MPLS WAN

Technology Design Guide
August 2014 Series

Table of Contents
Preface.........................................................................................................................................1
CVD Navigator..............................................................................................................................2
Use Cases................................................................................................................................... 2
Scope.......................................................................................................................................... 2
Proficiency................................................................................................................................... 2
Introduction..................................................................................................................................3
Related Reading........................................................................................................................... 3
Technology Use Cases................................................................................................................ 3
Use Case: Site-to-Site Communications Using MPLS L3VPN Services.................................. 3
Design Overview.......................................................................................................................... 4
WAN Design............................................................................................................................ 4
MPLS WAN Transport.............................................................................................................. 4
Ethernet WAN.......................................................................................................................... 4
WAN-Aggregation Designs...................................................................................................... 5
MPLS Static Design Model...................................................................................................... 5
MPLS Dynamic Design Model................................................................................................. 6
Dual MPLS Design Model........................................................................................................ 6
WAN Remote-Site Designs..................................................................................................... 7
WAN/LAN Interconnection....................................................................................................... 8
WAN Remote Sites—LAN Topology......................................................................................... 9
Layer 2 Access....................................................................................................................... 9
Distribution and Access Layer................................................................................................11
IP Multicast.............................................................................................................................13
Quality of Service.......................................................................................................................13
Deploying the WAN....................................................................................................................15
Overall WAN Architecture Design Goals..................................................................................... 15
IP Routing.............................................................................................................................. 15
LAN Access.......................................................................................................................... 15
High Availability...................................................................................................................... 15
Path Selection Preferences................................................................................................... 15
Quality of Service (QoS)........................................................................................................ 16
Design Parameters................................................................................................................ 16
Table of Contents

Deploying an MPLS WAN............................................................................................................ 17
Design Overview.........................................................................................................................17
WAN-Aggregation—MPLS CE Routers....................................................................................17
Remote Sites—MPLS CE Router Selection............................................................................. 18
Design Details....................................................................................................................... 20
Deployment Details.................................................................................................................... 23
Configuring the MPLS CE Router.......................................................................................... 23
Configuring the Remote-Site MPLS CE Router...................................................................... 35
Adding a Secondary MPLS Link on an Existing MPLS CE Router.......................................... 54
Configuring the Secondary Remote-Site Router.................................................................... 58
Deploying a WAN Remote-Site Distribution Layer....................................................................... 74
Deployment Details.....................................................................................................................74
Connecting the Single or Primary Remote-Site Router to the Distribution Layer.....................74
Connecting the Secondary Remote-Site Router to the Distribution Layer............................. 82
Deploying WAN Quality of Service..............................................................................................89
Deployment Details.................................................................................................................... 89
Configuring QoS.................................................................................................................... 89
Appendix A: Product List............................................................................................................95
Appendix B: Device Configuration Files.......................................................................................98
Appendix C: Changes.................................................................................................................99

Table of Contents

CVDs include two guide types that provide tested design details: • Technology design guides provide deployment details. This approach assures that the guides in a series are fully compatible with one another. allowing you to select solutions that solve an organization’s problems—without worrying about the technical complexity.Preface Cisco Validated Designs (CVDs) present systems that are based on common use cases or engineering priorities. To ensure the compatibility of designs in the CVD Foundation. WAN. information about validated products and software. CVD Foundation Series This CVD Foundation guide is a part of the August 2014 Series. Using the CVD Foundation simplifies system integration. and fully predictable deployment. complete system. For the most recent CVD Foundation guides. please visit the CVD Foundation web site. • Solution design guides integrate existing CVDs but also include product features and functionality across Cisco products and sometimes include information about third-party integration. the guides themselves are tested together. security. Both CVD types provide a tested starting point for Cisco partners or customers to begin designing and deploying systems. As Cisco develops a CVD Foundation series. please use the feedback form. Each series describes a lab-validated. Comments and Questions If you would like to comment on a guide or ask questions. Cisco engineers have comprehensively tested and documented each design in order to ensure faster. data center. more reliable. features. in the same network lab. you should use guides that belong to the same release. CVDs incorporate a broad set of technologies. The CVD Foundation series incorporates wired and wireless LAN. and applications that address customer needs. Preface August 2014 Series 1 . and network management technologies. and best practices for specific types of technology.

This section is a quick reference only. Scope VALIDATED DESIGN Campus Wired LAN Technology Design Guide GET VPN Technology Design Guide This guide covers the following areas of technology and products: • WAN design using Layer 3 MPLS services for central and remote sites VALIDATED DESIGN VPN WAN Technology Design Guide • Remote-site WAN redundancy options • Routing policy and control for WAN aggregation and remote sites • WAN quality of service (QoS) design and configuration For more information. Proficiency This guide is for people with the following technical proficiencies—or equivalent experience: • CCNP Routing and Switching—3 to 5 years planning. see the Introduction.CVD Navigator The CVD Navigator helps you determine the applicability of this guide by summarizing its key elements: the use cases. click the titles or visit the CVD Foundation web site. the proficiency or experience recommended. Use Cases This guide addresses the following technology use cases: • Site-to-Site Communications Using MPLS L3VPN Services— Many organizations are deploying Multiprotocol Label Switching (MPLS) WAN services in order to connect remote locations over private cloud Layer 3 VPN-based providermanaged MPLS networks. and troubleshooting local and widearea networks To view the related CVD guides. Related CVD Guides VALIDATED DESIGN For more information. see the “Use Cases” section in this guide. see the “Design Overview” section in this guide. CVD Navigator August 2014 Series 2 . the scope or breadth of the technology covered. For more details. and CVDs related to this guide. implementing. verifying.

regardless of location. Technology Use Cases For remote-site users to effectively support the business. the WAN must support the convergence of voice. The VPN WAN Technology Design Guide provides guidance and configuration for broadband or Internet transport in a both a primary or backup role.Introduction The MPLS WAN Technology Design Guide provides flexible guidance and configuration for Multiprotocol Label Switching (MPLS) transport. As organizations move into multinational or global business markets. • Support for Layer 2 or Layer 3 distribution switching designs • Support for IP multicast using Multicast VPN (mVPN) service provider-based offering • QoS for WAN traffic such as Voice over IP (VoIP) and business critical applications Introduction August 2014 Series 3 . Use Case: Site-to-Site Communications Using MPLS L3VPN Services This guide helps organizations deploy WAN services in order to connect remote locations over private cloud Layer 3 VPN-based provider managed MPLS services. Related Reading The Layer 2 WAN Technology Design Guide provides guidance and configuration for a VPLS or Metro Ethernet transport. The ability to easily scale bandwidth or to add additional sites or resilient links makes MPLS an effective WAN transport for growing organizations. Although most of the applications and services that the remote-site worker uses are centrally located. The ubiquity of carrier-provided MPLS networks makes it a required consideration for an organization building a WAN. they require a flexible network design that allows for country-specific access requirements and controls complexity. centrally managed infrastructure. video. the WAN architecture requires a flexible design. organizations require that the WAN provide sufficient performance and reliability. and data transport onto a single. This design guide enables the following network capabilities: • IP any-to-any WAN connectivity for up to 500 remote sites and one or two central hub site locations • Deployment of single or dual MPLS service providers for resiliency using single or dual routers in remote site locations • Static routing or dynamic Border Gateway Protocol (BGP) peering with the MPLS service provider for site-to-site communications. the WAN design must provide a common resource access experience to the workforce. To control operational costs. To reduce the time needed to deploy new technologies that support emerging business applications and communications.

This document shows you how to deploy the network foundation and services to enable the following: • MPLS WAN connectivity for up to 500 remote sites • Primary and secondary links to provide redundant topology options for resiliency • Wired LAN access at all remote sites WAN Design The primary focus of the design is to allow usage of the following commonly deployed WAN transports: • Multiprotocol Label Switching (MPLS) Layer 3 VPN (primary) • Multiprotocol Label Switching (MPLS) Layer 3 VPN (secondary) • Internet VPN (secondary) At a high level. and so on). The WAN leverages MPLS VPN as a primary WAN transport or as a backup WAN transport (to an alternate MPLS VPN primary). The usage of an Internet VPN transport to provide a redundant topology option for resiliency is covered in the VPN WAN Technology Design Guide. the WAN is an IP network. You can integrate this economical solution seamlessly over any existing infrastructure. the site uses network equipment scaled for high performance and redundancy. This peer-to-peer model allows enterprise subscribers to outsource routing information to service providers. or Ethernet. provides a design that enables highly available. such as IP. and optimized connectivity for multiple remote-site LANs. value-added services over a single infrastructure. The chosen architecture designates a primary WAN-aggregation site that is analogous to the hub site in a traditional hub-and-spoke design. ATM. In addition. Subscribers who need to transport IP multicast traffic can enable Multicast VPNs (MVPNs). OC-3. MPLS Layer 3 VPNs use a peer-to-peer VPN Model that leverages BGP to distribute VPN-related information. The WAN is the networking infrastructure that provides an IP-based interconnection between remote sites that are separated by large geographic distances. but they are not explicitly discussed. Introduction August 2014 Series 4 . Ethernet WAN Both of the WAN transports mentioned previously use Ethernet as a standard media type. and these transports can be easily integrated to the design. This site has direct connections to both WAN transports and high-speed connections to the selected service providers. MPLS WAN Transport Cisco IOS Software Multiprotocol Label Switching (MPLS) enables enterprises and service providers to build next-generation. which can result in significant cost savings and a reduction in operational complexity for enterprises.Design Overview This guide. secure. Frame Relay. Ethernet is becoming a dominant carrier handoff in many markets and it is relevant to include Ethernet as the primary media in the tested architectures. the MPLS WAN Technology Design Guide. Much of the discussion in this guide can also be applied to non-Ethernet media (such as T1/E1. intelligent networks that deliver a wide variety of advanced. DS-3. The primary WAN-aggregation site is coresident with the data center and usually the primary campus or LAN as well.

There are other various devices supporting WAN edge services. The various design models are contrasted in the following table. The WAN transport options include MPLS VPN used as a primary or secondary transport. When WAN edge routers are referred to in the context of the connection to a carrier or service provider.WAN-aggregation design models MPLS Static MPLS Dynamic Dual MPLS Remote sites Up to 50 Up to 100 Up to 500 WAN links Single Single Dual Edge routers Single Single Dual WAN routing protocol None (static) BGP (dynamic) BGP (dynamic) Transport 1 MPLS VPN A MPLS VPN A MPLS VPN A Transport 2 — — MPLS VPN B The characteristics of each design are discussed in the following sections. Each of the design models is shown with LAN connections into either a collapsed core/distribution layer or a dedicated WAN distribution layer. A similar method of connection and configuration is used for both. and these devices should also connect into the distribution layer. For each design model. The primary differences between the various designs are the usage of routing protocols and the overall scale of the architecture. Each MPLS carrier terminates to a dedicated WAN router with a primary goal of eliminating any single points of failure.WAN-Aggregation Designs The WAN-aggregation (hub) designs include two or more WAN edge routers. There are no functional differences between these two methods from the WAN-aggregation perspective. A single VPN hub router is used across both designs. In all of the WAN-aggregation designs. Table 1 . Introduction August 2014 Series 5 . This design guide documents multiple WAN-aggregation design models that are statically or dynamically routed with either single or dual MPLS carriers. Each transport connects to a dedicated CE router. All of the WAN edge routers connect into a distribution layer. they are typically known as customer edge (CE) routers. tasks such as IP route summarization are performed at the distribution layer. you can select several router platforms with differing levels of performance and resiliency capabilities.

MPLS Static and MPLS Dynamic design models (single MPLS carrier) Core Layer Distribution Layer Collapsed Core/ Distribution Layer MPLS CE Router MPLS CE Router MPLS Introduction MPLS 2183 Static Routing or BGP Dynamic Routing August 2014 Series 6 .MPLS Static Design Model The MPLS Static design model (Figure 1): • Supports up to 50 remote sites. • Has a single MPLS VPN carrier. Figure 1 . • Has a single MPLS VPN carrier. MPLS Dynamic Design Model The MPLS Dynamic design model (Figure 1): • Supports up to 100 remote sites. • Uses static routing with MPLS VPN carrier. • Uses BGP routing with MPLS VPN carrier.

Figure 3 .Dual MPLS Design Model The Dual MPLS design model (Figure 2): • Supports up to 500 remote sites. • Uses BGP routing with MPLS VPN carrier.Dual MPLS design model Core Layer Collapsed Core/ Distribution Layer MPLS CE Routers Distribution Layer MPLS CE Routers MPLS A BGP Dynamic Routing MPLS B MPLS A MPLS B 2184 BGP Dynamic Routing WAN Remote-Site Designs This guide documents multiple WAN remote-site designs. Figure 2 . • Is typically used with a dedicated WAN distribution layer. and they are based on various combinations of WAN transports mapped to the site specific requirements for service levels and redundancy. • Has multiple MPLS VPN carriers.WAN remote-site designs MPLS WAN Nonredundant MPLS-A MPLS-B Redundant Links & Routers MPLS-A MPLS-B 2117 MPLS Redundant Links Introduction August 2014 Series 7 .

The remote-site designs include single or dual WAN edge routers. These are always MPLS CE routers.
Most remote sites are designed with a single router WAN edge; however, certain remote-site types require a
dual router WAN edge. Dual router candidate sites include regional office or remote campus locations with large
user populations, or sites with business critical needs that justify additional redundancy to remove single points
of failure.
The overall WAN design methodology is based on a primary WAN-aggregation site design that can
accommodate all of the remote-site types that map to the various link combinations listed in the following table.
Table 2 - WAN remote-site transport options
WAN remote-site routers

WAN transports

Primary transport

Secondary transport

Single

Single

MPLS VPN A

Single

Dual

MPLS VPN A

MPLS VPN B

Dual

Dual

MPLS VPN A

MPLS VPN B

The modular nature of the network design enables you to create design elements that you can replicate
throughout the network.
Both WAN-aggregation designs and all of the WAN remote-site designs are standard building blocks in the
overall design. Replication of the individual building blocks provides an easy way to scale the network and allows
for a consistent deployment method.

WAN/LAN Interconnection
The primary role of the WAN is to interconnect primary site and remote-site LANs. The LAN discussion within
this guide is limited to how the WAN-aggregation site LAN connects to the WAN-aggregation devices and how
the remote-site LANs connect to the remote-site WAN devices. Specific details regarding the LAN components
of the design are covered in the Campus Wired LAN Technology Design Guide.
At remote sites, the LAN topology depends on the number of connected users and physical geography of the
site. Large sites may require the use of a distribution layer to support multiple access-layer switches. Other sites
may only require an access-layer switch directly connected to the WAN remote-site routers. The variants that
are tested and documented in this guide are shown in the following table.
Table 3 - WAN remote-site LAN options
WAN remote-site routers

WAN transports

LAN topology

Single

Single

Access only
Distribution/access

Single

Dual

Access only
Distribution/access

Dual

Dual

Access only
Distribution/access

Introduction

August 2014 Series

8

WAN Remote Sites—LAN Topology
For consistency and modularity, all WAN remote sites use the same VLAN assignment scheme, which is shown
in the following table. This design guide uses a convention that is relevant to any location that has a single access
switch and this model can also be easily scaled to additional access closets through the addition of a distribution
layer.
Table 4 - WAN remote-sites—VLAN assignment
VLAN

Usage

Layer 2 access

Layer 3 distribution/ access

VLAN 64

Data

Yes

VLAN 69

Voice

Yes

VLAN 99

Transit

Yes

Yes

(dual router only)

(dual router only)

VLAN 50

Router link (1)

Yes

VLAN 54

Router link (2)

Yes
(dual router only)

Layer 2 Access
WAN remote sites that do not require additional distribution layer routing devices are considered to be flat or
from a LAN perspective they are considered unrouted Layer 2 sites. All Layer 3 services are provided by the
attached WAN routers. The access switches, through the use of multiple VLANs, can support services such as
data and voice. The design shown in the following figure illustrates the standardized VLAN assignment scheme.
The benefits of this design are clear: all of the access switches can be configured identically, regardless of the
number of sites in this configuration.
Access switches and their configuration are not included in this guide. The Campus Wired LAN Technology
Design Guide provides configuration details on the various access switching platforms.
IP subnets are assigned on a per-VLAN basis. This design only allocates subnets with a 255.255.255.0 netmask
for the access layer, even if less than 254 IP addresses are required. (This model can be adjusted as necessary
to other IP address schemes.) The connection between the router and the access switch must be configured
for 802.1Q VLAN trunking with subinterfaces on the router that map to the respective VLANs on the switch. The
various router subinterfaces act as the IP default gateways for each of the IP subnet and VLAN combinations.
Figure 4 - WAN remote site—Flat Layer 2 LAN (single router)

MPLS

VLAN 64 - Data
VLAN 69 - Voice

802.1Q VLAN Trunk (64, 69)

Introduction

2118

No HSRP
Required

August 2014 Series

9

A similar LAN design can be extended to a dual-router edge as shown in the following figure. This design change
introduces some additional complexity. The first requirement is to run a routing protocol. You need to configure
Enhanced Interior Gateway Protocol (EIGRP) between the routers. For consistency with the primary site LAN, use
EIGRP process 100.
Because there are now two routers per subnet, a First Hop Redundancy Protocol (FHRP) must be implemented.
For this design, Cisco selected Hot Standby Router Protocol (HSRP) as the FHRP. HSRP is designed to allow for
transparent failover of the first-hop IP router. HSRP ensures high network availability by providing first-hop routing
redundancy for IP hosts configured with a default gateway IP address. HSRP is used in a group of routers for
selecting an active router and a standby router. When there are multiple routers on a LAN, the active router is the
router of choice for routing packets; the standby router is the router that takes over when the active router fails or
when preset conditions are met.
Figure 5 - WAN remote site—Flat Layer 2 LAN (dual router)

MPLS

MPLS

iBGP
EIGRP
VLAN99 - Transit

HSRP VLANs
Active HSRP Router

VLAN 64 - Data

802.1Q VLAN Trunk (64, 69, 99)

2119

VLAN 69 - Voice

Enhanced Object Tracking (EOT) provides a consistent methodology for various router and switching features to
conditionally modify their operation based on information objects available within other processes. The objects
that can be tracked include interface line protocol, IP route reachability, and IP service-level agreement (SLA)
reachability, as well as several others.
The IP SLA feature provides a capability for a router to generate synthetic network traffic that can be sent to a
remote responder. The responder can be a generic IP endpoint that can respond to an Internet Control Message
Protocol (ICMP) echo (ping) request, or can be a Cisco router running an IP SLA responder process, that can
respond to more complex traffic such as jitter probes. The use of IP SLA allows the router to determine endto-end reachability to a destination and also the roundtrip delay. More complex probe types can also permit the
calculation of loss and jitter along the path. IP SLA is used in tandem with EOT within this design.
In order to improve convergence times after an MPLS WAN failure, HSRP has the capability to monitor the
reachability of a next-hop IP neighbor through the use of EOT and IP SLA. This combination allows for a router
to give up its HSRP Active role if its upstream neighbor becomes unresponsive. This provides additional network
resiliency.

Introduction

August 2014 Series

10

Distribution and Access Layer Large remote sites may require a LAN environment similar to that of a small campus LAN that includes a distribution layer and access layer.1Q VLAN trunk configuration can easily accommodate an additional subinterface. The routing protocol runs between router subinterfaces assigned to the transit network. The dual router designs also warrant an additional component that is required for proper routing in certain scenarios. In these cases.WAN remote-site—IP SLA probe to verify upstream device reachability Detailed View IP SLA Probe as Tracked Object WAN IP SLA Probe WAN Interface Upstream Interface WAN WAN R1 EIGRP VLAN 99 . The primary WAN transport router then forwards the traffic out the same data interface to send it to the alternate WAN transport router. There are no hosts connected to the transit network. 69.1Q VLAN Trunk (64. 99) Active HSRP Router HSRP is configured to be active on the router with the highest priority WAN transport. the standby HSRP router associated with the lower priority (alternate) WAN transport becomes the active HSRP router. a traffic flow from a remote-site host might be sent to a destination reachable via the alternate WAN transport (for example. and in the dual router design. The appropriate method to avoid sending the traffic out the same interface is to introduce an additional link between the routers and designate the link as a transit network (VLAN 99). Introduction August 2014 Series 11 . These EtherChannel links are configured as 802. This is more effective than simply monitoring the status of the WAN interface. to provide a transit network for direct communication between the WAN routers. to support both a routed point-to-point link to allow EIGRP routing with the distribution switch.Voice 2142 802.Data VLAN 69 .1Q VLAN trunks.Transit HSRP VLANs VLAN 64 . EOT of IP SLA probes is implemented in conjunction with HSRP so that in the case of WAN transport failure. an MPLS A + MPLS B remote site communicating with an MPLS-B-only remote site). This topology works well with either a single or dual router WAN edge. This is referred to as hairpinning.Figure 6 . which then forwards the traffic to the proper destination. To implement this design. the routers should connect via EtherChannel links to the distribution switch. No additional router interfaces are required with this design modification because the 802. The IP SLA probes are sent from the MPLS CE router to the MPLS Provider Edge (PE) router in order to ensure reachability of the next hop router. and it is only used for router-router communication.

