Полезная информация


Table of Contents

Wide-Area Networking Overview

Wide-Area Networking Overview

Cisco IOS software provides a range of wide-area networking capabilities to fit almost every network environment need. Cisco offers cell relay via the Switched Multimegabit Data Service (SMDS), circuit switching via Integrated Services Digital Network (ISDN), packet switching via Frame Relay, and the benefits of both circuit and packet switching via Asynchronous Transfer Mode (ATM). LAN emulation (LANE) provides connectivity between ATM and other LAN types.

Cisco's dial backup capability provides continuous network access during WAN downtime. Dial-on-demand routing (DDR) provides access flexibility, using modems or ISDN to connect to a WAN. Dial-up connections can use Link Access Procedure, Balanced (LAPB), PPP, X.25, or can use Frame Relay encapsulation over X.25 or Frame Relay packet-switched networks.

The Wide-Area Networking Configuration Guide discusses the following software components:

This overview chapter gives a high-level description of each technology. For specific configuration information, refer to the appropriate chapter in this module.


ATM is a cell-switching and multiplexing technology designed to combine the benefits of circuit switching (constant transmission delay and guaranteed capacity) with those of packet switching (flexibility and efficiency for intermittent traffic).

Cisco provides ATM access in several ways, depending on the hardware available in the router:

In routers outside the Cisco 4500 series and the Cisco 7000 family, a serial interface can be configured for multiprotocol encapsulation over the Asynchronous Transfer Mode-Data Exchange Interface (ATM-DXI), as specified by RFC 1483. This standard describes two methods for transporting multiprotocol connectionless network interconnect traffic over an ATM network. One method allows multiplexing of multiple protocols over a single permanent virtual circuit (PVC). The other method uses different virtual circuits to carry different protocols. Our implementation supports transport of AppleTalk, Banyan VINES, Internet Protocol (IP), and Novell Internetwork Packet Exchange protocol (IPX) traffic.

In routers outside the Cisco 4500 series and the Cisco 7000 family, an ATM data service unit (ADSU) is required to do the following:

On the Cisco 7000 family routers, network interfaces reside on modular interface processors, which provide a direct connection between the high-speed Cisco Extended Bus (CxBus) and the external networks. Each AIP provides a single ATM network interface; the maximum number of AIPs that the Cisco 7000 supports depends on the bandwidth configured. The total bandwidth through all the AIPs in the system should be limited to 200 Mbps full duplex (two TAXI interfaces, or one SONET and one E3, or one SONET and one lightly used SONET, five E3s, or four T3s).

Cisco 4500 series routers support one OC-3c network processor module (NPM) or up to two slower E3/DS3 NPMs. Physical Layer Interface Modules (PLIMs) that support Synchronous Optical Network/Synchronous Digital Hierarchy (SONET/SDH) 155 Mbps are available for both single-mode and multimode fiber.

For a complete description of the 7000 family router and AIP, refer to the Hardware Installation and Maintenance publication for your specific router. For a complete description of the Cisco 4500 series router and the NPM, refer to the Cisco 4000 Series Hardware Installation and Maintenance manual. For information about installing the NPM, see the document called "Installing Network Processing Modules in the Cisco 4000 Series" (online, it is in the Cisco 4000 Series Configuration Notes).

ATM Environment

ATM is a connection-oriented environment. All traffic to or from an ATM network is prefaced with a virtual path identifier (VPI) and virtual channel identifier (VCI). A VPI-VCI pair is considered a single virtual circuit. Each virtual circuit is a private connection to another node on the ATM network. Each virtual circuit is treated as a point-to-point mechanism to another router or host and is capable of supporting bidirectional traffic.

Each ATM node is required to establish a separate connection to every other node in the ATM network that it needs to communicate with. All such connections are established by means of a PVC or a switched virtual circuit (SVC) with an ATM signaling mechanism. This signaling is based on the ATM Forum User-Network Interface (UNI) Specification V3.0.

Each virtual circuit is considered a complete and separate link to a destination node. Users can encapsulate data as needed across the connection. The ATM network disregards the contents of the data. The only requirement is that data be sent to the router's ATM processor card in a manner that follows the specific ATM adaptation layer (AAL) format.

An AAL defines the conversion of user information into cells. An AAL segments upper-layer information into cells at the transmitter and reassembles the cells at the receiver. AAL1 and AAL2 handle isochronous traffic, such as voice and video, and are not relevant to the router. AAL3/4 and AAL5 support data communications; that is, they segment and reassemble packets. Cisco supports both AAL3/4 and AAL5 on the Cisco 7000 family and the Cisco 4500 series. However, on the Cisco 4500 series, AAL3/4 is not supported at OC-3 rates; if AAL3/4 is configured on an OC-3c interface, you must limit the interface to E3 or DS3 rates by configuring a rate queue. See the "Configure the Rate Queue (Cisco 4500)" section for more information.

