Table Of Contents
Configuring an ADSL WAN Interface Card on Cisco 1700 Series Routers
Preventing Electrostatic Discharge Damage
Connecting an ADSL WIC to the Network
Configuring the ADSL Interface
Using POTS Splitters and Microfilters
Common Splitter and Microfilter Configurations
Telephone Company-Installed Splitter
Router and Telephone Using Separate Telephone Lines
Configuring Quality of Service Parameters
Low Latency Queuing (Priority Queuing with Class-Based Weighted Fair Queuing)
Multilink PPP over ATM with Link Fragmentation and Interleaving
Weighted Random Early Detection
ATM per-VC Queuing and VC Bundling
VC Bundling Configuration Example
ATM Cell Loss Priority Bit Marking
Configuring the SCC Clock Rate
Configuring FRF.5 and FRF.8 Internetworking Functions
Obtaining Technical Assistance
Configuring an ADSL WAN Interface Card on Cisco 1700 Series Routers
This document describes asymmetric digital subscriber line (ADSL) one-port wide area network (WAN) interface cards (WICs). These cards provide ADSL high-speed digital data transfer between a single customer premises equipment (CPE) subscriber and a central office.
ADSL WICs are available in three variations: ADSL over POTS (WIC-1ADSL), ADSL over POTS with Dying Gasp support (WIC-1ADSL-DG), and ADSL over ISDN with DyingGasp support (WIC-1ADSL-I-DG). The ADSL over POTS WICs are commonly used to provide ADSL services over ordinary telephone lines. The ADSL over ISDN WIC is used to provide ADSL services in those areas of the world which have extensive ISDN backbones already in place.
This document contains the following sections:
•
Connecting an ADSL WIC to the Network
•
•
•
•
•
•
Feature Overview
Figure 1 shows a data network with the card. Figure 2 shows a voice network with the card.
Figure 1 ADSL WIC in a Cisco 1700 Series Router Data Network
Figure 2 ADSL WIC in a Cisco 1700 Series Router Voice Network
On Cisco 1700 series routers, an ADSL WIC fits into a Cisco 1700 series router chassis. The card supports data and voice networks through the asynchronous transfer mode (ATM) protocol with the AAL5 format. ATM quality of service (QoS) for permanent virtual circuits (PVCs) is also supported. Figure 3, Figure 4, and Figure 5 show the various ADSL WICs.
Figure 3 ADSL over POTS WIC
Figure 4 ADSL over POTS with Dying Gasp WIC
Figure 5 ADSL over ISDN with Dying Gasp WIC
Memory Requirements
The memory requirements for running the full-featured Cisco 1700 router encryption images with the ADSL WICs are as follows:
•
•
Related Documents
The following documents provide additional information about installing and configuring ADSL WICs and configuring the router software:
•
•
•
•
The following document provides additional information about configuring QoS features and Frame Relay Forum (FRF) internetworking functions on ADSL WICs.
•
FCC Notice
The following FCC Notice applies to the Cisco 1700 series ADSL WIC:
WIC-1ADSL complies with FCC part 68 FCC ID:5B1USA-42011-DL-N
Safety Warnings
Safety warnings appear throughout this publication in procedures that can harm you if they are performed incorrectly. A warning symbol precedes each warning statement.
Warning Conventions
Power Supply Warnings
The following warnings apply when you are installing a card or working with the power supply:
Warning
Warning
Warning
Warning
Warning
Warning
Warning
The following warning applies when this product is used in Australia.
Warning
Electrical Warnings
The following warnings apply when you are working with electricity:
Warning
Warning
Warning
Warning
Warning
Warning
Warning
Warning
Warning
Follow these guidelines when working on equipment powered by electricity:
•
•
•
–
–
•
•
•
If an electrical accident occurs, proceed as follows:
•
•
•
•
Preventing Electrostatic Discharge Damage
Electrostatic discharge (ESD) can damage equipment and impair electrical circuitry. It can occur when printed circuit cards are improperly handled and can result in complete or intermittent failures. Always follow ESD prevention procedures when removing and replacing cards. Ensure that the router chassis is electrically connected to earth/ground. Wear an ESD-preventive wrist strap, ensuring that it makes good skin contact. Connect the clip to an unpainted surface of the chassis frame to safely channel unwanted ESD voltages to ground. To guard against ESD damage and shocks, the wrist strap and cord must be used properly. If no wrist strap is available, ground yourself by touching the metal part of the chassis.