1Q Trunk (yy. No HSRP is required when the design includes a distribution layer.1Q Trunk (yy.1Q Trunk (54.Router 1 Link 802.Transit 2007 802.Figure 7 . Figure 8 .Transit August 2014 Series 12 . 99) 802. zz) VLAN 50 . with VLANs trunked to access switches.Data VLAN yy .1Q Trunk (50) 802.1Q Trunk (ww. xx) 802. zz) VLAN ww . xx) 802.Router 1 Link VLAN 54 .Voice VLAN zz .Router 2 Link VLAN 99 .1Q Trunk (ww. 99) 802.1Q Trunk (50. 99) The distribution switch handles all access-layer routing.1Q Trunk (ww.1Q Trunk (54. A full distribution and access-layer design is shown in the following figure.Data VLAN xx .WAN remote site—Distribution and access layer (dual router) WAN 802.Router 2 Link VLAN 99 .Voice No HSRP Required Introduction 2144 VLAN 50 .WAN remote site—Connection to distribution layer WAN WAN 802. xx) VLAN 50 .Router 1 Link VLAN 54 .1Q Trunk (yy. zz) 802. 99) 802.1Q Trunk (50.

The RP is a control-plane operation that should be placed in the core of the network or close to the IP Multicast sources on a pair of Layer 3 switches or routers. and management traffic to the network. jitter. In reality. The design uses an Anycast RP implementation strategy. Within the architecture. and jitter parameters. the distribution layer performs the RP function. Quality of service (QoS) enables a multitude of user services and applications to coexist on the same network. Two RPs share the load for source registration and the ability to act as hot backup routers for each other. VLANs. delay. critical data applications. The benefit of this strategy from the WAN perspective is that all IP routing devices within the WAN use an identical configuration referencing the Anycast RPs. Even if you do not require QoS for your current applications. jitter. measurable. In designs without a core layer. IP Multicast routing begins at the distribution layer if the access layer is Layer 2 and provides connectivity to the IP Multicast RP. the network affects all traffic flows and must be aware of end-user requirements and services being offered. QoS enables a multitude of user services and applications. and delaysensitive data to coexist on the same network. you can use QoS for management and network protocols to protect the network functionality and manageability under normal and congested traffic conditions. and loss. Many sum this up as just “speeds and feeds. it must manage bandwidth. QoS is an essential function of the network infrastructure devices used throughout this architecture. the local router consults another router in the network that is acting as a rendezvous point (RP) to map the receivers to active sources so that they can join their streams. In order for the network to provide predictable. delay. end hosts must join a multicast group by sending an Internet Group Management Protocol (IGMP) message to their local multicast router. and loss parameters. IP Multicast is much more efficient than multiple individual unicast streams or a broadcast stream that would propagate everywhere. and packet loss. queuing. This strategy provides load sharing and redundancy in Protocol Independent Multicast sparse mode (PIM SM) networks.” While it is true that IP networks forward traffic on a best-effort basis by default. This functionality allows for the differentiation of applications. In a traditional IP Multicast design. Quality of Service Most users perceive the network as just a transport utility mechanism to shift data from point A to point B as fast as it can. Introduction August 2014 Series 13 . this type of routing only works well for applications that adapt gracefully to variations in latency. and congestion mechanisms as part of the integrated QoS to help ensure optimal use of network resources. prioritizing. time-sensitive applications are affected by jitter. The difference is that real-time applications require packets to be delivered within specified loss. there are wired and wireless connectivity options that provide advanced classification. IP PIM SM is enabled on all interfaces including loopbacks. either during the initial deployment or later with minimum system impact and engineering effort.IP Multicast IP Multicast allows a single IP data stream to be replicated by the infrastructure (routers and switches) and sent from a single source to multiple receivers. This design is fully enabled for a single global scope deployment of IP Multicast. and sometimes guaranteed services. including real-time voice. ensuring that each has the appropriate share of the network resources to protect the user experience and ensure the consistent operations of business critical applications. IP telephony Music On Hold (MOH) and IP video broadcast streaming are two examples of IP Multicast applications. However networks are multiservice by design and support real-time voice and video as well as data traffic. The goal of this design is to provide sufficient classes of service to allow you to add voice. interactive video. To receive a particular IP Multicast data stream. Even with unlimited bandwidth. delay. and subinterfaces. high-quality video.

QoS service class mappings Introduction Service class Per-hop behavior (PHB) Differentiated services code point (DSCP) IP precedence (IPP) Class of service (CoS) Network layer Layer 3 Layer 3 Layer 3 Layer 2 Network control CS6 48 6 6 Telephony EF 46 5 5 Signaling CS3 24 3 3 Multimedia conferencing AF41. This table is included as a reference. 30 3 3 Broadcast video CS5 40 4 4 Low-latency data AF21. 14 1 1 Scavenger CS1 8 1 1 Default “best effort” DF 0 0 0 August 2014 Series 14 . administration. 28. 42. 33 26. 12.The QoS classifications in the following table are applied throughout this design. 38 4 4 Real-time interactive CS4 32 4 4 Multimedia streaming AF31. 12. 23 18. 32. and maintenance (OAM) CS2 16 2 2 Bulk data AF11. 36. 13 10. 22. Table 5 . 43 34. 20. 22 2 2 Operation.

In the centralized Internet model. It is worth noting that sites with Internet/DMVPN for backup transport could potentially provide local Internet capability. a default route is advertised to the WAN remote sites in addition to the internal routes from the data center and campus. The use of the dual WAN transports is specifically tuned to behave in an active/standby manner. dual-link must be able to tolerate the loss of either an edge router or a WAN transport. The single WAN transport routing functions as follows. Path Selection Preferences There are many potential traffic flows based on which WAN transports are in use and whether or not a remote site is using a dual WAN transport. Deploying the WAN August 2014 Series 15 . for ease of troubleshooting and to prevent oversubscription of IP telephony Call Admission Control (CAC) limits • Provide site-site remote routing via the primary WAN-aggregation site (hub-and-spoke model) • Permit optimal direct site-site remote routing when carrier services allow (spoke-to-spoke model) • Support IP Multicast sourced from the primary WAN-aggregation site At the WAN remote sites. with traffic flowing along the same path in both directions. Symmetric routing simplifies troubleshooting because bidirectional traffic flows always traverse the same links. This type of configuration provides symmetric routing. for this design. This model is referred to as a centralized Internet model. MPLS VPN-connected site: • Connects to a site on the same MPLS VPN—The optimal route is direct within the MPLS VPN (traffic is not sent to the primary site). LAN Access All remote sites are to support both wired LAN access. High Availability The network must tolerate single failure conditions including the failure of any single WAN transport link or any single network device at the primary WAN-aggregation site. dual-link must be able tolerate the loss of either WAN transport.Deploying the WAN Overall WAN Architecture Design Goals IP Routing The design has the following IP routing goals: • Provide optimal routing connectivity from primary WAN-aggregation sites to all remote locations • Isolate WAN routing topology changes from other portions of the network • Ensure active/standby symmetric routing when multiple paths exist. • Connects to any other site—The route is through the primary site. there is no local Internet access for web browsing or cloud services. • Remote sites classified as single-router. • Remote sites classified as dual-router. only encrypted traffic to other DMVPN sites is permitted to use the Internet link. however.

• Connects to any other site—The route is through the primary site. Quality of Service (QoS) The network must ensure that business applications perform across the WAN during times of network congestion.Universal design parameters Network service IP address Domain name cisco.local Active Directory. These parameters are listed in the following table. DHCP server 10. DNS server. Table 6 . Design Parameters This design guide uses certain standard design parameters and references various network infrastructure services that are not located within the WAN.10 Cisco Secure Access Control System (ACS) 10. which is the preferred path in most conditions. the WAN edge QoS classification and treatment must align to the service provider offering to ensure consistent end-toend QoS treatment of traffic.The design assumes that one of the MPLS VPN WAN transports is designated as the primary transport.48.17 Deploying the WAN August 2014 Series 16 .4. MPLS VPN primary + MPLS VPN secondary dual-connected site: • Connects to a site on the same MPLS VPN—The optimal route is direct within the MPLS VPN (traffic is not sent to the primary site).4.48. When the WAN design uses a service provider offering with QoS.4.48.15 Network Time Protocol (NTP) server 10. Traffic must be classified and queued and the WAN connection must be shaped to operate within the capabilities of the connection.

The choice of whether to implement a single router or dual router is determined by the number of carriers that are required in order to provide connections to all of the remote sites. You should consider the following: the forwarding performance of the router using an Ethernet WAN deployment with broad services enabled. 20 Gbps. The amount of bandwidth required at the WAN-aggregation site determines which model of router to use. Cisco ASR 1000 Series Aggregation Services Routers represent the next-generation.to 16-mpps (millions of packets per second) packet-forwarding capabilities.WAN aggregation—MPLS CE router options Service Cisco 4451X ASR 1001 ASR 1002 ASR 1002-X Ethernet WAN with services 300 Mbps 500 Mbps 750 Mbps 1Gbps Software Redundancy Option None Yes Yes Yes Redundant power supply Option Default Default Default Supported Design Models All All All All Suggested Design Model MPLS Static MPLS Dynamic Dual MPLS Dual MPLS Suggested Number of Remote Sites 25 100 250 250+ Deploying an MPLS WAN August 2014 Series 17 . and the routers have all the elements of a true carrier-class routing product that serves both enterprise and service-provider networks. The Cisco ASR 1000 Series is fully modular from both hardware and software perspectives. and the number of remote sites. Table 7 .5 Gbps embedded service processor • Cisco 4451X Integrated Services Router All of the design models can be constructed using any of the MPLS CE routers listed in Table 7. 2. with the flexibility to support a wide range of 3.to 40-Gbps system bandwidth performance. This design uses the following routers as MPLS CE routers: • Cisco ASR 1002-X router configured with an embedded service processor (ESP) default bandwidth of 5 Gbps upgradable with software licensing options to 10 Gbps. and scaling. They are specifically designed for WAN aggregation. servicesintegrated Cisco routing platform. modular.5.Deploying an MPLS WAN Design Overview WAN-Aggregation—MPLS CE Routers The MPLS WAN designs are intended to support up to 500 remote sites with a combined aggregate WAN bandwidth of up to 1. and 36 Gbps • Cisco ASR 1002 router configured with an embedded service processor 5 (ESP5) • Cisco ASR 1001 router fixed configuration with a 2.0 Gbps. the router’s alignment with the suggested design model. The most critical devices are the WAN routers that are responsible for reliable IP forwarding and QoS.

and key to the initial deployment. Not all service modules are supported in Cisco 4451-X ISR. The Cisco 891 Integrated Services Router is recommended for use at single-router. 4. You can use this design with the CE router connected directly to the access layer. The 1941 is recommended for use at single-router. Cisco tested multiple integrated service router models as MPLS CE routers. The IP routing is straightforward and can be handled entirely by using static routes at the WAN-aggregation site and static default routes at the remote site. Dynamic routing makes it easy to add or modify IP networks at the remote site because any changes are immediately propagated to the rest of the network. single-link remote sites. MPLS VPN-connected sites require static routing in order to be handled by the carrier. and any changes or modifications require a change request to the carrier. enough module slots. Among those. Add the EHWIC-1GE-SFP-CU to the Cisco 2900 and 3900 Series Integrated Services Routers in order to provide the additional WAN-facing interface. However.WAN remote-site Cisco Integrated Services Router options 8911 19412 2911 2921 2951 3925 3945 4451-X 8 Mbps 25 Mbps 35 Mbps 50 Mbps 75 Mbps 100 Mbps 150 Mbps 1 Gbps On-board FE ports 1 (and 8-port LAN switch) 0 0 0 0 0 0 0 On-board GE ports4 1 2 3 3 3 3 3 4 Service module slots5 0 0 1 1 2 2 4 2 Redundant power supply option No No No No No Yes Yes Yes Ethernet WAN with services3 Notes: 1. single-link remote sites. Deploying an MPLS WAN August 2014 Series 18 . or you can use it to support a more complex LAN topology by connecting the CE router directly to a distribution layer. You also need to make sure that you have enough interfaces. 2. The single link MPLS WAN remote site is the most basic of building blocks for any remote location. there is significant value to configuring this type of site with dynamic routing and this approach is used for the MPLS Dynamic and Dual MPLS designs. dual-link remote-site requires four router interfaces when using a port-channel to connect to an access or distribution layer. 5. Some service modules are double-wide. Table 8 . and a properly licensed Cisco IOS Software image that supports the set of features that is required by the topology.Remote Sites—MPLS CE Router Selection The actual WAN remote-site routing platforms remain unspecified because the specification is tied closely to the bandwidth required for a location and the potential requirement for the use of service module slots. A single-router. There are many factors to consider in the selection of the WAN remote-site routers. The ability to implement this solution with a variety of potential router choices is one of the benefits of a modular design approach. and the expected performance is shown in the following table. 3. is the ability to process the expected amount and type of traffic. The MPLS CE routers at the WAN remote sites connect in the same manner as the MPLS CE routers at the WAN-aggregation site. The performance numbers are conservative numbers obtained when the router is passing IMIX traffic with heavy services configured and the CPU utilization is under 75 percent.

Deploying an MPLS WAN August 2014 Series 19 . The router can automatically detect failure of the primary link and reroute traffic to the secondary path.MPLS WAN dual-carrier remote site (dual-link options) MPLS VPN B MPLS VPN A Dual MPLS Design Model Only MPLS VPN B 2125 MPLS VPN A The dual-router.MPLS WAN remote site (single-router. By adding an additional link. you provide the first level of high availability for the remote site. single-link) MPLS VPN MPLS VPN Static Routing Dynamic Routing MPLS Dynamic MPLS Static 2124 Static Routing You can augment the basic single-link design by adding an alternate WAN transport that uses a secondary MPLS carrier and either connects on the same router or on an additional router. Figure 10 . dual-link design continues to improve upon the level of high availability for the site. The routing protocols are tuned to ensure the proper path selection. It is mandatory to run dynamic routing when there are multiple paths and the Dual MPLS or MPLS Dynamic design models are used. This design can tolerate the loss of the primary router because the secondary router reroutes traffic via the alternate path.The smaller scale MPLS Static design uses static routing and relies on the carrier to configure the additional required static routes on the PE routers. Tech Tip We recommend that you select the Dual MPLS or MPLS Dynamic designs if you intend to use resilient WAN links or want to be able to modify your routing configuration without carrier involvement. Figure 9 .

A static routing option is also included to support smaller scale requirements that do not require a dynamic routing protocol. Figure 11 . All devices use EtherChannel connections consisting of two port bundles. Other media types are commonly used (such as T1/E1). we decided to limit the focus of this design guide. The PE and CE routers are considered IP neighbors across this link. Tech Tip EIGRP and Open Shortest Path First (OSPF) Protocol are also effective as PE-CE routing protocols. Sites with only a single WAN transport (a single-homed site) do not require dynamic PE-CE routing. Documentation of additional variants is available in other guides. Cisco did not test the PE routers. you need to install and configure MPLS CE routers at every location. you typically need to use a routing protocol. Deploying an MPLS WAN August 2014 Series 20 .MPLS VPN (PE-CE connections) MPLS Carrier PE Direct Adjacencies Only Between CE and PE Routers PE CE Direct Adjacencies Only Between CE and PE Routers 2126 CE Both the PE and CE routers are required to have sufficient IP-routing information in order to provide end-toend reachability. and at every MPLS WAN-connected remote site. and these technologies are reliable and well understood. MPLS VPNs require a link between a PE router and a CE router. For an MPLS VPN WAN deployment. The Dual MPLS and MPLS Dynamic designs use dynamic PE-CE routing with BGP. media type. The PE routers propagate the routing information within the carrier network and in turn re-advertise the routes back to other CE routers. This propagation of routing information is known as dynamic PE-CE routing and it is essential when any sites have multiple WAN transports (often referred to as dual-homed or multi-homed). Due to the multiplicity of potential choices for transport. Static routing is used in the MPLS Static design model. CE routers are only able to communicate with other CE routers across the WAN via intermediate PE routers. This design provides both resiliency and additional forwarding performance. You can accomplish additional forwarding performance by increasing the number of physical links within an EtherChannel. but may not be universally available across all MPLS VPN carriers. and interface type. and their configurations are not included in this guide. The various CE routers advertise their routes to the PE routers.Design Details All WAN-aggregation MPLS CE routers connect to the same resilient switching device in the distribution layer. BGP is most commonly used for this purpose. and can rely on static routing because there is only a single path to any destination. WAN transport via Ethernet is the only media type tested and included in the configuration section. This design recommends dynamic PE-CE routing to provide consistency with configurations across both single-homed and dual-homed sites. To maintain this routing information. This also allows for easy transition from a single-homed to a dual-homed remote-site design by adding an additional link to an existing remote site. including the WAN-aggregation site.

you can reduce the amount of bandwidth. Figure 12 . processor. the number of IP prefixes or routes in the routing tables grows as well. and can scale to large networks. The IP routing details for the single and dual MPLS carrier WAN-aggregation topology with dynamic routing are shown in the following figure. and reduce convergence time associated with a link failure. Multiple routing protocols (EIGRP and BGP) are used to exchange routing information.MPLS Static Design—MPLS CE routing detail WAN Distribution EIGRP Static Routing MPLS CE Router MPLS 2128 Static Routing EIGRP Cisco chose EIGRP as the primary routing protocol because it is easy to configure. By performing IP summarization.At the WAN-aggregation site. has flexible summarization and filtering. and the routing protocol configurations are tuned from their default settings to influence traffic flows to their desired behavior. Deploying an MPLS WAN August 2014 Series 21 . and memory necessary to carry large route tables. Figure 13 .Dual MPLS and MPLS Dynamic designs—MPLS CE routing detail WAN Distribution WAN Distribution EIGRP EIGRP MPLS CE Routers iBGP eBGP eBGP MPLS A MPLS B eBGP MPLS Dual MPLS MPLS Dynamic 2127 MPLS CE Routers EIGRP The IP routing details for the single MPLS carrier WAN-aggregation topology with static routing are shown in the following figure. such as distribution layer links to the wide area or to a core. As networks grow. does not require a large amount of planning. an MPLS CE router must be connected both to the distribution layer and to its respective MPLS carrier. You should program IP summarization on links where logical boundaries exist.

In this design.With the advances in EIGRP. In this role. EIGRP Wide Metric support is on by default and backward compatible with existing routes. A dual-carrier MPLS design requires an iBGP connection between the CE routers to properly retain routing information for the remote sites. Named mode EIGRP includes features such as wide metrics. The EIGRP LAN process is configured at the WAN-aggregation site to connect to the primary site LAN distribution layer and at WAN remote sites with dual WAN routers or with distribution-layer LAN topologies. BGP Cisco chose BGP as the routing protocol for PE and CE routers to connect to the MPLS VPNs because it is consistently supported across virtually all MPLS carriers. this guide uses EIGRP named mode. In this design. For added security. supporting larger multi-gigabit links. Deploying an MPLS WAN August 2014 Series 22 . we use a private ASN (65511) as designated by the Internet Assigned Numbers Authority (IANA). The private ASN range is 64512 to 65534. The use of named mode EIGRP allows related EIGRP configurations to be centrally located in the configuration. you must select an Autonomous System Number (ASN). EIGRP neighbor authentication has been implemented to prevent unauthorized neighbor associations. BGP scales well and you can use it to advertise IP aggregate addresses for remote sites. the primary EIGRP process (AS 100) is referred to as EIGRP LAN and uses EIGRP named configuration. To use BGP. Tech Tip With EIGRP named mode configuration. BGP is straightforward to configure and requires little or no maintenance.

Configure BGP Deploying an MPLS WAN August 2014 Series 23 .242/32 10.32.10.4.32. Configure the WAN aggregation platform 3. Configure connectivity to the LAN 4. Connect to MPLS PE router 5.4.255.6/30 PROCESS Configuring the MPLS CE Router 1.Deployment Details How to Read Commands This guide uses the following conventions for commands that you enter at the command-line interface (CLI).204. The actual settings and values that you use are determined by your current network configuration. Table 9 .5 255. Redistribute WAN routes into EIGRP 6.48.5. Enter them as one command: police rate 10000 pps burst 10000 packets conform-action Commands that specify a value for a variable: ntp server 10.Parameters used in the deployment examples Hostname Loopback IP Address Port Channel IP Address CE-ASR1002-1 10.32.241/32 10.255.17 Commands with variables that you must define: class-map [highest class name] Noteworthy parts of system output (or of device configuration files) are highlighted: interface Vlan64 ip address 10. Configure the distribution switch 2.2/30 CE-ASR1001-2 10.4.0 The procedures in this section provide examples for some settings. Commands at a CLI or script prompt: Router# enable Commands to enter at a CLI prompt: configure terminal Long commands that line wrap are underlined.32.4.