An ATM connection is simply used to transfer raw bits of information to a destination router or host. The ATM router takes the common part convergence sublayer (CPCS) frame, carves it up into 53-byte cells, and sends these cells to the destination router or host for reassembly. In AAL5 format, 48 bytes of each cell are used for the CPCS data; the remaining 5 bytes are used for cell routing. The 5-byte cell header contains the destination VPI-VCI pair, payload type, cell loss priority (CLP), and header error control.

The ATM network is considered a LAN with high bandwidth availability. Each end node in the ATM network is a host on a specific subnet. All end nodes needing to communicate with one another must be within the same subnet in the network.

Unlike a LAN, which is connectionless, ATM requires certain features to provide a LAN environment to the users. One such feature is broadcast capability. Protocols wishing to broadcast packets to all stations in a subnet must be allowed to do so with a single call to Layer 2. To support broadcasting, the router allows the user to specify particular virtual circuits as broadcast virtual circuits. When the protocol passes a packet with a broadcast address to the drivers, the packet is duplicated and sent to each virtual circuit marked as a broadcast virtual circuit. This method is known as pseudobroadcasting.

Effective with Release 11.0, point-to-multipoint signaling allows pseudobroadcasting to be eliminated. On routers with point-to-multipoint signaling, the router can set up calls between itself and multiple destinations; drivers no longer need to duplicate broadcast packets. A single packet can be sent to the ATM switch, which replicates it to multiple ATM hosts.

Classical IP and ARP

Cisco implements classical IP and Address Resolution Protocol (ARP) over ATM as described in RFC 1577. RFC 1577 defines an application of classical IP and ARP in an ATM environment configured as a logical IP subnetwork (LIS). It also describes the functions of an ATM ARP server and ATM ARP clients in requesting and providing destination IP addresses and ATM addresses in situations when one or both are unknown. Our routers can be configured to act as an ARP client, or to act as a combined ARP client and ARP server.

The ATM ARP server functionality allows classical IP networks to be constructed with ATM as the connection medium. Without this functionality, you must configure both the IP network address and the ATM address of each end device with which the router needs to communicate. This static configuration task takes administrative time and makes moves and changes more difficult.

Cisco's implementation of the ATM ARP server functionality provides a robust environment in which network changes can be made more easily and more quickly than in a pure ATM environment. Cisco's ATM ARP client works with any ARP server that is fully compliant with RFC 1577.

The Cisco AIP

This section provides an overview of the ATM features, interfaces, microcode, and virtual circuits available on the AIP, currently supported on the Cisco 7000 family routers.

AIP Features

The AIP supports the following features:

Process-switched bridging over ATM supports AAL3/4-SMDS encapsulated packets only. All frames that originate at or are forwarded by the Cisco IOS software are sent as 802.3 bridge frames without frame check sequence (FCS)--that is, in RFC 1483 bridge frame formats with 0x0007 in the Protocol Identification (PID) field of the Subnetwork Access Protocol (SNAP) header. You can enable process-switched bridging for SMDS as described later in this chapter.
Fast-switched transparent bridging over ATM supports AAL5-SNAP encapsulated packets only. All bridged AAL5-SNAP encapsulated packets are fast switched. Fast-switched transparent bridging supports Ethernet, Fiber Distributed Data Interface (FDDI), and Token Ring packets sent in AAL5-SNAP encapsulation over ATM. You can enable fast-switched bridging for AAL5-SNAP as described later in this chapter.

AIP ATM Interface Types

All ATM interfaces are full duplex. You must use the appropriate ATM interface cable to connect the AIP with an external ATM network. Refer to the Asynchronous Transfer Mode Interface Processor (AIP) Installation and Configuration publication for descriptions of ATM connectors.

The AIP provides an interface to ATM switching fabrics for transmitting and receiving data at rates of up to 155 Mbps bidirectionally; the actual rate is determined by the physical layer interface module (PLIM). The PLIM contains the interface to the ATM cable. The AIP can support PLIMs that connect to the following physical layers:

For wide-area networking, ATM is currently being standardized for use in Broadband Integrated Services Digital Networks (BISDNs) by the International Telecommunication Union Telecommunication Standardization Sector (ITU-T) and the American National Standards Institute (ANSI). BISDN supports rates from E3 (34 Mbps) to multiple gigabits per second (Gbps).

Note The ITU-T carries out the functions of the former Consultative Committee for International Telegraph and Telephone (CCITT).

AIP Microcode

The AIP microcode is a software image that provides card-specific software instructions. An onboard read-only memory (ROM) component contains the default AIP microcode. The Cisco 7000 supports downloadable microcode, which enables you to upgrade microcode versions by loading new microcode images onto the Route Processor (RP), storing them in Flash memory, and instructing the AIP to load an image from Flash memory instead of the default ROM image. You can store multiple images for an interface type and instruct the system to load any one of them or the default ROM image with a configuration command. All processor modules of the same type will load the same microcode image from either the default ROM image or from a single image stored in Flash memory.

Although multiple microcode versions for a specific interface type can be stored concurrently in Flash memory, only one image can load at startup. The show controller cxbus command displays the currently loaded and running microcode version for the Switch Processor (SP) and for each IP. The show running-config command shows the current system instructions for loading microcode at startup.