Caution
Connecting an ADSL WIC to the Network
For this connection, use a standard lavender RJ-11 cable.
Note
The following steps tell how to connect the card, using the standard lavender RJ-11 cable; the steps also apply to the lavender crossover cable with the blue stripe.
Step 1
Step 2
Step 3
Figure 6 Connecting an ADSL WIC to a Wall Jack
Step 4
The following warning applies to routers that use a DC power supply:
Step 5
Configuring the ADSL Interface
Whenever you install a new WIC, or if you want to change the configuration of an existing interface, you must configure the interface. If you replace a card that was already configured, the router recognizes it and brings up the interface in the existing configuration.
Before you configure an interface, have the following information available:
•
•
Timesaver
You can configure the new interface and other router parameters by using any of the following methods:
•
•
•
These procedures are explained in the following sections. To change the settings shown in the examples, and for further information, refer to the IOS configuration guides and command references. If you have questions or need help, see the section "Obtaining Technical Assistance" later in this document.
Default Commands
The IOS software provides the following default configurations for ADSL-specific parameters.
The following default command sets the ADSL operating mode:
dsl operating-mode autoThe following command sets the ATM virtual circuit-per-virtual path (vc-per-vp) configuration for the router:
atm vc-per-vp 256Defaults for WIC-1ADSL-I-DG
For the WIC-1ADSL-I-DG, the default command operating-mode auto sets the carrier tone range from 33 to 56 to meet the requirements of the Deutsche Telekom U-R2 specification.
Alternately, to set the carrier tone range from 29 to 48, use the command
dsl operating-mode auto tone lowThis command, dsl operating-mode auto tone low, is not available on the WIC-1ADSL or WIC-1ADSL-DG.
Configuration Mode
You can configure the interfaces on your ADSL WIC manually by entering IOS commands on the command line. This method, called configuration mode, provides the greatest power and flexibility.
For further information about these commands, refer to the IOS configuration guides and command references.
Before you begin, disconnect all WAN cables from the router to keep it from running the AutoInstall process. The router tries to run AutoInstall whenever you power it on if there is a WAN connection on both ends and the router does not have a valid configuration file stored in NVRAM (for instance, when you add a new interface). It can take several minutes for the router to determine that AutoInstall is not connected to a remote Transmission Control Protocol/Internet Protocol (TCP/IP) host.
To enter configuration mode, follow this procedure:
Step 1
Step 2
Would you like to enter the initial dialog? [yes]:Enter no. You now enter the normal operating mode of the router.
Note
Step 3
Router> enablePassword:Configuration changes can be made only in enable mode. The prompt changes to the privileged EXEC (enable) prompt (Router#):
Router#Step 4
Router# config terminalRouter(config)#The router enters global configuration mode, indicated by the Router(config)# prompt.
Step 5
Router(config)# ip routingRouter(config)# appletalk routingRouter(config)# ipx routingFor complete information about global configuration commands, refer to the IOS configuration guides and command references.
Step 6
Router(config)# interface atm 0Router(config-if)#The prompt changes again to show that you are in interface configuration mode.
Note
Step 7
Router (config-if) dsl operating-mode mode
Caution
Step 8
Router(config-if)# ip address 1.10.16.16 255.255.0.0Router(config-if)# appletalk static cable-range 5-5Router(config-if)# appletalk zone ZZRouter(config-if)# ipx network B004Step 9
Router (config-if)# pvc 0/33Router (config-if-atm-vc)# protocol ip 1.10.16.16 broadcastRouter (config-if-atm-vc)# vbr-rt 160 160 1Router (config-if-atm-vc)# encapsulation aal5snapStep 10
Step 11
Router# show running-configTo see the configuration currently stored in NVRAM, enter the show startup-config command:
Router# show startup-configStep 12
Router# copy running-config startup-configBuilding configuration. . .[OK]Router#The router is now configured to boot in the new configuration.