255.252 ip pim sparse-mode logging event link-status carrier-delay msec 0 load-interval 30 no shutdown Step 2:  Configure EtherChannel member interfaces. Only the procedures required to support the integration of the WAN-aggregation router into the deployment are included. A Layer 3 portchannel interface connects to the distribution switch to the WAN-aggregation router and the internal routing protocol peers across this interface. Not all router platforms can support Link Aggregation Control Protocol (LACP) to negotiate with the switch. GigabitEthernet2/0/1 no switchport carrier-delay msec 0 Deploying an MPLS WAN August 2014 Series 24 . interface GigabitEthernet1/0/1 description CE-ASR1002-1 Gig0/0/0 ! interface GigabitEthernet2/0/1 description CE-ASR1002-1 Gig0/0/1 ! interface range GigabitEthernet1/0/1. Step 1:  Configure the Layer 3 port-channel interface and assign the IP address. Tech Tip As a best practice. interface Port-channel1 description CE-ASR1002-1 no switchport ip address 10.4. Configure the physical interfaces to tie to the logical port-channel by using the channel-group command.1 255. Also.Procedure 1 Configure the distribution switch Reader Tip This process assumes that the distribution switch has already been configured following the guidance in the Campus Wired LAN Technology Design Guide.32. The number for the port-channel and channel-group must match. use the same channel numbering on both sides of the link where possible. apply the egress QoS macro that was defined in the platform configuration procedure to ensure traffic is prioritized appropriately.255. The LAN distribution switch is the path to the organization’s main campus and data center. so you configure EtherChannel statically.

0.0 summary-address 10. enter the following configuration.248.4.0 exit-af-interface exit-address-family Deploying an MPLS WAN August 2014 Series 25 .255.0 255. Tech Tip It is a best practice to summarize IP routes from the WAN distribution layer towards the core. configure the layer 3 EIGRP interfaces connected to the LAN core to summarize the WAN network range.0 summary-address 10.4.128.255.160.255.0 255.channel-group 1 mode on logging event link-status logging event trunk-status logging event bundle-status no shutdown load-interval 30 macro apply EgressQoS Step 3:  Allow the routing protocol to form neighbor relationships across the port channel interface.0 255.32.4.5. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface Port-channel38 summary-address 10.240. router eigrp LAN address-family ipv4 unicast autonomous-system 100 no eigrp stub exit-address-family Step 5:  On the distribution layer switch.255.252. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface Port-channel1 no passive-interface authentication mode md5 authentication key-chain LAN-KEY exit-af-interface exit-address-family Step 4:  If it is necessary to disable EIGRP stub routing on the WAN distribution switch.0.0 255.0 summary-address 10.

A centralized authentication. username admin password c1sco123 enable secret c1sco123 service password-encryption aaa new-model Step 3:  By default. and accounting (AAA) service reduces operational tasks per device and provides an audit log of user access for security compliance and root cause analysis.255. This makes it easy to identify the device. there are features and services that are common across all WAN aggregation routers. you prevent the disclosure of plain text passwords when viewing configuration files.0. Procedure 2 Configure the WAN aggregation platform Within this design. Step 1:  Configure the device host name. By enabling password encryption. which provides only limited operational privileges. Step 4:  (Optional) Configure centralized user authentication.0 exit-af-interface exit-address-family Repeat this step as needed for additional WAN aggregation routers. The enable password secures access to the device configuration mode. Tech Tip It is a best practice to summarize IP routes from the WAN distribution layer towards the MPLS WAN. all management access to the network infrastructure devices (SSH and HTTPS) is controlled by AAA.0.0 255. The local login account and password provides basic access authentication to a router.5. These are system settings that simplify and secure the management of the solution. Deploying an MPLS WAN August 2014 Series 26 . As networks scale in the number of devices to maintain it poses an operational burden to maintain local user accounts on every device. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface Port-channel1 summary-address 10. hostname CE-ASR1002-1 Step 2:  Configure local login and password. authorization. configure the Layer 3 EIGRP interface connected to the WAN aggregation routers to summarize the WAN remote-site network range. HTTPS access to the router uses the enable password for authentication.Step 6:  On the distribution layer switch. When AAA is enabled for access control.

They use Secure Sockets Layer (SSL) and Transport Layer Security (TLS) to provide device authentication and data encryption. Both protocols are encrypted for privacy and the unsecure protocols. tacacs server TACACS-SERVER-1 address ipv4 10. Without this command.TACACS+ is the primary protocol used to authenticate management logins on the infrastructure devices to the AAA server. Secure HTTP (HTTPS) and Secure Shell (SSH) are secure replacements for the HTTP and Telnet protocols.4. With this command.15 key SecretKey ! aaa group server tacacs+ TACACS-SERVERS server name TACACS-SERVER-1 ! aaa authentication login default group TACACS-SERVERS local aaa authorization exec default group TACACS-SERVERS local aaa authorization console ip http authentication aaa Step 5:  Configure device management protocols. A local AAA user database is also defined in Step 2 on each network infrastructure device to provide a fallback authentication source in case the centralized TACACS+ server is unavailable. ip domain-name cisco. console log messages are displayed on the console after interactive CLI output is displayed or printed. are turned off. which allows the use of code upgrades using Prime Infrastructure via SSH-based SCP protocol.48. Telnet and HTTP. long timeout delays may occur for mistyped commands. When synchronous logging of unsolicited messages and debug output is turned on. Secure Copy Protocol is enabled. if the ip name-server is unreachable. you can continue typing at the device console when debugging is enabled.local ip ssh version 2 no ip http server ip http secure-server ip scp server enable line vty 0 15 transport input ssh transport preferred none Step 6:  Enable synchronous logging. Specify the transport preferred none on vty lines to prevent errant connection attempts from the CLI prompt. Secure management of the network device is enabled through the use of the SSH and HTTPS protocols. line con 0 transport preferred none logging synchronous Deploying an MPLS WAN August 2014 Series 27 .

255 ip pim sparse-mode Deploying an MPLS WAN August 2014 Series 28 . Allocate the loopback address from the IP address block that the router summarizes to the rest of the network.48. access-list 55 permit 10. snmp-server community cisco RO snmp-server community cisco123 RW Step 8:  If operational support is centralized in your network. ntp server 10. In this example. the loopback address is the best way to manage the router in-band.255. such as a radio clock or an atomic clock attached to a time server.0 0. Step 9:  Configure a synchronized clock.0/24 network will be able to access the device via SSH or SNMP. only devices on the 10. Layer 3 process and features are also bound to the loopback interface to ensure process resiliency. The loopback address is commonly a host address with a 32-bit address mask. interface Loopback 0 ip address 10.4.0. SNMPv2c is configured both for a read-only and a read-write community string.48.4.48.255 line vty 0 15 access-class 55 in ! snmp-server community cisco RO 55 snmp-server community cisco123 RW 55 Tech Tip If you configure an access-list on the vty interface you may lose the ability to use ssh to login from one router to the next for hop-by-hop troubleshooting. NTP then distributes this time across the organizations network.241 255. This allows the network infrastructure devices to be managed by a Network Management System (NMS). The loopback interface is a logical interface that is always reachable as long as the device is powered on and any IP interface is reachable to the network.Step 7:  Enable Simple Network Management Protocol (SNMP). you can cross-reference events in a network. An NTP network usually gets its time from an authoritative time source.32.4.255. You should program network devices to synchronize to a local NTP server in the network.4. you can increase network security by using an access list to limit the networks that can access your device. By configuring console messages.0.17 ! clock timezone PST -8 clock summer-time PDT recurring ! service timestamps debug datetime msec localtime service timestamps log datetime msec localtime Step 10:  Configure an in-band management interface. The local NTP server typically references a more accurate clock feed from an outside source. The Network Time Protocol (NTP) is designed to synchronize a network of devices. logs. and debug output to provide time stamps on output. Because of this capability.

Step 11:  Bind the device processes for SNMP. Enable IP Multicast routing on the platforms in the global configuration mode. either in a single network statement or in multiple network statements. end hosts must join a multicast group by sending an IGMP message to their local multicast router. the port-channel interface and the loopback must be EIGRP interfaces. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface default passive-interface exit-af-interface network 10. TACACS+ and NTP to the loopback interface address. IP Telephony MOH and IP Video Broadcast Streaming are two examples of IP Multicast applications.241 nsf exit-address-family Step 14:  Configure IP Multicast routing. IP Multicast allows a single IP data stream to be replicated by the infrastructure (routers and switches) and sent from a single source to multiple receivers. This provides optimal resiliency: snmp-server trap-source Loopback0 ip ssh source-interface Loopback0 ip pim register-source Loopback0 ip tacacs source-interface Loopback0 ntp source Loopback0 Step 12:  Configure IP unicast routing authentication key to be used for EIGRP neighbor authentication.255 eigrp router-id 10. PIM. In a traditional IP Multicast design. The network range must include both interface IP addresses. In this design. This design. SSH.4. The loopback may remain a passive interface.255. ip multicast-routing distributed Deploying an MPLS WAN August 2014 Series 29 . the distributed keyword is required. Using IP Multicast is much more efficient than using multiple individual unicast streams or a broadcast stream that would propagate everywhere. In order to receive a particular IP Multicast data stream.1.32. This design uses a best practice of assigning the router ID to a loopback address. uses Auto RP for a simple yet scalable way to provide a highly resilient RP environment.0 0. ip multicast-routing If you are using a Cisco ASR 1000 Series router. EIGRP is configured facing the LAN distribution or core layer.0. the local router consults another router in the network that is acting as an RP to map the receivers to active sources so they can join their streams. which is based on sparse mode multicast operation.4. key chain LAN-KEY key 1 key-string cisco Step 13:  Configure IP unicast routing using EIGRP named mode.

255. interface GigabitEthernet0/0/0 description WAN-D3750X Gig1/0/1 ! interface GigabitEthernet0/0/1 description WAN-D3750X Gig2/0/1 ! interface range GigabitEthernet0/0/0.252 ip pim sparse-mode no shutdown Step 2:  Configure EtherChannel member interfaces. Step 1:  Configure Layer 3 interface. In this step.255. The number for the port-channel and channel-group must match. Use the ip pim autorp listener command to allow for discovery across sparse mode links.Step 15:  Configure every Layer 3 switch and router to discover the IP Multicast RP with autorp. This configuration provides for future scaling and control of the IP Multicast environment and can change based on network needs and design. Configure the physical interfaces to tie to the logical port-channel by using the channel-group command. Not all router platforms can support LACP to negotiate with the switch. GigabitEthernet0/0/1 no ip address channel-group 1 no shutdown Step 3:  Configure the EIGRP interface. Allow EIGRP to form neighbor relationships across the interface to establish peering adjacencies and exchange route tables. so you configure EtherChannel statically.2 255. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface Port-channel1 no passive-interface authentication mode md5 authentication key-chain LAN-KEY exit-af-interface exit-address-family Deploying an MPLS WAN August 2014 Series 30 .4. ip pim sparse-mode Procedure 3 Configure connectivity to the LAN Any links to adjacent distribution layers should be Layer 3 links or Layer 3 EtherChannels.32. ip pim autorp listener Step 16:  Enable sparse mode multicast operation for all Layer 3 interfaces in the network. configure EIGRP authentication by using the authentication key specified in the previous procedure. interface Port-channel1 description WAN-D3750X ip address 10.

The IP addressing used between CE and PE routers must be negotiated with your MPLS carrier. We do not recommend the use of CDP on external interfaces.255. If you have a remote-site design that includes sites with dual WAN links. The example shows a Gigabit interface (1000 Mbps) with a subrate of 300 Mbps. then use the BGP option. This is the recommended approach. or do not wish to have your MPLS carrier make changes or modifications. The MPLS carrier is responsible for configuring static IP routing within the MPLS network. use the policed rate from the carrier. Typically a pointto-point netmask of 255.255.Procedure 4 Connect to MPLS PE router Step 1:  Assign the interface bandwidth.255.255.1 255.252 Step 3:  Administratively enable the interface and disable CDP.3.252 is used.168. If your remote-site design only uses single WAN links and you don’t anticipate adding or modifying IP networks at the remote sites. Deploying an MPLS WAN August 2014 Series 31 . interface GigabitEthernet0/0/3 no cdp enable no shutdown Procedure 5 Redistribute WAN routes into EIGRP The WAN-aggregation CE routers are configured either for dynamic routing with BGP or are statically routed. interface GigabitEthernet0/0/3 description MPLS PE Router bandwidth 300000 Tech Tip Command reference: bandwidth kbps (300 Mbps = 300. Or. then you can use the statically routed option. interface GigabitEthernet0/0/3 ip address 192.000 kbps) Step 2:  Assign the IP address and netmask of the WAN interface. if you are using a subrate service. The bandwidth value should correspond to the actual interface speed.

Tech Tip If you do not use dynamic routing with BGP. Depending on the actual design of your network. To accomplish this task. Table 10 . where certain routes are repeatedly installed and withdrawn from the device routing tables. only the bandwidth and delay values are used for metric calculation. you must create a route-map that matches any routes originating from the WAN indicated by a specific route tag. An inbound distribute-list with a route-map is used to limit which routes are accepted for installation into the route table. Site-specific routing details must be shared with your MPLS carrier. you might experience route flapping. then the MPLS carrier must configure a set of static routes on its PE routers for the WAN-aggregation site and for each of the remote sites. router eigrp LAN address-family ipv4 unicast autonomous-system 100 topology base default-metric 300000 100 255 1 1500 redistribute bgp 65511 exit-af-topology exit-address-family Step 2:  Configure route-map and inbound distribute-list for EIGRP. Option 1: BGP dynamic routing with MPLS carrier Step 1:  Redistribute BGP into EIGRP. This design uses mutual route redistribution: BGP routes are distributed into EIGRP and EIGRP routes are distributed into BGP (covered in Procedure 6). This method allows for dynamic identification of the various WAN routes. It is important to tightly control how routing information is shared between different routing protocols when you use this configuration. The WAN-aggregation MPLS CE routers are configured to only accept routes that do not originate from the MPLS or DMVPN WAN sources. By default. The specific route tags in use are shown below. A default metric redistributes the BGP routes into EIGRP. BGP-learned routes are implicitly tagged with their respective source AS and other WAN routes are explicitly tagged by their WAN-aggregation router (documented in a separate procedure). you may need to block more tags.Route tag information for WAN-aggregation MPLS CE routers Tag Route source Tag method action 65401 MPLS VPN A implicit block 65402 MPLS VPN B implicit block 300 Layer 2 WAN explicit accept 65512 DMVPN hub routers explicit block This example includes all WAN route sources in the reference design. otherwise. Deploying an MPLS WAN August 2014 Series 32 . Proper route control ensures the stability of the routing table.

168. ip route 10. It is a best practice to summarize the remote-site network ranges into a single route when possible. route-map BLOCK-TAGGED-ROUTES deny 10 match tag 65401 65402 65512 ! route-map BLOCK-TAGGED-ROUTES permit 20 ! router eigrp LAN address-family ipv4 unicast autonomous-system 100 topology base distribute-list route-map BLOCK-TAGGED-ROUTES in exit-af-topology exit-address-family Option 2: Static routing with service provider Step 1:  Configure static routes to remote sites’ LANs on the WAN-aggregation CE router.5.0 192. which includes the CE routers’ WAN interfaces. route-flapping may occur.2 Step 3:  Configure routes to the remote-site router loopback addresses. ip route 10.0 255.255. so you can use this to determine router reachability.255. A default metric redistributes these routes into EIGRP.0 255. and filtering. ip route 192.0 192.5. which can cause instability. for troubleshooting.255.168. tagging. only the bandwidth and delay values are used for metric calculation. router eigrp LAN address-family ipv4 unicast autonomous-system 100 topology base default-metric 300000 100 255 1 1500 redistribute static exit-af-topology exit-address-family Deploying an MPLS WAN August 2014 Series 33 .0 255.250. It is a best practice to summarize the PE-CE link ranges into a single route when possible.168.0 192.5.168.0.255.5.5.255.2 Step 4:  Configure EIGRP to advertise the remote-site static routes. Tech Tip If you configure mutual route redistribution without proper matching.2 Step 2:  It is desirable to advertise a route for the MPLS PE-CE links. A single summary route for the loopback range may be used when possible.255.0. By default.It is important when creating the route-map that you include a permit statement at the end in order to permit the installation of routes with non-matching tags.

168. The private ASN range is 64512 to 65534.32.4. It is most efficient if you summarize these routes before they are advertised to the CE router.0 mask 255. The MPLS carrier must provide their ASN (the ASN in the previous step is the ASN identifying your site).0. It is desirable to advertise a route for the PE-CE link. All EIGRP routes learned by the CE router. You can use this to determine router reachability.Procedure 6 Configure BGP If you are using BGP dynamic routing with the MPLS carrier. but you may be permitted to use a private ASN as designated by IANA.0.2 remote-as 65401 Step 3:  Redistribute EIGRP into BGP.255.242 next-hop-self Deploying an MPLS WAN August 2014 Series 34 . router bgp 65511 neighbor 10.168.4. You can consult with your MPLS carrier on the requirements for the ASN.0 in a network statement. Because the carrier PE router uses a different ASN. should be advertised into the WAN. you must explicitly specify 0.4.255. this configuration is considered an internal BGP (iBGP) connection.0.3. You must configure BGP with the MPLS carrier PE device.0.32. The CE router advertises only network routes to the PE via BGP when: • The route is specified in network statements and is present in the local routing table.242 remote-as 65511 neighbor 10. complete this procedure. configure a BGP link between the CE routers. router bgp 65511 network 0.3. router bgp 65511 network 192. router bgp 65511 no synchronization bgp router-id 10. This design uses iBGP peering using device loopback addresses.242 update-source Loopback0 neighbor 10. including routes from the core and for other WAN sites. Because BGP does not propagate a default route via redistribution.241 bgp log-neighbor-changes no auto-summary Step 2:  Configure eBGP. • The route is redistributed into BGP.32. which requires the update-source and next-hop-self-configuration options. so you should include this network in a network statement. Because the CE routers are using the same ASN. this configuration is considered an external BGP (eBGP) connection.32. you must use a BGP ASN.4. Step 1:  Enable BGP.252 neighbor 192. for troubleshooting. To complete this step.0 redistribute eigrp 100 Step 4:  If you have dual MPLS carriers.

Configure access-layer routing 6. Enable enhanced object tracking Use this process for the configuration of any of the following: • MPLS CE router for an MPLS WAN remote site (single router. Configure WAN routing 4. Connect router to access-layer switch 5. Configure BGP 11. Connect to the MPLS PE router PROCESS 3. Configure access-layer HSRP 8. dual-link design or for configuring the first router of the dual-router. dual-link design.Configuring the Remote-Site MPLS CE Router 1. Deploying an MPLS WAN August 2014 Series 35 . Configure remote-site DHCP 7. Configure the transit network 9. Configure the WAN remote router 2. single link) • MPLS WAN Dual Carrier remote site Use the following procedures when performing the initial configuration of a dual-connected MPLS CE in the single-router. Configure EIGRP (LAN side) 10.

Configure Access Layer HSRP 3. Single Link Remote-Site MPLS CE Router Configuration Procedures Remote-Site MPLS CE Router Dual Router. Configure EIGRP (LAN Side)) 11. Configure Transit Network 3 8. Connect to Distribution Layer 2. Configure BGP for Dual Router August 2014 Series 36 .Remote-site MPLS CE router configuration flowchart Remote-Site MPLS CE Router Single Router. Configure BGP for Dual Rout 9. Figure 14 . Dual Link 1. Configure the WAN Remote Router 2. Configure Remote-Site DHCP (Optional) YES NO RP 7. Configure EIGRP (LAN Side) Dual Router Design? NO YES Dual Router Design? 5. Dual Link (1st Router) Remote-Site MPLS CE Router Single Router. Enable ble Enhanced O bject Tra acking Object Tracking MPLS CE Spoke Router Configuration Complete Deploying an MPLS WAN e Second Configure Remote-Site Router MPLS CE Router Configuration Complete Configure Second C Remote-Site Router 2129 uter 10. Configure WAN Routing Distribution Layer Design? NO YES 4. Connect to MPLS PE Router 3. Connect nnect Router to Access Layer Switch Remote-Site Router to Distribution Layer Procedures 1. Configure Access Layer Routing 6. Configure Transit Network Router 4.The following flowchart provides details about the configuration process for a remote-site MPLS CE router.