For a complete description of microcode and downloading procedures, refer to the Asynchronous Transfer Mode Interface Processor (AIP) Installation and Configuration publication and the Configuration Fundamentals Configuration Guide.

AIP Virtual Circuits

A virtual circuit is a connection between remote hosts and routers. A virtual circuit is established for each ATM end node with which the router communicates. The characteristics of the virtual circuit that are established for the AIP when the virtual circuit is created include the following:

Each virtual circuit supports the following router functions:

By default, fast switching is enabled on all AIP interfaces. These switching features can be turned off with interface configuration commands. Autonomous switching must be explicitly enabled per interface.

The Cisco NPM

This section provides an overview of the ATM features, interfaces, and virtual circuits available on the NPM, currently supported on the Cisco 4500 series routers.

NPM Features

The NPM supports the following features:

An ATM adaptation layer (AAL) defines the conversion of user information into cells by segmenting upper-layer information into cells at the transmitter and reassembling them at the receiver. AAL1 and AAL2 handle isochronous traffic, such as voice and video, and are not relevant to the router. AAL3/4 and AAL5 support data communications by segmenting and reassembling packets. On the Cisco 4500 series, Cisco supports both AAL3/4 (except at OC-3 rates) and AAL5.

NPM ATM Interface Types

All ATM interfaces are full duplex. You must use the appropriate ATM interface cable to connect the NPM with an external ATM network. Refer to the Cisco 4000 Series Hardware Installation and Maintenance manual and the Installing NPMs in the Cisco 4000 Series manual for descriptions of ATM connectors.

The NPM provides an interface to ATM switching fabrics for transmitting and receiving data at rates of up to 155 Mbps bidirectionally; the actual rate is determined by the physical layer interface module (PLIM). The PLIM contains the interface to the ATM cable. The NPM can support PLIMs that connect to the following physical layers:

NPM Virtual Circuits

A virtual circuit is a point-to-point connection between remote hosts and routers. A virtual circuit is established for each ATM end node with which the router communicates. The characteristics of the virtual circuit that are established for the NPM when the virtual circuit is created include the following:

Each virtual circuit supports the following router functions:


Dial-on-demand routing (DDR) provides network connections across the Public Switched Telephone Network (PSTN). Traditionally, networks have been interconnected using dedicated lines for wide-area network (WAN) connections. With DDR, you can use modems, external channel service units (CSUs), Integrated Service Digital Network (ISDN) terminal adapters (TAs) or integrated ISDN interfaces, to establish low-volume, periodic network connections over public circuit-switched networks. You can also establish dial-up connections over X.25 or Frame Relay packet-switched networks by using LAPB, X.25, or Frame Relay encapsulations.

The following protocols can be routed over DDR: AppleTalk, Banyan VINES, CLNS, DECnet, IP, IPX, and XNS. For more information about IP, see Network Protocols Configuration Guide, Part 1; for more information about AppleTalk and IPX, see Network Protocols Configuration Guide, Part 2, and for more information about all other protocols, see Network Protocols Configuration Guide, Part 3.

Synchronous serial, asynchronous serial, and ISDN interfaces can be configured for DDR connections to one or more destination networks. On serial interfaces, when a packet is received for a remote network, the Cisco IOS software uses dialing commands to send the phone number of the destination network to a modem. The modem--a data communications equipment (DCE) device--then dials the destination DCE device and establishes a connection. On ISDN interfaces, DDR dialup connections are made through NT1 or CSU devices, for BRI and PRI respectively.

Figure 2 illustrates a typical DDR interconnection configuration.

Figure 2: DDR Interconnection

Beginning with Cisco IOS Release 11.2, our software includes two implementations of DDR:

In this release, most routed protocols are supported; however, Frame Relay, ISO CLNS, LAPB, and snapshot routing are not supported. The dialer profiles implementation supports dial backup, as described in the new "Configure Dialer Profiles" section of the "Configuring DDR" chapter of this publication.

The following sections describe important aspects of DDR:

Dialer Profiles

Dialer profiles allow the configuration of physical interfaces to be separated from the logical configuration required for a call, and also allow the logical and physical configurations to be bound together dynamically on a per-call basis. This release supports PPP and HIgh-Level Data Link Control (HDLC) encapsulation on the physical interface. All other settings are part of a logical configuration used and applied to the physical interface as needed for specific calls. Configuration of dial backup is also simplified if you use dialer profiles.

A dialer profile consists of the following elements:

All calls going to from the same destination subnetwork use the same dialer profile.

A dialer interface configuration includes all settings needed to reach a specific destination subnetwork (and any networks reached through it). Multiple dial strings can be specified for the same dialer interface, each dial string being associated with a different dialer map-class. The dialer map-class defines all the characteristics for any call to the specified dial string. For example, the map-class for one destination might specify ISDN speed 56 kbps; the map-class for a different destination might specify ISDN speed 64 kbps.

