Table Of Contents
Technology Overview
Introduction
Structure of an ISP Network
Characteristics of Access and Backbone Routers
POP Design
Driving for Density
Leased Lines—the Last Mile
Cisco 10000 Series Router: Specialized for Aggregation
Cisco 10000 Series Router Hardware View
High-Touch Feature Set
High Availability
Conclusion
Technology Overview
Introduction
The increasing need for Internet connectivity is placing huge demands on Internet service providers—not just in the number of connections that users require for Internet access, but also in the services that these users require on each connection. The Internet's explosive growth is driving the requirements for higher quality, faster connectivity, and more software features for an ever-growing number of customers.
One of the largest and fastest growing components of Internet access growth is business-to-business leased line connectivity. Leased line traffic is defined as DSO, T1/E1, and T3/E3 up to leased line OC-3. Independent analysts forecast this leased line market as a $30+ billion dollar business today, growing to $45 billion in 2004, as Figure 1-1 illustrates.
Figure 1-1 Leased Line Traffic
Aggregation of leased line traffic takes place at the "edge" of the network. The edge, the point at which the customer's enterprise network intersects with the ISP network, is rapidly becoming an area of strategic significance. At the edge, network subscribers attach to the ISP's network and service providers can apply services and aggregate leased line traffic. All of this has to be done without compromising performance.
To meet these unique requirements, ISPs need products that are optimized to provide the highest levels of density and availability, along with high-touch software features. The Cisco edge services router is designed to fit this unique position in an ISP's network. The key features are:
1.
Scalability and high bandwidth to meet increased customer demand for data, voice, and video transmission.
2.
Advanced security and reliability features to ensure continued uptime in the face of routine failures and increasingly sophisticated network attackers.
3.
High port density to meet continued growth in the number of customers.
4.
Advanced high-touch features that offer services beyond high-speed packet forwarding include quality of service (QoS) and value-added features in security, virtual networking, and other areas.
5.
A performance-optimized processing engine to handle high volume traffic aggregation.
The ISPs' demanding requirements for an edge services router are not easily met. Aggregating more and more leased lines places a tremendous strain on the system's processor performance. Similarly, enabling each new software service (such as QoS, MLPPP, or MPLS) places a new burden on the processor. To be successful, the edge services router must do the `impossible'—maintain high performance even as new network services are enabled and thousands of new leased lines are connected.
This chapter describes the following topics:
•
Structure of an ISP Network
•
Cisco 10000 Series Router: Specialized for Aggregation
•
Conclusion
Structure of an ISP Network
Physically, the Internet is composed of routers interconnected by communications links. Simple networks are built from a few general-purpose routers interconnected by links owned or leased by ISPs.
As networks become more complex, with greater numbers of elements, more structure is required. Elements become specialized in their applications, management and security become more important, physical location is a consideration, and the capability to handle higher densities of customers is critical.
Structure can be imposed on a complex network by assigning specific jobs to particular routers. A common approach in ISP networks is to divide assignments among routers in the following way:
•
Access routers provide individual subscribers with access to the network. They tend to focus more on large numbers of relatively low-speed ports connecting to subscribers.
•
Backbone routers provide Internet backbone transport. The emphasis is on achieving the highest possible forwarding rates on the fastest available interfaces.
Figure 1-2 illustrates the router specialization scheme in a typical ISP network.
Figure 1-2 Typical ISP Network
Most ISPs also impose physical structure on their networks by organizing them into points of presence (POPs). A POP is a physical location where a set of access and backbone routers is located. An ISP network usually consists of several POPs.
shows how the ISP network in Figure 1-2 might be physically structured at various POPs.
Figure 1-3 ISP POPs
Characteristics of Access and Backbone Routers
Although real networks are always more complicated than theoretical architectures, it is possible to make general distinctions between the two router types. Table 1-1 summarizes the principal differences between backbone and access routers.
