overview
provided programming guide. To watch the live video channels, users can use Cisco DMP and remote
control to navigate and access the channel. Video delivery over IP networks can be unicast or multicast,
depending on how IP/multicast is designed in campus network.
With larger displays in key physical locations, the ETV application becomes the primary
communications interface in the main site targeting large audiences. For example, live or VoD broadcast
for professional training, demonstrations, meetings, etc., targeting all the employees in a personal office,
conference rooms, or live news, etc.
Enterprise TV Components
To broadcast developed VoD or live video through ETV, the core Cisco DMS components used in the
desktop video application can be leveraged to integrate ETV in the network. The primary difference
between ETV and desktop video in this case would be Digital Media Player (DMP) instead of a PC as
an access end point for a large screen display targeting a larger audience. To broadcast television
network channels in the campus network, the Scientific Atlanta encoder must be integrated along with
other ETV components. It is recommended to deploy distributed television service in local campus and
not forward non-critical video traffic over the WAN infrastructure.
•
Scientific Atlanta Encoder—An encoder system that provides interworking function between analog
or digital television service provider and IP network. Encodes live video input and transforms into
MPEG-2 or MPEG-4 multicast stream.
•
Cisco Digital Media Player (DMP)—Key media access end point that connects to Cisco LCD TV
for large size displays. DMP provides capabilities to decode multi-format graphics and stream video
content received over unicast or multicast IP network.
Broadcasting Live TV or Video On-Demand Content
The communication flow between the DMP and the DMM and Video Portal function is similar to
desktop video applications. Proper planning, technologies, and equipment must be deployed in campus
network for successful live television video delivery. When designing the playlist in Cisco Enterprise
TV module, the network administrator must understand that it can support up to 99 live and on-demand
video channels broadcast in the campus network. The DMM administrator in the main site can use the
DMM-ETV software module to create customized TV navigation interfaces, such as adding corporate
logo and skins, programming video channel assignments, and configuring specific video channel
assignment to DMP deployed in specific campus location. Figure 7-4 shows a schematic of the
Enterprise TV video architecture.
Main and remote site architects must understand the codec type required for publishing video in the
campus network. Deployed digital encoders must follow the MPEG2 standard specification to stream
the video. It is recommended to deploy Scientific Atlanta 9032SD or 9050HD encoder to stream live
video stream to DMP for Enterprise TV application.
Small Enterprise Design Profile Reference Guide
7-7
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-4
Live Video Broadcast and VoD Using Enterprise TV Application
Video On-Demand (VoD)
Live Video Broadcast
Studio
Content
Sever
Live TV
Live
Video
Feed
DME
Set-Top-Box
DMS
Admin
DMM
Scientific Atlantic
Encoders
CVP
Live
Video
TV
ADS/LDAP
Portal Access
RTSP/HTTP/
Multicast
HDTV
HDTV
227856
HDTV
Digital Signage Application Overview
Cisco's Digital Signage solution is a comprehensive solution for the publishing of dynamic and
on-demand signage using digital media displays deployed locally or regionally in small enterprise
network over an IP network. The key benefits of digital signage over traditional static signs are that the
digital content can be exchanged and updated more dynamically, using digital media tools to make the
content more relevant and interactive. Publishing corporate messages, local announcements, or
emergency alerts through Cisco digital signage becomes more effective and with better investment
return compared to traditional models.
The Cisco digital signage application is a Web-based media management and publishing application that
creates a playlist with a set of content that is required to be published to a single or group of DMPs in a
network. The digital signage application use standard HTTP protocols to communicate with centrally
deployed DMM in the main site serverfarm and single or distributed Web servers to pull and publish the
real-time information to display on large Cisco LCD TVs. With flexible user administration, the DMM
administrator is empowered to create groups of users with different privileges who can develop, publish,
and manage the signage content; e.g., user group like Web admin, IT admin, and security admin can
create content assets, control display properties, etc.
Small Enterprise Design Profile Reference Guide
7-8
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Digital Signage Components
The requirements of the integrated digital media components depend on which content needs to be
published through a signage module. The Cisco digital signage application provides the flexibility for
the DMM administrator or Web administrator to develop multi-functional, integrated content that can
play key messages, stream VoD files from a content server, and keep connected with external
information. For example, administrator or employees can publish a single Adobe flash file that is
composed of static text information, embedded with education short VoD stream and live news
information with RSS feed. The basic digital signage components are:
•
Digital Media Manager (DMM)—Centralized management appliance in main site governs the
content and communicate with local or remotely deployed critical desktop video components, e.g.,
CVP, HTTP server, etc.
•
Cisco Digital Media Player (DMP)—The Cisco DMP is a highly reliable IP-based hardware
endpoint for video decoding and playback of digital media content—including high-definition live
broadcasts and VoD, Flash animations, text tickers, and other Web content—across digital displays.
DMP is a critical component of the digital signage and ETV applications allow for the networking
of digital displays and the broadcasting of live and on-demand media. The current DMP portfolio
includes the Cisco DMP 4305G for standard signage and ETV and the Cisco DMP 4400G for
high-end signage and ETV.
•
Cisco LCD Professional Series—For an end-to-end Cisco digital media solution, the Cisco LCD
professional series displays is a high definition LCD display that can be centrally managed through
DMM.
•
Web Server/Content Repository—Stores all HTML, VoDs, flash files referenced by HTTP server or
Video Portal server. Multiple files can be played on same DMP; based on Web application design,
the program triggers the content request and it is pulled by DMP from a source server.
Video Surveillance System Architecture
The Cisco Video Surveillance solution relies on an IP network infrastructure to link all components. The
design of a highly available hierarchical network has been proven and tested for many years and allows
applications to converge on an intelligent and resilient infrastructure.
Figure 7-5 shows the main components of the Cisco Physical Security solution, including video
surveillance, physical access control, incident response and integration with third-party systems.
Small Enterprise Design Profile Reference Guide
7-9
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-5
Cisco Physical Security Components
Video and Device
Capture
Video Surveillance, Access Control
and Incident Management
Response
IP
IP Surveillance
Cameras
Safety and
Security
Physical Access
Manager Multiservices
Platform
Desktop
IP
IP
Third Party
Analog and IP
Cameras
Cisco Video
Surveillance
Operations
Manager (VSOM)
Open API - Third
Party Systems
IPICS
Multiservices
Platform
ISR - Video Media
Management and
Storage
Radios, Mobile Phones,
IP Phones
Digital Signage
227677
Physical Access
Gateway
Some of the benefits of Cisco’s Video Surveillance solution include the following:
•
Access to video at any time from any network location, enabling real-time incident response and
investigation.