These are system settings that simplify and secure the management of the solution. As networks scale in the number of devices to maintain. By enabling password encryption. authorization.48.15 key SecretKey ! aaa group server tacacs+ TACACS-SERVERS server name TACACS-SERVER-1 ! aaa authentication login default group TACACS-SERVERS local aaa authorization exec default group TACACS-SERVERS local aaa authorization console ip http authentication aaa Step 5:  Configure device management protocols. Step 4:  (Optional) Configure centralized user authentication. there are features and services that are common across all WAN remote-site routers. https access to the router uses the enable password for authentication. Secure Copy Deploying an MPLS WAN August 2014 Series 37 . all management access to the network infrastructure devices (SSH and HTTPS) is controlled by AAA. username admin password c1sco123 enable secret c1sco123 service password-encryption aaa new-model Step 3:  By default. Secure management of the network device is enabled through the use of the SSH and HTTPS protocols. Step 1:  Configure the device host name to make it easy to identify the device.Procedure 1 Configure the WAN remote router Within this design. The enable password secures access to the device configuration mode. The local login account and password provides basic access authentication to a router that provides only limited operational privileges. tacacs server TACACS-SERVER-1 address ipv4 10. you prevent the disclosure of plain text passwords when viewing configuration files. A centralized authentication. A local AAA user database is also defined in Step 2 on each network infrastructure device in order to provide a fallback authentication source in case the centralized TACACS+ server is unavailable. Both protocols are encrypted for privacy and the unsecure protocols. and accounting (AAA) service reduces operational tasks per device and provides an audit log of user access for security compliance and root cause analysis.4. Secure HTTP (HTTPS) and Secure Shell (SSH) are secure replacements for the HTTP and Telnet protocols. They use Secure Sockets Layer (SSL) and Transport Layer Security (TLS) to provide device authentication and data encryption. Telnet and HTTP. TACACS+ is the primary protocol used to authenticate management logins on the infrastructure devices to the AAA server. it can be an operational burden to maintain local user accounts on every device. are turned off. When AAA is enabled for access control. hostname [hostname] Step 2:  Configure the local login and password.

0/24 network will be able to access the device via SSH or SNMP. if the ip name-server is unreachable. you can continue typing at the device console when debugging is enabled.0. In this example. ip domain-name cisco. only devices on the 10.0 0. Configure SNMPv2c both for a read-only and a readwrite community string.local ip ssh version 2 no ip http server ip http secure-server ip scp server enable line vty 0 15 transport input ssh transport preferred none Step 6:  Enable synchronous logging. you can increase network security by using an access list to limit the networks that can access your device.4. which allows the use of code upgrades using Prime Infrastructure via SSH-based SCP protocol.4.255 line vty 0 15 access-class 55 in ! snmp-server community cisco RO 55 snmp-server community cisco123 RW 55 Tech Tip If you configure an access-list on the vty interface you may lose the ability to use ssh to log in from one router to the next for hop-by-hop troubleshooting. snmp-server community cisco RO snmp-server community cisco123 RW Step 8:  If your network operational support is centralized. When synchronous logging of unsolicited messages and debug output is turned on. The Network Time Protocol (NTP) is designed to synchronize a network of devices. This allows the network infrastructure devices to be managed by a Network Management System (NMS). access-list 55 permit 10. such as a radio clock or an atomic clock attached to a time server. Deploying an MPLS WAN August 2014 Series 38 . Step 9:  Configure a synchronized clock. With this command. line con 0 transport preferred none logging synchronous Step 7:  Enable Simple Network Management Protocol (SNMP). Specify the transport preferred none on vty lines to prevent errant connection attempts from the CLI prompt. long timeout delays may occur for mistyped commands.48.48.0.Protocol is enabled. An NTP network usually gets its time from an authoritative time source. Without this command. console log messages are displayed on the console after interactive CLI output is displayed or printed.

Layer 3 process and features are also bound to the loopback interface to ensure process resiliency. The loopback interface is a logical interface that is always reachable as long as the device is powered on and any IP interface is reachable to the network. In a traditional IP Multicast design. PIM.48.255.255. Using IP Multicast is much more efficient than multiple individual unicast streams or a Broadcast stream that would propagate everywhere.4. interface Loopback 0 ip address [ip address] 255. logs. The loopback address is commonly a host address with a 32-bit address mask. Because of this capability. which is based on sparse mode multicast operation. the loopback address is the best way to manage the router in-band. To receive a particular IP Multicast data stream. end hosts must join a multicast group by sending an IGMP message to their local multicast router. By configuring console messages. IP Multicast allows a single IP data stream to be replicated by the infrastructure (routers and switches) and sent from a single source to multiple receivers. You should program network devices to synchronize to a local NTP server in the network. IP Telephony MOH and IP Video Broadcast Streaming are two examples of IP Multicast applications. ntp server 10. the local router consults another router in the network that is acting as an RP to map the receivers to active sources so they can join their streams. and debug output to provide time stamps on output.NTP then distributes this time across the organization’s network.17 ntp update-calendar ! clock timezone PST -8 clock summer-time PDT recurring ! service timestamps debug datetime msec localtime service timestamps log datetime msec localtime Step 10:  Configure an in-band management interface. SSH.255 ip pim sparse-mode Step 11:  Bind the device processes for SNMP. Auto RP is used to provide a simple yet scalable way to provide a highly resilient RP environment. TACACS+ and NTP to the loopback interface address for optimal resiliency: snmp-server trap-source Loopback0 ip ssh source-interface Loopback0 ip pim register-source Loopback0 ip tacacs source-interface Loopback0 ntp source Loopback0 Step 12:  Configure IP Multicast routing. The local NTP server typically references a more accurate clock feed from an outside source. ip multicast-routing Deploying an MPLS WAN August 2014 Series 39 . Enable IP Multicast routing on the platforms in the global configuration mode. Allocate the loopback address from a unique network range that is not part of any other internal network summary range. In this design. you can cross-reference events in a network.

ip pim autorp listener Step 14:  Enable sparse mode multicast operation for all Layer 3 interfaces in the network.3.252.000 kbps Step 2:  Assign the IP address and netmask of the WAN interface.255. The bandwidth value should correspond to the actual interface speed.Step 13:  Configure every Layer 3 switch and router to discover the IP Multicast RP with autorp. interface [interface type] [number] ip address [IP address] [netmask] Step 3:  Administratively enable the interface and disable CDP.255. you’d use a point-to-point netmask of 255.252 no cdp enable no shutdown Deploying an MPLS WAN August 2014 Series 40 . Or. The IP addressing used between CE and PE routers must be negotiated with your MPLS carrier. Typically.168. The example shows a Gigabit interface (1000 Mbps) with a subrate of 10 Mbps.255. Use the ip pim autorp listener command to allow for discovery across sparse mode links.255. ip pim sparse-mode Procedure 2 Connect to the MPLS PE router Step 1:  Assign the interface bandwidth.9 255. This configuration provides for future scaling and control of the IP Multicast environment and can change based on network needs and design. if you are using a subrate service. interface [interface type] [number] bandwidth [bandwidth (kbps)] Tech Tip Command Reference: bandwidth kbps 10 Mbps = 10. The use of CDP on external interfaces is not recommended. you should use the policed rate from the carrier. interface [interface type] [number] no cdp enable no shutdown Example interface GigabitEthernet0/0 bandwidth 10000 ip address 192.

Option 1: BGP dynamic routing with MPLS carrier Step 1:  Enable BGP. then the MPLS carrier must configure a set of static routes on its PE routers for the WAN-aggregation site and for each of the remote sites. Tech Tip If you do not use dynamic routing with BGP. You can use this to determine router reachability. The MPLS carrier must provide their ASN (the ASN in the previous step is the ASN identifying your site). Configure BGP with the MPLS carrier PE device. The CE router advertises only network routes to the PE via BGP in the following cases: • The route is specified in network statements and is present in the local routing table. This is the recommended approach. It is desirable to advertise a route for the PE-CE link. so you should include this network in a network statement. To complete this step. • The route is redistributed into BGP (not applicable in the remote-site use case). a BGP ASN is required. If you have a remote-site design that includes sites with dual WAN links or you do not want to have your MPLS carrier make changes or modifications. you must configure BGP to advertise the loopback network for the router. You might be able to reuse the same value used on the MPLS VPN CE from the WAN-aggregation site. Similarly. The MPLS carrier is responsible for configuring static IP routing within the MPLS network. If your remote-site design only uses single WAN links and you don’t anticipate adding or modifying IP networks at the remote site. Deploying an MPLS WAN August 2014 Series 41 . Site-specific routing details must be shared with your MPLS carrier. Because the carrier PE router uses a different ASN. then you can use the statically routed option. use the BGP option. Consult with your MPLS carrier on the requirements for the ASN. for troubleshooting. router bgp 65511 no synchronization bgp router-id [IP address of Loopback0] bgp log-neighbor-changes no auto-summary Step 2:  Configure eBGP.Procedure 3 Configure WAN routing The remote-site CE routers are configured either for dynamic routing with BGP or are statically routed. and assumes that the WAN-aggregation CE router has already been configured for BGP. this configuration is considered an external BGP (eBGP) connection.

You must advertise the remote-site LAN networks. This properly configures the static routes to the remote site.255.3.248. ip route 0.255.0.251. You must add a separate network statement for the loopback address. The IP assignment for the remote sites was designed so that all of the networks in use can be summarized within a single aggregate route.255 network [DATA network] mask [netmask] network [VOICE network] mask [netmask] aggregate-address [summary IP address] [summary netmask] summary-only neighbor [IP address of PE] remote-as [carrier ASN] Example router bgp 65511 no synchronization bgp router-id 10. Deploying an MPLS WAN August 2014 Series 42 .255.10 Step 2:  For the MPLS carrier for each remote site.3.10 remote-as 65401 no auto-summary Option 2: Static routing with service provider This option has remote sites using static routing to the MPLS WAN to forward all traffic to the WAN-aggregation site. the WAN-aggregation CE router must have a corresponding static host route for that site’s loopback address.206 mask 255.0 network 10.0 aggregate-address 10.3.5. you must list each individually.8. which offers a measure of resiliency.0 mask 255.255.8 mask 255.255.0.0 255. Step 1:  Enter a default route for traffic forwarded to the WAN-aggregation site.168.0. If any LAN network is present in the route table.255 network 10.255.12.251.255.13.168.5. the aggregate is advertised to the MPLS PE. If the various LAN networks cannot be summarized.168. provide the remote-site specific IP range and the chosen loopback IP address for the router.252 network 10.5.206 bgp log-neighbor-changes network 192.255.0.0 0.255.255.0 summary-only neighbor 192. Tech Tip For each remote site with static routing.255.255. The aggregate address configured below suppresses the more specific routes.0 mask 255.0 192.255. router bgp 65511 network [PE-CE link network] mask [PE-CE link netmask] network [Loopback network] mask 255.

In the access-layer design. so you configure EtherChannel statically. a simple Layer 2 trunk between the router and switch is used. The number for the port-channel and channel-group must match. For complete access-layer configuration details. Layer 2 EtherChannels are used to interconnect the CE router to the access layer in the most resilient method possible. Connect the router EtherChannel uplinks to separate switches in the access layer switch stack. the remote sites use collapsed routing. interface Port-channel1 description EtherChannel link to RS206-A2960X no shutdown Step 2:  Configure EtherChannel member interfaces on the router.1Q trunk interfaces to the LAN access layer. Doing the configuration in this order allows for minimal configuration and reduces errors because most of the commands entered to a port-channel interface are copied to its members’ interfaces and do not require manual replication. skip to the “Deploying a WAN Remote-Site Distribution Layer” chapter of this guide. to separate redundant modules for additional resiliency. with 802. The VLAN numbering is locally significant only. fixed-configuration switch.Procedure 4 Connect router to access-layer switch Reader Tip This guide includes only the additional steps to complete the distribution-layer configuration. If you are using a remote-site distribution layer. The physical interfaces that are members of a Layer 2 EtherChannel are configured prior to configuring the logical port-channel interface. Configure the physical interfaces to tie to the logical port-channel by using the channel-group command. Option 1: Layer 2 EtherChannel from router to access-layer switch Step 1:  Configure port-channel interface on the router. Deploying an MPLS WAN August 2014 Series 43 . Not all router platforms can support LACP to negotiate with the switch. GigabitEthernet0/2 no ip address channel-group 1 no shutdown Step 3:  Configure EtherChannel member interfaces on the access-layer switch. If your access-layer device is a single. interface GigabitEthernet0/1 description RS206-A2960X Gig1/0/24 ! interface GigabitEthernet0/2 description RS206-A2960X Gig2/0/24 ! interface range GigabitEthernet0/1. or in the case of the Cisco Catalyst 4507R+E distribution layer. see the Campus Wired LAN Technology Design Guide.

the interface type is port-channel. and the number must match the channel group configured in Step 3. interface GigabitEthernet1/0/24 description Link to RS206-3925-1 Gig0/1 interface GigabitEthernet2/0/24 description Link to RS206-3925-1 Gig0/2 ! interface range GigabitEthernet1/0/24. When using EtherChannel. Prune the VLANs allowed on the trunk to only the VLANs that are active on the access-layer switch. Set DHCP Snooping and Address Resolution Protocol (ARP) inspection to trust. Also.99 switchport mode trunk ip arp inspection trust spanning-tree portfast trunk logging event link-status logging event trunk-status ip dhcp snooping trust no shutdown load-interval 30 The Cisco Catalyst 3750 Series Switch requires the switchport trunk encapsulation dot1q command. interface GigabitEthernet0/2 description RS202-A3560X Gig1/0/24 no ip address no shutdown Deploying an MPLS WAN August 2014 Series 44 . which allows the router to provide the Layer 3 services to all the VLANs defined on the access-layer switch.69. Use an 802. interface Port-channel1 description EtherChannel link to RS206-3925-1 switchport trunk allowed vlan 64. Option 2: Layer 2 trunk from router to access-layer switch Step 1:  Enable the physical interface on the router. Not all connected router platforms can support LACP to negotiate with the switch. so you configure EtherChannel statically.1Q trunk for the connection.Configure two physical interfaces to be members of the EtherChannel. apply the egress QoS macro that was defined in the LAN switch platform configuration procedure to ensure traffic is prioritized appropriately. GigabitEthernet2/0/24 switchport channel-group 1 mode on logging event link-status logging event trunk-status logging event bundle-status load-interval 30 macro apply EgressQoS Step 4:  Configure EtherChannel trunk on the access-layer switch.

interface GigabitEthernet1/0/24 description Link to RS201-2911 Gig0/2 switchport trunk allowed vlan 64. After you have enabled the physical interface or port-channel. When using a centralized DHCP server. Use an 802. interface [type][number]. which allows the router to provide the Layer 3 services to all the VLANs defined on the access-layer switch. If the remote-site router is the first router of a dual-router design. This requires a modified IP configuration on each subinterface. Procedure 5 Configure access-layer routing Option 1: Layer 2 EtherChannel or Layer 2 trunk Step 1:  Create subinterfaces and assign VLAN tags.1Q trunk for the connection. then HSRP is configured at the access layer.N.N.N.Step 2:  Configure the trunk on the access-layer switch. This is the preferred method. This design uses an IP addressing convention with the default gateway router assigned an IP address and IP mask combination of N. An alternate option for local DHCP server configuration is shown in the following procedure. The subinterface number does not need to equate to the 802.1 255.[sub-interface number] encapsulation dot1Q [dot1q VLAN tag] Step 2:  Configure IP settings for each subinterface.[sub-interface number] description [usage] ip address [LAN network 1] [LAN network 1 netmask] ip helper-address 10.255. you can map the appropriate data or voice subinterfaces to the VLANs on the LAN switch.69 switchport mode trunk ip arp inspection trust spanning-tree portfast trunk logging event link-status logging event trunk-status ip dhcp snooping trust no shutdown load-interval 30 macro apply EgressQoS The Cisco Catalyst 3750 Series Switch requires the switchport trunk encapsulation dot1q command.255. Set DHCP Snooping and Address Resolution Protocol (ARP) inspection to trust.N is the IP network and 1 is the IP host. but making them the same simplifies the overall configuration.10 ip pim sparse-mode Deploying an MPLS WAN August 2014 Series 45 .48. Prune the VLANs allowed on the trunk to only the VLANs that are active on the access-layer switch.1Q tag. interface [type][number].4. The subinterface portion of the configuration should be repeated for all data or voice VLANs.0 where N. routers with LAN interfaces connected to a LAN using DHCP for endstation IP addressing must use an IP helper.

255.1 10. you should use a resilient DHCP solution.48.4. Options for resilient DHCP at the remotesite include using IOS on a distribution-layer switch stack or implementing a dedicated DHCP server solution.1 255. ip dhcp excluded-address 10.68.local dns-server 10.0 255. Deploying an MPLS WAN August 2014 Series 46 . Tech Tip If you intend to use a dual-router remote-site design.64 description Wired Data encapsulation dot1Q 64 ip address 10.48.5.255. and other endpoint devices.5.11.19 ip dhcp pool DHCP-Wired-Data network 10.255.5.0 default-router 10. Step 2:  Configure a DHCP scope for data endpoints. Step 1:  Remove the previously configured ip helper-address commands for any interface that uses a local DHCP server. excluding DHCP assignment for the first 19 addresses in the subnet.4. This procedure uses a local DHCP service on the router in order to assign basic network configuration for IP phones.1 domain-name cisco. users’ laptop and desktop computers.10 ip pim sparse-mode Procedure 6 Configure remote-site DHCP (Optional) The previous procedure assumes the DHCP service has been configured centrally and uses the ip helperaddress command to forward DHCP requests to the centralized DHCP server.4. such as a centralized DHCP server.4.Example interface GigabitEthernet0/2.4.0 ip helper-address 10.10 Step 3:  Configure a DHCP scope for voice endpoints. If you choose to run a local DHCP server on the remote-site router instead of centralizing the DHCP service. wireless access points.4. complete this procedure.255.5. excluding DHCP assignment for the first 19 addresses in the subnet.

local dns-server 10.5.5. The router with the higher standby priority value is elected as the HSRP active router.Step 4:  Voice endpoints require an option field to tell them where to find their initial configuration.48.10 Procedure 7 through Procedure 11 are only relevant for the dual-router design.1 . however. The DR is normally elected based on the highest IP address. The HSRP active router is the MPLS CE router connected to the primary MPLS carrier.0 default-router 10. complete this procedure. The PIM designated router (DR) should be on the HSRP active router. you configure the HSRP active router with a standby priority that is higher than the HSRP standby router. In this design. without waiting for a scenario where there is no router in the HSRP active state. Different vendors use different option fields. ip dhcp excluded-address 10.5.WAN remote-site HSRP parameters (dual-router design) Router HSRP role Virtual IP address (VIP) Real IP address HSRP priority PIM DR priority MPLS CE (primary) Active . Tech Tip The HSRP priority and PIM DR priority are shown in the previous table to be the same value.3 105 105 The assigned IP addresses override those configured in the previous procedure.11. assigning the HSRP active router a lower real IP address than the HSRP standby router requires a modification to the PIM configuration. so the default gateway IP address remains consistent across locations with single or dual routers.5. The preempt option allows a router with a higher priority to become the HSRP active.5. You can influence the PIM DR election by explicitly setting the DR priority on the LAN-facing subinterfaces for the routers.4.255. In this procedure. Cisco uses DHCP option 150).19 ip dhcp pool DHCP-Wired-Voice network 10.5.2 110 110 MPLS CE (secondary) or DMVPN spoke Standby . Procedure 7 Configure access-layer HSRP If you are using a dual-router design. Deploying an MPLS WAN August 2014 Series 47 .255.1 .5. and the HSRP standby router is the router connected to the secondary MPLS carrier or backup link. Table 11 . so the number may vary based on the voice product you choose (for example. you are not required to use identical values.1 10. The relevant HSRP parameters for the router configuration are shown in the following table. and it has no awareness of the HSRP configuration.0 255. You need to configure HSRP in order to enable the use of a virtual IP (VIP) address as a default gateway that is shared between two routers. The dual-router access-layer design requires a modification for resilient multicast.1 domain-name cisco.