Table 1-1 Access Router Compared to Backbone Router
Variable
|
Backbone Router
|
Access Router
|
Packet-per-second throughput
|
Extremely high
|
High
|
Packet processing feature set
|
Minimal, focused on fast forwarding
|
High-touch value-added features
|
Interface types
|
Modest number of very high-speed interfaces
|
Large number of relatively low-speed interfaces
|
Traffic patterns
|
Any interface to any interface
|
Predominantly subscriber-to-trunk and trunk-to-subscriber (also called "north south")
|
The differences listed in Table 1-1 are not absolute, and often a particular router can fulfill either role. However, as Internet traffic continues to grow, the demands for access routers to handle increased density and backbone routers to handle greater throughput become more important. These density and performance requirements can be met more efficiently with a platform designed for this specific purpose—an aggregation router.
Aggregation Router Definition
An aggregation router is an access router that aggregates large numbers of leased lines from ISP customers into a few trunk lines for entry onto the Internet backbone. An effective aggregation router must be designed specifically for very high throughput performance on a large number of relatively low-speed leased line interfaces with high-touch, value-added network services enabled on each connection.
POP Design
An ISP can simplify network design and maintenance by using a cookie cutter design for its POPs. In this approach, all POPs have a similar structure, with variations for the size or specific needs of each site. A typical POP design is shown in Figure 1-4. There are as many POP designs as there are ISPs.
Figure 1-4 Typical POP Design
The design shown in Figure 1-4 includes the following desirable characteristics:
•
Backbone and access routers are separate, so that backbone router configuration can be kept relatively stable over time. Backbone routers are not affected when individual customers are added to or removed from access routers, or when individual customers add or remove value-added services.
•
Two backbone routers are used in each POP to enhance network availability.
•
Access routers are equipped with redundant trunk connections, again allowing enhanced network availability.
•
Access routers can be added as new subscribers are added to the network.
An edge services router should be flexible enough to serve as an access router in a variety of POP architectures.
Driving for Density
A significant factor affecting the design of an access router is the drive for increased density—an increase in the number of subscriber ports that can be terminated on a single router. Increased density has several advantages for the ISP:
•
Floor space in a POP can be very expensive. Real estate costs are lowered by reducing the number of racks required to terminate a large number of customer connections.
•
Network management is simplified by deploying a smaller number of larger routers. Fewer individual routers to configure, manage, and monitor results in a more efficient operation.
•
Increasing the number of interfaces on the access router improves the statistical performance of a network. Packet networks are usually designed to take advantage of statistical multiplexing, capitalizing on the fact that not all links are busy all of the time. Having more available interfaces reduces the chances that a fluke burst of traffic from several sources at one time will cause temporary network congestion.
The appropriate degree of statistical multiplexing is not a "one size fits all" proposition. ISPs have different views on the trade-off between a network that is rich in bandwidth resources and does not rely on multiplexing, versus one that uses a lower-cost infrastructure but risks occasional congestion. For this reason, an edge services router should be configurable with different combinations of interface cards to create a wide range of multiplexing ratios.
Leased Lines—the Last Mile
The access router serves as the ISP's front line, connecting directly to routers on its customers' premises. However, there is usually a complex circuit-switched infrastructure that transports the leased line signal the "last mile" between the customer premises and the ISP POP.
One common method for constructing the last-mile network for new installations is based on metropolitan-area fiber-optic ring technology. Figure 1-5 shows a simple network that transports 1.544-Mbps DS1 and 44.736-Mbps DS3 signals from a subscriber site, across a fiber-optic ring, to an ISP POP.
Figure 1-5 SONET Metropolitan-Area Transport Network
Most new fiber-optic networks are based on Synchronous Optical Network (SONET) standards in North America or on Synchronous Digital Hierarchy (SDH) standards in much of the rest of the world. SONET/SDH technology is important in transport networks providing leased line connectivity to subscriber routers for two reasons:
•
The high capacity of fiber-optic cables provides a critical advantage by reducing the mass of copper wiring that would otherwise have to be deployed.
•
SONET/SDH provides a high-density, industry-wide standard interface between network transport equipment and the equipment that uses the transported signals, such as aggregation routers.
