ATM Services (AXSM) Software Configuration Guide and Command Reference for MGX Switches, Release 4 - AXSM Configuration Guide [Cisco MGX 8800 Series Switches]

Table Of Contents

AXSM Configuration Guide

AXSM Configuration Concepts

Adding ATM Ports

Channelization on the AXSM-XG

Inverse Multiplexing over ATM

Partitioning Port Resources Between Controllers

Selecting the Port Signaling Protocol

Assigning Static ATM Addresses to Destination Ports

Configuring ILMI on a Port

Configuring ILMI Traps and Signaling

Configuring ILMI Automatic Configuration

Configuring ILMI Dynamic Addressing

Starting ILMI with the Default or Existing Values

Configuring AXSM Line Clock Sources

Procedures for PNNI Links

Verifying PNNI Communications

Verifying PNNI Trunk Communications

Verifying End-to-End PNNI Communications

Configuring SPVCs and SPVPs

Configuring the Slave Side of SPVCs and SPVPs

Configuring the Master Side of SPVCs and SPVPs

Configuring SPVC/SPVP Overrides on Single-Ended Connections

Deleting SPVCs and SPVPs

Defining a Feeder Port

Defining Destination Addresses for Static Links

Configuring Point-to-Multipoint SPVCs and SPVPs

Obtaining the NSAP for a Party

Rerouting a P2MP Party

Deleting a P2MP Party Configuration

AXSM Configuration Procedures

Before You Begin the Configuration Procedures

Inverse Multiplexing over ATM Configuration Procedures

Configuring IMA

Administratively Enabling and Disabling IMA

Testing IMA

NNI Trunk Configuration Procedure with MPLS and PNNI Partitions

UNI Port Configuration Procedure with MPLS and PNNI Partitions

SVC Configuration Procedure

SPVC and SPVP Configuration Procedure

PNNI Virtual Trunk Configuration Procedure

Cisco IGX Feeder to Cisco MGX 8850 Configuration Procedure

Cisco IGX Feeder Removal Procedure

AXSM-XG Channelization Configuration Procedure

Configuring DS3 Paths on a SONET Path Example Procedure

PNNI Feeder Configuration Procedure

Cisco BPX PNNI Trunk Configuration Procedure

AINI Link Configuration Procedure

IISP Link Configuration Procedure

XLMI Link Configuration Procedure

AXSM Configuration Guide

This chapter describes how to configure the AXSM card and provides procedures for adding ATM ports and connections to the physical lines. The types of links and connections presented in this chapter are listed in Table 2-1.

Table 2-1 AXSM Link and Connection Types

AXSM Link or Connection Type

Description

MPLS and PNNI trunks

MPLS and PNNI trunks connect Cisco MGX switches to other Cisco MGX switches.

MPLS and PNNI UNI ports

MPLS and PNNI UNI ports connect Cisco MGX switches to CPE.

Switched Virtual Circuits (SVCs)

SVCs are temporary connections that are brought up and torn down upon request from CPE.

Soft Permanent Virtual Circuits (SPVCs)

SPVCs are permanent connections that can be rerouted if a link fails.

PNNI virtual trunks

PNNI virtual trunks are used to traverse public networks. The virtual trunk endpoints are on separate networks, but the path between the networks is treated like a single link.

Inverse Multiplexing over ATM (IMA)

Inverse Multiplexing over ATM (IMA) is a protocol that runs on the AXSM-32-T1E1-E. IMA allows you to combine multiple T1 or E1 interfaces into a single, high-speed IMA interface.

Channelized paths

Channelization is possible on AXSM-XG cards of the Cisco MGX8950. Channelization makes it possible to implement multiple SONET paths on a single line. It also makes it possible to implement multiple DS3 paths on a single SONET path.

Cisco MGX 8850 PXM1-based feeder trunks

Feeder trunks link a feeder switch, such as a Cisco MGX 8230 or
MGX 8250 switch, to a Cisco MGX 8850 PXM45-based switch. The feeder switch concatenates relatively low speed traffic and feeds it over a higher speed interface to the Cisco MGX 8850 switch, which provide the link to the ATM network core.

Cisco BPX PNNI trunks

Cisco BPX PNNI trunks provide PNNI links between Cisco MGX 8850 and MGX 8950 switches and Cisco BPX switches that support PNNI. The Cisco BPX switch supports PNNI when connected to the Cisco SES PNNI Controller.

ATM Inter-Network Interface (AINI) links

AINI links enable connectivity between two independent PNNI networks and block the PNNI database exchange so the two networks remain independent.

Interim Inter-switch Protocol (IISP) links

IISP links enable connectivity between two independent PNNI networks and block the PNNI database exchange so the two networks remain independent. IISP is the predecessor to AINI and should be used only when AINI is not supported on one or both ends of the network link.

Extended Link Management Interface (XLMI) links

XLMI links connect PNNI networks to AutoRoute networks. XLMI links enable the expansion of AutoRoute networks using PNNI, and they facilitate migration from AutoRoute networking to PNNI.

Point-to-Multipoint SPVCs and SPVPs

Point-to-multipoint (P2MP) connections enable a single master endpoint to support several slave endpoints.

Tip You can get configuration information for any command by entering the command without parameters in the CLI.

Caution Before you can configure any ATM connections, you must first complete the general switch configuration procedures described in Cisco MGX 8850 (PXM1E/PXM45), Cisco MGX 8950, and Cisco MGX 8830 Software Configuration Guide, Release 4 .