1 standby 1 priority 110 standby 1 preempt standby 1 authentication md5 key-string c1sco123 ! interface GigabitEthernet0/2.[sub-interface number] ip address [LAN network 1 address] [LAN network 1 netmask] ip pim dr-priority 110 standby version 2 standby 1 ip [LAN network 1 gateway address] standby 1 priority 110 standby 1 preempt standby 1 authentication md5 key-string c1sco123 Step 2:  Repeat this procedure for all data or voice subinterfaces.255.2 255.0 ip helper-address 10. interface [type][number].13.13.5.Step 1:  Configure HSRP.5.0 ip helper-address 10.255.2 255.1 standby 1 priority 110 standby 1 preempt standby 1 authentication md5 key-string c1sco123 Deploying an MPLS WAN August 2014 Series 48 .10 ip pim dr-priority 110 ip pim sparse-mode standby version 2 standby 1 ip 10.64 description Data encapsulation dot1Q 64 ip address 10.12.255.255.4.48.5.48.4.10 ip pim dr-priority 110 ip pim sparse-mode standby version 2 standby 1 ip 10.69 description Voice encapsulation dot1Q 69 ip address 10.12.5. Example: Layer 2 link interface GigabitEthernet0/2 no ip address no shutdown ! interface GigabitEthernet0/2.

This design uses a best practice of assigning the router ID to a loopback address. so HSRP and DHCP are not required.[sub-interface number] encapsulation dot1Q [dot1q VLAN tag] ip address [transit net address] [transit net netmask] ip pim sparse-mode Example interface GigabitEthernet0/2. In this design. The transit network is configured between the two routers. complete this procedure.99 description Transit Net encapsulation dot1Q 99 ip address 10. interface [type][number].252 ip pim sparse-mode Step 2:  On the access-layer switch. You must configure a routing protocol between the two routers. complete this procedure. All interfaces except the transit-network subinterface should remain passive.5. all LAN-facing interfaces and the loopback must be EIGRP interfaces. This network is used for router-router communication and to avoid hairpinning.255. Step 1:  On the primary MPLS CE router.1 255. vlan 99 name Transit-net Step 3:  Add the transit network VLAN to the existing access-layer switch trunk. There are no end stations connected to this network. This ensures that the HSRP active router has full reachability information for all WAN remote sites. Step 1:  On the router. The transit network should use an additional subinterface on the router interface that is already being used for data or voice. Do not include the WAN interface (MPLS PE-CE link interface) as an EIGRP interface. configure the transit network interface. The network range must include all interface IP addresses either in a single network statement or in multiple network statements.255. add the transit network VLAN. configure the EIGRP LAN process facing the access layer.8.Procedure 8 Configure the transit network If you are using a dual-router design. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface default passive-interface Deploying an MPLS WAN August 2014 Series 49 . interface GigabitEthernet1/0/24 switchport trunk allowed vlan add 99 Procedure 9 Configure EIGRP (LAN side) If you are using a dual-router design.

only the WAN bandwidth and delay values are used for metric calculation. A default metric redistributes the BGP routes into EIGRP.exit-af-interface af-interface [Transit link interface] no passive-interface exit-af-interface network [network] [inverse mask] eigrp router-id [IP address of Loopback0] exit-address-family Step 2:  Configure EIGRP neighbor authentication. Neighbor authentication enables the secure establishment of peering adjacencies and exchange route tables over the DMVPN tunnel interface. By default. Deploying an MPLS WAN August 2014 Series 50 . router eigrp LAN address-family ipv4 unicast autonomous-system 100 topology base default-metric [WAN bandwidth] [WAN delay] 255 1 1500 redistribute bgp 65511 exit-af-topology exit-address-family Tech Tip Command Reference: default-metric bandwidth delay reliability loading mtu bandwidth—Minimum bandwidth of the route in kilobytes per second delay—Route delay in tens of microseconds. key chain LAN-KEY key 1 key-string cisco ! router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface [interface type] [number] authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface exit-address-family Step 3:  On the router. redistribute BGP into the EIGRP LAN process.

1.255. By default.99 authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface topology base default-metric 100000 100 255 1 1500 redistribute bgp 65511 exit-af-topology network 10.0 0.4.255. this configuration is considered an internal BGP (iBGP) connection.255 network 10. In the dual-MPLS design. In certain cases. configure iBGP and enable the next-hop-self configuration option. Unless the remote site has been specifically designed for this type of routing behavior. You must complete this step on both remote-site MPLS CE routers. this means that MPLS-A routes will be advertised to MPLS-B and vice-versa. The dual-carrier MPLS design requires that a BGP link is configured between the CE routers. This design uses iBGP peering using the transit network.251. when a link to a MPLS hub has failed. You must use a route-map and an as-path access-list filter.206 exit-address-family Procedure 10 Configure BGP If you are using a dual-router design. remote sites will advertise themselves as a transit autonomous system. Deploying an MPLS WAN August 2014 Series 51 .255. the iBGP session will not be established until you complete the transit network and EIGRP (LAN-side) steps.Example router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface default passive-interface exit-af-interface af-interface GigabitEthernet0/2.255 eigrp router-id 10. providing access between the two carriers. with a high bandwidth connection.0. Step 1:  On both remote-site MPLS CE routers. BGP readvertises all BGP-learned routes. complete this procedure.0 0. which requires the next-hop-self-configuration option. Because the CE routers are using the same ASN.0. Note. router bgp 65511 neighbor [iBGP neighbor Transit Net IP] remote-as 65511 neighbor [iBGP neighbor Transit Net IP] next-hop-self Step 2:  Configure BGP to prevent the remote site from becoming a transit AS.0. You need to apply this route-map on both remote-site MPLS CE routers.255. it is a best practice to disable the site from becoming a transit site.

ip as-path access-list 1 permit _65401$ ! route-map PREFER-MPLS-A permit 10 match as-path 1 set local-preference 200 ! route-map PREFER-MPLS-A permit 20 Tech Tip The regular expression _65401$ corresponds to routes originated from the AS 65401 (MPLS-A). This type of filter allows for only the locally originated routes to be advertised. Step 4:  Add a loopback network for the secondary router. router bgp 65511 neighbor [IP address of PE] route-map NO-TRANSIT-AS out ip as-path access-list 10 permit ^$ ! route-map NO-TRANSIT-AS permit 10 match as-path 10 Tech Tip The regular expression ^$ corresponds to routes originated from the remote-site. the BGP local preference is 200 for the primary MPLS carrier. This requires the use of a route-map and an as-path access-list filter. router bgp 65511 network [Secondary router loopback network] mask 255. router bgp 65511 neighbor [IP address of PE] route-map PREFER-MPLS-A in Apply a route-map inbound to the neighbor for the primary MPLS carrier only. Routes originated from the secondary MPLS carrier continue to use their default local preference of 100. and it is likely that the first path selected will remain the active path unless the routing protocol detects a failure. This allows BGP to selectively modify the routing information for routes originated from this AS.Each router will apply this outbound to the neighbor for its respective MPLS carrier.255. Step 3:  Tune BGP routing to prefer the primary MPLS carrier. BGP uses a well-known rule set in order to determine the “best path” when the same IP route prefix is reachable via two different paths.255 Deploying an MPLS WAN August 2014 Series 52 . Accomplishing the design goal of deterministic routing and primary/secondary routing behavior necessitates tuning BGP. In this example. The MPLS dual-carrier design in many cases provides two equal cost paths.255.

complete this procedure. Having the HSRP router remain as the active router can lead to undesired behavior. and then send standard ICMP echo (ping) probes at 15-second intervals. track 50 ip sla 100 reachability Step 3:  Link HSRP with the tracked object.Procedure 11 Enable enhanced object tracking If you are using a dual-router design. The HSRP active router remains the active router unless the router is reloaded or fails. HSRP can monitor the tracked object status. and if the PE router becomes unreachable. the tracked object status is Up. If the status is down.3. the destination address is the same as the BGP neighbor address configured in Procedure 3. If you are using the MPLS PE router as the probe destination. If the probe is successful. If the decrease is large enough. This is sub-optimal routing.168. The HSRP active router (primary MPLS CE) can use the IP SLA feature to send echo probes to its MPLS PE router.10 source-interface GigabitEthernet0/0 timeout 1000 Deploying an MPLS WAN August 2014 Series 53 . Step 1:  Enable the IP SLA probe. so that the HSRP standby router can preempt and become the HSRP active router. interface [interface type] [number]. the HSRP active router would learn an alternate path through the transit network to the HSRP standby router and begin to forward traffic across the alternate path. the HSRP priority is decremented by the configured priority. The object being tracked is the reachability success or failure of the probe. Responses must be received before the timeout of 1000 ms expires. the HSRP standby router preempts. and you can address it by using EOT. This procedure is valid only on the router connected to the primary transport (MPLS VPN).[sub-interface number] standby 1 track 50 decrement 10 Example interface GigabitEthernet 0/2.69 standby 1 track 50 decrement 10 ! track 50 ip sla 100 reachability ! ip sla 100 icmp-echo 192. if it fails. All data or voice subinterfaces should enable HSRP tracking.64 standby 1 track 50 decrement 10 interface GigabitEthernet 0/2. ip sla 100 icmp-echo [probe destination IP address] source-interface [WAN interface] threshold 1000 timeout 1000 frequency 15 ip sla schedule 100 life forever start-time now Step 2:  Configure EOT based on the IP SLA probe. then the router can lower its HSRP priority. the tracked object status is Down. If the primary MPLS VPN transport were to fail.

PROCESS threshold 1000 frequency 15 ip sla schedule 100 life forever start-time now Adding a Secondary MPLS Link on an Existing MPLS CE Router 1. The following procedures assume that the configuration of an MPLS CE router for an MPLS WAN remote site (single-router. Configure BGP for Dual-link link Site Complete Deploying an MPLS WAN 2130 NO August 2014 Series 54 . Only the additional procedures to add an additional MPLS link to the running MPLS CE router are included here. Connect to MPLS PE Router uter 2. Configure BGP for dual-link design This process includes the additional steps necessary to complete the configuration of an MPLS CE router for an MPLS WAN dual-carrier remote site (single-router.Flowchart for adding an MPLS backup configuration MPLS CE Router Configuration Complete Adding Second MPLS Link on Existing MPLS CE Router Configuration Procedures YES Add MPLS Secondary Link? 1. The following figure provides details on how to add a second MPLS backup link on an existing remote-site MPLS CE router. single-link) has already been completed and BGP dynamic routing has been configured. Connect to MPLS PE router 2. dual-link). Figure 15 .

252 is used.255. interface [interface type] [number] bandwidth [bandwidth (kbps)] Tech Tip Command Reference: bandwidth kbps 10 Mbps = 10. a point-to-point netmask of 255.Procedure 1 Connect to MPLS PE router This procedure applies to the interface used to connect the secondary or additional MPLS carrier.000 kbps Step 2:  Assign the IP address and netmask of the WAN interface. interface [interface type] [number] no cdp enable no shutdown Example interface GigabitEthernet0/1 bandwidth 10000 ip address 192. use the policed rate from the carrier. Typically.255. The example shows a Gigabit interface (1000 Mbps) with a subrate of 10 Mbps. It is not recommended that you use the Cisco Discovery Protocol on external interfaces. interface [interface type] [number] ip address [IP address] [netmask] Step 3:  Administratively enable the interface and disable Cisco Discovery Protocol.168.255.255. If you are using a subrate service. Step 1:  Assign an interface bandwidth value that corresponds to the actual interface speed.4.252 ip pim sparse-mode no cdp enable no shutdown Deploying an MPLS WAN August 2014 Series 55 . The IP addressing used between CE and PE routers must be negotiated with your MPLS carrier.13 255.

and it is likely that the first path selected will remain the active path unless the routing protocol detects a failure. The MPLS carrier must provide their ASN (the ASN in this step is the ASN identifying your site). It is desirable to advertise a route for the PE-CE link. this configuration is considered an external BGP (eBGP) connection. BGP readvertises all BGP-learned routes. router bgp 65511 neighbor [IP address of PE] route-map NO-TRANSIT-AS out neighbor [IP address of PE 2] route-map NO-TRANSIT-AS out ip as-path access-list 10 permit ^$ ! route-map NO-TRANSIT-AS permit 10 match as-path 10 Tech Tip The regular expression ^$ corresponds to routes originated from the remote-site. so you should include this network in a network statement. Step 3:  Tune BGP routing to prefer the primary MPLS carrier. Unless the remote site has been specifically designed for this type of routing behavior. router bgp 65511 neighbor [IP address of PE] route-map PREFER-MPLS-A in Deploying an MPLS WAN August 2014 Series 56 . you need to use a route-map and an as-path access-list filter. BGP uses a well-known rule set in order to determine the “best path” when the same IP route prefix is reachable via two different paths. This type of filter allows for only the locally originated routes to be advertised. To do this. You can use it to determine router reachability. BGP must be configured with the MPLS carrier PE device.Procedure 2 Configure BGP for dual-link design Step 1:  Configure eBGP to add an additional eBGP neighbor and advertise the PE-CE link. In the dual-MPLS design. providing access between the two carriers. Apply this route-map outbound to the neighbors for both MPLS carriers. remote sites will advertise themselves as a transit autonomous system. with a high bandwidth connection. Because the carrier PE router uses a different ASN. The MPLS dual-carrier design in many cases provides two equal cost paths. In certain cases. router bgp 65511 network [PE-CE link 2 network] mask [PE-CE link 2 netmask] neighbor [IP address of PE 2] remote-as [carrier ASN] Step 2:  Configure BGP to prevent the remote site from becoming a transit AS. Accomplishing the design goal of deterministic routing and primary/secondary routing behavior necessitates tuning BGP. By default. This requires the use of a route-map and an as-path access-list filter. when a link to a MPLS hub has failed. for troubleshooting. The remote-site LAN networks are already advertised based on the configuration already completed in the “Configuring the Remote-Site MPLS CE Router” process. this means that MPLS-A routes are advertised to MPLS-B and vice-versa. it is a best practice to disable the site from becoming a transit site.

168.Step 4:  Apply a route map inbound to the neighbor for the primary MPLS carrier only.14 route-map NO-TRANSIT-AS out neighbor 192.168. This allows BGP to selectively modify the routing information for routes originated from this AS.4.14 route-map NO-TRANSIT-AS out ! ip as-path access-list 1 permit _65401$ ip as-path access-list 10 permit ^$ ! route-map NO-TRANSIT-AS permit 10 match as-path 10 ! route-map PREFER-MPLS-A permit 10 match as-path 1 set local-preference 200 ! route-map PREFER-MPLS-A permit 20 Deploying an MPLS WAN August 2014 Series 57 .168. ip as-path access-list 1 permit _65401$ ! route-map PREFER-MPLS-A permit 10 match as-path 1 set local-preference 200 ! route-map PREFER-MPLS-A permit 20 Tech Tip The regular expression _65401$ corresponds to routes originated from the AS 65401 (MPLS-A).168.252 neighbor 192.3. Example router bgp 65511 network 192.4.255. Routes originated from the secondary MPLS carrier continue to use their default local preference of 100.12 mask 255.3.255. Apply this route-map inbound to the neighbor for the primary MPLS carrier only. In this example.168. the BGP local preference is 200 for the primary MPLS carrier.14 route-map PREFER-MPLS-A in neighbor 192.4.14 remote-as 65402 neighbor 192.

complete this procedure in order to configure the secondary router in the MPLS WAN remote site.Configuring the Secondary Remote-Site Router PROCESS 1. C Connectt R Router 5. Configure the WAN remote router 2. Co Configure August 2014 Series 58 . Connect to MPLS PE Router 3. Configure WAN Routing NO O Distribution Layer Design? YES Y Remote-Site Router to Distribution Layer (Router 2) ccess Layer Switch 4. Configure EIGRP (LAN Side) 6. Connect Router to Access 1 t to Distribution Layer 1. Figure 16 . Connect router to access-layer switch 5.Router 2) Configuration Procedures 1. The following flowchart provides details about how to configure a secondary remote-site MPLS CE router. Configure Access Layer Routing 2. Configure the transit network 8. Connect to the MPLS PE router 3. Configure WAN routing 4.Remote-site MPLS CE router 2 configuration flowchart Remote-Site MPLS CE Router Dual Router. Configure Transit Network Site Complete Deploying an MPLS WAN Site Complete 2131 8 gu e EIGRP G ((LAN Side) S de) 8. Dual Link (2nd Router) Remote-Site Router (Dual Router . dual-link design. Configure Access Layer HSRP 7. Configure access-layer routing 6. Configure access-layer HSRP 7. Configure EIGRP (LAN side) If you are using a dual-router. Configure the WAN Remote Router 2.

HTTPS access to the router uses the enable password for authentication. all management access to the network infrastructure devices (SSH and HTTPS) is controlled by AAA. This makes it easy to identify the device. As networks scale in the number of devices to maintain it poses an operational burden to maintain local user accounts on every device. The enable password secures access to the device configuration mode. and accounting (AAA) service reduces operational tasks per device and provides an audit log of user access for security compliance and root cause analysis.Procedure 1 Configure the WAN remote router Within this design. Step 1:  Configure the device host name. Secure HTTP (HTTPS) and Secure Shell (SSH) are secure replacements for the HTTP and Telnet protocols. you prevent the disclosure of plain text passwords when viewing configuration files. A local AAA user database is also defined in Step 2 on each network infrastructure device to provide a fallback authentication source in case the centralized TACACS+ server is unavailable.4. tacacs server TACACS-SERVER-1 address ipv4 10. They use Secure Sockets Layer (SSL) and Transport Layer Security (TLS) to provide device authentication and data encryption. authorization. hostname [hostname] Step 2:  Configure local login and password. username admin password c1sco123 enable secret c1sco123 service password-encryption aaa new-model By default. Step 3:  (Optional) Configure centralized user authentication. A centralized authentication.15 key SecretKey ! aaa group server tacacs+ TACACS-SERVERS server name TACACS-SERVER-1 ! aaa authentication login default group TACACS-SERVERS local aaa authorization exec default group TACACS-SERVERS local aaa authorization console ip http authentication aaa Step 4:  Configure device management protocols. By enabling password encryption. When AAA is enabled for access control. and this access provides only limited operational privileges. The local login account and password provides basic access authentication to a router. Deploying an MPLS WAN August 2014 Series 59 . These are system settings that simplify and secure the management of the solution.48. TACACS+ is the primary protocol used to authenticate management logins on the infrastructure devices to the AAA server. there are features and services that are common across all WAN remote-site routers.

only devices on the 10. Telnet and HTTP. ip domain-name cisco. Both protocols are encrypted for privacy.4. if the ip name-server is unreachable. When synchronous logging of unsolicited messages and debug output is turned on. which allows the use of code upgrades using Prime Infrastructure via the SSH-based SCP protocol. are turned off. line con 0 transport preferred none logging synchronous Step 6:  Enable Simple Network Management Protocol (SNMP). Specify the transport preferred none on vty lines to prevent errant connection attempts from the CLI prompt. SCP is enabled.Secure management of the network device is enabled through the use of the SSH and HTTPS protocols. access-list 55 permit 10. you can increase network security by using an access list to limit the networks that can access your device. In this example. This allows the network infrastructure devices to be managed by a Network Management System (NMS). and the unsecure protocols. snmp-server community cisco RO snmp-server community cisco123 RW Step 7:  If operational support is centralized in your network. Deploying an MPLS WAN August 2014 Series 60 .local ip ssh version 2 no ip http server ip http secure-server ip scp server enable line vty 0 15 transport input ssh transport preferred none Step 5:  Enable synchronous logging. Without this command. console log messages are displayed on the console after interactive CLI output is displayed or printed.48.4.255 line vty 0 15 access-class 55 in ! snmp-server community cisco RO 55 snmp-server community cisco123 RW 55 Tech Tip If you configure an access-list on the vty interface you may lose the ability to use ssh to login from one router to the next for hop-by-hop troubleshooting. With this command.0 0.0. long timeout delays may occur for mistyped commands. SNMPv2c is configured both for a read-only and a read-write community string. you can continue typing at the device console when debugging is enabled.0/24 network will be able to access the device via SSH or SNMP.0.48.