POPs designed before the introduction of the standard channelized SONET/SDH interface required many racks of data service units (DSUs) to terminate T1 service over traditional copper wiring. The edge services router in today's POP can provide a high density of terminations for DS1 and DS3 connections. A single line card can terminate hundreds of DS1 circuits, all carried on a single fiber.
Cisco 10000 Series Router: Specialized for Aggregation
The increasingly stringent requirements for edge services routing—more services, more leased line connections, greater throughput—are addressed by Cisco Systems with the Cisco 10000 series router. The Cisco 10000 series router is specifically designed for use in aggregation applications by offering the following:
•
Compliance with Network Equipment Building Systems (NEBS) standards to make the Cisco 10008 suitable for central office deployment
•
High interface density to minimize rack space requirements
•
High port count for subscriber connections to minimize the number of routers that must be monitored and configured
•
High-touch features such as extended access lists and QoS classification for value-added services
•
Redundant components for high availability
•
Optimized performance, including parallel processing, for fast throughput despite additional packet analyses required for intelligent services offerings
The remainder of this section describes the Cisco 10000 series router and features that provide industry-leading capabilities for leased-line aggregation.
Cisco 10000 Series Router Hardware View
The Cisco 10000 series router is a chassis-based product that meets all requirements for deployment in central office environments. The product has several major units:
•
A NEBS-compliant chassis equipped with redundant cooling and power management.
•
An engine card called the performance routing engine (PRE) that provides packet forwarding and routing services. A second PRE can be added for redundancy.
•
Two available chassis that can be filled with line cards of the following types:
–
Cisco 10008 chassis—8 line card slots
–
Cisco 10005 chassis—5 line card slots
Figure 1-6 shows the layout of components and key specifications of the Cisco 10008 chassis.
Figure 1-6 Cisco 10008 Chassis Layout
Figure 1-7 shows the layout and key specifications of the Cisco 10005 chassis.
Figure 1-7 Cisco 10005 Chassis Layout
Central Office Features
Several features make the Cisco 10000 series router particularly well-suited to central office installations:
•
Both the Cisco 10008 and 10005 chassis have redundant cooling and power management, ensuring reliability.
•
48 VDC and 110/220 VAC power options are provided, ensuring that the Cisco 10008 and 10005 chassis can be powered by existing central office power facilities.
•
Chassis height ensures high density:
–
The Cisco 10008 chassis is 21.75 inches high, making it possible to fit three Cisco 10008 routers into a standard 7-foot rack.
–
The Cisco 10005 chassis is 11.75 inches high, allowing six Cisco 10005 routers to be mounted in a standard 7-foot rack.
Additional Cisco 10008 ESR Central Office Features
The Cisco 10008 chassis is well-suited for central office environments:
•
Complies with recommendations contained in Telcordia's NEBS generic requirements. Products that pass NEBS testing can be deployed safely and efficiently in a central office environment.
•
Adheres to the 12-inch depth form-factor commonly used by central office transmission equipment. This provides additional flexibility for installation in central offices. For example, in an environment that uses racks with a 30-inch depth, 12-inch routers placed in both the front and rear of the rack makes it possible to fit six Cisco 10008 routers in a standard 7-foot rack.
Interface Types
Deployment at the edge of the network requires several specialized interfaces. Theoretically, any of these interfaces could be used for connections on either the subscriber side or the Internet backbone side of the router. However, in typical installations, different interface types are used for these two applications.
For subscriber-side connections, the Cisco 10000 series router supports:
•
Channelized DS3 line cards
•
Channelized OC-12 line cards
•
Channelized STM-1 line cards
•
Unchannelized E3/T3 line cards
The channelized interface cards support full-rate (unchannelized) DS3 connections, as well as channelization to DS1 (1.544 Mbps), E1 (2.048 Mbps), and nxDS0 (nx64 kbps). All four interface cards also support "subrate" DS3, in which the rate of data transfer across a DS3 can be reduced to limit peak access rate. Subrate modes allow the system to interoperate with other equipment.