AXSM Configuration Concepts

This section describes the following AXSM configuration concepts and general procedures:

•Adding ATM Ports

•Channelization on the AXSM-XG

•Inverse Multiplexing over ATM

•Partitioning Port Resources Between Controllers

•Selecting the Port Signaling Protocol

•Assigning Static ATM Addresses to Destination Ports

•Configuring ILMI on a Port

•Configuring AXSM Line Clock Sources

•Procedures for PNNI Links

•Configuring SPVCs and SPVPs

•Defining a Feeder Port

•Defining Destination Addresses for Static Links

•Configuring Point-to-Multipoint SPVCs and SPVPs

The descriptions and procedures in this section use AXSM commands and show the syntax for AXSM commands. See Chapter 3, "AXSM Command Reference" for descriptions of the AXSM commands and parameters.

See Table 1-2 in "Introduction" for a list of the AXSM model numbers, back cards, and the number of possible connections.

Some of the procedures in this section use PXM commands and PNNI commands. Refer to the Cisco MGX 8830, Cisco MGX 8850 (PXM45 and PXM1E), and Cisco MGX 8950 Command Reference, Release 4 for descriptions of the PXM and PNNI commands and parameters.

For more information on port signaling, refer to the Cisco MGX 8850 (PXM1E/PXM45), Cisco MGX 8950, and Cisco MGX 8830 Software Configuration Guide, Release 4 .

For more information on ATM address planning, refer to the Cisco MGX and SES PNNI Network Planning Guide.

For information on additional ILMI management procedures, refer to the Cisco MGX 8850 (PXM1E/PXM45), Cisco MGX 8950, and Cisco MGX 8830 Software Configuration Guide, Release 4 . See the Chapter 3, "AXSM Command Reference" for descriptions of the ILMI commands and parameters.

Adding ATM Ports

The line ports correspond to line connectors on the switch back cards. Bringing up a line establishes minimal connectivity between two nodes. When you add an ATM port to a line, you enable ATM communications over the line.

Each line can support UNI, NNI, VNNI , EVNNI, or EVUNI ports. UNI ports are used for lines that connect to PBXs, ATM routers, and other ATM devices that connect to the core ATM network through the switch. NNI ports are used for trunks that connect to other core ATM network devices, such as another MGX 8850 or MGX 8950 switch. VNNI ports support virtual trunk connections between two ATM end stations. EVNNI and EVUNI are enhanced virtual trunks for network and user connections respectively. UNI, NNI, VNNI , EVNNI, or EVUNI are explained in more detail in the Logical Ports section of Introduction.

Table 2-1 shows the relationship between cards, bays, lines, and logical interface numbers.

Figure 2-1 Relationship Between Cards, Bays, Lines, and Logical Interface Numbers

You must configure one ATM port for each line or trunk to enable ATM communications over that link. You define the port type (UNI, NNI, VNNI, EVNNI, or EVUNI) when you add the ATM port to the line or trunk.

Note For information on adding ports on a channelized path on an AXSM-XG, see the AXSM-XG Channelization Configuration Procedure section in this chapter.

To add an ATM port to a line, use the following procedure.

Step 1 Establish a configuration session using a user name with GROUP1 privileges or higher.

Step 2 Get the line number on which you will add the port. To display a list of the lines and line numbers, enter the following command:
MGX8850.10.AXSM.a > dsplns 
Tip Remember that you cannot configure a line until you have brought it up as described in the Cisco MGX 8850 (PXM1E/PXM45), Cisco MGX 8950, and Cisco MGX 8830 Software Configuration Guide, Release 4 .

Step 3 Verify that the line and port number you want to use is not configured. To display a list of the ports configured on the AXSM card, enter the following command:
MGX8850.10.AXSM.a > dspports 
This command displays all ports on the AXSM card in the ifNum (interface number) column. The interfaces listed include UNI, NNI, VNNI, EVNNI, and EVUNI ports. Pay attention to the port numbers already in use. When you add a port, you must specify a port number that is unique on the AXSM card. For example, if port number 2 is assigned to line 2.1 (bay 2, line 1), you cannot use port 2 on any other line on that AXSM card.

Step 4 To add an ATM port to a line, enter the following command:
MGX8850.10.AXSM.a > addport <bay.line> <guaranteedRate> <maxrate> <sctID> <ifType> 
[-vpi vpi] [-minvpi minvpi] [-maxvpi maxvpi]
The following example command defines a line port as a UNI T3 line:
MGX8850.10.AXSM.a > addport 1 1.1 96000 96000 1 1
The following example command defines a line port as an OC48 NNI trunk:
MGX8850.10.AXSM.a > addport 2 2.1 5651328 5651328 2 2
Step 5 To display a list of the ports configured on the AXSM card, enter the following command:
MGX8850.10.AXSM.a > dspports 
This command displays all configured ports on the AXSM card. Port numbers are listed in the ifNum (interface number) column. If you want to view information on a particular port, note the number of that port.

Step 6 To display the port configuration, enter the following command:
MGX8850.10.AXSM.a > dspport <ifNum>
Replace <ifNum> with the number assigned to the port during configuration. The following example shows the report for this command.
MGX8850.1.AXSM.a > dspport 1
  Interface Number               : 1
  Line Number                    : 1.1
  Admin State                    : Up        Operational State   : Up
  Guaranteed bandwidth(cells/sec): 1412830   Number of partitions: 1
  Maximum bandwidth(cells/sec)   : 1412830   Number of SPVC      : 0
  ifType                         : NNI       Number of SPVP      : 0
  Port SCT Id                    : 2 
  VPI number(VNNI only)          : 0         Number of SVC       : 2
Tip To change the port configuration, enter the cnfport command, or enter the delport command to delete a port configuration. You can also activate and deactivate ports using the upport and dnport commands.

Channelization on the AXSM-XG

Channelization is possible on AXSM-XG cards of the Cisco MGX8950. Channelization makes it possible to implement multiple paths on a single line. These paths can carry an ATM payload by itself, or they can carry DS3 and the DS3 will carry the ATM payload.