AutoInstall
The AutoInstall process is designed to configure the router automatically after it connects to your WAN. For AutoInstall to work properly, a TCP/IP host on your network must be configured to provide the configuration files. The TCP/IP host can reside anywhere on the network if the following two conditions are met:
•
•
This functionality is coordinated by your system administrator at the TCP/IP host site. You should not try to use AutoInstall unless the required files are installed on the TCP/IP host.
Follow this procedure to prepare your router for the AutoInstall process:
Step 1
Step 2
The router loads the operating system image from Flash memory. If the remote end of the WAN connection is connected and properly configured, the AutoInstall process begins.
Step 3
Router# copy running-config startup-configBuilding configuration. . .[OK]Router#
Note
System Configuration Dialog
You can configure the router manually, using the System Configuration dialog (also called the Setup facility). Unlike configuration mode, the System Configuration dialog prompts you for each response.
Before you begin, disconnect all WAN cables from the router to keep it from trying to run the AutoInstall process. The router tries to run AutoInstall whenever you power it on if there is a WAN connection on both ends and the router does not have a configuration file stored in NVRAM. It can take several minutes for the router to determine that AutoInstall is not connected to a remote TCP/IP host.
This section shows a sample configuration using the System Configuration dialog. You should enter values appropriate for your router and network. To change the settings shown in the examples, and for further information, refer to the IOS configuration guides and command references.
Many prompts in the System Configuration dialog include default answers, shown in square brackets following the question. Enter your response, or press Return to accept the default answer.
You can request help at any time by entering a question mark (?) at the System Configuration dialog prompt.
Follow this procedure to configure the router, using the System Configuration dialog:
Step 1
Step 2
Would you like to enter the initial dialog? [yes]:Press Return or enter yes to enter the System Configuration dialog.
Note
Step 3
Step 4
Step 5
Step 6
Step 7
Use this configuration? [yes/no]: yesBuilding configuration...Use the enabled mode 'configure' command to modify this configuration.Press RETURN to get started!The configuration is saved. If you added an interface, the router reboots in the new configuration when you press Return.
You can configure the router for ATM and ADSL parameters using the following scenarios:
•
•
•
•
•
The scenario configurations above are identical to those for the Cisco 827-4V router. To configure these scenarios on the Cisco 1720, 1721, 1751, and 1760 routers with the ADSL WIC, refer to the Cisco 827 Routers Software Configuration Guide.
You can also access the information online at the following location:
http://www.cisco.com/univercd/cc/td/doc/product/access/acs_fix/827/index.htm.
Example Configuration
The following example shows a Cisco 1751 router configured for bridging on the ATM interface with an ADSL over POTS card:
Current configuration :!version 12.2no parser cacheno service single-slot-reload-enableservice timestamps debug uptimeservice timestamps log uptimeno service password-encryption!hostname 1751-uut1!interface ATM0/0mtu 4000ip address 1.0.0.1 255.0.0.0atm vc-per-vp 256no atm ilmi-keepalivepvc 0/16 ilmipvc 88/88encapsulation aal5snap!bundle-enabledsl operating-mode autobridge-group 1!interface FastEthernet0/0ip address 6.0.0.1 255.0.0.0speed autohalf-duplexbridge-group 1no cdp enable!ip classlessno ip http server!bridge 1 protocol ieee!line con 0line aux 0line vty 0 4login!endUsing POTS Splitters and Microfilters
POTS splitters and microfilters are used on telephone lines to ensure voice- and data-call quality. This section describes splitters and microfilters and tells how and when to use them with the Cisco 1700 series routers. POTS splitters result in the best data and voice performance when the router and the telephone are used on the same telephone line.
POTS Splitters
A POTS splitter (also called a splitter) is installed on a telephone line that is connected to both data (high-frequency) and voice (low-frequency) devices. The splitter routes the high-frequency and low-frequency signals on the telephone line to the correct device. Signals intended for the router can disrupt voice calls; signals intended for voice calls can affect router operation.