•
Transfer of control and monitoring to any other point in the network in an emergency situation.
•
Ability to manage devices and alarms from a centralized location.
•
Ability for products from various vendors to interoperate in the same network.
•
An open, standards-based infrastructure that enables the deployment and control of new security
applications.
The main components of the Cisco Video Surveillance solution include the following:
•
Cisco Video Surveillance Media Server—The core component of the network-centric Video
Surveillance Manager solution. This software manages, stores, and delivers video from a wide range
of cameras and encoders over an IP network
•
Cisco Video Surveillance Operations Manager—The Operations Manager authenticates and
manages access to video feeds. It is a centralized administration tool for management of Media
Servers, Virtual Matrixes, cameras, encoders, and viewers and for viewing network-based video.
•
Cisco Video Surveillance IP Cameras—The high-resolution digital cameras are designed for
superior performance in a wide variety of environments.
•
Cisco Video Surveillance Virtual Matrix—The Virtual Matrix monitors video feeds in command
center and other 24-hour monitoring environments. It allows operators to control the video being
displayed on multiple local and remote monitors.
•
Cisco Video Surveillance Encoding Server—This all-in-one appliance encodes, distributes,
manages, and archives digital video feeds for analog cameras. Each server encodes up to 64
channels and provides up to 12 TB of storage.
•
Cisco Video Surveillance Storage System—This complementary component allows the Media
Server’s internal storage to be expanded with direct attached storage (DAS) and storage area
networks (SANs). The Storage System allows video to be secured and accessed locally or remotely.
The following subsections describe the components used for this solution.
Small Enterprise Design Profile Reference Guide
7-10
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Cisco Video Surveillance Media Server
The Cisco Video Surveillance Media Server (VSMS) is the core component in the Cisco Video
Surveillance Manager solution and performs the following networked video surveillance system
functions:
•
Collection and routing of video from a wide range of third-party cameras and video encoders over
an IP network
•
Event-tagging and recording of video for review and archival purposes
•
Secure local, remote, and redundant video archive capabilities
In Figure 7-6, the Media Server is responsible for receiving video streams from different IP cameras and
encoders and replicating them as necessary to different viewers.
Figure 7-6
Video Surveillance Media Server (VSMS)
Media Server
(VSMS)
IP
IP
IP Cameras
Operations
Manager
(VSOM)
IP
D
A
Analog Cameras
OM Viewer
OM Viewer
227678
Encoder
By using the power and advanced capabilities of today’s IP networks, the Media Server software allows
third-party applications, additional users, cameras, and storage to be added over time. This system
flexibility and scalability supports the following:
•
Hundreds of simultaneous users viewing live or recorded video
•
Standard video compression algorithms such as MJPEG, MPEG-2, MPEG-4, and H.264
simultaneously via a single Media Server
•
Conservation of storage using events and loop-based archival options
•
Integration with other security applications
Cisco Video Surveillance Operations Manager
Working in conjunction with the Cisco Video Surveillance Media Server, the Cisco Video Surveillance
Operations Manager (VSOM) enables organizations to quickly and effectively configure, manage, and
view video streams throughout the enterprise. Figure 7-7 shows the Operations Manager main screen,
which is accessed through a Web browser.
Small Enterprise Design Profile Reference Guide
7-11
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-7
Video Surveillance Operations Manager
The Operations Manager meets the diverse needs of administrators, systems integrators, and operators
by providing the following:
•
Multiple Web-based consoles to configure, manage, display, and control video throughout a
customer’s IP network.
•
The ability to manage a large number of Cisco Video Surveillance Media Servers, Cisco Video
Surveillance Virtual Matrixes, cameras and users.
•
Customizable interface, ideal for branded application delivery.
•
Encoder and camera administration.
•
Scheduled and event-based video recording.
•
Interface to Media Server and Virtual Matrix software for pushing predefined views to multiple
monitors.
•
User and role management.
•
Live and archived video views.
•
Friendly user interface for PTZ controls and presets, digital zoom, and instant replay.
•
Event setup and event notifications.
•
“Record Now” feature while viewing live video
Cisco Video Surveillance IP Cameras
Cisco 2500 Series Video Surveillance IP Camera
The Cisco 2500 Series Video Surveillance IP camera is a high resolution standard-definition,
feature-rich digital camera designed for secure performance in a wide variety of environments. The
camera supports MPEG-4 and MJPEG compressions with up to 30 frames per second.
Contact closure and two-way audio allow integration with microphones, speakers, and access control
systems. By providing wired and wireless models, the Cisco 2500 IP camera provides an ideal platform
for integration and operation as an independent device or as part of the Cisco Video Surveillance
network. Figure 7-8 shows both the wired and wireless models of the 2500 IP Camera.
Small Enterprise Design Profile Reference Guide
7-12
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-8
Cisco 2500 Series IP Cameras
The 2500 Series IP camera provides the following features:
•
The camera employs powerful digital imaging technology, allowing it to capture high-quality
images in a wide variety of indoor and outdoor lighting conditions. It uses a progressive scan
image-sensor with global electronic shuttering to ensure natural color rendition, and minimal
motion blurring.
•
The wireless IP camera model supports 1X2 Multiple Input Multiple Output (MIMO)
communication, which provides better data throughput and higher link range than single antenna
designs. The wireless IP camera offers strong wireless security using Wi-Fi Protected Access
(WPA)/WPA2 and supports various network protocols for 802.1x authentication.
•
Power over Ethernet (PoE) 802.3af or DC power through an optional external power supply.
•
Support for the Cisco Media API, an open, standards-based interface that allows integration with
compatible video surveillance management systems.
•
Support for 802.1x authentication on both the wired and wireless models.
Cisco 4000 Series Video Surveillance IP Camera
The Cisco Video Surveillance 4000 Series IP Cameras employ true high-definition (HD) video and
H.264 compression, streaming up to 30 frames per second at 1080p (1920 x 1080) resolution. The Cisco
4000 IP Camera series also supports contact closure and two-way audio allow integration with
microphones, speakers, and access control systems.