CLI commands are available for performing the following functions:

•Configuring a line to operate in channelized mode

•Displaying a single path

•Displaying all paths

•Bringing a path up

•Taking a path down

•Configuring path parameters

•Displaying the status of a path

•Displaying path alarms

•Displaying path performance monitoring counters

•Displaying path ATM cell counters

When a line is brought up initially, there is one path with a width of 48. To implement channelization, you set the path to the desired width (1, 3, or 16). A width of 48 results in one path only (clear channel).

You provision a channelized path by setting the administrative status to up. For any path with a width greater than one, this results in provisioning ATM service.

Channelization of DS3 is possible only on STS-1 paths. You can provision an STS-1 path to carry ATM directly or to carry a channelized DS3 path. You do this by setting the payload parameter to either ATM or DS3.

Specifying the payload as DS3 results in the creation of a DS3 path. Then, bringing the DS3 path up automatically provisions ATM on the DS3 path.

Note For specific procedures on configuring channelized paths on the AXSM-XG, see the AXSM-XG Channelization Configuration Procedure section in this chapter.

Inverse Multiplexing over ATM

Inverse Multiplexing over ATM (IMA) is a protocol that runs on the AXSM-32-T1E1-E. IMA allows you to combine multiple T1 or E1 interfaces into a single, high-speed IMA interface.

These combinations of multiple links are called IMA groups. IMA groups are comprised of IMA links.

The AXSM-32-T1E1-E supports a maximum of 32 IMA groups; 16 groups in the top bay and 16 groups in the bottom bay. All the IMA links in an IMA grou must be in the same bay.

IMA is also supported on the following Cisco MGX 8850 and Cisco MGX 8830 cards:

•PXM1E-16-T1E1 (supports a maximum of 16 IMA groups in the bottom bay only)

•MGX-AUSM-8-T1 (supports a maximum of 8 IMA groups)

•MGX-AUSM-8-E1/B (supports a maximum of 8 IMA groups)

Note For information on PXM1E IMA, refer to the Cisco MGX 8850 (PXM1E/PXM45), Cisco MGX 8950, and Cisco MGX 8830 Software Configuration Guide, Release 4 .

Note For information on AUSM IMA, refer to the Cisco AUSM Software Configuration Guide and Command Reference for MGX 8850 (PXM1E) and MGX 8830, Release 3.

SCTs number 54 and 55 provide support for IMA groups. However they only support IMA groups with up to 4 lines. You must create your own SCTs for IMA groups with more than 4 lines.

The Cisco MGX 8850 and Cisco MGX 8830 switches support IMA Versions 1.0 and 1.1.

Note For specific procedures on configuring IMA, see the Inverse Multiplexing over ATM Configuration Procedures section in this chapter.

Partitioning Port Resources Between Controllers

After you add a line or trunk port, you need to define how the port resources are used by the PNNI and MPLS controllers. You can assign all resources to one controller, or you can divide the port resources between both controllers. You can assign the following port resources to controllers:

•Range of VPI values

•Range of VCI values

•Guaranteed percent of bandwidth for ingress and egress directions

•Minimum and maximum number of connections

Note You can and should use the partition definition to control how available connections are distributed within the switch. Each switch, card, and port supports a maximum number of connections. Although you can enable the maximum number of connections on all ports, two or three very busy ports could use all available connections and disable communications on all other ports.

The port resources are defined as a group in a controller partition, which is dedicated to a single port controller. You must define one controller partition for each controller type you want to support, and you must configure one resource partition for each port that uses a controller. Figure 2-2 presents a simplified view of the relationship between the port controller, controller partition, and resource partitions.

Figure 2-2 Relationship of Port Controller, Controller Partition, and Resource Partitions

Figure 2-2 shows that the single controller partition connects to the port controller and to the resource partitions. After you create a port, you must create a resource partition for that port, select either the MPLS or the PNNI controller, and define which ATM resources the port will use. You do not have to create the controller partition, as it is automatically created when you create the first resource partition. It is important that the same controller partition, and therefore the same partition ID be used for all resource partitions of the same type on the same AXSM card. For example, the controller is identified by the controller ID and the controller partition is identified by the partition ID. The resource partitions are identified by specifying the partition ID in combination with the port ID (interface number).

Important VPI/VCI Range Issues

When configuring a partition, be sure to configure the VPI/VCI ranges to meet your actual usage requirements. It is important that you do not configure the entire VPI/VCI range for a single partition. The ability to seamlessly add new partitions in the future depends on configuring only the necessary ranges for each partition.

The Cisco recommended ranges for a single partition are as follows:

•For a VPI on a UNI port where the available range is 0-255, the recommended configured range is 0-140.

•For a VPI on a PNNI port where the range is 0-4096, the recommended configured range is 0-2500 or about 60%.

Caution When adding or configuring a PNNI partition, do not configure the entire VPI/VCI range for one partition. In the future, if you migrate from a PNNI only service to a PNNI/MPLS service with multiple partitions, you will need the additional VPI/VCI ranges to be able to add a new partition. If you configure all of the available ranges for the PNNI partition, you will not be able to add a new MPLS partition without bringing down the port using the dnport command to change the PNNI VPI/VCI ranges. Bringing down a port on a live network is usually not an option.

To create a resource partition for a port, use the following procedure.

Step 1 Establish a configuration session using a user name with GROUP1 privileges or higher.

Note You must add the PNNI controller and add a port before you create a resource partition for a port. For instructions on adding the controller, refer to the Cisco MGX 8850 (PXM1E/PXM45), Cisco MGX 8950, and Cisco MGX 8830 Software Configuration Guide, Release 4 . For instructions on adding ports, see the "Adding ATM Ports" section in this chapter.

