Nortel Networks CG030601 User Manual

Configuration Guide

Contivity Secure IP Services Gateway

Frame Relay on Contivity Secure IP Services Gateway

Introduction to Frame Relay

Frame relay is a high-speed, packet-switching WAN protocol that connects
geographically dispersed LANs. A public network provider usually offers frame relay.
However, private organizations can acquire and manage their own frame relay networks
as well.

Frame relay is a connection-oriented protocol. This means it relies on end-to-end paths
between devices connected across the network. It implements these connections using
permanent virtual circuits (PVCs).

Frame relay assumes that networks use transmission lines with low error rates, such as
digital transmission media. Therefore, frame relay provides only basic error detection
with no error recovery. This minimizes the processing required for each packet, allowing
frame relay networks to operate at high speeds with few network delays.

Because frame relay performs only basic error checking, end stations running upper-layer
protocols such as the Internet Protocol (IP) are responsible for resending packets that did
not transmit correctly the first time.

Permanent virtual circuits

A permanent virtual circuit (PVC) is a dedicated logical path that connects two devices
over a network. When configured, a PVC is always available to the connected devices; a
PVC does not require setup before data can travel across the network, nor does it need to
be disconnected after data has passed. Because many PVCs can coexist for one physical
line, devices can share the bandwidth of the transmission line.

Frame Relay packets

The structure of a frame relay packet is shown in the following figure.

Figure 1 Frame Relay packet

The packet’s header field includes the following components:

Data link connection identifier (DLCI)

The DLCI is the virtual circuit identification number. The frame relay network uses the
DLCI to direct basic data flow. You configure the DLCI for PVCs.

CG030601 2.00 July 2003 Page: 1 of 14

Configuration Guide

Contivity Secure IP Services Gateway

Frame Relay on Contivity Secure IP Services Gateway

Command/response bit (C/R)

ITU-T (formerly CCITT) standards do not use this bit.

Forward explicit congestion notification (FECN) and backward explicit congestion
notification (BECN)

The FECN and BECN indicate congestion on the network. For information about how the
frame relay software uses these bits, see “Congestion Control”.

Discard eligibility (DE)

The DE bit allows the router to mark specific frames as low priority (discard eligible)
before transmitting them to the frame relay network.

Extended address bit (EA)

The EA bit signals whether the next byte is part of the address. This bit indicates the last
byte of the DLCI.

Management protocols

Frame relay is an access protocol that runs between a CES or data terminal equipment
(DTE) and a switch or data communications equipment (DCE). The CES and the switch
use the Data Link Control Management Interface (DLCMI) to exchange information
about the interface and the status of each virtual circuit.

DLCMI supports three standard data link management specifications: LMI, ANSI T1.617
Annex D, and CCITT (now ITU-T) Q.933 Annex A.

• The networking industry first developed the local management interface (LMI)

specification. The LMI approach is asymmetric; the router sends a status-inquiry
message to the network, signaling that the router’s connection to the network is
functioning. The network replies with a status response.

• ANSI modified the LMI specification and incorporated it as Annex D to ANSI

standard T1.617. The ANSI method is generally similar to the LMI approach.

• The CCITT (now ITU-T) modified the ANSI standard and adopted it as Annex A

to Q.933. The CCITT Annex A specification is similar to Annex D, but it uses an
international numbering scheme.

Be sure to configure the frame relay interface on the CES to use the same management
protocol as the switched network to which it is connected.

CG030601 2.00 July 2003 Page: 2 of 14

Configuration Guide

Contivity Secure IP Services Gateway

Frame Relay on Contivity Secure IP Services Gateway

Address resolution for PVCs

Address resolution for PVCs maps a remote network address such as an IP address to a
local DLCI number. IP uses the Address Resolution Protocol (ARP). ARP dynamically
generates an ARP table of addresses and DLCI numbers by sending messages back and
forth to each network node to gather address information. This process increases
broadcast traffic across the network.

Committed information rate

The committed information rate (CIR) is the rate at which the network supports data
transfer under normal operations. Its name is descriptive: you have a contract with your
carrier, who has committed to providing a given throughput, here called the committed
information rate. The CIR is measured in bits per second. You configure this value that
the carrier provides per virtual circuit.