The Cisco 4000 Series includes two models: the CIVS-IPC-4300 and CIVS-IPC-4500. These cameras
have identical feature sets, with the exception of the additional digital signal processor capabilities
specifically designed to support real-time video analytics at the edge on the CIVS-IPC-4500. On this
model, applications and end users have the option to run multiple analytics packages without
compromising video streaming performance on the camera.
Figure 7-9 shows a Cisco 4000 IP Camera with an optional DC Auto Iris Lens.
Small Enterprise Design Profile Reference Guide
7-13
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-9
Cisco 4000 Series IP Camera
The 4000 Series IP camera provides the following features:
•
True high-definition video—The camera streams crisp and clear 1080p (1920 x 1080) video at 30
frames per second while maintaining surprisingly low network bandwidth.
•
Progressive scan video—The camera captures each frame at its entire resolution using progressive
scan rather than interlaced video capture, which captures each field of video.
•
Embedded security and networking—The camera provides hardware-based Advanced Encryption
Standard (AES).
•
IP Multicast for enhanced bandwidth management.
•
Event notification—The camera can examine designated areas for activity and notify users or other
applications when it detects activity that exceeds a predefined sensitivity and threshold.
•
True day/night functionality that includes an infrared (IR) filter that automatically switches to night
mode in low light scenes.
•
The camera supports Power-over-Ethernet (PoE) 802.3af, 12 VDC or 24 VAC power through an
optional external power supply.
•
The camera can be installed with a fixed mount or with an optional external pan/tilt mount and
motorized zoom lens.
For more information, refer to the IP Video Surveillance Deployment Guide at the following URL:
http://www.cisco.com/en/US/docs/solutions/Enterprise/Video/IPVS/IPVS_DG/IPVS_DG.pdf
Deploying Digital Signage in Small Enterprise Network
To ensure a successful digital media deployment, network, display, and management planning must be
done prior to deploying digital signage in the small enterprise network. A well-planned digital signage
network design provides flexibility to incrementally deploy desktop video and ETV digital media
applications without making major infrastructure changes. As described earlier, digital signage uses
standard HTTP protocols to pull and publish the signage content on displays. Network bandwidth
consumption for digital signs varies widely as it depends on playlist and content types. This document
provides the best practices to design and configure the digital signage with content developed with rich
text, flash, and animation and located on distributed Web servers to increase network efficiency.
Centralized Management Model
Cisco DMM is highly scalable appliance server that can be deployed centrally in main site location to
manage up to 1000 DMPs deployed in local and regional small enterprise campus network. Cisco DMP
deployed in main or remote sites can communicate with centralized DMM over LAN and WAN network
Small Enterprise Design Profile Reference Guide
7-14
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
using standard HTTP as the control protocol to receive Web or content server re-direction information
to display content. Deploying DMM in a centralized location allows the DMM administrators in main
site to manage all registered DMP in various ways:
•
Add and archive digital content and assign metadata and keywords.
•
Create and manage play lists, ticker alerts, messages, closed captions, and promotional interstitials.
•
Preview digital signage content and manage approval workflow.
•
Ability to pre-configure the playlist and schedule for instant and future deployments.
•
Take WAN optimization solution advantage and provide tight integration with Cisco ACNS and
Cisco Content Engines.
•
Manage user administrator accounts and permissions.
Centralizing DMP management and publishing signage content centrally at a main site allows the DMM
administrator to advertise consistent information and messaging throughout the network. To minimize
WAN network utilization, the remote site network administrator must leverage internal storage or their
local Web server to store and advertise the local, regional, and department news and information.
However the main site and Internet news must be communicated over the WAN.
The network architect must consider integrating enterprise-class Cisco Application and Content
Networking System (ACNS) that uses caching technology and offers higher scalability and reliable
video delivery solution in the network to improve end user experience and application response time.
When integrated with the Cisco Wide-Area Application Services (WAAS) solution, it helps in
optimizing WAN bandwidth utilization significantly with local redirection and high data compression
over the WAN.
Figure 7-10 is a validated design to integrate digital signage with centralized management in the main
site with distributed DMP in the network.
Small Enterprise Design Profile Reference Guide
7-15
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-10
Centralized DMM with Distributed DMP in Small Enterprise Network
Main Site
Admin
Lobby
HDTV
Cafe
HDTV
Serverfarm
HDTV
Conference
Room
DMS
Admin
DMM
HDTV
ADS/LDAP
Web Server
HTTP
Remote Site
Admin
Lobby
HDTV
Cafe
HDTV
HDTV
Server Room
229300
Web Server
Web
Admin
Distributed Content Storage Model
Digital signage content is typically wrapped with HTML or Adobe Flash applications that provide
greater flexibility for a Web administrator to display more types of content from various sources on a
single page. When deploying large numbers of DMPs with rich and static digital signage content, the
unicast communication between DMP and the distributed content server may waste network bandwidth
by retrieving the same content for continuous display. Hence it becomes important for the network
architect to understand the content distribution and network level requirements to optimally deploy
signage application in a campus network.
Small Enterprise Design Profile Reference Guide
7-16
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Depending on DMP scalability, overall network capacity, and the bandwidth allocation for digital
signage application, Cisco DMS offers the following three distributed content storage solutions for
Cisco DMP to pull and display the static content from the local network instead of downloading all of it
through the WAN network:
•
Cisco DMS-Content Distribution (CD)—Is an ideal solution for a small-scale enterprise network
with few Cisco DMPs. Cisco DMM can push the static HTML or flash content via FTP or SFTP
protocol to Cisco DMM on internal storage or to external storage device like USB drive and redirect
Cisco DMP to access internal storage. This solution helps to minimize WAN bandwidth utilization.
•
Local Content Server (Web or CIFS)—Storing the digital media content on a single local content
server, like HTTP or CIFS sever, gives the network administrator more flexibility and management
of the content distribution solution. The Web administrator can dynamically add, modify, and store
the updated HTML or Flash file on a centralized server for Cisco DMP to retrieve and display. On
the next HTTP request from DMP, the refreshed copy is displayed automatically. This solution
provides more flexibility compared to Cisco DMS-CD solution, as it can dynamically update
content without updating and managing content on each individual Cisco DMP.
•
Cisco ACNS—Highly scalable and intelligent large size video content distribution to remote
locations. Hierarchical content distribution system at main and remote sites distributes single copy
of pre-recorded video to ACNS edge at the remote sites. To increase WAN network efficiency,
Cisco ACNS leverages the caching technology and provides unicast VoD delivery in local LAN
networks to end users instead of downloading one copy for each user over the WAN network.