Step 2 Determine the port number to which you want to assign the resource partition. To display a list of the ports, enter the following command:
MGX8850.10.AXSM.a > dspports 
This command displays all ports on the AXSM card in the ifNum (interface number) column.

Step 3 To create a resource partition, enter the following command:
MGX8850.10.AXSM.a > addpart <ifNum> <partId> <ctrlrId> <egrminbw> <egrmaxbw> <ingminbw> 
<ingmaxbw> <minVpi> <maxVpi> <minVci> <maxVci> <minConns> <maxConns> 
Step 4 To display a list showing the resource partition you created, enter the following command:
MGX8850.10.AXSM.a > dspparts
Step 5 To display the configuration of a specific resource partition, note the interface and partition numbers, and enter the following command:
MGX8850.10.AXSM.a > dsppart <ifNum> <partId>
The following example shows the report provided by the dsppart command.
MGX8850.1.AXSM.a > dsppart 1 2
  Interface Number               : 1
  Partition Id                   : 2        Number of SPVC: 0
  Controller Id                  : 2        Number of SPVP: 0
  egr Guaranteed bw(.0001percent): 1000000  Number of SVC : 2
  egr Maximum bw(.0001percent)   : 1000000
  ing Guaranteed bw(.0001percent): 1000000
  ing Maximum bw(.0001percent)   : 1000000
  min vpi                        : 0
  max vpi                        : 4095
  min vci                        : 32
  max vci                        : 65535
  guaranteed connections         : 0
  maximum connections            : 5000
Selecting the Port Signaling Protocol

The default signaling protocol for all new ports is UNI none. If you plan to use this protocol on a line, you can accept this default and skip this section. However, if you plan to use a different protocol on the line, such as NNI or PNNI, you must select the correct protocol using the following procedure.

Step 1 Establish a configuration session using a user name with GROUP1 privileges or higher.

Step 2 Enter the following command to display a list of the ports you can configure:
MGX8850.7.PXM.a > dsppnports
Step 3 Enter the following command to bring down the port you want to configure:
MGX8850.7.PXM.a > dnpnport <portid>
A port is automatically brought up when you add it. You must bring down the port before you can change the port signaling protocol. Replace <portid> using the format slot[:bay].line[:ifNum].

Step 4 To confirm the port is down, enter the dsppnports command. The following example shows the report that appears.
MGX8850.7.PXM.a > dsppnports
Summary of total connections
(p2p=point to point,p2mp=point to multipoint,SpvcD=DAX spvc,SpvcR=Routed spvc)
Type   #Svcc:   #Svpc:   #SpvcD:  #SpvpD:  #SpvcR:  #SpvpR:  #Total:
p2p:   0        0        0        0        0        0        0      
p2mp:  0        0        0        0        0        0        0      
                                                       Total=0      
Summary of total configured SPVC endpoints
Type   #SpvcCfg: #SpvpCfg:
p2p:   1         0      
p2mp:  0         0      
Per-port status summary
PortId         IF status         Admin status       ILMI state       #Conns
7.35           up                up                 Undefined        0       
7.36           up                up                 Undefined        0       
7.37           up                up                 Undefined        0       
7.38           up                up                 Undefined        0       
Type <CR> to continue, Q<CR> to stop: 
1:1.1:1        down              down               Disable          0       
2:2.2:1        up                up                 Disable          0 
Step 5 To select the port signaling protocol, enter the following command:
MGX8850.7.PXM.a > cnfpnportsig <portid> [-univer {uni30|uni31|uni40|none}] [-nniver 
{iisp30|iisp31|pnni10}] [-unitype {public|private}] [-addrplan {both|aesa|e164}] [-side 
{user|network}] [-vpi <vpi>] [-sigvci <signalling-vci>] [-rccvci <routing-vci>] [-cntlvc 
<ip>]
The only required parameter for this command is the <portid> parameter, but the command serves no purpose if you do not enter at least one option with it. If you include some options with the command and omit others, the omitted option remains set to the last configured value.

Tip With some commands, you can refer to a port using only the interface number, while other commands require you to enter a complete port identification number, which includes the slot, bay, line, and interface numbers. When entering commands at the PXM switch prompt, you always need to specify the complete port identification number. When entering commands at the AXSM switch prompt, you can enter only the interface number, because the interface number is unique on the AXSM card and identifies the slot, bay, and line for the port.

Note The selection of UNI or NNI is made when the port is added with the addport command. You cannot use the -univer and -nniver options to change the port type.

The following example illustrates how to configure an NNI port to use PNNI Version 1.0 signaling.
MGX8850.7.PXM.a > cnfpnportsig 1:1.1:1 -nniver pnni10
Step 6 Enter the following command to define the local routing switch feeder port as a non-OAM segment endpoint:
MGX8850.7.PXM.a > cnfoamsegep <portid>
Replace <portid> using the format slot:bay.line:ifNum.

Note This step is required to enable testing with the tstdelay command.
Step 7 Enter the following command to bring up the port you just configured:
MGX8850.7.PXM.a > uppnport <portid>
Replace <portid> using the format slot:bay.line:ifNum.

Step 8 To verify the status of the port, enter the dsppnports command.