For subscriber-side optical connections, the Cisco 10000 series router supports:
•
OC-3 packet over SONET line cards
•
OC-3 ATM line cards
Both of these optical interfaces provide efficient, high-performance bandwidth with STS-3/STM-1 connections for throughput of up to 155.52 Mbps over SONET/ITU-T Synchronous Digital Hierarchy (SDH) interfaces.
For connections to the backbone network, the Cisco 10000 series router supports:
•
STS-12c (unchannelized OC-12) packet over SONET line cards
•
Gigabit Ethernet line cards
•
OC-12 ATM line cards
Density Summary
In a typical application, the Cisco 10008 ESR is configured with two cards connecting to the network backbone and six cards connecting to subscriber sites. Table 1-2 compares T1 density possible for connecting to subscriber sites using channelized T3, channelized OC-12, or channelized STM-1 line cards.
Table 1-2 Overall T1 Density for Line Cards in a Cisco 10008 Chassis
Subscriber Interface Card Type
|
T1 Density (Sessions)
|
Per 10008 Chassis
|
Per Rack (12-inch depth)
|
Per Rack (30-inch depth)
|
Six-port channelized T3 card (DS1)
|
1008
|
3024
|
6048
|
Channelized OC-12 card (DS1)
|
2016
|
6048
|
12096
|
Channelized STM-1 card (E1)
|
1512
|
4536
|
9072
|
In the same application, the Cisco 10005 ESR is configured with one card connecting to the network backbone and four cards connecting to subscriber sites. Table 1-3 compares T1 density possible for connecting to subscriber sites using channelized T3, channelized OC-12, or channelized STM-1 line cards.
Table 1-3 Overall T1 Density for Line Cards in a Cisco 10005 Chassis
Subscriber Interface Card Type
|
T1 Density (Sessions)
|
Per 10005 Chassis
|
Per Rack (24-inch depth)
|
Six-port channelized T3 card (DS1)
|
672
|
4032
|
Channelized OC-12 card (DS1)
|
1344
|
8064
|
Channelized STM-1 card (E1)
|
1008
|
6048
|
High-Touch Feature Set
In its role as a gateway to an ISP network, the Cisco 10000 series router is uniquely positioned to protect the network and also to offer new, value-added features to ISP customers.
Underlying the Cisco 10000 series router feature set is the Cisco IOS software. This system software supports standard routing protocols and network configuration and monitoring. It also supports a variety of network interfaces.
Some areas that have received special attention in the development of the Cisco 10000 series ESR IOS software include:
•
Value-added services, such as MPLS-based virtual private networks (MPLS VPNs), quality of service (QoS) differentiation, and advanced multicast support
•
Flexible access rate mechanisms, such as Multilink PPP and DS3 subrate
•
Network security features such as extended access lists and reverse path forwarding checks
The following sections summarize some of the situations in which advanced router features are critical to ISP networks in today's evolving Internet market.
Virtual Private Networks
Internet access is an important networking application for many ISPs, but it is not the only one. Many ISP customers also want the ability to build private networks, with paths among geographically diverse sites and limited access to the Internet. Private networks allow enhanced security by limiting access to the Internet from a few protected locations in an enterprise network.
In the past, private networks were built by enterprise information technology departments either by leasing dedicated circuits between sites or by using a virtual circuit technology such as Frame Relay or ATM. In each case, the enterprise operating the private network incurred substantial expense building and maintaining the network. ISPs often built several wide-area networks—one for Internet traffic, a second for Frame Relay, and a third for ATM.
An ISP can reduce network costs by moving all traffic onto one common network and then creating virtual private networks (VPNs) built on top of a common Internet backbone. The ISP can maintain one network instead of several, which makes it economical to offer a total private network package to businesses that do not want to manage their corporate wide-area network.
There are several techniques and technologies available for creating IP-based VPNs. Two technologies, Multiprotocol Label Switching (MPLS) and IPSec tunneling, are the leading choices for IP VPNs for service providers. The Cisco 10000 series ESR supports MPLS VPNs natively in the box.