Most splitters must be installed by the telephone company; however, some splitters can be installed by the customer. If you are not sure what type of splitter to use, contact your service provider.
Figure 7 is an example of a type of POTS splitter that is installed at the customer premises by the customer. Other types of POTS splitters are installed by the telephone company on an exterior wall of the customer premises.
Figure 7 POTS Splitters
Microfilters
Microfilters are installed on telephones to improve voice-call quality when voice and data equipment are using the same telephone line (twisted pair). You should use microfilters with the Cisco 1700 series routers only when the two following conditions exist:
•
•
Figure 8 shows one type of microfilter.
Figure 8 Microfilter
Common Splitter and Microfilter Configurations
This section describes the most common scenarios for using splitters and microfilters with the Cisco 1700 series routers. The scenarios are listed from most common to least common.
Telephone Company-Installed Splitter
This scenario is described below and illustrated in Figure 9.
•
•
•
•
Figure 9 Telephone Company-Installed Splitter
Customer-Installed Splitter
This scenario is described below and illustrated in Figure 10.
•
•
•
•
•
•
Figure 10 Customer-Installed Splitter
Router and Telephone Using Separate Telephone Lines
This scenario is described below and illustrated in Figure 11.
•
•
Figure 11 No Splitter, Optional Microfilter
Configuring Quality of Service Parameters
This section discusses quality of service (QoS) parameters that can be configured for the 1700 series platforms when using the ADSL WIC. The following features are included:
•
•
•
•
•
Low Latency Queuing (Priority Queuing with Class-Based Weighted Fair Queuing)
Low latency queuing (LLQ) allows strict priority queuing (PQ) to class-based weighted fair queuing (CBWFQ). This priority queuing allows delay-sensitive data such as voice packets to be de-queued and sent before other packet traffic, reducing jitter in voice conversations. To configure LLQ, enter the priority command under the CBWFQ configuration.
Configuration Example
The following example shows a Cisco 1751 router configured with LLQ.
hostname zorrousername ruby-1 password 7 36497A4872384A!class-map match-all VOIPmatch ip dscp 32class-map CRITICALmatch access-group 100!policy-map 1751_ADSLclass CRITICALpriority 48class VOIPbandwidth 64set ip precedence 6!interface Loopback1ip address 10.0.0.10 255.255.255.252!interface ATM0/0dsl operating-mode autono ip addressno atm ilmi-keepalive!int atm0/0.1 point-to-pointpvc 0/33vbr-rt 320 320 30tx-ring-limit 3protocol ppp Virtual-Template1!interface Virtual-Template1bandwidth 320ip unnumbered Loopback1ip mroute-cacheservice-policy output 1751_ADSLppp multilinkppp multilink fragment-delay 4ppp multilink interleave!access-list 100 permit udp any any precedence critical!dial-peer voice 201 voipdestination-pattern 3640200session target ipv4:10.0.0.11ip qos dscp cs4 mediaip qos dscp cs4 signallingDiffServ
DiffServ addresses the clear need for relatively simple and coarse methods of categorizing traffic into different classes and applying QoS parameters to those classes. DiffServ supports class-based marking.
Cisco Express Forwarding (CEF) mode is required for DiffServ support. To enable CEF, enter the ip cef command.