Step 9 To display the configuration of the PNNI port, enter the following command:
MGX8850.7.PXM.a > dsppnport <portid>
Replace <portid> using the format slot:bay.line:ifNum. The following example shows the report for this command.
MGX8850.7.PXM.a > dsppnport 1:1.1:1
Port:           1:1.1:1               Logical Id:     16848897
IF status:      up                    Admin Status:     up      
UCSM:           enable   
Auto-config:    enable                Addrs-reg:        enable   
IF-side:        network               IF-type:          nni     
UniType:        private               version:          pnni10      
Input filter:   0                     Output filter:    0
minSvccVpi:            0              maxSvccVpi:         4095
minSvccVci:           35              maxSvccVci:        65535
minSvpcVpi:            1              maxSvpcVpi:         4095
       #SpvcCfg:  #SpvcActive:      #SpvpCfg:  #SpvpActive:
p2p :       0             0               0            0
p2mp:       0             0               0            0
       #Svcc:     #Svpc:            Total:
p2p :       0         0                0
p2mp:       0         0                0
                                          Total :            0
Assigning Static ATM Addresses to Destination Ports

When a CPE does not support ILMI, the switch cannot automatically determine the CPE address. To enable communications with the CPE, you must assign a static ATM address to the port leading to the CPE. The static address must match the address used by the CPE. When assigning the static address, you can use command options to define how widely the static address is advertised within the switch network. Use the following procedure to define a static address for a UNI port.

Step 1 Establish a configuration session using a user name with GROUP1 privileges or higher.

Step 2 To locate the port to which you want to add an address, enter the dsppnports command.

Step 3 Enter the following command to turn off automatic address registration (it is enabled by default) on the port that will use the static address:
MGX8850.7.PXM.a > cnfaddrreg <portid> no
Replace portid using the format slot:bay.line:ifNum.

Step 4 Specify an ATM address for the port using the following command:
MGX8850.7.PXM.a > addaddr <portid> <atm-address> <length> [-type int] [-proto local] 
[-plan {e164 | nsap}] [-scope scope] [-redistribute {yes | no}]
Note The addaddr command is used to specify static addresses for UNI links to CPE and to define destination addresses for AINI and IISP static links. The command format above shows the options that apply when defining static addresses for CPE.

Replace <portid> with the ID you used with the cnfaddreg command described earlier.

The following example assigns an ATM address to port 9:1.2:2:
MGX8850.7.PXM.a > addaddr 1:2.1:3 47.1111.1111.1111.1111.1111.1111.1111.1111.1111.11 160 
Step 5 To verify that the new address has been assigned, enter the dspatmaddr command as shown in the following example:
MGX8850.7.PXM.a > dspatmaddr 2:2.2:1
Port Id: 2:2.2:1
Configured Port Address(es) :
 47.1111.1111.1111.1111.1111.1111.1111.1111.1111.11
length: 160    type: internal      proto: local
scope: 0       plan: nsap_icd      redistribute: false
Configuring ILMI on a Port

ILMI is optional on most ports. Use ILMI on a port to do any of the following tasks:

•Use ILMI automatic configuration, which negotiates ATM communication parameters

•Use ILMI address registration, which negotiates an ATM address for an attached CPE using an ILMI prefix assigned to the port

•Enable CWM auto-discovery on a link, which allows CWM to search for and discover Cisco Systems switches that it can manage

•Create a PNNI link to a BXM card on a Cisco BPX

ILMI is enabled by default on all ports and remains in a down state until ILMI is started. There are two ways to start ILMI on a port.

1. To configure and start ILMI with a single command, enter the cnfilmi command.

2. To start ILMI using the default values, use the upilmi command.

The sections that follow describe how to perform the following tasks:

•Configure ILMI traps and signaling and start ILMI

•Configure ILMI automatic configuration

•Configure ILMI dynamic addressing

•Start ILMI with the default trap and signaling parameters

Configuring ILMI Traps and Signaling

The default ILMI configuration uses the standard ILMI signaling VPI and VCI, sets three ILMI signaling timers, and enables the distribution of ILMI management messages (traps) to SNMP managers such as CWM. If the defaults are acceptable, you can start ILMI on the port using the upilmi command. To change the defaults and start ILMI, use the following procedure.

Note When ILMI is configured and started at one end of a link, it must be configured and started at the other end of the link before the link will operate properly.

Step 1 Establish a configuration session using a user name with GROUP1 privileges or higher.

Step 2 If you want to preview the current ILMI configuration for a port, enter the dspilmis command. The following example shows the dspilmis command report.
MGX8850.1.AXSM.a > dspilmis
    Sig. rsrc  Ilmi  Sig  Sig Ilmi  S:Keepalive T:conPoll K:conPoll   
    Port Part State  Vpi  Vci Trap  Interval    Interval  InactiveFactor
    ---- ----  ---- ---- ---- --- ------------  ---------- ----------
    1    2    On     0    16    On          1          5          4 
    2    2   Off     0    16    On          1          5          4 
    3    2   Off     0    16    On          1          5          4 
The example above shows that ILMI is enabled on port 1 (ILMI State = On) and is disabled on ports 2 and 3 (ILMI State = Off). All other ILMI parameters are set to the default values.

Note The ILMI state displayed by the dspilmis command is the configuration state, not the operational state, which appears when you enter the dsppnports or dsppnilmi commands.

Step 3 Enter the cnfilmi command as follows:
MGX8850.10.AXSM.a > cnfilmi -if <ifNum> -id <partitionID> [-ilmi <ilmiEnable>] [-vpi 
<vpi>] [-vci <vci>] [-trap <ilmiTrapEnable>] [-s <keepAliveInt>] [-t 
<pollingIntervalT491>] [-k <pollInctFact>]
Step 4 To confirm your configuration changes, enter the dspilmis command.

Configuring ILMI Automatic Configuration

The Cisco MGX 8850 switches support the automatic configuration feature of ILMI 4.0, which allows two devices that share a link to share their configurations and negotiate a common set of communication parameters. For example, if two network devices share a link and are configured for different maximum VCIs on a partition, the automatic configuration feature can determine and select the highest VCI supported by both nodes. To use ILMI automatic configuration, the devices at each end of the link must support this ILMI 4.0 feature.

Note A link between two nodes will not operate correctly if the ILMI automatic configuration feature is enabled at one end and disabled at the other.

To enable or disable automatic configuration on a port, enter the cnfautocnf command as described in the following procedure.