Each dialer interface uses a dialer pool, a pool of physical interfaces ordered on the basis of the priority assigned to each physical interface. A physical interface can belong to multiple dialer pools, contention being resolved by priority. ISDN BRI and PRI interfaces can set a limit on the minimum and maximum number of B channels reserved by any dialer pools. A channel reserved by a dialer pool remains idle until traffic is directed to the pool.

When dialer profiles are used to configure DDR, a physical interface has no configuration settings except encapsulation and the dialer pools the interface belongs to.

Legacy DDR Features

Legacy DDR provides options designed to enable specific applications and provide enhanced WAN optimization. The options are described in the following sections:

In addition, DDR provides the following default features to enhance calling and switching:

Dial Backup

Dial backup provides protection against WAN downtime by allowing you to configure a backup serial line circuit-switched connection. Dial backup software keeps the secondary line inactive-- data terminal ready (DTR) inactive--until one of the following conditions is met:

When the software detects a lost Carrier Detect signal from the primary line device or finds that the line protocol is down, it activates DTR on the secondary line. At that time, the data communications equipment (DCE) must be set to dial the remote site. When the connection is made, the routing protocol defined for the serial line takes over the task of transmitting traffic over the dialup line.

Bandwidth on Demand

The bandwidth on demand option provides additional bandwidth by placing additional calls to a single destination if the load for the interface exceeds a specified weighted value. Parallel communication links are established based on traffic load. The number of parallel links that can be established to one location is not limited. Dialer rotary groups can be configured to support this option.

Snapshot Routing

Snapshot routing, which is available on serial and ISDN lines, is a method whereby the Cisco IOS software can learn remote routes dynamically and then keep the routes available for a period of time while regular routing updates are not being exchanged. Such a period might occur when a remote site is not dialed into the local site or when a remote site has a dedicated connection to the local site but cannot afford the overhead of exchanging routing updates. Snapshot routing allows you to avoid configuring static routes when using dial-on-demand routing (DDR). It also eliminates the overhead required for sending periodic updates over dedicated serial lines.

Fast Call Rerouting for ISDN

When DDR calls using an ISDN interface are not accepted, the dialer is able to place the call again or proceed to other calls almost immediately, and does not have to wait for the dialer wait-for-carrier timer to expire. The ISDN software learns within a few seconds that a call was not accepted and always informs the dialer software, thus greatly reducing delays.

This feature is automatically enabled for all ISDN interfaces when the Cisco IOS software begins to run.

You can still modify the dialer wait-for-carrier timer for DDR interfaces, and the show dialer command still shows the destination number, if connected.

DDR Fast Switching

In the past, only process switching was available on interfaces configured for DDR. Process switching provided an acceptable level of performance because DDR was used on low-speed lines. Now, however, fast switching is required to take advantage of ISDN Primary Rate Interface (PRI) and multiple Basic Rate Interface (BRI) platforms.

Fast switching is enabled by default on all interfaces configured for DDR. It is enabled for two routed protocols, IP and IPX, and for two encapsulations, High-Level Data Link Control (HDLC) and Point-to-Point Protocol (PPP).

Fast switching can be disabled and reenabled on a protocol-by-protocol basic on a DDR interface. For information about disabling and reenabling fast switching of protocols, see the "Disable and Reenable DDR Fast Switching" section in the "Configuring DDR" chapter.

Controlling Access for DDR

DDR supports a variety of security and access control methods including the following:

Packets that are permitted entry according to the access list are identified as interesting or packets of interest. Packets that are not permitted entry or are denied entry by an access list are deemed uninteresting.
A router or access server activates the dial-on-demand feature when it receives an interesting packet destined for a location that can be reached over a dialed connection through a Public Switched Telephone Network (PSTN). After the Cisco IOS software routine dials the destination phone number and establishes a connection, packets can be transmitted. When the transmission is complete and a configured period of line time has elapsed during which no interesting traffic exists on the line, the line is automatically disconnected.

Note The Transmission Control Protocol/Internet Protocol (TCP/IP) routing protocols Intermediate System-to-Intermediate System (IS-IS), Border Gateway Protocol (BGP), and Open Shortest Path First (OSPF) are not recommended for use with DDR because they require an acknowledgment for routing updates. Because DDR lines are brought up as needed, DDR might not be active and available to send responses at the times the updates are sent.

Note Access lists must be defined before you can use DDR. If no access lists are defined, access is implicitly denied. See the Network Protocols Configuration Guide, Part 1 for information about the IP access lists with the tcp keyword specified. See the "Configuring AppleTalk" chapter for information about AppleTalk static routes defined, and the "Configuring Novell IPX" chapter for information about the IPX access lists; these chapters appear in the Network Protocols Configuration Guide, Part 2.

Placing Calls Using DDR

DDR can use following methods to make outbound calls:

Dialer Interfaces

Dialer profiles provide a way to configure logical dialer interfaces separately from the physical interfaces which use the logical configuration. Dialer interfaces can support both inbound and outbound calls. See the Dialer Profiles section of the "Configuring DDR" chapter for information.