Configuration Example
The following example shows a Cisco 1751 router configured with DiffServ:
access-list 102 permit udp host 16.0.0.4 host 15.0.0.5access-list 103 permit udp host 16.0.0.4 host 13.0.0.5ip cefclass-map match-all traffic-INTRAmatch access-group 102class-map match-all traffic-INTERmatch access-group 103class-map match-all traffic-dscp1match ip dscp 1class-map match-any traffic-prec3match ip dscp 24match ip dscp 25match ip dscp 26match ip dscp 27policy-map ADSL-outclass traffic-INTRAbandwidth percent 8class traffic-dscp1set ip dscp 5class traffic-prec3set ip precedence 2class traffic-INTERbandwidth percent 8class class-defaultfair-queue!interface ATM0/0dsl operating-mode autono ip addressno atm ilmi-keepalive!interface ATM0/0.1 point-to-pointdescription COLLEGAMENTOmtu 576ip address 1.0.0.1 255.0.0.0pvc 99/99protocol ip 2.0.0.2 broadcastvbr-nrt 142 142 1tx-ring-limit 3oam-pvc 0oam retry 5 5 1encapsulation aal5snapservice-policy out ADSL-out!dial-peer voice 201 voipdestination-pattern 3640200session target ipv4:14.0.0.3playout-delay maximum 300ip qos dscp cs4 mediaip qos dscp cs4 signalingCommitted Access Rate
Committed access rate (CAR) allows you to limit bandwidth transmission rates to traffic sources and destinations and allows you to specify policies for handling traffic that both conforms to and breaches the specified bandwidth allocations.
CEF mode is required for CAR support. To enable CEF, enter the ip cef command.
To enable CAR, enter the rate-limit command under the atm interface.
Configuration Example
The following example shows a Cisco 1751 router configured with CAR:
ip cefinterface ATM0/0.1 point-to-pointdsl operating-mode automtu 576ip address 10.0.0.10 255.255.255.0rate-limit output 368000 2000 2000 conform-action set-dscp-transmit 40 exceed-action set-dscp-transmit 48pvc 0/33protocol ip 10.0.0.9 broadcastvbr-rt 160 160 1encapsulation aal5snap!Multilink PPP over ATM with Link Fragmentation and Interleaving
This feature allows multilink PPP (MLPPP) encapsulation over a single slow link to fragment and interleave packets to a small enough size that the delay requirements of delay-sensitive traffic will be met.
Fragment size at the MLPPP bundle can be configured by using the virtual-template interface bandwidth command and the ppp multilink fragment-delay command. The ideal fragment size for MLPPP over ATM should allow the fragments to fit into an exact multiple of ATM cells. These commands calculate fragment size using the following formula:
fragment size = bandwidth x fragment-delay / 8.
For example, if the MLPPP ATM header is 10 bytes and the AAL5 packet overhead is 8 bytes, the fragment size for MLPPP over ATM can be calculated as follows:
fragment size = 48 x # of cells - 10 - 8.
In this case, 2 cells per fragment are desirable, so the fragment size is calculated at 78 bytes.
The total bandwidth usable on this interface is 75 percent of the value declared in the bandwidth command. To change this default value, enter the max-reserved-bandwidth command.
LLQ must be enabled when you configure MLPPP with link fragmentation and interleaving.
Note
MLPPP + LFI Configuration
The following example shows a Cisco 1751 router configuration with MLPPP + LFI:
hostname zorrousername ruby-1 password 7 36497A4872384A!class-map match-all VOIPmatch ip dscp 32class-map CRITICALmatch access-group 100!policy-map 1751_ADSLclass CRITICALpriority 48class VOIPpriority 64set ip precedence 6!interface ATM0/0dsl operating-mode autono ip addressno atm ilmi-keepalive!int atm0/0.1 point-to-pointpvc 0/33vbr-rt 160 160 1tx-ring-limit 3protocol ppp Virtual-Template1!interface Loopback1ip address 10.0.0.10 255.255.255.255interface Virtual-Template1bandwidth 320ip unnumbered Loopback1ip mroute-cacheservice-policy output 1751_ADSLppp multilinkppp multilink fragment-delay 4ppp multilink interleave!access-list 100 permit udp any any precedence criticalWeighted Random Early Detection
You can set a queuing technique on a device's interface to manage how packets are handled when an interface starts to become congested. The queuing technique available for congestion avoidance is called weighted random early detection (WRED). WRED is IP precedence and differentiated services code point (DSCP) value aware.
WRED allows the interface to start dropping packets from selected flows when traffic begins to exceed the interface's traffic thresholds, but before congestion occurs. If the dropped packets are TCP packets, the TCP source recognizes that packets are being dropped, and then lowers its transmission rate. The lowered transmission rate reduces the traffic to the interface, thus avoiding congestion. Because TCP retransmits dropped packets, no actual data loss occurs.