Validated Content Distribution Model
To provide a simplified, scalable, and cost-effective content distribution and management solution in
small enterprise network architecture, it is recommended to leverage local Web or CISF servers in the
main and remote sites to store and publish local digital signage content. Cisco DMM can be programmed
to redirect local DMPs to a local Web server and remote DMPs to pull the content from a local Web
server. Such distributed content storage design minimizes the critical WAN bandwidth usage to publish
local information. However, the WAN network may still be utilized to access global signage
information, such as county or state level education and emergency news that can be broadcast by Web
server from main site, and similarly real-time news ticker from the Internet can be embedded in major
content that provides constant world-wide news updates.
Small enterprise network architects and Web administrators must perform pre-deployment exercise to
assess the type of local versus distributed content (text/graphics/VoD/RSS) embedded in signage and the
number of DMPs to be deployed in remote sites. This assessment provides WAN bandwidth guidelines
to integrate digital signage in the network. As described earlier, Cisco WAN optimization solution like
ACNS and WAAS must be integrated in the network if it demands higher WAN bandwidth. Figure 7-11
depicts the unicast communication flow between Cisco DMP deployed across the network and Web
servers located in intranet and Internet domains.
Small Enterprise Design Profile Reference Guide
7-17
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-11
Distributed Content Server Communication
Intranet Content Server Communication
Admin
Lobby
HDTV
Internet Content Server Communication
Cafe
HDTV
Serverfarm
Admin
HDTV
HDTV
HDTV
Main Site
Cafe
HDTV
Serverfarm
Conference
Room
Web Server
Lobby
HDTV
Conference
Room
HDTV
Internet
Main Site
WAN
WAN
Remote Site
Admin
Remote Site
Lobby
HDTV
Cafe
HDTV
Admin
HDTV
Lobby
HDTV
Cafe
HDTV
HDTV
Server Room
Web Server
229301
Web
Admin
Implementing Network Services for Digital Signage
Prior to integrating digital signage applications, the network architect must make network services ready
with the best practices for resilient and seamless operation and integration. Building the network as a
highly-available platform is a foundational requirement for applications like IP telephony and digital
media solutions as they demand constant bandwidth and network availability. Deploying centralized
DMM with a distributed content server spans the digital communication beyond the campus boundary;
hence it is recommended to deploy digital media solutions based on these network design principles:
•
Low latency—Deploy the high-speed campus network that offers lower latency for real-time
applications like voice and video.
•
High availability—To increase network resiliency it is recommended to deploy the network with
redundant modules, systems, and power supplies offering non-stop communication and sub-second
network recovery during minor or major network failure events.
Small Enterprise Design Profile Reference Guide
7-18
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
•
QoS—Enhance user experience and content quality with robust QoS policies at the campus network
edge.
•
Confidentiality—Protect digital media end points, appliances, and data with centralized
authentication and encryption.
This section provides access layer design and configuration guidelines to deploy Cisco DMP at the
campus network edge and DMM in a centralized serverfarm in a main site. For more information on
design and implementation guideline for building a strong and resilient core and foundational campus
network, refer to Chapter 2, “Small Enterprise Design Profile(SEDP)—Network Foundation Design.”
Deploying Cisco DMP in the Access Layer Network
Cisco DMP is an enterprise managed and trusted end point in the campus access layer, hence the network
administrator must apply the common security and QoS policy for DMP as defined for other trusted
endpoints like IP phones. In a typical deployment scenario, a single access layer switch may be
connected to several other trusted and un-trusted endpoints, hence it becomes an important task for
administrator to provide secure and suitable network services.
Cisco access layer switches provides the flexibility to deploy Cisco DMP in two different modes, manual
deployment and plug-and-play Auto Smartport (ASP) macro.
Manual Deployment
Network administrator must manually implement the following three major network services to
successfully integrate DMP in the network:
•
Assigning unique Layer 2 VLAN
•
Implement network edge security
•
Implement network edge QoS
Assigning Unique Layer 2 VLAN
To provide secured and simplified digital signage communication, the DMP must be assigned a unique
broadcast domain. De-coupling DMP with other trusted and un-trusted end points makes DMP more
secure during any attacks and is easier to manage and troubleshoot. When single access layer connects
to multiple DMPs, then all the DMPs can be assigned on the same Layer 2 VLAN. Like any other logical
network partition design, it is recommended to use unique Layer 2 VLAN for DMP that are physically
deployed on different Cisco access layer switches. Figure 7-12 provides recommended Cisco DMP
Layer 2 segmentation guidelines in the access-layer:
Small Enterprise Design Profile Reference Guide
7-19
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Cisco DMP-Layer 2 VLAN Segmentation
Access
Lobby
DMP
Access
Cafe
DMP-1
VLAN 10
Cafe
DMP-2
Cafe
DMP-3
VLAN 20
227859
Figure 7-12
Cisco DMP cannot transmit or receive 802.1Q tagged frames, hence it is recommended to change default
switchport mode from dynamic to access mode. The following is a sample configuration to enable
VLAN in the database and apply VLAN on the DMP physical port:
2960-S
cr24-2960-DO(config-if)#interface FastEthernet0/7
cr24-2960-DO(config-if)# description CONNECTED TO LOBBY DMP
cr24-2960-DO(config-if)# switchport mode access
cr24-2960-DO(config-if)# spanning-tree portfast
cr24-2960-DO(config-if)# switchport access vlan 10
3750-X
cr25-3750-DO(config-if)#interface range GigabitEthernet 1/0/1 - 3
cr25-3750-DO(config-if-range)# switchport mode access
cr25-3750-DO(config-if-range)# spanning-tree portfast
cr25-3750-DO(config-if-range)# switchport access vlan 20
For flexible and scalable DMP deployment, it is recommended that the DMP edge port be in Layer 2
mode even when the access layer switch is deployed in a multilayer or routed access network design.
Implement Network Edge Security
Cisco DMP player requires communication with certain critical systems in the network, such as an IP
gateway, Cisco DMM, Web servers, SNMP, and NTP. To display the digital signage content and
synchronize with management servers, Cisco DMP receives more data from the network than
transmitting to the network.