V.25bis for Synchronous Interfaces

Cisco IOS software supports connections from the synchronous serial interface to any DCE device that supports V.25bis. These devices include ISDN TAs for ISDN B channel connections. V.25bis is an International Telecommunication Union Telecommunication Standardization Sector (ITU-T) recommendation for initiating calls using in-band signaling. Depending on the type of modem or CSU/DSU you are using, ITU-T V.25bis options might be supported.

The V.25bis specification describes two modes for establishing or receiving calls: the direct call mode and the addressed call mode. The Cisco IOS software supports connections using the addressed call mode and synchronous, bit-oriented operation. The addressed call mode allows control signals and commands to be sent over the DCE data interface to establish and terminate calls. These commands are packaged in High-Level Data Link Control (HDLC) synchronous data frames.

Devices used by the router or access server for dialing out must support certain hardware signals in addition to V.25bis. When the router or access server drops DTR, the device must disconnect any calls that are currently connected. When the device connects to the remote end, Data Carrier Detect (DCD) must be automatically asserted.

Note For many V.25bis devices, raised DCD requires a special cable to cross over DCD and Data Set Ready (DSR) signals, because the V.25bis specification requires DSR to be raised when a connection is established.

Our routers and access servers support connections over serial lines connected by non-V.25bis modems, using data terminal ready (DTR) Electronic Industries Association (EIA) signaling only.

DTR Dialing for Synchronous Interfaces

Routers and access servers also support connections from synchronous serial lines through non-V.25bis modems. Cisco devices connected by non-V.25bis modems must use data terminal ready (DTR) EIA signaling only.

Chat Scripts on Asynchronous Interfaces

A chat script is a string of text that defines the login "conversation" that occurs between two systems. It consists of expect-send pairs that define the string that the local system expects to receive from the remote system and what the local system should send as a reply.

On asynchronous lines, our software supports chat scripts that send commands for modem dialing and logging on to remote systems. To dial a call on an asynchronous line, a chat script must be defined. If multiple chat scripts are defined, regular expressions are used for powerful pattern matching to select between many scripts. See the "Regular Expressions" appendix in the Access Services Command Reference for information about regular expressions.

Note On Cisco routers, only the auxiliary port supports asynchronous lines.

Frame Relay

Cisco's Frame Relay implementation currently supports routing on IP, DECnet, AppleTalk, Xerox Network Service (XNS), Novell IPX, International Organization for Standards (ISO) Connectionless Network Service (CLNS), Banyan VINES, and transparent bridging.

Although Frame Relay access was originally restricted to leased lines, dial-up access is now supported. For more information, see the "Configure DDR over Frame Relay" section for dialer profiles or for legacy DDR in the "Configuring DDR" chapter.

To install software on a new router or access server by downloading software from a central server over an interface that supports Frame Relay, see the "Loading Images and Configuration Files" chapter in the Configuration Fundamentals Configuration Guide.

To configure access between Systems Network Architecture (SNA) devices over a Frame Relay network, see the "Configuring SNA Frame Relay Access Support" chapter in the Bridging and IBM Networking Configuration Guide.

The Frame Relay software provides the following capabilities:

Switched virtual circuits (SVCs) allow access through a Frame Relay network by setting up a path to the destination endpoints only when the need arises and tearing down the path when it is no longer needed.
Frame Relay switching is used when all traffic arriving on one DLCI can be sent out on another DLCI to the same next hop address. In such cases, the Cisco IOS software does not have to examine the frames individually to discover the destination address, and as a result, the processing load on the router decreases.


Cisco implements the physical layer protocols for the ISDN Basic Rate Interface (BRI) and the ISDN Primary Rate Interface (PRI) on the following routers:

The BRI interface includes one ISDN Basic Rate connection.The Basic Rate connection consists of a D channel and two B channels, both of which are full-duplex, 64-kbps channels.

For detailed technical information about Cisco's implementation, see the description of the Cisco ISDN MIB in the Cisco Management Information Base (MIB) User Quick Reference.

Figure 3 represents the general relationships between circuit-switched access methods (asynchronous, synchronous, and ISDN) and DDR and dial backup. It also summarizes the steps you use to get the appropriate line up and working. However, this module describes only ISDN.

Figure 3: Configuring ISDN Access


The ISDN specifications describe a planned digital network that will provide a wide and evolving variety of services and use digital transmission and switching technologies to provide worldwide, integrated access. ISDN is an effort to standardize user services, user-network interfaces, and network and internetwork capabilities. Among the services ISDN is planned to support are integrated text, voice, graphics, music, video, and data communications.

ISDN standards define services, common procedures, and a single set of interface rules so that any device can gain access to an ISDN network. ISDN standards describe a three-layer protocol architecture, similar but not identical to the OSI reference model's physical, data link, and network layers.

ISDN Channels

The data or D channel is used for call setup control and network connection teardown. Call setup involves the data link and network connection. D channel communication is from the router to the ISDN switch.

The bearer or B channels contain user data. The B channels are treated as 64-kbps serial lines and support HDLC and PPP encapsulation. The interface configuration is propagated to each of the B channels. Although each channel is treated as a separate line, the B channels cannot be configured separately. (However, if you use the new DDR dialer profiles, you can specify the number of B channels reserved for a specific dialer pool. See the "Configuring DDR" chapter for more information.)