Note
Configuration Example
The following example shows a Cisco 1751 configured with WRED:
random-detect-group 1751_DSLexponential-weighting-constant 5precedence 2 96 256 100precedence 5 192 256 100!interface ATM0/0mtu 1000no ip addressatm vc-per-vp 256no atm ilmi-keepalivepvc 0/16 ilmi!bundle-enable!dsl operating-mode auto!interface ATM0/0.1 point-to-pointdescription configuring limi manage in pvc brings down the atm protocolmtu 4000ip address 1.0.0.1 255.0.0.0pvc 88/88random-detect attach 1751_DSLprotocol ip 2.0.0.2vbr-rt 320 320 30no ilmi manageoam-pvc 40oam retry 3 5 1encapsulation aal5snap!endATM per-VC Queuing and VC Bundling
Parameters can be applied to individual VCs either by using VC classes or by directly applying them to the bundle members. Parameters applied to an individual VC supersede bundle-level parameters. Parameters applied directly to a VC take precedence over the same parameters applied within a class to the VC at the bundle-VC configuration level.
All of the QoS features are supported in per-virtual circuit (VC) and VC bundling mode. The default is per-VC queuing mode.
VC bundling allows individual VCs going to the same destination to be grouped together. Traffic mapping to each VC is based on traffic protocol criteria such as IP precedence. To enable VC bundling, enter the bundle command under the ATM interface.
VC Bundling Configuration Example
The following example shows a VC bundling configuration:
vc-class atm atm-bundlebroadcastoam-pvc manage 1oam retry 3 3 1encapsulation aal5snapprotocol ip inarp broadcastoam-bundle manage 1!vc-class atm vipvbr-rt 256 256 20precedence 5-7bump implicitno protect vcno protect group!vc-class atm highvbr-rt 256 256 20precedence 2-4bump implicitno protect vcno protect group!vc-class atm normalvbr-rt 256 256 20precedence 0-1bump explicit 2no protect vcno protect group!interface ATM0/0description COLLEGAMENTOno ip addressatm vc-per-vp 256no atm ilmi-keepalive!bundle-enabledsl operating-mode auto!interface ATM0/0.1 point-to-pointdescription COLLEGAMENTOip address 2.0.0.2 255.255.0.0bundle MMAclass-bundle atm-bundlepvc-bundle vip 0/33class-vc vippvc-bundle high 0/34class-vc highpvc-bundle normal 0/35class-vc normalATM Cell Loss Priority Bit Marking
When congestion occurs in an ATM network, ATM cells are discarded. One way to control which cells are discarded is to use the cell loss priority (CLP) bit in the ATM header of each cell. The CLP bit may be set to either 1 or 0. Those cells that have the CLP bit set to 1 are always discarded before any of the cells with the CLP bit set to 0.
The ATM CLP bit marking feature allows you to control the CLP setting on Cisco routers. The marking of the CLP bit is implemented on a per-packet basis so that the CLP bit of every ATM cell that belongs to a particular packet is set to either 0 or 1.
Configuration Example
The following is an example of enabling ATM CLP bit marking using the set atm-clp command and modular QoS command-line interface. In this example, all output packets that have an IP Precedence value of 0 are sent with the CLP set to 1. Note that IP CEF must be on when using ATM CLP bit marking.
ip cefclass-map match-all prec0match ip precedence 0policy-map ATM_OUTclass prec0set atm-clpinterface ATM0/0dsl operating-mode autopvc 0/33service-policy output ATM_OUTCompressed RTP
The Real-Time Transport Protocol (RTP), as described in RFC 1889, is used to carry real-time data for voice and video applications. For a typical Voice over IP (VoIP) application, the payload portion of the packet can be smaller than the header. For instance, using the G.729 codec, the payload is 20 bytes, but the IP, User Data Protocol (UDP), and RTP header is 40 bytes. It is inefficient to send the IP, UDP, and RTP header across a slow link without compressing it. The Compressed Real-Time Transport Protocol (cRTP) feature, as defined in RFC 2508, addresses this inefficiency by making the VoIP packet headers smaller.