MPLS Virtual Private Networks
MPLS VPNs, which are available in a special software release train, offer all of the value of traditional VPNs. Furthermore, because MPLS VPNs are created in Layer 3, they are more scalable and easier to configure and manage than Layer 2 VPNs.
MPLS VPNs offer:
•
Privacy and security equal to that provided by Layer 2 VPNs by limiting the distribution of a VPN's routes to only those routers that are members of the VPN
•
Seamless integration with customer intranets
•
Increased scalability over current Layer 2 VPN implementations
•
Easy management of VPN membership and provisioning of new VPNs for rapid deployment
•
Scalable any-to-any connectivity for extended intranets and extranets that encompass multiple businesses
MPLS VPN Application
With MPLS, the VPN is encoded in the MPLS label applied to each incoming packet by the provider's edge router (such as the Cisco 10000 series router). Labeled packets are forwarded across an ISP network according to forwarding rules specific to each label. This creates multiple virtual networks on one network infrastructure.
Forwarding rules associated with the labels on packets prevent the packets from being forwarded outside the bounds of the virtual network. These rules can also allow packets to be forwarded between the virtual network and the Internet under controlled circumstances. Figure 1-8 shows a simple application of MPLS that creates two VPNs.
Figure 1-8 MPLS VPNs
As illustrated in Figure 1-8, the following events take place:
1.
The enterprise router at headquarters (Enterprise A) sends normal IP packets destined for a remote site in the corporate network (1).
2.
An ISP access router tags packets from Enterprise A with a special label (2).
3.
Each packet traverses the ISP backbone with a label indicating that the packet should be routed on the Enterprise A virtual network (3).
4.
At the egress from the ISP network, an IPS access router removes the label and forwards each original packet to the Enterprise A sales office (4).
Each Cisco 10000 series router can support 1000 or more distinct VPNs, allowing ISPs to plan for large-scale deployments.
Quality of Service Features
The traditional Internet access model is based on a one-size-fits-all approach, in which all packets in the network are treated equally. As network connectivity becomes a more critical resource for companies, ISPs respond by offering value-added services that yield different degrees of access to network services.
For example, many ISPs now offer:
•
Business class data service (differentiated from economy class)
•
Voice over IP services
Business Class Data Service
In the Internet today, all packets are treated equally, whether they belong to a consumer paying low monthly rates and downloading the latest MP3 music files, or to a large corporation transferring a business-to-business e-commerce order for a thousand custom-manufactured industrial engines. When network congestion happens, packets can be delayed or lost without regard to who sent them, how much traffic they are sending, or how much the owners of the data are willing to pay for reliable transport.
ISPs who employ quality of service (QoS) differentiation can offer business class data services, ensuring that more important traffic has an improved chance of reaching its destination, no matter how much congestion is being caused by less important traffic.
The Cisco 10000 series router offers several features that allow ISPs to offer QoS differentiation. These features include:
•
Flexible packet classification using access lists, IP type of service (IP ToS), differentiated services code point (DSCP), other keying techniques.
•
Packet marking that updates the IP ToS or DSCP.
•
Weighted random early detection (WRED) to allow certain packets to enjoy a lower loss probability at congested routers.
•
Class-based queuing and traffic shaping to guarantee bandwidth to certain classes of traffic.
Voice-over-IP Services
A second important application that requires special QoS treatment is the transport of real-time packet data, particularly traffic associated with packet telephony or voice over IP (VoIP).
Successful deployment of a VoIP service requires careful attention to latency through the network, given that small variations in delay caused by network congestion can annoy listeners and can cause high error rates in fax and modem traffic.
The Cisco 10000 series router provides features that allow ISPs to offer controlled-latency services including:
•
Priority queuing mechanisms that allow some classes of traffic to be configured for the lowest possible latency in spite of network congestion
•
Careful design that ensures latency-sensitive packets cannot get held up in queues behind long backlogs of latency-insensitive traffic
Access Rate Control
As network bandwidths increase, more and more leased line users are moving from nx64 kbps circuits to DS1 (1.544-Mbps) circuits, and from DS1 to DS3 (44.736-Mbps) circuits, with some subscribers going to direct optical connections at OC-3 (155 Mbps) or faster rates.