Note A single switch type must be configured for the router as a whole.

In North America and Japan, the PRI is a rotary group of 23 B channels (T1) at the combined rate of 1.544 Mbps. Elsewhere, PRI is a rotary group of 30 B channels (E1) at a combined rate of 2.048 Mbps. On the MBRI and PRI, you can create a rotary group from a number of BRI or PRI interfaces.

Network-Customer Premises Boundary

In North America, the boundary between the ISDN network and the BRI on routers in the Cisco 4500 series and below is represented by customer premises equipment known as network termination type 1 equipment (NT1). In North America, an NT1 is required for each BRI. Outside North America, the NT1 is supplied as part of the telecommunications services.

In North America, the boundary between the ISDN network and the PRI on Cisco 7000 family routers is represented by customer premises equipment known as a channel service unit (CSU). In North America, a CSU is required for each PRI. Outside North America, the CSU is supplied as part of the telecommunications services. Figure 4 illustrates the boundary between customer premises and the ISDN network in North America and in other locations.

Figure 4: Customer Premises and ISDN Network Boundary

The ISDN data link layer interface (used for call setup) that is provided by the Cisco IOS software conforms to the specification defined by the ITU-T recommendation Q.921. The ISDN network layer interface (used for call control) provided by the software conforms to the specifications for specific switch types defined by the ITU-T recommendation Q.931.

For a list of ISDN switch types that the ISDN interface supports, see the section "Select the ISDN Switch Type" in the chapter "Configuring ISDN".

LAN Emulation (LANE)

Cisco's implementation of LANE makes an ATM interface look like one or more Ethernet interfaces.

LANE is an ATM service defined by the ATM Forum specification LAN Emulation over ATM, ATM_FORUM 94-0035. This service emulates the following LAN-specific characteristics:

LANE service provides connectivity between ATM-attached devices and connectivity with LAN-attached devices. This includes connectivity between ATM-attached stations and LAN-attached stations and also connectivity between LAN-attached stations across an ATM network.

Because LANE connectivity is defined at the MAC layer, upper protocol layer functions of LAN applications can continue unchanged when the devices join emulated LANs. This feature protects corporate investments in legacy LAN applications.

An ATM network can support multiple independent emulated LAN networks. Membership of an end system in any of the emulated LANs is independent of the physical location of the end system. This characteristic enables easy hardware moves and location changes. In addition, the end systems can also move easily from one emulated LAN to another, whether or not the hardware moves.

LAN emulation in an ATM environment provides routing between emulated LANs for supported routing protocols and high-speed, scalable switching of local traffic.

The ATM LANE system has three servers that are single points of failure. These are the LECS (Configuration Server), the LES (emulated LAN server), and the BUS (the broadcast and unknown server). Beginning with Release 11.2, LANE fault tolerance or Simple LANE Service Replication on the emulated LAN provides backup servers to prevent problems if these servers fail.

The fault tolerance mechanism that eliminates these single points of failure is described in the "Configuring LANE" chapter. Although this scheme is proprietary, no new protocol additions have been made to the LANE subsystems.

LANE Components

Any number of emulated LANs can be set up in an ATM switch cloud. A router can participate in any number of these emulated LANs.

LANE is defined on a LAN client-server model. The following components are implemented in this release:

A LANE client emulates a LAN interface to higher layer protocols and applications. It forwards data to other LANE components and performs LANE address resolution functions.
Each LANE client is a member of only one emulated LAN. However, a router can include LANE clients for multiple emulated LANs: one LANE client for each emulated LAN of which it is a member.
If a router has clients for multiple emulated LANs, the Cisco IOS software can route traffic between the emulated LANs.
The LANE server for an emulated LAN is the control center. It provides joining, address resolution, and address registration services to the LANE clients in that emulated LAN. Clients can register destination unicast and multicast MAC addresses with the LANE server. The LANE server also handles LANE ARP (LE ARP) requests and responses.
Our implementation has a limit of one LANE server per emulated LAN.
The LANE broadcast-and-unknown server sequences and distributes multicast and broadcast packets and handles unicast flooding.
In this release, the LANE server and the LANE broadcast-and-unknown server are combined and located in the same Cisco 7000 family or Cisco 4500 series router; one combined LANE server and broadcast-and-unknown server is required per emulated LAN.
The LANE configuration server contains the database that determines which emulated LAN a device belongs to (each configuration server can have a different named database). Each LANE client consults the LANE configuration server just once, when it joins an emulated LAN, to determine which emulated LAN it should join. The LANE configuration server returns the ATM address of the LANE server for that emulated LAN.
One LANE configuration server is required per LANE ATM switch cloud.
The LANE configuration server's database can have the following four types of entries:

  • Emulated LAN name-ATM address of LANE server pairs

  • LANE client MAC address-emulated LAN name pairs

  • LANE client ATM template-emulated LAN name pairs

  • Default emulated LAN name

Note Emulated LAN names must be unique on an interface. If two interfaces participate in LANE, the second interface may be in a different switch cloud.