When configuring the CIR, consider the following:

CIR of 0

You can contract with a carrier for a CIR of 0, which yields best-effort service at low cost.
The carrier transmits data, but does not commit to providing a specified throughput. To
configure a CIR of 0, set both the throughput (which is the CIR) and the committed burst
(Bc) to 0, and set the excess burst (Be) to a value greater than 0. For more information
about burst rates, see the next section, “Committed burst rate and excess burst rate.”

Maximum CIR

The maximum CIR should not be greater than the speed of the access line on the slower
end of a virtual circuit. In a big pipe/little pipe topology likely CIRs at the remote sites
would be 32 Kb/s, 56 Kb/s, or 64 Kb/s. If you configure CIRs for these virtual circuits at
the central site, you can use CIR enforcement (described in the next section) to prevent
the big pipe from sending traffic that exceeds the PVC CIRs.

Committed burst rate and excess burst rate

The committed burst rate (Bc) defines the number of bits that the CES can transmit over a
specified time interval (Tc) when congestion is occurring. The excess burst (Be) defines
the number of extra bits that the CES attempts to send over the Tc when there is no
congestion. Both the Bc and the Be are values that you configure.

The sum of the Bc and the Be is the maximum amount of traffic that can travel across the
network per Tc when there is no congestion. If you set the Be to a value greater than zero,
the CES can send traffic exceeding the CIR. To enforce the CIR, that is, to limit traffic
that the CES can send to the amount of the CIR, set the Be to 0.

CG030601 2.00 July 2003 Page: 3 of 14

Configuration Guide

Contivity Secure IP Services Gateway

Frame Relay on Contivity Secure IP Services Gateway

Congestion control

Network congestion can degrade network performance. Congestion occurs when a node
receives more frames than it can process, or sends more frames than the transmission line
can handle. The frame relay network informs the nodes of congestion so that they can
reduce the amount of traffic across the network.

In the frame relay packet header, there are two bits that the network sets to alert nodes of
network congestion. These bits, as defined by the frame relay specification, are the
forward explicit congestion notation (FECN) bit and the backward explicit congestion
notation (BECN) bit.

If the network detects congestion, it alerts the CES in the same direction as the received
frame by changing the frame’s FECN bit from 0 to 1. For nodes in the opposite direction
of the received frame, it changes the frame’s BECN bit from 0 to 1.

CG030601 2.00 July 2003 Page: 4 of 14

Configuration Guide

Contivity Secure IP Services Gateway

Frame Relay on Contivity Secure IP Services Gateway

Configuring Frame Relay on CES

Configuring frame relay on the CES can be accomplished through the command line
interface (CLI) or through the Web GUI.

To enable Frame Relay through the CLI

1. To configure a physical interface on a slot and connector, navigate to the top-level
box prompt and enter:

Interface serial <slot_number>/<interface_number>

slot_number is the number of the slot on which the link module is located.
interface_number is the number of the module on which the slot is located.

After you configure a physical interface, the CLI returns a prompt that specifies your
current working location

.

For example, the following command configures a serial interface on slot 6, interface

1.

CES(config)#interface serial 6/1
CES(config-if)#

2. Configure description and circuit number. The circuit number is only relevant for
administration purposes only. (Not used by system for identification of circuit.)

CES(config-if)#description “frame relay 61”
CES(config-if)#circuit-id 1

3. Configure line rate and interface filter.

CES(config-if)# filter “deny all”

quotes are needed for filter names with 2 words or more)

(

CES(config-if)# data-rate 1536

(parameter is entered in kilobits, only valid parameters are multiples of 64k starting at
64k and ending at 2048k)

4. Configure frame-relay connection type with the following command:

Frame-relay connection-type <connection-type>

<connection-type> can be:

direct for placing 2 FR CES ports back-to-back
switched (default) to connection to a frame-relay cloud.

Example:

CES(config-if)#frame-relay connection-type switched

CG030601 2.00 July 2003 Page: 5 of 14