For subscribers who are starting to outgrow a single 1.544-Mbps DS1 circuit, this poses a problem. It is a big jump from DS1 at 1.544 Mbps to DS3 at 44.736 Mbps.
The Cisco 10000 series router offers two features to help bridge this gap:
•
High-Performance Multilink PPP (MLP)
•
Subrate DS3
High-Performance Multilink PPP
You can progress from one DS1 to several DS1s (or from one E1 to several E1s) through the use of the high-performance Multilink PPP feature on the Cisco 10000 series ESR. This industry-standard protocol (IETF RFC 1990) uses special packet headers and procedures to distribute a single stream of packets onto several parallel links and put the stream back together at the receiving end.
The Cisco 10000 series router allows up to 10 DS1 or E1 links to be combined into a parallel path that is up to ten times faster than a single DS1 or E1. Multilink PPP is implemented with special microcode in the parallel express forwarding (PXF) network processor (see the "Forwarding Path" section on page 3-3) for high performance and scalability in the central office aggregation application.
To terminate MLP connections at the customer premises, CPE routers must be configured with MLP support. In Figure 1-9, two CPE routers are each connected to the Cisco 10000 series router by two parallel DS1 links.
Figure 1-9 Multilink PPP Application
Subrate DS3
A second way to get to an intermediate rate is to start with a DS3 and slow it down. Instead of using MLP to combine several DS1 circuits to get a rate that is between the rates for DS1 and DS3, subrating a DS3 interface limits the bandwidth on the DS3.
Although the rate at which bits are clocked across a DS3 is fixed at 44.736 Mbps, vendors of data service units (DSUs) have created various hardware mechanisms to limit the rate of user traffic that can be sent across a DS3. These rate-limiting mechanisms:
•
Allow ISPs to offer graded rates of access to their networks, along with a flexible pricing structure.
•
Are usually simple and reliable, although special hardware is required on both ends of the DS3 link.
The Cisco 10000 series router line cards can provide subrate DS3 support compatible with PA-T3 and PA-2T3 port adapters on the Cisco 7200 and Cisco 7500 series, and also with products from DS3 DSU vendors. Each unchannelized DS3 attached to the Cisco 10000 series router can be configured for rates ranging from 1 Mbps up to the full rate of the DS3, 44.736 Mbps.
Subrate DS3 provides an additional benefit in flexibility. After the DS3 circuit is installed, ISPs can upgrade customer access rates with a simple software reconfiguration of the line cards at each end of the link. Figure 1-10 shows a possible subrate DS3 configuration.
Figure 1-10 Subrate DS3 Configuration
Network Security
ISPs place high emphasis on network security, because they want to defend their networks and their customers' computers from sophisticated denial of service attacks. Many standard security features are available on the Cisco 10000 series router through the use of Cisco IOS Release 12.0(S) commands. These security features include:
•
Password encryption
•
Authentication, authorization, and accounting (AAA) support
•
Router authentication
•
Kerberos authentication and client support on Telnet
Some additional special measures are implemented in the Cisco 10000 series router to resist denial of service attacks. These are discussed in the following sections.
Access Lists
The Cisco 10000 series router implements high-performance access lists (standard and extended), allowing providers to specify the traffic to be forwarded through the router. A new algorithm called turbo ACL has an improved access list so that the router can process large lists with a minimum throughput penalty.
Reverse Path Forwarding Check
Many common denial of service attacks involve forged IP source addresses. The packets appear to be coming from a source that either does not exist or exists at some other point in the network. By using forged source addresses, attackers are better able to hide the attacking machines' identities, making it more challenging to find the culprits.
The Cisco 10000 series router implements a feature called reverse path forwarding (RPF) check, which can be used with both unicast and multicast traffic. The feature checks all packets forwarded through the router to ensure that each one has a plausible source address. The RPF check supplements the usual verifications performed on the destination address and other fields in the IP header.
The RPF feature does not affect the router's packet forwarding rate. Hence, network administrators are not forced to disable it to improve throughput, which is sometimes necessary if a security check impairs performance.