Cisco Catalyst integrated security feature must be deployed on the physical port to protect DMP from
being attacked by viruses or unauthorized hosts. Based on the protocol and data communication
characteristics of Cisco DMP, it is recommended to apply the following set of security configurations to
protect the network and the DMP from unknown traffic floods and attacks:
Access
interface FastEthernet0/7
! Block transmitting all unknown unicast traffic
switchport block unicast
! Enable port-security on this port
switchport port-security
! Default, allow single-host to access this port
switchport port-security maximum 1
! Block receiving BPDU from this port
spanning-tree bpduguard enable
Small Enterprise Design Profile Reference Guide
7-20
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Implement Network Edge QoS
It is important to implement differential service treatment for digital media applications over non-critical
network traffic in the network. Depending on the digital media applications and the distributed content,
appropriate QoS services must be implemented at the network edge that connects media endpoints and
in the serverfarm where typically centralized management and content servers are deployed. As
described earlier, the Cisco DMP primarily use standard HTTP protocol to communicate with
centralized DMM management server and the distributed Web server to publish the digital signage
content.
By default, HTTP packets between digital media endpoints are set with default DSCP values and rely on
intermediate network devices to classify the traffic and provide advanced QoS techniques to protect the
digital media communication between DMP and other back end systems. Since the communication and
publish content is delivered using HTTP protocol, it becomes challenging to distinguish between HTTP
control traffic versus digital content in the campus network. Following RFC 4594 QoS deployment
guidelines, the unicast control plane communication between Cisco DMM and DMP system can be
classified as signaling traffic and must be marked an appropriate DSCP value and assigned a proper
queue. Figure 7-13 provides QoS references to deploy digital media application in a campus network:
Digital Signage QoS Reference Chart
Application
Class
PHB
Admission
Control
Congestion Managment
and
Congestion Avoidance
Cisco Video Applications
VoIP Telephony
EF
Required
Priority Queue (PQ)
Broadcast Video
CS5
Required
Optional (PQ)
Cisco DMS (Live Streams)/Enterprise TV/IPVS
Realtime Interactive
CS4
Required
Optional (PQ)
Cisco TelePresence
Multimedia Conferencing
AF4
Required
BW Queue + DSCP WRED
Cisco CUPC/CUVA/CI IP Phone 7985G
Multimedia Streaming
AF3
Recommended
BW Queue + DSCP WRED
Cisco DMS (VoDs)
Network Control
CS6
BW Queue
Call-Signaling
CS3*
BW Queue
OAM
CS2
BW Queue
Transactional Data
AF2
BW Queue + DSCP WRED
Bulk Data
AF1
BW Queue + DSCP WRED
Best Effort
DF
Best
Effort+ RED
Default
Queue
Best
Effort
Scavenger
DF
CS1
Best Effort
Min BW Queue
(Deferential)
DMM/DMP Control Traffic
Cisco WebEx/CU MeetingPlace
Effort
YouTube/Xbox Best
Live/iTunes/BitTorent/etc.
227860
Figure 7-13
*Cisco classification method slightly differs from RFC 4594. CS3 and CS5 definition are interchanged.
Applying Ingress QoS Policy on Access Layer
Network QoS policies must be set at the campus access edge to mark recommended DSCP bit for control
or management traffic between DMM and DMP. HTML or Flash based digital signage content can
remain in same best-effort class. The control traffic can be identified from digital signage content based
on TCP flow between Cisco DMM and DMP. Figure 7-14 provides QoS marking guidelines between
Cisco DMP, Cisco DMM appliance, and content server:
Small Enterprise Design Profile Reference Guide
7-21
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-14
QoS Marking Between Digital Signage Components
DMP
HDTV
DSCP=CS3
DSCP=DF
DSCP=DF
Internet
Web Server
DMM-DMP Control Traffic
DMP-Content Server Traffic
227861
DMM
Based on TCP and static Cisco DMM and DMP player information, the following configuration
guideline must be implement QoS policy on access layer switches that connect to Cisco DMP and Cisco
DMM in a centralized serverfarm:
! Classify DMP and DMM HTTP traffic with extended ACL
ip access-list extended DMS-SIGNALING
remark DMM-DMP-MGMT
permit tcp host <DMP-IP-Address> host <DMM-IP-Address>
permit tcp host <DMM-IP-Address> host <DMP-IP-Address>
!
class-map DMS-SIGNALING
match access-group name DMS-SIGNALING
!
policy-map DMS-Policy
class DMS-SIGNALING
set dscp cs3 ! Explicit mark DSCP CS3
!
interface FastEthernet0/7
description CONNECTED TO LOBBY DMP
mls qos trust dscp
service-policy input DMS-Policy!Apply ingress service-policy
!
interface FastEthernet0/10
description CONNECTED TO Cisco DMM Appliance
mls qos trust dscp
service-policy input DMS-Policy!Apply ingress service-policy
cr24-3560r-DO#show mls qos interface fas0/7 | inc policy-map|dscp
Attached policy-map for Ingress: DMS-Policy
trust state: trust dscp
trust mode: trust dscp
Additional ingress QoS policies, such as policers, can be implemented on access switches if the network
administrator is concerned about securing the restricting ports to consume higher bandwidth.
Small Enterprise Design Profile Reference Guide
7-22
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Applying Egress QoS Policy on Access Layer
Ingress QoS policy helps the network to distinguish between HTTP control and digital media content
traffic within the campus backbone. Similar QoS techniques are required to provide differential services
between control and digital media content traffic exiting the port connected to Cisco DMP and DMM
appliance on the access layer switches. For global egress policy for trusted and un-trusted device, it is
recommended to share the egress bandwidth to each hardware queue and enable prioritization for the
low-latency traffic:
cr24-3560r-DO(config)#interface FastEthernet0/7
cr24-3560r-DO(config-if)# srr-queue bandwidth share 1 30 35 5 ! Enable BW share
cr24-3560r-DO(config)#priority-queue out! Enable Priority-Queueing
cr24-3560r-DO#show mls qos interface fast0/7 queueing
FastEthernet0/7
Egress Priority Queue : enabled
Shaped queue weights (absolute) : 25 0 0 0
Shared queue weights : 1 30 35 5
The port bandwidth limit : 100 (Operational Bandwidth:100.0)
The port is mapped to qset : 1
To deploy QoS in a small enterprise campus network design, refer to the “Deploying QoS in Network”
section on page 2-51.