Step 1 Establish a configuration session using a user name with GROUP1 privileges or higher.

Step 2 To display the automatic configuration status of a port, enter the dsppnport command. For example:
MGX8850.7.PXM.a > dsppnport 1:1.1:1
Port:           1:1.1:1               Logical Id:     16848897
IF status:      up                    Admin Status:     up      
UCSM:           enable   
Auto-config:    enable                Addrs-reg:        enable   
IF-side:        network               IF-type:          nni     
UniType:        private               version:          pnni10      
Input filter:   0                     Output filter:    0
minSvccVpi:            0              maxSvccVpi:         4095
minSvccVci:           35              maxSvccVci:        65535
minSvpcVpi:            1              maxSvpcVpi:         4095
       #SpvcCfg:  #SpvcActive:      #SpvpCfg:  #SpvpActive:
p2p :       0             0               0            0
p2mp:       0             0               0            0
       #Svcc:     #Svpc:            Total:
p2p :       0         0                0
p2mp:       0         0                0
                                          Total :            0
The Auto-config field shows whether the automatic configuration feature is enabled or disabled.

Step 3 If you want to enable or disable automatic configuration, bring down the port to be configured with the dnpnport command. For example:
MGX8850.7.PXM.a > dnpnport 1:1.1:1
Step 4 To enable or disable the automatic configuration feature, enter the cnfautocnf command as follows:
MGX8850.7.PXM.a > cnfautocnf <portid> <yes | no>
Replace portid with the port address using the format slot:bay.line:ifnum.

Enter yes to enable automatic configuration or enter no to disable automatic configuration. The default is yes.

Step 5 Up the port you configured with the uppnport command. For example:
MGX8850.7.PXM.a > uppnport 1:1.1:1
Step 6 To verify the change, re-enter the dsppnport command.

Configuring ILMI Dynamic Addressing

Dynamic ATM addressing is enabled by default on all Cisco MGX 8850 ports. Once ILMI is started, ILMI can negotiate ATM addresses for CPE connected to the port. To determine the ATM address for the CPE, the switch uses a 13-byte ILMI prefix that is assigned to the port, a 6-byte end system ID, and a 1-byte selector byte. The end system ID and selector byte are defined on the end system. Depending on the end system configuration, the end system ID may correspond with the interface MAC address. For dynamic addressing to work, the remote device must support it. ILMI versions 3.x and 4.0 support dynamic address registration.

The default ILMI prefix matches the PNNI node prefix and the SPVC prefix, both of which are described in the Cisco MGX and SES PNNI Network Planning Guide. If you change the PNNI node prefix, the SPVC prefix and the ILMI prefix remain unchanged. If you change the SPVC prefix, the ILMI prefix will change with it, as long as no ILMI prefix is assigned directly to the port. To eliminate the possibility of having a future SPVC prefix change affect dynamic addressing on a port, assign one or more ILMI prefixes to the port.

The following procedure describes how to enable or disable dynamic addressing and how to assign an ILMI address prefix to a port.

Note The Cisco MGX 8850 switches support up to 255 ILMI prefixes per AXSM card, and these prefixes can be assigned to one port or distributed among the ports.

Step 1 Establish a configuration session using a user name with GROUP1 privileges or higher.

Step 2 To display the dynamic addressing status of a port, use the dsppnport command. For example:
MGX8850.7.PXM.a > dsppnport 1:1.1:1
Port:           1:1.1:1               Logical Id:     16848897
IF status:      up                    Admin Status:     up      
UCSM:           enable   
Auto-config:    enable                Addrs-reg:        enable   
IF-side:        network               IF-type:          nni     
UniType:        private               version:          pnni10      
Input filter:   0                     Output filter:    0
minSvccVpi:            0              maxSvccVpi:         4095
minSvccVci:           35              maxSvccVci:        65535
minSvpcVpi:            1              maxSvpcVpi:         4095
       #SpvcCfg:  #SpvcActive:      #SpvpCfg:  #SpvpActive:
p2p :       0             0               0            0
p2mp:       0             0               0            0
       #Svcc:     #Svpc:            Total:
p2p :       0         0                0
p2mp:       0         0                0
                                          Total :            0
The Auto-reg field shows whether the dynamic addressing feature is enabled or disabled.

Step 3 To view the ILMI prefixes assigned to a port, enter the dspprfx command as follows:
MGX8850.7.PXM.a > dspprfx <portid>
Replace portid with the port address using the format slot:bay.line:ifnum. For example:
MGX8850.7.PXM.a > dspprfx 1:1.1:1
INFO:  No Prefix registered
In the example above, no ILMI prefixes have been assigned to the port, so the port will use the prefix configured for the SPVC prefix.

Step 4 If you want to change the dynamic addressing configuration, bring down the port to be configured with the dnpnport command. For example:
MGX8850.7.PXM.a > dnpnport 1:1.1:1
Step 5 To enable or disable dynamic address registration, enter the following command:
MGX8850.7.PXM.a > cnfaddrreg <portid> <yes | no>
Enter yes to enable dynamic address configuration or enter no to disable it. The default is yes.

Step 6 Enter the following command to define an ATM prefix for a port:
MGX8850.7.PXM.a > addprfx <portid> <atm-prefix>
Replace portid using the format slot:bay.line:ifNum.

Replace atm-prefix with the 13-byte ATM address prefix that you want the dynamically assigned address to use. Specify the address prefix using 26 hexadecimal digits. The range for each digit is 0 through F (0 through 9, A, B, C, D, E, and F).

Note The address prefix you choose should conform to the address plan for your network. For more information on address planning, refer to the Cisco MGX and SES PNNI Network Planning Guide.

Tip Each hexadecimal digit represents 1 nibble (four bits), and each pair of hexadecimal digits represents a byte. There are 13 pairs of hexadecimal digits in the prefix, or 26 total digits.