255 ip pim sparse-mode Bind the device processes for SNMP. Deploying an MPLS WAN August 2014 Series 61 . such as a radio clock or an atomic clock attached to a time server. By configuring console messages. Layer 3 process and features are also bound to the loopback interface to ensure process resiliency. Auto RP is used to provide a simple yet scalable way to provide a highly resilient RP environment. In this design. and debug output to provide time stamps on output. The local NTP server typically references a more accurate clock feed from an outside source. PIM. end hosts must join a multicast group by sending an IGMP message to their local multicast router. which is based on sparse mode multicast operation. An NTP network usually gets its time from an authoritative time source.255. IP Telephony MOH and IP Video Broadcast Streaming are two examples of IP Multicast applications. TACACS+ and NTP to the loopback interface address for optimal resiliency.Step 8:  Configure a synchronized clock. To receive a particular IP Multicast data stream. you can cross-reference events in a network. You should program network devices to synchronize to a local NTP server in the network. Using IP Multicast is much more efficient than multiple individual unicast streams or a Broadcast stream that would propagate everywhere. The Network Time Protocol (NTP) is designed to synchronize a network of devices. Because of this capability.255. logs.48.17 ntp update-calendar ! clock timezone PST -8 clock summer-time PDT recurring ! service timestamps debug datetime msec localtime service timestamps log datetime msec localtime Step 9:  Configure an in-band management interface. interface Loopback 0 ip address [ip address] 255. the local router consults another router in the network that is acting as an RP to map the receivers to active sources so they can join their streams. ntp server 10. The loopback address is commonly a host address with a 32-bit address mask. In a traditional IP Multicast design. SSH. NTP then distributes this time across the organizations network. IP Multicast allows a single IP data stream to be replicated by the infrastructure (routers and switches) and sent from a single source to multiple receivers. the loopback address is the best way to manage the router in-band. Allocate the loopback address from a unique network range that is not part of any other internal network summary range. The loopback interface is a logical interface that is always reachable as long as the device is powered on and any IP interface is reachable to the network.4. snmp-server trap-source Loopback0 ip ssh source-interface Loopback0 ip pim register-source Loopback0 ip tacacs source-interface Loopback0 ntp source Loopback0 Step 10:  Configure IP Multicast routing.

interface [interface type] [number] bandwidth [bandwidth (kbps)] Tech Tip Command Reference: bandwidth kbps 10 Mbps = 10. It is not recommend that you use Cisco Discovery Protocol on external interfaces. You must negotiate the IP addressing used between CE and PE routers with your MPLS carrier. ip pim autorp listener Step 12:  Enable sparse mode multicast operation for all Layer 3 interfaces in the network.255.255.255. Use the ip pim autorp listener command to allow for discovery across sparse mode links. use the policed rate from the carrier.252 no cdp enable no shutdown Deploying an MPLS WAN August 2014 Series 62 . Typically.168.Enable IP Multicast routing on the platforms in the global configuration mode. The example shows a Gigabit interface (1000 Mbps) with a subrate of 10 Mbps. ip multicast-routing Step 11:  Configure every Layer 3 switch and router to discover the IP Multicast RP with autorp. This configuration provides for future scaling and control of the IP Multicast environment and can change based on network needs and design.252 is used.9 255.4. ip pim sparse-mode Procedure 2 Connect to the MPLS PE router Step 1:  Assign an interface bandwidth value that corresponds to the actual interface speed. interface [interface type] [number] ip address [IP address] [netmask] Step 3:  Administratively enable the interface and disable Cisco Discovery Protocol. interface [interface type] [number] no cdp enable no shutdown Example interface GigabitEthernet0/0 bandwidth 25000 ip address 192. a pointto-point netmask of 255. If you are using a subrate service.255.000 kbps Step 2:  Assign the IP address and netmask of the WAN interface.

this configuration is considered an external BGP (eBGP) connection. The MPLS carrier must provide their ASN (the ASN in the previous step is the ASN identifying your site). The aggregate address configured below suppresses the more specific routes. you must list the loopbacks of both the primary and secondary routers as BGP networks. so you should include this network in a network statement. router bgp 65511 no synchronization bgp router-id [IP address of Loopback0] bgp log-neighbor-changes no auto-summary Step 2:  Configure eBGP. You must configure BGP with the MPLS carrier PE device. Because the carrier PE router uses a different ASN. To complete this step. the aggregate is advertised to the MPLS PE. If the various LAN networks cannot be summarized. The CE router advertises only network routes to the PE via BGP in the following cases: • The route is specified in network statements and is present in the local routing table.255.255. router bgp 65511 network [PE-CE link network] mask [PE-CE link netmask] network [Primary router loopback network] mask 255.255.255. The IP assignment for the remote sites was designed so that all of the networks in use can be summarized within a single aggregate route. To ensure reachability of the loopback interfaces in a dual-router design. you must list each individually. It is desirable to advertise a route for the PE-CE link. you must add a network statement for the loopback address of the secondary MPLS CE router.255 network [Secondary router loopback network] mask 255. If any LAN network is present in the route table. This is required for loopback resiliency. • The route is redistributed into BGP (not applicable in the remote-site use case). You might be able to reuse the same value used on the MPLS VPN CE from the WAN-aggregation site. Consult with your MPLS carrier on the requirements for the ASN.Procedure 3 Configure WAN routing Step 1:  Enable BGP. for troubleshooting. Tech Tip On the primary MPLS CE router. The remote-site routers have in-band management configured via the loopback interface. You can use this to determine router reachability. you must use a BGP ASN.255 network [DATA network] mask [netmask] network [VOICE network] mask [netmask] aggregate-address [summary IP address] [summary netmask] summary-only neighbor [IP address of PE] remote-as [carrier ASN] Deploying an MPLS WAN August 2014 Series 63 . which offers a measure of resiliency. The remote-site LAN networks must be advertised.

4. Because the CE routers are using the same ASN.255.0 mask 255.10 remote-as 65402 neighbor 192.255. router bgp 65511 neighbor [IP address of PE 2] route-map NO-TRANSIT-AS out ip as-path access-list 10 permit ^$ ! route-map NO-TRANSIT-AS permit 10 match as-path 10 Tech Tip The regular expression ^$ corresponds to routes originated from the remote-site. with a high bandwidth connection. By default.255.0 255.252. when a link to an MPLS hub has failed. the iBGP session will not be established until you complete the transit network and EIGRP (LAN side) steps.5. In certain cases.5. Note. Each router applies this route map outbound to the neighbor for its respective MPLS carrier.168.255.255.8. You need to apply this route map on both remote-site MPLS CE routers.255.10 route-map NO-TRANSIT-AS out no auto-summary ! ip as-path access-list 10 permit ^$ Deploying an MPLS WAN August 2014 Series 64 .8. Example: MPLS CE Router (secondary) router bgp 65511 no synchronization bgp router-id 10.5. BGP readvertises all BGP learned routes.248.255.0 summary-only neighbor 10.4. This type of filter allows for only the locally originated routes to be advertised.13.168. The dual-carrier MPLS design requires that a BGP link is configured between the CE routers.5.0 network 10.0 mask 255. this configuration is considered an internal BGP (iBGP) connection.168.252 network 10.255 network 10. You must use a route-map and an as-path access-list filter.206 mask 255.255.255. In the dual-MPLS design.1 remote-as 65511 neighbor 10.4.1 next-hop-self neighbor 192.206 mask 255.255. router bgp 65511 neighbor [iBGP neighbor Transit Net IP] remote-as 65511 neighbor [iBGP neighbor Transit Net IP] next-hop-self Step 4:  Configure BGP to prevent the remote site from becoming a transit AS.8.8 mask 255.5. it is a best practice to disable the site from becoming a transit site. this means that MPLS-A routes will be advertised to MPLS-B and vice-versa.206 bgp log-neighbor-changes network 192.252.12.255.255.0 aggregate-address 10.Step 3:  Configure iBGP between the remote-site MPLS CE routers.255. Unless the remote site has been specifically designed for this type of routing behavior.255. remote sites will advertise themselves as a transit autonomous system.255 network 10.251. providing access between the two carriers.

or in the case of the Cisco Catalyst 4507R+E distribution layer. In the access-layer design. The VLAN numbering is locally significant only. interface Port-channel2 description EtherChannel link to RS206-A2960X no shutdown Step 2:  Configure EtherChannel member interfaces on the router. interface GigabitEthernet0/1 description RS206-A2960X Gig1/0/23 ! interface GigabitEthernet0/2 description RS206-A2960X Gig2/0/23 ! interface range GigabitEthernet0/1.! route-map NO-TRANSIT-AS permit 10 match as-path 10 Procedure 4 Connect router to access-layer switch Reader Tip This guide includes only the additional steps to complete the distribution-layer configuration. the remote sites use collapsed routing. Layer 2 EtherChannels are used to interconnect the CE router to the access layer in the most resilient method possible. Configure the physical interfaces to tie to the logical port-channel by using the channel-group command. Not all router platforms can support LACP to negotiate with the switch. so you configure EtherChannel statically. If the access-layer device is a single. to separate redundant modules for additional resiliency. Option 1: Layer 2 EtherChannel from router to access-layer switch Step 1:  Configure a port-channel interface on the router. then skip to the “Deploying a WAN Remote-Site Distribution Layer” chapter of this guide.1Q trunk interfaces to the LAN access layer. with 802. If you are using a remote-site distribution layer. GigabitEthernet0/2 no ip address channel-group 2 no shutdown Step 3:  Configure EtherChannel member interfaces on the access-layer switch Connect the router EtherChannel uplinks to separate switches in the access layer switch stack. see the Campus Wired LAN Technology Design Guide. Deploying an MPLS WAN August 2014 Series 65 . a simple Layer 2 trunk between the router and switch is used. fixed-configuration switch. For complete access-layer configuration details. The number for the port-channel and channel-group must match.

GigabitEthernet2/0/23 switchport macro apply EgressQoS channel-group 2 mode on logging event link-status logging event trunk-status logging event bundle-status Step 4:  Configure EtherChannel trunk on the access-layer switch. Deploying an MPLS WAN August 2014 Series 66 . Not all connected router platforms can support LACP to negotiate with the switch.69. This allows the router to provide the Layer 3 services to all the VLANs defined on the access-layer switch. Use an 802. Doing the configuration in this order allows for minimal configuration and reduces errors because most of the commands entered to a port-channel interface are copied to its members’ interfaces and do not require manual replication. apply the egress QoS macro that was defined in the platform configuration procedure in order to ensure traffic is prioritized appropriately.1Q trunk for the connections. Also. so you configure EtherChannel statically. Set DHCP Snooping and Address Resolution Protocol (ARP) inspection to trust. Prune the VLANs allowed on the trunk to only the VLANs that are active on the access-layer switch.The physical interfaces that are members of a Layer 2 EtherChannel are configured prior to configuring the logical port-channel interface. interface Port-channel2 description EtherChannel link to RS206-3925-2 switchport trunk allowed vlan 64.99 switchport mode trunk ip arp inspection trust spanning-tree portfast trunk ip dhcp snooping trust no shutdown The Cisco Catalyst 3750 Series Switch requires the switchport trunk encapsulation dot1q command. the interface type is port-channel. When using EtherChannel. Configure two or more physical interfaces to be members of the EtherChannel. interface GigabitEthernet1/0/23 description Link to RS206-3925-2 Gig0/1 interface GigabitEthernet2/0/23 description Link to RS206-3925-2 Gig0/2 ! interface range GigabitEthernet1/0/23. It is recommended that they are added in multiples of two. and the number must match the channel group configured in Step 2.

N. and then set DHCP Snooping and Address Resolution Protocol (ARP) inspection to trust.69. but making them the same simplifies the overall configuration. interface [type][number]. This design uses an IP addressing convention with the default gateway router assigned an IP address and IP mask combination of N.N. you can map the appropriate data or voice subinterfaces to the VLANs on the LAN switch.1Q trunk for the connection. When using a centralized DHCP server. This allows the router to provide the Layer 3 services to all the VLANs defined on the access-layer switch.1Q tag. Step 1:  Create subinterfaces and assign VLAN tags. The subinterface number does not need to equate to the 802.Option 2: Layer 2 trunk from router to access-layer switch Step 1:  Enable the physical interface on the router. The actual interface IP assignments are configured in the following procedure. interface GigabitEthernet0/2 description RS206-A2960X Gig1/0/23 no ip address no shutdown Step 2:  Configure the trunk on the access-layer switch.255.1 255. Step 3:  Configure IP settings for each subinterface. Deploying an MPLS WAN August 2014 Series 67 .[sub-interface number] encapsulation dot1Q [dot1q VLAN tag] Step 2:  Repeat the subinterface portion of the previous step for all data or voice VLANs.255.N. routers with LAN interfaces connected to a LAN using DHCP for endstation IP addressing must use an IP helper. interface GigabitEthernet1/0/23 description Link to RS206-3925-2 Gig0/2 switchport trunk allowed vlan 64.99 switchport mode trunk ip arp inspection trust spanning-tree portfast trunk logging event link-status logging event trunk-status ip dhcp snooping trust no shutdown load-interval 30 macro apply EgressQoS The Cisco Catalyst 3750 Series Switch requires the switchport trunk encapsulation dot1q command. After the physical interface or port-channel has been enabled. Use an 802.N is the IP network and 1 is the IP host. Procedure 5 Configure access-layer routing This remote-site MPLS CE router is the second router of a dual-router design. and HSRP is configured at the access layer.0 where N. Prune the VLANs allowed on the trunk to only the VLANs that are active on the access-layer switch.

48.4. The actual interface IP assignments will be configured in the following procedure.48.10 ip pim sparse-mode Example: Layer 2 Trunk interface GigabitEthernet0/2 no ip address no shutdown ! interface GigabitEthernet0/2.10 ip pim sparse-mode Deploying an MPLS WAN August 2014 Series 68 .10 ip pim sparse-mode Example: Layer 2 EtherChannel interface Port-channel2 no ip address no shutdown ! hold-queue 150 in ! interface Port-channel2.[sub-interface number] description [usage] ip helper-address 10.64 description Data encapsulation dot1Q 64 ip helper-address 10.4.4.48.10 ip pim sparse-mode ! interface GigabitEthernet0/2.69 description Voice encapsulation dot1Q 69 ip helper-address 10.48.10 ip pim sparse-mode ! interface Port-channel2.4. interface [type][number].This remote-site MPLS CE router is the second router of a dual-router design and HSRP is configured at the access layer.48.64 description Data encapsulation dot1Q 64 ip helper-address 10.69 description Voice encapsulation dot1Q 69 ip helper-address 10.4.

and it has no awareness of the HSRP configuration.1 . The relevant HSRP parameters for the router configuration are shown in the following table.3 105 105 The dual-router access-layer design requires a modification for resilient multicast. The router with the higher standby priority value is elected as the HSRP active router. The preempt option allows a router with a higher priority to become the HSRP active. Step 1:  Configure HSRP. The DR is normally elected based on the highest IP address.2 110 110 MPLS CE (secondary) or DMVPN Spoke Standby . you are not required to use identical values. interface [type][number].Procedure 6 Configure access-layer HSRP Configure HSRP to use a virtual IP (VIP) as a default gateway that is shared between two routers. The PIM designated router (DR) should be on the HSRP active router. You can influence the PIM DR election by explicitly setting the DR priority on the LAN-facing subinterfaces for the routers.WAN remote-site HSRP parameters (dual-router design) Router HSRP role Virtual IP address (VIP) Real IP address HSRP priority PIM DR priority MPLS CE (primary) Active . In this design. you configure the HSRP active router with a standby priority that is higher than the HSRP standby router. Table 12 .[sub-interface number] ip address [LAN network 1 address] [LAN network 1 netmask] ip pim dr-priority 105 standby version 2 standby 1 ip [LAN network 1 gateway address] standby 1 priority 105 standby 1 preempt standby 1 authentication md5 key-string c1sco123 Deploying an MPLS WAN August 2014 Series 69 . assigning the HSRP active router a lower real IP address than the HSRP standby router requires a modification to the PIM configuration. however. Tech Tip The HSRP priority and PIM DR priority are shown in the previous table to be the same value. In this procedure. without waiting for a scenario where there is no router in the HSRP active state.1 . The HSRP active router is the MPLS CE router connected to the primary MPLS carrier. and the HSRP standby router is the router connected to the secondary MPLS carrier or backup link.

255.5.255.1 standby 1 priority 105 standby 1 preempt standby 1 authentication md5 key-string c1sco123 ! interface Port-channel2.4.5.255.10 ip pim dr-priority 105 ip pim sparse-mode standby version 2 standby 1 ip 10.12.4.12.4. Example: MPLS CE Router (Secondary) with Layer 2 EtherChannel interface Port-channel2 no ip address no shutdown ! interface Port-channel2.69 description Voice encapsulation dot1Q 69 ip address 10.1 Deploying an MPLS WAN August 2014 Series 70 .13.5.0 ip helper-address 10.Step 2:  Repeat this procedure for all data or voice subinterfaces.3 255.5.48.48.10 ip pim dr-priority 105 ip pim sparse-mode standby version 2 standby 1 ip 10.64 description Data encapsulation dot1Q 64 ip address 10.12.1 standby 1 priority 105 standby 1 preempt standby 1 authentication md5 key-string c1sco123 Example: MPLS CE Router (Secondary) with Layer 2 Trunk interface GigabitEthernet0/2 no ip address no shutdown ! interface GigabitEthernet0/2.5.255.3 255.13.0 ip helper-address 10.48.3 255.10 ip pim dr-priority 105 ip pim sparse-mode standby version 2 standby 1 ip 10.0 ip helper-address 10.255.5.12.255.64 description Data encapsulation dot1Q 64 ip address 10.

69 description Voice encapsulation dot1Q 69 ip address 10.5.255.252 ip pim sparse-mode Deploying an MPLS WAN August 2014 Series 71 .4.8. The transit network should use an additional subinterface on the router interface that is already being used for data or voice.3 255.255.2 255. interface [interface type][number].255.5.13. Step 1:  On the secondary MPLS CE router.48. You use this network for router-router communication and to avoid hairpinning. configure the transit network interface.255.5.10 ip pim dr-priority 105 ip pim sparse-mode standby version 2 standby 1 ip 10.[sub-interface number] encapsulation dot1Q [dot1q VLAN tag] ip address [transit net address] [transit net netmask] ip pim sparse-mode Example interface GigabitEthernet0/2.0 ip helper-address 10.1 standby 1 priority 105 standby 1 preempt standby 1 authentication md5 key-string c1sco123 Procedure 7 Configure the transit network Configure the transit network between the two routers. so HSRP and DHCP are not required.99 description Transit Net encapsulation dot1Q 99 ip address 10.13.standby 1 priority 105 standby 1 preempt standby 1 authentication md5 key-string c1sco123 ! interface GigabitEthernet0/2. There are no end stations connected to this network.

All interfaces except the transit-network subinterface should remain passive.Procedure 8 Configure EIGRP (LAN side) You must configure a routing protocol between the two routers. Step 1:  On the router. In this design. configure the EIGRP LAN process facing the access layer. Do not include the WAN interface (MPLS PE-CE link interface) as an EIGRP interface. all LAN-facing interfaces and the loopback must be EIGRP interfaces. This design uses a best practice of assigning the router ID to a loopback address. The network range must include all interface IP addresses either in a single network statement or in multiple network statements. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface default passive-interface exit-af-interface af-interface [Transit link interface] no passive-interface exit-af-interface network [network] [inverse mask] eigrp router-id [IP address of Loopback0] exit-address-family Step 2:  Configure EIGRP neighbor authentication. This ensures that the HSRP active router has full reachability information for all WAN remote sites. key chain LAN-KEY key 1 key-string cisco ! router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface [interface type] [number] authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface exit-address-family Deploying an MPLS WAN August 2014 Series 72 . Neighbor authentication enables the secure establishment of peering adjacencies and exchange route tables over the DMVPN tunnel interface.

redistribute BGP into the EIGRP LAN process.255. A default metric redistributes the BGP routes into EIGRP.255 eigrp router-id 10.99 authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface topology base default-metric 100000 100 255 1 1500 redistribute bgp 65511 exit-af-topology network 10.255.252. By default. Example router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface default passive-interface exit-af-interface af-interface GigabitEthernet0/2.4.Step 3:  On the router. only the WAN bandwidth and delay values are used for metric calculation.206 exit-address-family Deploying an MPLS WAN August 2014 Series 73 . router eigrp LAN address-family ipv4 unicast autonomous-system 100 topology base default-metric [WAN bandwidth] [WAN delay] 255 1 1500 redistribute bgp 65511 exit-af-topology exit-address-family Tech Tip Command Reference: default-metric bandwidth delay reliability loading mtu bandwidth—Minimum bandwidth of the route in kilobytes per second delay—Route delay in tens of microseconds.1.0.0 0.

Deploying a WAN Remote-Site
Distribution Layer

PROCESS

Deployment Details
Connecting the Single or Primary Remote-Site Router to the
Distribution Layer
1. Connect router to distribution layer
2. Configure EIGRP (LAN side)
3. Configure the transit network
4. Configure BGP
If you are configuring an MPLS WAN remote-site that uses a single-router, single link design or a dual-router,
dual-link design, complete this process. This process includes all required procedures in order to connect either
the single-router in a single-link design or the primary router in a dual-link design to a LAN distribution layer.
Both distribution-layer remote-site options are shown in the following figure.
Figure 17 - WAN remote site—Connection to distribution layer
WAN

WAN

R1

802.1Q Trunk
(50)

802.1Q Trunk (xx-xx)

R2

802.1Q Trunk
(50, 99)

802.1Q Trunk (64, 69)

VLAN 50 - Router 1 Link

Deploying a WAN Remote-Site Distribution Layer

802.1Q Trunk
(54, 99)

802.1Q Trunk (xx-xx)

VLAN 50 - Router 1 Link
VLAN 54 - Router 2 Link
VLAN 99 - Transit

August 2014 Series

2185

R1

74

Procedure 1

Connect router to distribution layer

Reader Tip
This guide includes only the additional steps to complete the distribution-layer
configuration. For complete distribution-layer configuration details, see the Campus
Wired LAN Technology Design Guide.