Auto Smartport Macro Deployment
Cisco access layer switches provide a zero-touch or plug-n-play type network provisioning solution by
dynamically detecting connected endpoints and automatically applying best practices and recommended
configurations. Cisco Auto Smartport macro helps network architects to reduce the number of challenges
in implementation when deploying complex configurations. Cisco validated design comprehensively
validates multiple set of tools and technologies to solve the critical business problems. Cisco Auto
Smartport leverages validated and recommended network configuration and parameters that
dynamically provision the network without any user intervention. Implementing recommended and
validated configuration parameters helps enterprise network administrators to automatically provide
network and device security and optimize application performance.
Cisco Auto Smartport leverages several Layer-2 protocol techniques to dynamically detect the endpoint
platform that intelligently triggers the function and applies the configuration from pre-defined
recommended templates. To further increase operational efficiency, Cisco Auto Smartport removes all
dynamically applied configurations when the device is un-plugged or removed from the network. The
following is the list of Layer 2 technologies and endpoint types that Cisco Auto Smartport macros use
to dynamically apply configurations:
•
Layer 2 Technologies
– Cisco Discovery Protocol (CDP)
– IEEE AB - LLDP
– 802.1x
– MAC-Authentication Bypass (MAB)
– Layer 2 Source MAC address
– Ethernet OUI
•
Supported Endpoint Platforms
– IP Phones—Cisco and Avaya IP Phone
Small Enterprise Design Profile Reference Guide
7-23
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
– Wireless Access-Points—Cisco AP 11xx series
– IP Video Surveillance—Cisco IPVS 25xx and 4xxx series camera
– Digital Media Player—Cisco DMP 4x00 series players
This section focuses on providing guidelines for the basic configuration needed to enable Cisco Auto
Smartport to dynamically provision Cisco DMP in the network.
Cisco Auto Smartport macro leverages a simple shell function to execute the pre-defined configuration
template embedded in the switch for each type of supported endpoint. Cisco DMP configuration gets
executed dynamically based on the port event triggers. The following output provides the Auto
Smartport event trigger and dynamic configuration guideline when it detects Cisco DMP on the physical
port:
cr26-3750#show shell functions CISCO_DMP_AUTO_SMARTPORT
function CISCO_DMP_AUTO_SMARTPORT () {
!!Provision this configuration when Link up event is triggered
!!and Cisco DMP is detected:
if [[ $LINKUP -eq YES ]]; then
conf t
interface $INTERFACE
macro description $TRIGGER
switchport access vlan $ACCESS_VLAN
switchport mode access
switchport block unicast
mls qos trust dscp
spanning-tree portfast
switchport port-security
switchport port-security maximum 1
switchport port-security violation shutdown
spanning-tree bpduguard enable
priority-queue out
exit
end
fi
!!Remove dynamic configuration when Link Down event is triggered.
if [[ $LINKUP -eq NO ]]; then
conf t
interface $INTERFACE
no macro description
no switchport access vlan $ACCESS_VLAN
no switchport block unicast
no switchport port-security
no switchport port-security maximum 1
no switchport port-security violation shutdown
no mls qos trust dscp
no spanning-tree portfast
no spanning-tree bpduguard enable
no priority-queue out
if [[ $AUTH_ENABLED -eq NO ]]; then
no switchport mode access
fi
exit
end
fi
}
Comparing the Auto Smartport macro configuration with a recommended manual configuration, it can
be seen that the majority of the recommended manual configuration is in the macro template. Some of
the advanced QoS parameters, such as MQC-based DSCP marking, may have to be manually configured.
Small Enterprise Design Profile Reference Guide
7-24
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Implementing Auto SmartPort Macro
Enterprise network administrators must enable Cisco Auto SmartPort Macro function along with basic
network parameters on access layer switches to dynamically detect and provision the configuration for
various types of endpoints. Enabling Cisco Auto Smartport macro function globally enables all the
physical ports and provisions and un-provisions the network configuration for the Cisco DMP based on
shell triggers and functions. It also provides the flexibility to disable the Auto Smartport function on a
per-port basis where the static configuration is required. The following is the simple global configuration
to enable Cisco Auto Smartport processing for all the physical ports:
3750-X
cr26-3750(config)#macro auto global processing
cr26-3750#show macro auto interface | inc Auto
Global Auto Smart Port Status
Auto Smart Ports Enabled
As described earlier, Cisco Auto Smartport can leverage multiple Layer 2 technologies to detect the
endpoints, by default Auto SmartPort use the pre-defined MAC address-group range to dynamically
detect the Cisco DMP based on Ethernet OUI address. To deploy Cisco DMP based on Ethernet OUI,
no additional configuration is required. To enable secure access-control solution, Cisco Secure ACS and
MAB can be integrated to authenticate Cisco DMP based on registered MAC address in the ACS
database.
cr26-3750#show macro auto address-group CISCO_DMP_EVENT
MAC Address Group Configuration:
Group Name
OUI
MAC ADDRESS
-------------------------------------------------------------------------CISCO_DMP_EVENT 0023.AC
Because some of the network parameters like VLAN IDs are unique in the network, the network
administrator needs to determine the common VLAN ID to deploy for a common set of end points. For
example, when Cisco DMP is detected on Switch1, then it must dynamically detect the media player and
assign it to appropriate VLAN and execute the network configuration template. By default, when Cisco
Auto Smartport detects the Cisco DMP device on the port it configures the port in access-mode and
assigned it a default VLAN ID = 1. After applying following single global configuration, the Auto
Smartport automatically assigns all the Cisco DMP in to user-defined VLAN.
cr26-3750(config)#macro auto device media-player ACCESS_VLAN=58
cr26-3750#show macro auto device media-player | inc Device|ACCESS
Device:media-player
Configurable Parameters:ACCESS_VLAN
Defaults Parameters:ACCESS_VLAN=1
Current Parameters:ACCESS_VLAN=58
Tuning Auto SmartPort Macro
Cisco Auto Smartport is optimized in detecting the end points and provisioning the configuration for
rapid and error-free deployments. The shell function that performs the configuration provisioning and
un-provisioning task is based on physical link up and down events. In initial deployment, the endpoints
can be detected using different Layer 2 techniques and all dynamically provisioned configurations can
be saved in configuration files. Due to its nature the configuration is removed and then the same
configuration is re-applied when the link goes down temporarily for any common reason, e.g., link flap,
endpoint is power cycled, etc.