Fast-Path Internet Control Message Protocol
Most denial of service attacks are directed against host computers or web servers and use routers to access the target. However, attacks are also launched against the router through operations that are not normally optimized for throughput. The flooding that results can consume large amounts of router memory or processor cycles.
Internet Control Message Protocol (ICMP) is the network layer Internet protocol (documented in RFC 792) that reports errors and provides other information relevant to IP packet processing. The Cisco 10000 series router implements many of the normal ICMP response functions in the high-speed forwarding path, to ensure that floods of ping packets or other ICMP messages cannot preempt more important router functions.
In addition, messages that are sent from the high-speed forwarding path to the router's internal processor can be categorized by priority. This helps ensure that the router cannot become so busy responding to an overload of unimportant traffic that it neglects essential packets, such as keepalives and route updates, that keep the network operating.
High Availability
Availability corresponds to the probability that a network or network element will function properly at a given moment. From the outset, achieving high availability has been a critical design goal for the Cisco 10000 series router.
Cisco 10000 series router achieves high availability in the following ways:
•
Highly integrated designs reduce the number of electronic components and attendant connections that can fail.
•
Redundant elements allow a system to continue operating in spite of component failures.
•
Software detects and corrects failures with decreased latency and increased accuracy.
The Cisco 10000 series router incorporates several mechanisms to reduce its downtime:
•
The chassis has redundant cooling and power distribution, which accommodates failure of internal power supplies, external power sources, or mechanical air-movers, with no disruption to the operation of the box.
•
The principal data paths on the backplane are implemented with point-to-point wiring between line cards and the routing engine, preventing failures on one line card from interrupting traffic on other line cards. This approach also reduces problems caused by adding or removing cards during operation.
•
The chassis is designed so that you can add a second PRE for redundancy—a backup for the card that provides the packet processing and forwarding capabilities for the system.
•
You can protect SONET external links using the SONET automatic protection switching mechanism (also known as 1+1 APS). APS uses redundant pairs of cards and external fiber connections to protect against failures of either the line cards or the fiber links.
Route Processor Redundancy Plus
When the Cisco 10000 series router is equipped with a second Performance Routing Engine (PRE):
•
One PRE is designated the primary PRE and is responsible for running routing protocols and forwarding traffic
•
The other PRE is designated the secondary PRE, and acts as a backup to the primary PRE
Route Processor Redundancy Plus (RPR+) software selects the primary PRE and the secondary PRE during system initialization. RPR+ software also boots Cisco IOS software and tracks changes in the configuration of the primary PRE.
If the primary PRE fails, RPR+ ensures that the secondary is ready to take over operation of the system with minimal disruption. RPR+ software running on each PRE monitors the health of the other PRE so that if the primary PRE fails, the secondary PRE can take over, or if the secondary PRE fails, the network operator can be notified that the system is no longer fully redundant.
Conclusion
The Cisco 10000 series router is an advanced Layer 3 aggregation router that meets the needs of today's ISPs, but also provides the flexibility to satisfy future requirements. It can aggregate thousands of leased line connections, contribute processor-intensive IP network services, and still satisfy the performance and availability requirements of today's Internet market.
The Cisco 10000 series router offers the following advantages to an ISP:
•
Compliance with central office requirements.
•
Conservation of POP space.
•
Redundancy and hot-swapping features.
•
Scalability and line card flexibility.
•
Cost effective software-based upgrades that do not require new hardware purchases.
•
A flexible Layer 3 feature set that facilitates the delivery of new network services.
•
High-density subscriber termination.
To ISP customers, the Cisco 10000 series router offers the following advantages:
•
High availability and performance.
•
Differentiated services.
•
Quality of service differentiation for newly evolving applications such as voice-over-IP (VoIP), video transmission, and virtual private networks (VPNs).
To ISPs and their customers, the Cisco 10000 series router offers reliability, availability, and the capacity to handle future growth. It arises from and contributes to the synergy among developers, providers, and users that is necessary in today's telecommunications market.