Layer 2 EtherChannels are used to interconnect the CE router to the distribution layer in the most resilient
method possible. This connection allows for multiple VLANs to be included on the EtherChannel as necessary.
Step 1:  Configure a port-channel interface on the router.
interface Port-channel1
description EtherChannel link to RS200-D4500X-VSS
no shutdown
Step 2:  Configure the port channel subinterfaces and assign IP addresses.
After you have enabled the interface, map the appropriate subinterfaces to the VLANs on the distributionlayer switch. The subinterface number does not need to equate to the 802.1Q tag, but making them the same
simplifies the overall configuration.
The subinterface configured on the router corresponds to a VLAN interface on the distribution-layer switch.
Traffic is routed between the devices with the VLAN acting as a point-to-point link.
interface Port-channel1.50
description R1 routed link to distribution layer
encapsulation dot1Q 50
ip address 10.5.0.1 255.255.255.252
ip pim sparse-mode
Step 3:  On the router, configure EtherChannel member interfaces.
Configure the physical interfaces to tie to the logical port-channel by using the channel-group command. The
number for the port-channel and channel-group must match. Not all router platforms can support LACP to
negotiate with the switch, so you configure EtherChannel statically.
interface GigabitEthernet0/1
description RS200-D4500X-VSS Ten1/1/1
!
interface GigabitEthernet0/2
description RS200-D4500X-VSS Ten2/1/1
!
interface range GigabitEthernet0/1, GigabitEthernet0/2
no ip address
channel-group 1
no shutdown

Deploying a WAN Remote-Site Distribution Layer

August 2014 Series

75

Step 4:  On the distribution-layer switch, configure the VLAN.
vlan 50
name R1-link
Step 5:  On the distribution-layer switch, configure Layer 3.
Configure a VLAN interface, also known as a switch virtual interface (SVI), for the new VLAN added. The SVI is
used for point-to-point IP routing between the distribution layer and the WAN router.
interface Vlan50
ip address 10.5.0.2 255.255.255.252
ip pim sparse-mode
no shutdown
Step 6:  On the distribution-layer switch, configure EtherChannel member interfaces.
Connect the router EtherChannel uplinks to separate switches in the distribution layer.
If you are using a Cisco Catalyst 4507R+E chassis in the distribution layer, connect the uplinks to separate
redundant modules. This provides additional resiliency.
The physical interfaces that are members of a Layer 2 EtherChannel are configured prior to configuring the
logical port-channel interface. Doing the configuration in this order allows for minimal configuration and reduces
errors because most of the commands entered to a port-channel interface are copied to its member interfaces
and do not require manual replication.
Configure two or more physical interfaces to be members of the EtherChannel. It is recommended that they
are added in multiples of two. Also, apply the egress QoS macro that was defined in the platform configuration
procedure to ensure traffic is prioritized appropriately.
Not all connected router platforms can support LACP to negotiate with the switch, so you configure EtherChannel
statically.
interface GigabitEthernet1/0/1
description Link to RS200-3925-1 Gig0/1
interface GigabitEthernet2/0/1
description Link to RS200-3925-1 Gig0/2
!
interface range GigabitEthernet1/0/1, GigabitEthernet2/0/1
switchport
channel-group 1 mode on
logging event link-status
logging event trunk-status
logging event bundle-status
load-interval 30
macro apply EgressQoS

Deploying a WAN Remote-Site Distribution Layer

August 2014 Series

76

Step 7:  On the distribution-layer switch. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface default passive-interface exit-af-interface af-interface [Routed link interface] no passive-interface exit-af-interface network [network] [inverse mask] eigrp router-id [IP address of Loopback0] exit-address-family Step 2:  Configure EIGRP neighbor authentication. key chain LAN-KEY key 1 key-string cisco ! router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface [interface type] [number] authentication mode md5 authentication key-chain LAN-KEY Deploying a WAN Remote-Site Distribution Layer August 2014 Series 77 . Neighbor authentication enables the secure establishment of peering adjacencies and exchange route tables over the DMVPN tunnel interface. In this design. Step 1:  On the router. This allows the router to provide the Layer 3 services to all the VLANs defined on the distribution-layer switch. the interface type is port-channel. When using EtherChannel. . interface Port-channel1 description EtherChannel link to RS200-3925-1 switchport trunk allowed vlan 50 switchport mode trunk spanning-tree portfast trunk no shutdown The Cisco Catalyst 3750 Series Switch requires the switchport trunk encapsulation dot1q command.1Q trunk for the connection. all distribution-layer-facing subinterfaces and the loopback must be EIGRP interfaces. Procedure 2 Configure EIGRP (LAN side) You must configure a routing protocol between the router and distribution layer. The network range must include all interface IP addresses either in a single network statement or in multiple network statements. and the number must match the channel group configured in Step 3. All other interfaces should remain passive. configure the EIGRP LAN process facing the distribution layer. configure an EtherChannel trunk. This design uses a best practice of assigning the router ID to a loopback address. Use an 802. Prune the VLANs allowed on the trunk to only the VLANs that are active on the distribution-layer switch.

redistribute BGP into the EIGRP LAN process. enter the following configuration. router eigrp LAN address-family ipv4 unicast autonomous-system 100 topology base default-metric [WAN bandwidth] [WAN delay] 255 1 1500 redistribute bgp 65511 exit-af-topology exit-address-family On the distribution-layer switch VLAN interface.no passive-interface exit-af-interface exit-address-family Step 3:  On the router. If the additional IP networks are outside the existing remote-site summary range. router bgp 65511 network [network and mask] Deploying a WAN Remote-Site Distribution Layer August 2014 Series 78 . A default metric redistributes the BGP routes into EIGRP. advertise the new IP network in BGP. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface Vlan50 summary-address [Summary address] exit-af-interface exit-address-family Step 5:  On the router. EIGRP is already configured on the distribution-layer switch. only the WAN bandwidth and delay values are used for metric calculation. Step 6:  Configure a BGP network statement with a mask matching the distribution-layer EIGRP summary route. The VLAN interface that connects to the router must be configured for EIGRP neighbor authentication and as a non-passive EIGRP interface. enable EIGRP. you will need to add an EIGRP summary on the distribution switch. By default. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface Vlan50 authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface exit-address-family Step 4:  If it is necessary to define additional IP networks on the distribution-layer switch.

24.255.128 ! router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface default passive-interface exit-af-interface af-interface Vlan50 summary-address 10.0.4.255.255.254 eigrp stub connected summary redistributed nsf exit-address-family Deploying a WAN Remote-Site Distribution Layer August 2014 Series 79 .0 0.255.0.0 mask 255.1 255.255.0 authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface topology base exit-af-topology network 10.7.0 0.255 eigrp router-id 10.255.251.5.0.0 255.5.255.0 0.255.5.248.255.5.0 ! router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface default passive-interface exit-af-interface af-interface Port-channel1.5.0.24.0.Example (router configuration) router bgp 65511 network 10.200 exit-address-family Example (distribution switch configuration) interface Vlan153 description Server Room RS-200 ip address 10.1.26.255 eigrp router-id 10.248.50 authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface topology base default-metric 100000 100 255 1 1500 redistribute bgp 65511 exit-af-topology network 10.255 network 10.

vlan 99 name Transit-net Step 4:  Add the transit network VLAN to the existing distribution-layer switch EtherChannel trunk.Procedure 3 Configure the transit network If you are using a dual-router design. Step 1:  Configure iBGP between the remote-site MPLS CE routers. Configure the transit network between the two routers. complete this procedure. complete this procedure. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface Port-channel1. enable EIGRP on the transit network interface. The transit network must be a non-passive EIGRP interface. Deploying a WAN Remote-Site Distribution Layer August 2014 Series 80 . this configuration is considered an internal BGP (iBGP) connection. Step 1:  On the router. The transit network uses Layer 2 pass-through on the distribution-layer switch. You use this network for router-router communication and to avoid hairpinning.99 authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface exit-address-family Step 3:  On the distribution-layer switch. There are no end stations connected to this network.255. so no SVI is required.5. interface Port-channel1 switchport trunk allowed vlan add 99 Procedure 4 Configure BGP If you are using a dual-router design. This design uses iBGP peering and requires the next-hop-self-configuration option.9 255. Because the CE routers are using the same ASN.255. configure the transit network VLAN.252 ip pim sparse-mode Step 2:  On the router. configure the transit net interface.99 description Transit Net encapsulation dot1Q 99 ip address 10. The dual-carrier MPLS design requires that a BGP link is configured between the CE routers. The transit network should use an additional subinterface on the EtherChannel interface that is already used to connect to the distribution layer. interface Port-channel1.0. so HSRP and DHCP are not required.

BGP readvertises all BGP learned routes. providing access between the two carriers. with a high bandwidth connection. router bgp 65511 neighbor [iBGP neighbor Transit Net IP] remote-as 65511 neighbor [iBGP neighbor Transit Net IP] next-hop-self Step 2:  Configure BGP to prevent the remote site from becoming a transit AS. This type of filter allows for only the locally originated routes to be advertised. You must use a route-map and an as-path access-list filter. the iBGP session will not be established until you complete the transit network and EIGRP (LAN side) steps. Accomplishing the design goal of deterministic routing and primary/secondary routing behavior necessitates tuning BGP. This requires the use of a route-map and an as-path access-list filter. this means that MPLS-A routes are advertised to MPLS-B and vice-versa. ip as-path access-list 1 permit _65401$ ! route-map PREFER-MPLS-A permit 10 match as-path 1 set local-preference 200 ! route-map PREFER-MPLS-A permit 20 Deploying a WAN Remote-Site Distribution Layer August 2014 Series 81 . when a link to an MPLS hub has failed. The MPLS dual-carrier design in many cases provides two equal cost paths. You need to apply this route-map on both remote-site MPLS CE routers. router bgp 65511 neighbor [IP address of PE] route-map PREFER-MPLS-A in Step 4:  Apply a route-map inbound to the neighbor for the primary MPLS carrier only. In certain cases. Note. Unless the remote site has been specifically designed for this type of routing behavior. By default. In the dual-MPLS design. Each router applies this outbound to the neighbor for its respective MPLS carrier. remote sites advertise themselves as a transit autonomous system. BGP uses a well-known rule set in order to determine the “best path” when the same IP route prefix is reachable via two different paths. Step 3:  Tune BGP routing to prefer the primary MPLS carrier.You must complete this step on both remote-site MPLS CE routers. it is a best practice to disable the site from becoming a transit site. router bgp 65511 neighbor [IP address of PE] route-map NO-TRANSIT-AS out ip as-path access-list 10 permit ^$ ! route-map NO-TRANSIT-AS permit 10 match as-path 10 Tech Tip The regular expression ^$ corresponds to routes originated from the remote-site. and it is likely that the first path selected will remain the active path unless the routing protocol detects a failure.

Routes originated from the secondary MPLS carrier continue to use their default local preference of 100. Figure 18 .255.Router 1 Link VLAN 54 . Step 5:  Add a loopback network for the secondary router. PROCESS router bgp 65511 network [Secondary router loopback network] mask 255. 99) 802. the BGP local preference is 200 for the primary MPLS carrier. zz) 2132 802.1Q Trunk (yy.255 Connecting the Secondary Remote-Site Router to the Distribution Layer 1.Router 2 Link VLAN 99 . xx) R2 Deploying a WAN Remote-Site Distribution Layer August 2014 Series 82 . distribution layer-remote-site design is shown in the following figure.Tech Tip The regular expression _65401$ corresponds to routes originated from the AS 65401 (MPLS-A). complete this process. dual-link design.1Q Trunk (ww.WAN remote site—Connection to distribution layer WAN R1 VLAN 50 . Configure EIGRP (LAN side) If you are using dual-carrier design for the MPLS WAN remote site. In this example. This allows BGP to selectively modify the routing information for routes originated from this AS.1Q Trunk (54. The dual-router. Connect router to distribution layer 2.255.Transit 802. This process connects the distribution layer to the second router of the dual-router. 99) 802.1Q Trunk (50. This design uses a separate routed link from the second router of the dual-router scenario to the LAN distribution-layer switch.

54 description R2 routed link to distribution layer encapsulation dot1Q 54 ip address 10. map the appropriate subinterfaces to the VLANs on the distributionlayer switch. After you have enabled the interface. so you configure EtherChannel statically. Traffic is routed between the devices with the VLAN acting as a point-to-point link. interface Port-channel2.1Q tag.255. configure a port-channel interface. interface GigabitEthernet0/1 description RS200-D4500X-VSS Ten1/1/2 ! interface GigabitEthernet0/2 description RS200-D4500X-VSS Ten2/1/2 ! Deploying a WAN Remote-Site Distribution Layer August 2014 Series 83 .255. This guide only includes the additional steps to complete the distribution layer configuration. interface Port-channel2. The subinterface number does not need to equate to the 802. Step 1:  On the secondary router. configure the transit network interface. configure the EtherChannel member interfaces. This connection allows for multiple VLANs to be included on the EtherChannel as necessary. interface Port-channel2 description EtherChannel link to D4500X-VSS no shutdown Step 2:  Configure the port channel subinterfaces and assign IP address.252 ip pim sparse-mode Step 3:  On the router. The subinterface configured on the router corresponds to a VLAN interface on the distribution-layer switch.5. Not all router platforms can support LACP to negotiate with the switch.99 description Transit Net encapsulation dot1Q 99 ip address 10.5. but making them the same simplifies the overall configuration.Procedure 1 Connect router to distribution layer Reader Tip Please refer to the Campus Wired LAN Technology Design Guide for complete distribution layer configuration details.255.10 255.252 ip pim sparse-mode Step 4:  On the router.5 255. The number for the port-channel and channel-group must match.255.0.0. Configure the physical interfaces to tie to the logical port-channel using by the channel-group command. Layer 2 EtherChannels are used to interconnect the CE router to the distribution layer in the most resilient method possible.

GigabitEthernet0/2 no ip address channel-group 2 no shutdown Step 5:  On the distribution-layer switch. Doing the configuration in this order allows for minimal configuration and reduces errors because most of the commands entered to a port-channel interface are copied to its member interfaces and do not require manual replication.6 255. GigabitEthernet2/0/2 switchport channel-group 2 mode on logging event link-status logging event trunk-status logging event bundle-status load-interval 30 macro apply EgressQoS Deploying a WAN Remote-Site Distribution Layer August 2014 Series 84 . configure EtherChannel member interfaces. It is recommended that they are added in multiples of two. interface GigabitEthernet1/0/2 description Link to RS200-3925-2 Gig0/1 interface GigabitEthernet2/0/2 description Link to RS200-3925-2 Gig0/2 ! interface range GigabitEthernet1/0/2. so you configure EtherChannel statically.252 ip pim sparse-mode no shutdown Step 7:  On the distribution-layer switch. configure Layer 3. apply the egress QoS macro that was defined in the platform configuration procedure to ensure traffic is prioritized appropriately. for the new VLAN added.5. interface Vlan54 ip address 10.255. Configure a VLAN interface. Connect the router EtherChannel uplinks to separate switches in the distribution layer switches or stack. The SVI is used for point-to-point IP routing between the distribution layer and the WAN router. Also.0. to separate redundant modules for additional resiliency. and in the case of the Cisco Catalyst 4507R+E distribution layer. also known as a switch virtual interface (SVI).255. vlan 54 name R2-link Step 6:  On the distribution-layer switch. Not all connected router platforms can support LACP to negotiate with the switch. The physical interfaces that are members of a Layer 2 EtherChannel are configured prior to configuring the logical port-channel interface. Configure two or more physical interfaces to be members of the EtherChannel.interface range GigabitEthernet0/1. configure a VLAN.

The network range must include all interface IP addresses either in a single network statement or in multiple network statements. In this design. Use an 802.99 switchport mode trunk spanning-tree portfast trunk no shutdown The Cisco Catalyst 3750 Series Switch requires the switchport trunk encapsulation dot1q command. configure the EIGRP LAN process facing the distribution layer. the interface type is port-channel. interface Port-channel2 description EtherChannel link to RS200-3925-2 switchport trunk allowed vlan 54. When using EtherChannel. configure an EtherChannel trunk. Step 1:  On the router. all distribution-layer-facing subinterfaces and the loopback must be EIGRP interfaces.Step 8:  On the distribution-layer switch.1Q trunk for the connection. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface default passive-interface exit-af-interface af-interface [Routed link interface] no passive-interface exit-af-interface af-interface [Transit link interface] no passive-interface exit-af-interface network [network] [inverse mask] eigrp router-id [IP address of Loopback0] exit-address-family Deploying a WAN Remote-Site Distribution Layer August 2014 Series 85 . and the number must match the channel group configured in Step 4. Procedure 2 Configure EIGRP (LAN side) You must configure a routing protocol between the router and distribution layer. This design uses a best practice of assigning the router ID to a loopback address. Prune the VLANs allowed on the trunk to only the VLANs that are active on the distribution-layer switch. All other interfaces should remain passive. This allows the router to provide the Layer 3 services to all the VLANs defined on the distribution-layer switch.

The VLAN interface that connects to the router must be configured as a non-passive EIGRP interface. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface Vlan54 authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface exit-address-family Step 5:  If it is necessary to define additional IP networks on the distribution-layer switch. enable EIGRP. enter the following configuration. you will need to add an EIGRP summary on the distribution switch. key chain LAN-KEY key 1 key-string cisco ! router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface [interface type] [number] authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface exit-address-family Step 3:  On the router. only the WAN bandwidth and delay values are used for metric calculation.Step 2:  Configure EIGRP neighbor authentication. By default. EIGRP is already configured on the distribution-layer switch. Neighbor authentication enables the secure establishment of peering adjacencies and exchange route tables over the DMVPN tunnel interface. redistribute BGP into the EIGRP LAN process. If the additional IP networks are outside the existing remote-site summary range. router eigrp LAN address-family ipv4 unicast autonomous-system 100 topology base default-metric [WAN bandwidth] [WAN delay] 255 1 1500 redistribute bgp 65511 exit-af-topology exit-address-family Step 4:  On the distribution-layer switch VLAN interface. A default metric redistributes the BGP routes into EIGRP. router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface Vlan54 summary-address [New Summary address] exit-af-interface exit-address-family Deploying a WAN Remote-Site Distribution Layer August 2014 Series 86 .

0.0 0.24. Configure a BGP network statement with a mask matching the distribution-layer EIGRP summary.5.255 network 10.255.252 ip pim sparse-mode ! router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface default passive-interface exit-af-interface af-interface Vlan50 summary-address 10.5.248.200 exit-address-family Example (Distribution switch configuration) interface Vlan54 ip address 10.252.0 ! router eigrp LAN address-family ipv4 unicast autonomous-system 100 af-interface default passive-interface exit-af-interface af-interface Port-channel1.255.0 0.255.54 authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface af-interface Port-channel1.248.255.255 eigrp router-id 10.0 mask 255.255.0.255.0.5.255.0 255.0.0 authentication mode md5 authentication key-chain LAN-KEY Deploying a WAN Remote-Site Distribution Layer August 2014 Series 87 .Step 6:  On the router.24.5.0.99 authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface topology base default-metric 500000 100 255 1 1500 redistribute bgp 65511 exit-af-topology network 10.255.6 255. router bgp 65511 network [network and mask] Example (router configuration) router bgp 65511 network 10. advertise the new IP network in BGP.

255.0 0.0 authentication mode md5 authentication key-chain LAN-KEY no passive-interface exit-af-interface topology base exit-af-topology network 10.1.5.255.4.0 255.7.no passive-interface exit-af-interface af-interface Vlan54 summary-address 10.24.254 eigrp stub connected summary redistributed nsf exit-address-family Deploying a WAN Remote-Site Distribution Layer August 2014 Series 88 .0.255 eigrp router-id 10.248.5.