LANE Operation and Communication

Communication among LANE components is ordinarily handled by several types of switched virtual circuits (SVCs). Some SVCs are unidirectional; others are bidirectional. Some are point-to-point and others are point-to-multipoint. Figure 5 illustrates the various virtual channel connections (VCCs)--also known as virtual circuit connections--that are used in LANE configuration. In this figure, LE server stands for the LANE server, LECS stands for the LANE configuration server, and BUS stands for the LANE broadcast-and-unknown server.

Figure 5: LANE VCC Types

The following section describes various processes that occur, starting with a client requesting to join an emulated LAN after the component routers have been configured.

Client Joining a Emulated LAN

The following process normally occurs after a LANE client has been enabled:

The client sets up a connection to the LANE configuration server--a bidirectional point-to-point Configure Direct virtual channel connection (VCC)--to find the ATM address of the LANE server for its emulated LAN.
LANE clients find the LANE configuration server by using the following methods in the listed order:
Using the same VCC, the LANE configuration server returns the ATM address and the name of the LANE server for the client's emulated LAN.
The client sets up a connection to the LANE server for its emulated LAN (a bidirectional point-to-point Control Direct VCC) to exchange control traffic.
Once a Control Direct VCC is established between a LANE client and a LANE server, it remains up.
The server for the emulated LAN sets up a connection to the LANE configuration server to verify that the client is allowed to join the emulated LAN--a bidirectional point-to-point Configure Direct (server) VCC. The server's configuration request contains the client's MAC address, its ATM address, and the name of the emulated LAN. The LANE configuration server checks its database to determine whether the client can join that LAN; then it uses the same VCC to inform the server whether the client is or is not allowed to join.
If allowed, the LANE server adds the LANE client to the unidirectional point-to-multipoint Control Distribute VCC and confirms the join over the bidirectional point-to-point Control Direct VCC. If disallowed, the LANE server rejects the join over the bidirectional point-to-point Control Direct VCC.
Sending LE ARP packets for the broadcast address sets up the VCCs to and from the broadcast-and-unknown server.

Address Resolution

As communication occurs on the emulated LAN, each client dynamically builds a local LANE ARP (LE ARP) table. A client's LE ARP table can also have static, preconfigured entries. The LE ARP table maps MAC addresses to ATM addresses.

Note LE ARP is not the same as IP ARP. IP ARP maps IP addresses (Layer 3) to Ethernet MAC addresses (Layer 2); LE ARP maps emulated LAN MAC addresses (Layer 2) to ATM addresses (also Layer 2).

When a client first joins an emulated LAN, its LE ARP table has no dynamic entries and the client has no information about destinations on or behind its emulated LAN. To learn about a destination when a packet is to be sent, the client begins the following process to find the ATM address corresponding to the known MAC address:

For unknown destinations, the client sends a packet to the broadcast-and-unknown server, which forwards the packet to all clients via flooding. The broadcast-and-unknown server floods the packet because the destination might be behind a bridge that has not yet learned this particular address.

Multicast Traffic

When a LANE client has broadcast or multicast traffic, or unicast traffic with an unknown address to send, the following process occurs:

This VCC branches at each ATM switch. The switch forwards such packets to multiple outputs. (The switch does not examine the MAC addresses; it simply forwards all packets it receives.)

Typical LANE Scenarios

In typical LANE cases, one or more Cisco 7000 family routers, or Cisco 4500 series routers are attached to a Cisco LightStream ATM switch. The LightStream ATM switch provides connectivity to the broader ATM network switch cloud. The routers are configured to support one or more emulated LANs. One of the routers is configured to perform the LANE configuration server functions. A router is configured to perform the server function and the broadcast-and-unknown server function for each emulated LAN. (One router can perform the server function and the broadcast-and-unknown server function for several emulated LANs.) In addition to these functions, each router also acts as a LANE client for one or more emulated LANs.

This section presents two scenarios using the same four Cisco routers and the same Cisco LightStream ATM switch. Figure 6 illustrates a scenario in which one emulated LAN is set up on the switch and routers. Figure 7 illustrates a scenario in which several emulated LANs are set up on the switch and routers.

The physical layout and the physical components of an emulated network might not differ for the single and the multiple emulated LAN cases. The differences are in the software configuration for the number of emulated LANs and the assignment of LANE components to the different physical components.

Single Emulated LAN Scenario

In a single emulated LAN scenario, the LANE components might be assigned as follows:

Figure 6 illustrates this single emulated LAN configured across several routers.

Figure 6: Single Emulated LAN Configured on Several Routers

Multiple Emulated LAN Scenario

In the multiple LAN scenario, the same switch and routers are used, but multiple emulated LANs are configured. See Figure 7.

Figure 7: Multiple Emulated LANs Configured on Several Routers

In the following scenario, three emulated LANs are configured on four routers:

In this scenario, once routing is enabled and network level addresses are assigned, Router 1 and Router 2 can route between the man and the eng emulated LANs, and Router 3 and Router 4 can route between the man and the mkt emulated LANs.