Step 7 Up the port you configured with the uppnport command. For example:
MGX8850.7.PXM.a > uppnport 1:1.1:1
Step 8 To verify the proper ATM prefix configuration for a port, re-enter the dspprfx command.

Step 9 To see a dynamically assigned address that uses the prefix, enter the dspilmiaddr <port> command.

Starting ILMI with the Default or Existing Values

The upilmi command starts ILMI on a port with the existing ILMI configuration, which is the default configuration when ILMI has never been configured on that port. Although ILMI starts automatically when you configure it with the cnfilmi command, you might have to bring down ILMI with the dnilmi command to make a configuration change such as adding an ILMI prefix. To start or restart ILMI with the upilmi command, use the following procedure.

Step 1 Establish a configuration session using a user name with GROUP1 privileges or higher.

Step 2 If you do not know the interface number and partition ID for the port on which you are starting ILMI, enter the dspparts command as shown in the following example.
MGX8850.1.AXSM.a > dspparts
if  part Ctlr egr     egr     ingr    ingr    min max   min   max  min   max
Num ID   ID   GuarBw  MaxBw   GuarBw  MaxBw   vpi vpi   vci   vci  conn  conn
             (.0001%)(.0001%)(.0001%)(.0001%)
-----------------------------------------------------------------------------
 1   2    2 1000000 1000000 1000000 1000000    0 4095    32 65535      0   5000
 2   2    2 1000000 1000000 1000000 1000000    0 4095    32 65535      0   5000
 3   2    2 1000000 1000000 1000000 1000000    0  255    32 65535      0   1000
Tip To see the relationship between interface numbers and lines, enter the dspports command.

Step 3 To start ILMI on a port, enter the upilmi command as follows:
MGX8850.10.AXSM.a > upilmi <ifNum> <partId>
Replace ifNum with the interface number for the port, and replace partId with the partition number assigned to the port. For example:
MGX8850.10.AXSM.a > upilmi 2 1
Step 4 To display the ILMI status of all the ports on an AXSM card, enter the dspilmis command. For example:
MGX8850.1.AXSM.a > dspilmis
    Sig. rsrc  Ilmi  Sig  Sig Ilmi  S:Keepalive T:conPoll K:conPoll   
    Port Part State  Vpi  Vci Trap  Interval    Interval  InactiveFactor
    ---- ----  ---- ---- ---- --- ------------  ---------- ----------
    1    2    On     0    16    On          1          5          4 
    2    2   Off     0    16    On          1          5          4 
    3    2   Off     0    16    On          1          5          4 
The ILMI State column displays the configured state for ILMI, which is On if ILMI is enabled and Off if ILMI is disabled (use dsppnports or dsppnilmi to see the operational state).

Configuring AXSM Line Clock Sources

To configure the switch to receive a clock source on an AXSM line, you must do the following tasks:

•Connect a line between the AXSM and the node with the clock source.

•Activate the line.

•Create a logical port (subport) for the clock signal.

•Create a resource partition.

Refer to the Cisco MGX 8850 (PXM1E/PXM45), Cisco MGX 8950, and Cisco MGX 8830 Software Configuration Guide, Release 4 for information on how to activate a line. Procedures for creating ports and resource partitions appear in this chapter. The following procedure describes how to configure an AXSM clock source after the line and port have been configured.

Step 1 Establish a configuration session using a user name with GROUP1 privileges or higher.

Step 2 To set a primary or secondary AXSM clock source, enter the following command:
MGX8850.7.PXM.a > cnfclksrc <priority> [shelf.]<slot:bay.line:ifnum>
Tip To get the correct slot:bay.line:ifnum specification, use the port ID displayed by the dsppnports command.

Step 3 To configure an additional clock source, repeat Step 2 using the correct parameters for the additional source.

The following command example shows how to configure a secondary clock source for subport (logical port) 10 on line 1 of the AXSM card in the upper bay of slot 3. Note the placement of the periods and colons.
MGX8850.7.PXM.a > cnfclksrc secondary 3:1.1:10
Procedures for PNNI Links

This section describes AXSM configuration procedures that apply only to PNNI links. The following subsections explain the following tasks:

•Verifying PNNI Communications

•Configuring SPVCs and SPVPs

•Deleting SPVCs and SPVPs

•Defining a Feeder Port

Verifying PNNI Communications

After setting up trunks or when problems occur, use the procedures in this section to determine if PNNI is operating. The next section describes how to verify PNNI communications on a single trunk. The following section describes how to verify PNNI communications between two nodes, which can be separated by multiple PNNI links.

Verifying PNNI Trunk Communications

After you configure both ends of a PNNI trunk, it should be ready to support SVCs and any SPVCs or SPVPs that are configured. To verify that the trunk is functioning, use the following procedure.

Step 1 Establish a CLI session using a user name at any access level. When both ends of the trunk are connected to MGX 8850 or MGX 8950 switches, you can start the CLI session at either end.

Step 2 If you do not know the line number you are validating, you can view the port and line numbers by entering the dsppnports command.

The first three numbers identify the slot, bay, and line. For example, port 10:2.1:3 represents slot 10, bay 2, line 1. The remaining number is the interface number assigned with the addport command.