After you have configured the class-map command.Deploying WAN Quality of Service When configuring the WAN-edge QoS. Apply the shaping and queuing policy to a physical interface Procedure 1 Create the QoS Maps to Classify Traffic The class-map command defines a traffic class and identifies traffic to associate with the class name. Configure shaping and queuing policy 5. class-map match-any [class-map name] match dscp [dcsp value] [optional additional dscp value(s)] Deploying WAN Quality of Service August 2014 Series 89 . Use the following steps to configure the required WAN class maps and matching criteria. Step 1:  For each of the six WAN classes of service listed in Table 13. You use the following two forms of the match command: match dscp and match protocol. In this case. Define a policy map that defines the queuing policy 4. Create the QoS Maps to Classify Traffic 2. such as DSCP and protocols. This keyword is followed by the name you want to assign to the class of service. you are defining how traffic egresses your network. the match-any keyword indicates that the maps match any of the specified criteria. and bandwidth allocations align to the service provider offering to ensure consistent QoS treatment end to end. create a class map for DSCP matching. These class names are used when configuring policy maps that define actions you wish to take against the traffic type. The class-map command sets the match logic. Deployment Details PROCESS Configuring QoS 1. You do not need to explicitly configure the default class. marking. to match with the match command. It is critical that the classification. Create the policy map that marks BGP traffic 3. you define specific values.

operations. Citrix. create a class map for BGP protocol matching. Use NBAR to match BGP by protocol. administration and maintenance (OAM) traffic. cs2 3 — default Best effort other 25 random Step 2:  If you are using a WAN-aggregation MPLS CE router or a WAN remote-site MPLS CE router that is using BGP.QoS classes of service Class of service Traffic type DSCP values Bandwidth % Congestion avoidance VOICE Voice traffic ef 10 (PQ) — INTERACTIVEVIDEO Interactive video (such as video conferencing) cs4. af41 23 (PQ) — CRITICAL-DATA Highly interactive (such as Telnet. class-map match-any [class-map name] match ip protocol [protocol name] Example class-map match-any VOICE match dscp ef ! class-map match-any INTERACTIVE-VIDEO match dscp cs4 af41 ! class-map match-any CRITICAL-DATA match dscp af31 cs3 ! class-map match-any DATA match dscp af21 ! class-map match-any SCAVENGER match dscp af11 cs1 ! class-map match-any NETWORK-CRITICAL match dscp cs6 cs2 ! class-map match-any BGP-ROUTING match protocol bgp Deploying WAN Quality of Service August 2014 Series 90 . BGP traffic is not explicitly tagged with a DSCP value. cs3 15 DSCP-based DATA Data af21 19 DSCP-based SCAVENGER Scavenger af11.Table 13 . cs6. cs1 5 — NETWORKCRITICAL Routing protocols. and Oracle thin clients) af31.

This specification is accomplished with the use of a policy map. Although the class map you created in the previous step matches all BGP traffic to the class named BGP. marking. One additional default class defines the minimum allowed bandwidth available for best-effort traffic. policy-map [policy-map-name] Step 2:  Apply the previously created class map. priority percent [percentage] Deploying WAN Quality of Service August 2014 Series 91 . complete this procedure. and queuing of network-critical traffic on egress to the WAN. each class within the policy map invokes an egress queue. Most providers perform this remapping by matching on DSCP values cs6 and cs2. To ensure proper treatment of BGP routing traffic in the WAN. The NETWORK-CRITICAL policy map is defined in order to ensure the correct classification. This is implicitly included within classdefault as shown in Procedure 4. you must configure a policy map to assign the required DSCP value to all BGP traffic. Then. policy-map MARK-BGP class BGP-ROUTING set dscp cs6 Procedure 3 Define a policy map that defines the queuing policy This procedure applies to all WAN routers.Tech Tip You do not need to configure a best-effort class. The WAN policy map references the class names you created in the previous procedures and defines the queuing behavior along with the maximum guaranteed bandwidth allocated to each class. you must assign a DSCP value of cs6. the network-critical traffic is typically remapped by the service provider into the critical data class. and then assign it a DSCP value of cs6. Step 1:  Create a policy map. class [class-name] Step 3:  (Optional) Assign the maximum guaranteed bandwidth for the class. After the traffic has been transmitted to the service provider. Step 1:  Create the parent policy map. Procedure 2 Create the policy map that marks BGP traffic If you are using a WAN-aggregation MPLS CE router or a WAN remote-site MPLS CE router that uses BGP. and associates a specific traffic class to that queue. The local router policy maps define seven classes while most service providers offer only six classes of service. bandwidth percent [percentage] Step 4:  (Optional) Define the priority queue for the class. assigns a percentage of bandwidth.

service-policy [policy-map-name] Step 6:  (Optional) Define the congestion mechanism. including class-default. Deploying WAN Quality of Service August 2014 Series 92 .Step 5:  (Optional) Apply the child service policy. Example policy-map WAN class VOICE priority percent 10 class INTERACTIVE-VIDEO priority percent 23 class CRITICAL-DATA bandwidth percent 15 random-detect dscp-based class DATA bandwidth percent 19 random-detect dscp-based class SCAVENGER bandwidth percent 5 class NETWORK-CRITICAL bandwidth percent 3 service-policy MARK-BGP class class-default bandwidth percent 25 random-detect Tech Tip Although these bandwidth assignments represent a good baseline. it is important to consider your actual traffic requirements per class and adjust the bandwidth settings accordingly. This is an optional step only for the NETWORK-CRITICAL class of service with the MARK-BGP child service policy. random-detect [type] Step 7:  Repeat Step 2 through Step 6 for each class in Table 13.

policy-map [policy-map-name] Step 2:  Configure the shaper. To ensure the offered load to the service provider does not exceed the contracted rate that results in the carrier discarding traffic. ensure that the value matches the contracted bandwidth rate from your service provider. service-policy [policy-map-name] Example This example shows a router with a 20-Mbps link on interface GigabitEthernet0/0 and a 10-Mbps link on interface GigabitEthernet0/1. You can repeat this procedure multiple times to support devices that have multiple WAN connections attached to different interfaces. You configure a QoS service policy on the outside Ethernet interface. Instead. This is called a hierarchical Class-Based Weighted Fair Queuing (HCBWFQ) configuration.Procedure 4 Configure shaping and queuing policy With WAN interfaces using Ethernet as an access technology. and this parent policy includes a shaper that then references a second or subordinate (child) policy that enables queuing within the shaped rate. the demarcation point between the enterprise and service provider may no longer have a physical-interface bandwidth constraint. Step 1:  Create the parent policy map. embed the interface name within the name of the parent policy map. a specified amount of access bandwidth is contracted with the service provider. class [class-name] shape [average | peak] [bandwidth (bps)] Step 3:  Apply the child service policy. you need to configure shaping on the physical interface. As a best practice. When you configure the shape average command. This shaping is accomplished with a QoS service policy. policy-map WAN-INTERFACE-G0/0 class class-default shape average 20000000 service-policy WAN ! policy-map WAN-INTERFACE-G0/1 class class-default shape average 10000000 service-policy WAN Deploying WAN Quality of Service August 2014 Series 93 . This procedure applies to all WAN routers.

you must apply the parent policy that you configured in the previous procedure. Step 1:  Select the WAN interface. You can repeat this procedure multiple times to support devices that have multiple WAN connections attached to different interfaces.Procedure 5 Apply the shaping and queuing policy to a physical interface To invoke shaping and queuing on a physical interface. This procedure applies to all WAN routers. interface [interface type] [number] Step 2:  Apply the WAN QoS policy in the outbound direction. service-policy output [policy-map-name] Example interface GigabitEthernet0/0 service-policy output WAN-INTERFACE-G0/0 ! interface GigabitEthernet0/1 service-policy output WAN-INTERFACE-G0/1 Deploying WAN Quality of Service August 2014 Series 94 .

4 EHWIC. 4SM. SEC. AVC and WAAS-Express C1941-AX/K9 15. 1 SM. AX license with. AVC.5G-VPNK9 IOS-XE 15. AX license with. 3 DSP. AX licenses with. ISR-WAAS with 2500 connection RTU ISR4451-X-AX/K9 IOS-XE 15. 1 SM. DATA. SEC. 256MB CF. IP Base.Appendix A: Product List WAN Aggregation Functional Area Product Description Part Numbers Software WAN-aggregation Router Aggregation Services 1002X Router ASR1002X-5G-VPNK9 Aggregation Services 1002 Router ASR1002-5G-VPN/K9 Aggregation Services 1001 Router ASR1001-2. and WAAS/vWAAS with 1300 connection RTU C2921-AX/K9 Cisco ISR 2911 w/ 3 GE. 4EHWIC. 3GE. 256MB CF. DATA. 2SM. DATA.2SM. 4 EHWIC.3(3)M3 securityk9 feature set datak9 feature set uck9 feature set Unified Communications Paper PAK for Cisco 3900 Series SL-39-UC-K9 Cisco ISR 2951 w/ 3 GE.4(2)S Advanced Enterprise feature set Cisco ISR 4451-X Security Bundle w/SEC license PAK ISR4451-X-SEC/K9 IOS-XE 15. SEC.8G FLASH. AVC. 1GB DRAM. AX license with. 1GBDRAM. 4DSP. 4G DRAM. DATA. 1GBDRAM. IP Base.4 EHWIC. SEC. and WAAS/vWAAS with 1300 connection RTU C2951-AX/K9 Cisco ISR 2921 w/ 3 GE. 256MBCF. 256MB CF. AVC. and WAAS/vWAAS with 2500 connection RTU C3945-AX/K9 Cisco ISR 3925 w/ SPE100 (3GE. 3 DSP. SEC. AXlicenses with. 256MBCF. DATA.3NIM. 2 DSP. IP Base. 2 EHWIC slots. 256MB CF. SEC. AVC. AX license with: DATA. IP Base.4(2)S securityk9 feature set WAN Remote Site Functional Area Product Description Part Numbers Software Modular WAN Remote-site Router Cisco ISR 4451 w/ 4GE. WAAS/vWAAS with 2500 connection RTU C3925-AX/K9 15. AX license with. 2. IP Base. 1GB DRAM. DATA. 4EHWIC.3(3)M3 securityk9 feature set datak9 feature set Cisco 891 Router CISCO891W-AGN-A-K9 WAAS-Express/AVC/Advanced IP with upgrade up to 1GB DRAM for Cisco 800 Series Routers FL-C800-APP 15. 1GB DRAM. SEC. AVC and WAAS/vWAAS with 1300 connection RTU C2911-AX/K9 Unified Communications Paper PAK for Cisco 2900 Series SL-29-UC-K9 Cisco ISR 1941 Router w/ 2 GE. IP Base. 4DSP.5GB DRAM.4(2)S securityk9 feature set appxk9 feature set Cisco ISR 3945 w/ SPE150. IP Base.3(3)M3 securityk9 feature set datak9 feature set Fixed WAN Remote-site Router Appendix A: Product List August 2014 Series 95 . AVC. 2 SM.

3. 520Gbps WS-X45-SUP7L-E Cisco Catalyst 4500E 48 Ethernet 10/100/1000 (RJ45) PoE+.01EZ3) IP Base feature set August 2014 Series 96 .0.3at PoE+ 10/100/1000 (RJ-45) WS-X4748-RJ45V+E Cisco Catalyst 4500E Series 4507R+E 7-slot Chassis with 48Gbps per slot WS-C4507R+E Cisco Catalyst 4500E Supervisor Engine 7L-E.1EZ3) IP Base feature set 3.UPoE ports WS-X4748-UPOE+E Cisco Catalyst 4500E 48 Ethernet 10/100/1000 (RJ45) PoE+ ports WS-X4648-RJ45V+E Cisco Catalyst 3850 Series Stackable 48 Ethernet 10/100/1000 PoE+ ports WS-C3850-48F Cisco Catalyst 3850 Series Stackable 24 Ethernet 10/100/1000 PoE+ Ports WS-C3850-24P Cisco Catalyst 3850 Series 2 x 10GE Network Module C3850-NM-2-10G Cisco Catalyst 3850 Series 4 x 1GE Network Module C3850-NM-4-1G Cisco Catalyst 3650 Series 24 Ethernet 10/100/1000 PoE+ and 2x10GE or 4x1GE Uplink WS-C3650-24PD Cisco Catalyst 3650 Series 24 Ethernet 10/100/1000 PoE+ and 4x1GE Uplink WS-C3650-24PS Cisco Catalyst 3650 Series Stack Module C3650-STACK Cisco Catalyst 3750-X Series Stackable 48 Ethernet 10/100/1000 PoE+ ports WS-C3750X-48PF-S Cisco Catalyst 3750-X Series Stackable 24 Ethernet 10/100/1000 PoE+ ports WS-C3750X-24P-S Cisco Catalyst 3750-X Series Two 10GbE SFP+ and Two GbE SFP ports network module C3KX-NM-10G Cisco Catalyst 3750-X Series Four GbE SFP ports network module C3KX-NM-1G Cisco Catalyst 2960-X Series 24 10/100/1000 Ethernet and 2 SFP+ Uplink WS-C2960X-24PD Cisco Catalyst 2960-X FlexStack-Plus Hot-Swappable Stacking Module C2960X-STACK Cisco Catalyst 3650 Series 24 Ethernet 10/100/1000 PoE+ and 4x1GE Uplink WS-C3650-24PS Stackable Access Layer Switch Standalone Access Layer Switch Appendix A: Product List 3.2.2(1)E3 IP Base feature set 15.3.3.3.3E(15.0.5.3SE(15.0(2)EX5 LAN Base feature set 3.1EZ3) IP Base feature set 15.LAN Access Layer Functional Area Product Description Part Numbers Software Modular Access Layer Switch Cisco Catalyst 4500E Series 4507R+E 7-slot Chassis with 48Gbps per slot WS-C4507R+E 3. 928Gbps WS-X45-SUP8-E Cisco Catalyst 4500E 12-port 10GbE SFP+ Fiber Module WS-X4712-SFP+E Cisco Catalyst 4500E 48-Port 802.3SE(15.1.3SE(15.1E3) IP Base feature set 3.1XO1) IP Base feature set Cisco Catalyst 4500E Supervisor Engine 8-E. Unified Access.1XO(15.

LAN Distribution Layer Functional Area Product Description Part Numbers Software Modular Distribution Layer Virtual Switch Pair Cisco Catalyst 6800 Series 6807-XL 7-Slot Modular Chassis C6807-XL Cisco Catalyst 6500 VSS Supervisor 2T with 2 ports 10GbE and PFC4 VS-S2T-10G 15.3.3at PoE+ 10/100/1000 (RJ-45) WS-X4748-RJ45V+E Fixed Distribution Layer Virtual Switch Pair Cisco Catalyst 4500-X Series 32 Port 10GbE IP Base Frontto-Back Cooling WS-C4500X-32SFP+ 3.3E(15.3SE(15.1(2)SY3 IP Services feature set 3.5.0.3E(15.5.1E3) Enterprise Services feature set 15.1E3) Enterprise Services feature set Stackable Distribution Layer Switch Cisco Catalyst 3850 Series Stackable Switch with 12 SFP Ethernet WS-C3850-12S 3.1(2)SY3 IP Services feature set Cisco Catalyst 6500 4-port 40GbE/16-port 10GbE Fiber Module w/DFC4 WS-X6904-40G-2T Cisco Catalyst 6500 4-port 10GbE SFP+ adapter for WX-X6904-40G module CVR-CFP-4SFP10G Cisco Catalyst 6500 CEF720 48 port 10/100/1000mb Ethernet WS-X6748-GE-TX Cisco Catalyst 6500 Distributed Forwarding Card 4 WS-F6K-DFC4-A Cisco Catalyst 6500 Series 6506-E 6-Slot Chassis WS-C6506-E Cisco Catalyst 6500 VSS Supervisor 2T with 2 ports 10GbE and PFC4 VS-S2T-10G Cisco Catalyst 6500 4-port 40GbE/16-port 10GbE Fiber Module w/DFC4 WS-X6904-40G-2T Cisco Catalyst 6500 4-port 10GbE SFP+ adapter for WX-X6904-40G module CVR-CFP-4SFP10G Cisco Catalyst 6500 48-port GigE Mod (SFP) WS-X6748-SFP Cisco Catalyst 6500 Distributed Forwarding Card 4 WS-F6K-DFC4-A Cisco Catalyst 6500 24-port GigE Mod (SFP) WS-X6724-SFP Cisco Catalyst 6500 Distributed Forwarding Card 4 WS-F6K-DFC4-A Extensible Fixed Distribution Layer Virtual Switch Pair Cisco Catalyst 6800 Series 6880-X Extensible Fixed Aggregation Switch (Standard Tables) C6880-X-LE Cisco Catalyst 6800 Series 6880-X Multi Rate Port Card (Standard Tables) C6880-X-LE-16P10G Modular Distribution Layer Virtual Switch Pair Cisco Catalyst 4500E Series 4507R+E 7-slot Chassis with 48Gbps per slot WS-C4507R+E Cisco Catalyst 4500E Supervisor Engine 7-E. 848Gbps WS-X45-SUP7-E Cisco Catalyst 4500E 12-port 10GbE SFP+ Fiber Module WS-X4712-SFP+E Cisco Catalyst 4500E 48-Port 802.2.2.1EZ3) IP Services feature set Cisco Catalyst 3850 Series 4 x 1GE Network Module C3850-NM-4-1G Cisco Catalyst 3850 Series 2 x 10GE Network Module C3850-NM-2-10G Cisco Catalyst 3750-X Series Stackable 12 GbE SFP ports WS-C3750X-12S-E Cisco Catalyst 3750-X Series Two 10GbE SFP+ and Two GbE SFP ports network module C3KX-NM-10G Cisco Catalyst 3750-X Series Four GbE SFP ports network module C3KX-NM-1G Appendix A: Product List 15.2(1)E3 IP Services feature set August 2014 Series 97 .

Appendix B: Device Configuration Files To view the configuration files from the CVD lab devices that we used to test this guide.com/fw/240-14b Appendix B: Device Configuration Files August 2014 Series 98 . please go to the following URL: http://cvddocs.

• We added EIGRP Named mode configurations.Appendix C: Changes This appendix summarizes the changes Cisco made to this guide since its last edition. • We added EIGRP Authentication configurations. • We added the ip scp server enable command. Appendix C: Changes August 2014 Series 99 .

Any examples.Feedback Please use the feedback form to send comments and suggestions about this guide. Inc. WITHOUT LIMITATION. “DESIGNS”) IN THIS MANUAL ARE PRESENTED “AS IS. USERS ARE SOLELY RESPONSIBLE FOR THEIR APPLICATION OF THE DESIGNS. WITHOUT LIMITATION. CONSEQUENTIAL. San Jose. CISCO AND ITS SUPPLIERS DISCLAIM ALL WARRANTIES. To view a list of Cisco trademarks.com/go/offices. THE DESIGNS DO NOT CONSTITUTE THE TECHNICAL OR OTHER PROFESSIONAL ADVICE OF CISCO. LOST PROFITS OR LOSS OR DAMAGE TO DATA ARISING OUT OF THE USE OR INABILITY TO USE THE DESIGNS. USERS SHOULD CONSULT THEIR OWN TECHNICAL ADVISORS BEFORE IMPLEMENTING THE DESIGNS. STATEMENTS. THE DESIGNS ARE SUBJECT TO CHANGE WITHOUT NOTICE. FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT OR ARISING FROM A COURSE OF DEALING. The use of the word partner does not imply a partnership relationship between Cisco and any other company. IN NO EVENT SHALL CISCO OR ITS SUPPLIERS BE LIABLE FOR ANY INDIRECT. (1110R) B-0000245-1 08/14 . Cisco and the Cisco logo are trademarks or registered trademarks of Cisco and/or its affiliates in the U. and figures included in the document are shown for illustrative purposes only. OR TRADE PRACTICE. Addresses. and other countries. CA Asia Pacific Headquarters Cisco Systems (USA) Pte. Any Internet Protocol (IP) addresses used in this document are not intended to be actual addresses. THE WARRANTY OF MERCHANTABILITY. go to this URL: www. Third-party trademarks mentioned are the property of their respective owners. SPECIAL. Inc. Any use of actual IP addresses in illustrative content is unintentional and coincidental. RESULTS MAY VARY DEPENDING ON FACTORS NOT TESTED BY CISCO.” WITH ALL FAULTS. Ltd. OR INCIDENTAL DAMAGES. INCLUDING.cisco. © 2014 Cisco Systems. The Netherlands Cisco has more than 200 offices worldwide. INCLUDING. USAGE.cisco. and fax numbers are listed on the Cisco Website at www. All rights reserved. command display output. Singapore Europe Headquarters Cisco Systems International BV Amsterdam. INFORMATION.com/go/trademarks. AND RECOMMENDATIONS (COLLECTIVELY. ITS SUPPLIERS OR PARTNERS. phone numbers. Americas Headquarters Cisco Systems.S. EVEN IF CISCO OR ITS SUPPLIERS HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. SPECIFICATIONS. ALL DESIGNS.