The basicpremise of cRTP is that although several fields in the IP, UDP, and RTP header change from packet to packet, the differences in these fields from packet to packet are constant. The compression scheme in cRTP encodes the header to reduce the size of the information to be transmitted. With cRTP, a 40-byte IP, UDP, and RTP header of a VoIP packet can be compressed to 2 to 4 bytes per packet, yielding approximately 11.2 kbps of bandwidth for a G.729 codec call with RTP.
cRTP can be applied to an ATM link through cRTP for MLP over ATM, or through cRTP for PPP over ATM.
Configuration Example
The following are examples of cRTP for MLP over ATM, and cRTP for PPP over ATM. The ip rtp header-compression command sets cRTP.
cRTP Using MLP over ATM
interface Loopback1ip address 10.0.0.9 255.255.255.255!interface ATM0/0no ip addressno atm ilmi-keep-alive!dsl operating-mode auto!interface ATM0/0.1 point-to-pointpvc 0/33ip 10.0.0.10vbr-rt 320 320 30tx-ring-limit 3protocol ppp Virtual-Template1!interface Virtual-Template1bandwidth 320ip unnumbered Loopback1ip tcp header-compression iphc-formatservice-policy output ADSL-2ppp multilinkppp multilink fragment-delay 4ppp multilink interleaveip rtp header-compression iphc-formatcRTP Using PPP over ATM
interface Loopback1ip address 10.0.0.9 255.255.255.255!interface ATM0/0no ip addressno atm ilmi-keep-alive!
dsl operating-mode auto!interface ATM0/0.1 point-to-pointpvc 0/33protocol ip 10.0.0.10vbr-rt 320 320 30tx-ring-limit 3protocol ppp Virtual-Template1!interface Virtual-Template1bandwidth 320ip unnumbered Loopback1ip tcp header-compression iphc-formatservice-policy output ADSL-2ip rtp header-compression iphc-formatTunable Transmission Ring
The transmission (tx) ring is the first-in, first-out (FIFO) buffer used to hold frames before transmission at the DSL driver level. The tx ring defines the maximum number of packets that can wait for transmission at Layer 2.
The tx ring complements the ability of LLQ to minimize jitter and latency of voice packets. For maximum voice quality, a low tx ring setting should be used. For maximum data throughput, a high tx ring setting should be used.
You can configure the size of the tx ring for each PVC. The default value is 60. However, the value of the setting can be from 2 through 60.
Note
For example, when the tx ring limit is configured as 3 and LLQ is configured on the PVC, the worst case delay for a voice packet is the time required to transmit three data packets. When the buffering is reduced by configuring the tx ring limit, the delay experienced by voice packets is reduced by a combination of the tx ring and LLQ mechanism.
Note
Configuration Example
The following example is a configuration of the tx ring limit on an ATM PVC interface. To enable the tx ring limit, enter the tx-ring-limit command.
class-map match-all VOIPmatch ip dscp 32class-map CRITICALmatch access-group 100!policy-map 1751_ADSLclass CRITICALpriority 48class VOIPbandwidth 64set ip precedence 6!interface Loopback1ip address 10.0.0.10 255.255.255.252!interface ATM0/0dsl operating-mode autono ip addressno atm ilmi-keepalive!interface ATM0/0.1pvc 0/33vbr-rt 320 320 30tx-ring-limit 3protocol ppp Virtual-Template1!interface Virtual-Template1bandwidth 320ip unnumbered Loopback1ip mroute-cacheservice-policy output 1751_ADSLppp multilinkppp multilink fragment-delay 4ppp multilink interleave!access-list 100 permit udp any any precedence critical!dial-peer voice 201 voipdestination-pattern 3640200session target ipv4:10.0.0.11ip qos dscp cs4 mediaip qos dscp cs4 signallingMLP Bundling
Multilink PPP (MLP), standardized in RFC 1990, is similar to load balancing techniques in that it sends packets across the individual links in a round-robin fashion. However, MLP adds three significant capabilities:
•
•
•
Guest