PPP for Wide-Area Networking

The Point-to-Point Protocol (PPP), described in RFCs 1661 and 1332, encapsulates network layer protocol information over point-to-point links. You can configure PPP on asynchronous serial, HSSI, ISDN, and synchronous serial physical interfaces:

When PPP encapsulation is enabled on physical interfaces, PPP is also enabled on calls placed by the dialer interfaces that use the physical interfaces.

The current implementation of PPP supports the following options:

The software always sends option 5 and negotiates for options 3 and 4 if so configured. All other options are rejected.

Cisco supports the following upper-layer protocols: AppleTalk, Bridging, CLNS, DECnet, IP, IPX, VINES, and XNS.

The software provides PPP as an encapsulation method. It also provides the Challenge Handshake Authentication Protocol (CHAP) and Password Authentication Protocol (PAP) on serial interfaces running PPP encapsulation. The following sections describe the tasks to configure PPP routing features.

Our PPP implementation supports the following features:


Cisco's implementation of the SMDS protocol is based on cell relay technology as defined in the Bellcore Technical advisories, which are based on the IEEE 802.6 standard. We provide an interface to an SMDS network using DS-1 or DS-3 high-speed transmission facilities. Connection to the network is made through a device called an SDSU--an SMDS channel service unit/digital service unit (CSU/DSU) developed jointly by Cisco Systems and Kentrox. The SDSU attaches to a Cisco router or access server through a serial port. On the other side, the SDSU terminates the line.

Cisco's implementation of SMDS supports the IP, DECnet, AppleTalk, XNS, Novell IPX,
Banyan VINES, and OSI internetworking protocols, and transparent

Cisco's implementation of SMDS also supports SMDS encapsulation over an Asynchronous Transfer Mode (ATM) interface. For more information and for configuration tasks, see the "Configuring ATM" chapter.

Routing of AppleTalk, DECnet, IP, IPX, and ISO CLNS is fully dynamic; that is, the routing tables are determined and updated dynamically. Routing of the other supported protocols requires that you establish a static routing table of SMDS neighbors in a user group. Once this table is set up, all interconnected routers and access servers provide dynamic routing.

When configuring IP routing over SMDS, you may need to make adjustments to accommodate split horizon effects. Refer to the "Configuring IP Routing Protocols" chapter in the Network Protocols Configuration Guide, Part 1 for information about how our software handles possible split horizon conflicts. By default, split horizon is disabled for SMDS networks.

Cisco's SMDS implementation includes multiple logical IP subnetworks support as defined by
RFC 1209. This RFC describes routing IP over an SMDS cloud in which each connection is considered a host on one specific private network, and points to cases where traffic must transit from network to network.

Cisco's implementation of SMDS also provides the Data Exchange Interface (DXI) Version 3.2 with heartbeat. The heartbeat mechanism periodically generates a heartbeat poll frame.

When a multicast address is not available to a destination, pseudobroadcasting can be enabled to broadcast packets to those destinations using a unicast address.

LAPB and X.25

X.25 is one of a group of specifications published by the International Telecommunication Union Telecommunication Standardization Sector (ITU-T); these specifications are international standards that are formally called Recommendations. The ITU-T Recommendation X.25 defines how connections between data terminal equipment (DTE) and data communications equipment (DCE) are maintained for remote terminal access and computer communications. The X.25 specification defines protocols for two layers of the Open Systems Interconnection (OSI) reference model. The data link layer protocol defined is Link Access Procedure, Balanced (LAPB). The network layer is sometimes called the packet level protocol (PLP), but is commonly (although less correctly) referred to as the X.25 protocol.

The ITU-T updates its Recommendations periodically. The specifications dated 1980 and 1984 are the most common versions currently in use. Additionally, the International Standards Organization (ISO) has published ISO 7776:1986 as an equivalent to the LAPB standard, and ISO 8208:1989 as an equivalent to the ITU-T 1984 X.25 Recommendation packet layer. Cisco's X.25 software follows the ITU-T 1984 X.25 Recommendation, except for its Defense Data Network (DDN) and Blacker Front End (BFE) operation, which follow the ITU-T 1980 X.25 Recommendation.

Note The ITU-T carries out the functions of the former Consultative Committee for International Telegraph and Telephone (CCITT). The 1988 X.25 standard was the last published as a CCITT Recommendation. The first ITU-T Recommendation is the 1993 revision.

In addition to providing remote terminal access, Cisco's X.25 software provides transport for LAN protocols--IP, DECnet, XNS, ISO CLNS, AppleTalk, Novell IPX, Banyan VINES, and Apollo Domain--and bridging. For information about these protocols, refer to the Network Protocols Configuration Guide, Part 1, Network Protocols Configuration Guide, Part 2, and Network Protocols Configuration Guide, Part 3.

Briefly, Cisco IOS X.25 software provides the following capabilities:

Cisco's X.25 implementation does not support fast switching.

Copyright 1989-1998 © Cisco Systems Inc.