Small Enterprise Design Profile Reference Guide
7-25
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
To make configuration persistent during such a link flap, the following single global configuration can
be applied to retain the dynamic configuration during a link flap:
cr26-3750(config)#macro auto sticky
Implementing Cisco Digital Media Player
Once the recommended network edge configuration on the access layer is implemented, the Cisco DMP
is ready to be deployed. Since Cisco DMP uses an embedded Linux OS which is not accessible directly
to end users, the basic network parameters must be provisioned using Web-based Cisco DMP-Device
Manager (DM). The Cisco DMP-DM is divided into three major configuration modes:
•
Settings (Network/Browser/Storage)
•
Display
•
DMP Administration
The network administrator must configure the basic network parameters to deploy Cisco DMP in
production network; the DMM administrator from the main site can apply the global display and
management parameters to DMP without intervening network administrator for any advanced
configuration task.
This chapter provides the following deployment guidelines to successfully deploy Cisco DMP and Cisco
DMM appliance server communication in the network:
•
Assigning IP address to Cisco DMP
•
Registering Cisco DMP to Cisco DMM database
Assigning IP Address to Cisco DMP
The default IP setting on Cisco DMP is to dynamically acquire IP address and gateway information from
DHCP server. It is recommended to assign a unique static IP address to DMP as it provides flexibility
to the network administrator to provide secured DMP-DM GUI access with ACLs and the ability to
provide distinguished QoS treatment for control traffic between Cisco DMP and DMM appliance server.
Figure 7-15 is a simple IP configuration task that the network administrator must perform to change the
default IP address method from DHCP to static IP address mode:
Figure 7-15
Assigning DMP Static IP Address
Small Enterprise Design Profile Reference Guide
7-26
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Registering Cisco DMP to Cisco DMM Database
The first step to enable communication between digital signage components is to register the Cisco DMP
into the centralized Cisco DMM appliance database. Cisco DMM appliance requires basic information
from Cisco DMM to register-MAC and IP address firmware version and storage information.
Centralized DMM appliance can register large numbers of DMPs across the enterprise network and it
may become an operational challenge to manage each DMP player. DMM administrator can create a
logical DMP group along with the range of IP subnet; DMM-DSM automatically groups the registered
DMPs based on assigned IP address. The following are some of the advantages in deploying DMP group
in small enterprise network architecture:
•
Organize registered DMPs into a single logical group.
•
Instead of managing each individual DMP, the DMP group allows the DMM administrator to
manage a group of DMPs collectively.
•
Accelerate digital signage content deployment and instruction to the DMP group instead of
individual players.
•
Like content management, common display attributes to all DMPs in the DMP group.
The Cisco DMS solution provides flexibility to the DMM administrator for manual or automatic Cisco
DMP registration into the database. Depending on the number of Cisco DMP deployed across the
network, either of the registration process can be selected.
•
Manual DMP Registration—The DMM administrator must manually enter Cisco DMP MAC and
IP address information into the DMM database. For better operational management and
troubleshooting, the DMP must be assigned to an logical DMP groups.
•
Auto DMP Registration—Is a highly scalable, simplified, and error-free DMP registration solution
for large deployments. To auto-register and auto-group the DMP in user-defined groups, the range
of IP subnets or CIDR ranges must be specified to scan the Cisco DMP players in the network.
Multiple IP subnets can be configured for single DMP group. The DMM appliance transmits TCP
based unicast packet with pre-defined port number 7777 to scan Cisco DMP, which cannot be
modified by the user. Hence for successful auto-registration process, it is recommended to make
sure TCP port 7777 is not filtered anywhere in the network. The DMM admin can control the DMM
appliance to trigger on-demand or schedule scan to locate DMP in the network for auto registration.
The DMM admin must complete these three steps sequentially to successfully deploy DMM groups and
auto DMP registration in the DMM-DSM:
Step 1
Configure DMM Group and assign an IP subnet.
Small Enterprise Design Profile Reference Guide
7-27
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-16
Step 2
Configure DMP Discovery Application and assign an IP subnet.
Figure 7-17
Step 3
Configuring DMP Group Using DMM-DSM
Configuring DMP Discovery Application
Trigger the DMP discovery with manual action or schedule to discover in future.
Small Enterprise Design Profile Reference Guide
7-28
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Manual DMP Discovery Trigger
Figure 7-18
Triggering DMP Discovery Manually
Refreshing the window in few seconds will reflect the dynamically discovered Cisco DMP that gets
automatically registered and grouped as depicted in Figure 7-19. The default name of the auto-registered
DMP is the same as their MAC address; the DMM admin must change to reflect with proper name or
location name.
Figure 7-19
Dynamically Discovered DMP Discovery Triggered Manually
Scheduling DMP Discovery
Depending on the number of DMP groups, network ranges, and DMP players in the network, the DMP
discovery may take some time. In large deployments, it is recommended to schedule Cisco DMP
discovery during non-business hours. Scheduling DMP discovery in the network is identical to
scheduling the digital signage publishing time. To schedule DMP discovery using Cisco DMM-DSM:
1.
Click on Schedules ->Play in Future and select discovery month and date
2.
Select the DMP group to be discovered and click on Add an Event button.
3.
Select DMP group from Select Group Tab and click OK.
4.
Select Advanced Task from Task Type Tab and click on Select Advanced Tasks Tab.
5.
Select DMP Discovery from Types and the Action Name and click OK.
6.
Configure Start and Stop Action run time and optionally configure repeat value to dynamically
discover Cisco DMP based on schedule.
Implementing Digital Signage
Upon successfully discovering and registering the Cisco DMP in the DMM appliance database, the
DMM administrator can start preparing to implement digital signage in the network. The provision
checklist must include exact content path and schedule to display the content on individual or group
Small Enterprise Design Profile Reference Guide
7-29
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
DMP in the network. As described earlier, the content must be stored on a Web or CISF server that is
physically located on the same campus network as a Cisco DMP. Hence when creating the playlists, it
is recommended to specify the URL path where the content is stored.
Publishing digital signage requires three simple configuration step on DMM-DSM as depicted in
Figure 7-20.
Figure 7-20
Digital Signage Deployment Steps
Executing each step populates content information in DMM in the common repository, provides
flexibility to compile playlist with distributed content, and schedules the digital signage publishing time
by mapping the playlist to individual or group DMPs. The network topology in Figure 7-21 is used as a
reference point to configure each deployment step.