Step 3 Enter the dsppnni-link command as follows:
MGX8850.7.PXM.a > dsppnni-link
The dsppnni-link command displays a report for every PNNI link on the switch. The following example shows the report for a switch with a single PNNI link.
MGX8850.7.PXM.a > dsppnni-link
node index   : 1
Local port id:   16848897          Remote port id:   17438721
Local Phy Port Id: 1:1.1:1
   Type. lowestLevelHorizontalLink     Hello state....... twoWayInside
   Derive agg...........         0     Intf index...........  16848897
   SVC RCC index........         0     Hello pkt RX.........        10
                                       Hello pkt TX.........         9
   Remote node name.......MGX8850
   Remote node id.........56:160:47.00918100000000107b65f33c.00107b65f33c.01
   Upnode id..............0:0:00.000000000000000000000000.000000000000.00
   Upnode ATM addr........00.000000000000000000000000.000000000000.00
   Common peer group id...00:00.00.0000.0000.0000.0000.0000.00
In the dsppnni-link command report, there should be an entry for the port for which you are verifying communications. The Local Phy Port Id field in this entry displays the port id in the same format shown in the dsppnports command report. The Hello state reported for the port should be twoWayInside and the Remote note ID should display the remote node ATM address after the second colon.

In the example above, the report shown is for port 1:1.1:1. The Hello state is twoWayInside, and the ATM address of the node at the other end of the link is 47.00918100000000107b65f33c.00107b65f33c.01. This link is ready to support connections between the two switches.

Tip If the Hello state for the link is oneWayInside, that side is trying to communicate. Check the status at the other end. Remember that the configuration at each end of the trunk must be compatible with that on the other end. For example, if ILMI auto configuration is configured at one end and not at the other, the Hello state cannot change to twoWayInside or twoWayOutside.

Verifying End-to-End PNNI Communications

When connections between two nodes travel over multiple trunks, use the following steps to verify that the PNNI communications path is operational.

Step 1 Establish a CLI session using a user name at any access level. When both ends of the communications path are connected to MGX 8850 or MGX 8950 switches, you can start the CLI session at either end.

Step 2 To display information on all accessible nodes, enter the dsppnni-node-list command as shown in the following example:
MGX8850.7.PXM.a > dsppnni-node-list
node #  node id                                            node name
------- -------------------------------------------------- ----------
    1   56:160:47.00918100000000001a531c2a.00001a531c2a.01 MGX8850      
node #  node id                                            node name
------- -------------------------------------------------- ----------
    2   56:160:47.00918100000000036b5e2bb2.00036b5e2bb2.01 8850_NY      
If a switch appears in this list, you have verified communications with it.

Step 3 To display additional information on the local switch, enter the dsppnni-node command. For example.
MGX8850.7.PXM.a > dsppnni-node
node index: 1                      node name: MGX8850        
   Level...............        56     Lowest..............      true
   Restricted transit..       off     Complex node........       off
   Branching restricted        on
   Admin status........        up     Operational status..        up
   Non-transit for PGL election..       off
   Node id...............56:160:47.00918100000000001a531c2a.00001a531c2a.01
   ATM address...........47.00918100000000001a531c2a.00001a531c2a.01
   Peer group id.........56:47.00.9181.0000.0000.0000.0000.00
Step 4 To display additional information on remote switches, enter the dsppnni-reachable-addr command as follows:
MGX8850.7.PXM.a > dsppnni-reachable-addr network
scope...............         0     Advertising node number         2
Exterior............     false
ATM addr prefix.....47.0091.8100.0000.0003.6b5e.2bb2/104
Advertising nodeid..56:160:47.00918100000000036b5e2bb2.00036b5e2bb2.01
Node name...........8850_NY
The remote node ATM address appears in the Advertising nodeid row. The information before the first colon (56) is the PNNI level, the information between the first and second colons (160) is the ATM address length, and the remainder of the node ID is the ATM address for the remote node.

Tip If you cannot verify communications with a remote node, try verifying communications across each of the links between the nodes as described in the previous section, "Verifying PNNI Trunk Communications."

Configuring SPVCs and SPVPs

SPVCs and SPVPs are created between two ATM ports, and each SPVC and SPVP has two endpoints. The master endpoint is responsible for routing and rerouting functions. The slave endpoint is responsible for responding to requests from the master endpoint during connection setup and rerouting. Both endpoints are configured on the switch or switches to which the ATM CPE connects. Such endpoints can be in the same switch or in different switches. One endpoint of an SPVC or SPVP can exist on an MSSBU switch, while the endpoint can exist on different Cisco ATM equipment, or on ATM equipment from another vendor.

The master and slave relationships exist for each SPVC or SPVP, and apply only to the SPVC or SPVP connection. For example, you can have one SPVC with a master on Node A and a slave on Node B, and then create another with the Master on Node B and the slave on Node A. It is good practice to distribute the master side of SPVCs and SPVPs among the network nodes so that route processing is distributed.

Cisco MGX 8850 PXM1E-based and PXM45-based switches support two types of SPVCs/SPVPs:

•Single-ended SPVCs

•Double-ended SPVCs

Single-ended SPVCs are defined at the master endpoint and do not require configuration of a slave endpoint. The primary benefit of single-ended SPVCs is that they are easier to configure. After configuration, the master endpoint configures and brings up the slave endpoint. In order for this feature to work correctly, the destination endpoint must support single-ended SPVCs. Single-ended SPVCs are non-persistent. Non-persistent SPVCs will attempt to route on the specified path first. If the configured path is unavailable, the non-persistent SPVC will attempt to route over another available path.

Note The AXSM supports only the origination of single-ended SPVCs. This means that you can configure master endpoints for single-ended SPVCs that terminate on other card types, such as the FRSM12. If both SPVC endpoints must terminate on AXSM cards, you must create a double-ended SPVC.

Double-ended SPVCs and SPVPs require separate configuration of the master and slave endpoints. The slave endpoint must be configured first because this step generates a slave address that must be entered during master endpoint configuration Double-ended SPVCs are persistent, because they will follow only the specified path. If that path is unavailable, the persistent SPVC/SPVP will not route.

The following sections describe how to configure slave and master SPVC and SPVP connections.

Tip The configuration of SPVCs and SPVPs is very similar. The diff