Small Enterprise Design Profile Reference Guide
7-30
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-21
Digital Signage Network Topology
Main Site
Admin
Lobby
HDTV
Cafe
HDTV
Serverfarm
DMS
Admin
DMM
HDTV
Conference
Room
HDTV
acme.com
Web/CISF
Server
WAN
Remote Site
Admin
Lobby
HDTV
Cafe
HDTV
HDTV
Server Room
acme.sales.com
Web/CISF
Server
Step 1
229306
Chapter 7
Adding Asset in Media Library.
Cisco DMM-DSM builds an asset of digital media content in a common Media Library database.
Figure 7-22 provides information about the variety of digital media content types supported and can be
stored in two major locations—remote Web/CISF content server or it must be uploaded to local DMM
appliance server. As described earlier, the recommended content distribution model is to keep content
distributed on remote servers that reside on the same LAN as Cisco DMPs. Cisco DMM appliance must
not be used as a content server. Each asset type must be entered one at a time in the media library. Each
asset is executed serially within the playlist; the Cisco DMM-DSM also provides flexibility to publish
digital signage content in random order.
Small Enterprise Design Profile Reference Guide
7-31
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-22
Supported Asset Types
Execute the following steps to add distributed digital signage (non-video) content asset into Media
Library:
1.
Click on Add Media Asset button.
2.
Click on Single tab.
3.
Click URL in radio button as a Source and type the exact URL to access content (e.g.,
http://creek.rosenelemetary.com/default.html). Prior to deploying, the content must be tested and
verified by applying same URL from local internet browser.
4.
Do not click on download check box. This will prevent downloading provided HTTP URL content
to the local DMM hard-drive.
5.
Select File Type from drop down window based on URL extension.
6.
Enter estimated or planned playback time for this asset.
7.
Select the media category from Category window in which this URL fits in.
8.
Optionally, provide description and content owner/developer information.
9.
Click on Save button to review the entered information.
10. Click the Close button to exit the window.
Step 2
Creating Playlist.
Playlist is a user-defined logical group that is packaged with the compiled list of distributed digital
signage assets which are being added in the shared Media Library database. For example, a playlist can
include the intranet home page of the main and remote site and corporate document catalog developed
in Adobe Flash.
Cisco DMM-DSM provides the flexibility to develop playlists in two different modes—Standalone and
Cisco Digital Media Designer (DMD). When the digital signage content is designed and developed
outside the DMM-DMD, then DMM administrator must create a standalone playlist.
Cisco Digital Media Designer (DMD) is a Java-based, powerful, drag and drop design tool to create
customized digital signage content. Cisco DMM offers pre-designed assets that can be leverage to create
personalized design. DMD also offers flexibility to customized horizontal and vertical screen display
orientation.
Small Enterprise Design Profile Reference Guide
7-32
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-23
Digital Signage Playlist Options
This section provides guideline to implement a standalone playlist. Multiple playlists can be created and
associated to an individual or group of DMPs. When designing the asset in the playlist, it is important
to remember that the content publishing time and mappings to DMP groups are applied based on
per-playlist and not on per-asset basis. The order of playlist execution is done serially based on specified
time. Figure 7-24 depicts the playlist order and duration time on per-cycle basis.
Figure 7-24
Digital Signage Display Order and Duration
In a best practice, the playlist can be created based on individual main or remote site departments that
can provide flexibility to announce the news or any other department specific content at specific time
without impacting the playlist for other DMP groups deployed in different departments. Execute the
following steps to compile the distributed digital signage asset and estimated or planned run time for
each individual asset.
Click on Create Playlist button and follow step-by-step instruction provided in Figure 7-25 to create a
complied and distributed digital signage content in playlist.
Small Enterprise Design Profile Reference Guide
7-33
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Figure 7-25
Step 3
Compiling Assets with Standalone Playlist
Scheduling Playlist.
After successfully completing the above two steps the URL of distributed digital signage content is
added in media library database and the playlist is compiled with the list of signage content that needs
to be played for DMP group. The DMM administrator can send the playback command in two different
modes—Instant Play and Future Play.
Instant Play or “Play Now” sends the playback command to selected DMP groups and all the DMPs can
start displaying digital signage content immediately. Instant Play option provides an option for
immediate signage content deployment; it can also be used to publish the newly added or updated
signage asset in an already playing playlist. When Cisco DMP receives the new and updated playlist
command from centralized DMM appliance, it aborts playing the previous playlist commands and
immediately starts the display based on new information.
Future Play or “Play in Future” gives flexibility to the DMM administrator to pre-deploy digital signage
in the DMM appliance database and schedule to play the playlist on a pre-determined month, date, or
specific time. For example, you could have next month’s café lunch menu automatically published in
the last week of the current month.
Figure 7-26 depicts tab selection to schedule the playlist in both deployments modes.
Figure 7-26
Playlist Scheduling Modes
Small Enterprise Design Profile Reference Guide
7-34
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Digital Signage Application Overview
Implementing Instant Play Mode
Execute the step-by-step procedure as depicted in Figure 7-27 to deploy digital signage instantly in the
network. Cisco DMM-DSM provides the flexibility to select single, multiple, or all the DMPs using
shift-key to instantaneously publish digital signage in large deployments.
Figure 7-27
Publishing Instant Digital Signage
Implementing Future Play Mode
Click on Play in Future tab to schedule a future digital signage content publishing time. Click on Add
an Event button from the bottom window and follow the step-by-step configuration guideline as
displayed in Figure 7-28 to schedule future signage deployment.
Figure 7-28
Scheduling Signage Deployment
Small Enterprise Design Profile Reference Guide
7-35
Chapter 7
Small Enterprise Design Profile(SEDP)—Digital Media and Video Surveillance Design
Summary
Summary
The Cisco Digital Media Suite (DMS) is a comprehensive, scalable, and network-centric solution that
offers next-generation enterprise network for enabling business communication through social video,
digital signage, and IPTV systems. The emerging video technology can help transform the organization
to learn, grow, communicate, and collaborate between employees, partners, and customers. Cisco DMS
modules help increase customer experience, communication, and training to improve productivity at
reduced cost. Cisco DMS enables Web 2.0 technologies to users to share innovative ideas and expertise
through wide range of video collaboration techniques.
Small Enterprise Design Profile Reference Guide
7-36
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement