Cisco Voice Switch Service Module (VXSM) Configuration Guide Release 5.4
VXSM Troubleshooting

Table Of Contents

VXSM Troubleshooting

Collecting Troubleshooting Data

Troubleshooting Procedures

Obtaining Information on Current Voice Calls—H.248

Examples

Obtaining Information on Current Voice Calls—xGCP

Bearer Tracing Feature

Bearer Trace Operation

Bearer Trace Types

PCM Traces

Echo Canceller Traces—Sout

Packet Traces

Voice Playout Unit Traces

T.38 Traces

MSG Traces

Bearer Tracing Connectivity

Bearer Trace CLI Requirements

Considerations and Limitations

Configuring the Bearer Tracing Feature

Configuration Summary

Detailed Configuration

Bearer Trace Configuration Examples

Trace Files

Server File Requirements

FileNames

Troubleshooting a Bearer Trace Operation

Bearer Trace Command Summary

Troubleshooting Commands


VXSM Troubleshooting


This chapter describes how to trouble shoot and resolve known types of problems on the Cisco MGX 8850 VXSM card. The following topics are covered in this chapter:

Collecting Troubleshooting Data

Troubleshooting Procedures

Obtaining Information on Current Voice Calls—H.248

Obtaining Information on Current Voice Calls—xGCP

Bearer Tracing Feature

Troubleshooting Commands

Collecting Troubleshooting Data

This section provides procedures for collecting troubleshooting data for when a VXSM card fails. These procedures use PXM45, VXSM, RPM-XF and AXSM commands. See Chapter 5, "VXSM CLI Commands" in this document for detailed descriptions of the VXSM CLI commands. Refer to the following documents for descriptions of the PXM45, RPM-XF and AXSM CLI commands:

Cisco MGX 8850 (PXM1E/PXM45), MGX 8950, and MGX 8830 Command Reference, Release 5.2.

Cisco ATM Services (AXSM) Software Configuration Guide and Command Reference for MGX Switches, Release 5.2

Cisco MGX Route Processor Module (RPM-XF) Installation and Configuration Guide, Release 4

Table 8-1 shows the initial steps you can take using CLI commands to trouble shoot any problem on the VXSM. These steps enable you to collect data and failure reports.

Table 8-1 Initial Troubleshooting Steps Using CLI Commands

Run these CLI commands on the PXM45:

1. dspcds

2. dspcd

3. dsplog

4. dsperr

5. dspversion

6. dsprev

7. dspclksrc

 
Run these CLI commands on the VXSM:

1. dspcd

2. dspversion


Troubleshooting Procedures

Table 8-2 provides a list of known types of problems, their possible causes, and their possible solutions. The problems are listed alphabetically by topic.

Table 8-2 VXSM Troubleshooting Procedures 

Problem
Possible Cause
Possible Solutions

Access to Card—
Cannot use the cc command to access VXSM card.

VSXM card is not in the active or standby state. If the VSXM card is not in the active or standby state, you cannot use the cc command to access the card.

Verify that the VXSM card is in either the active or standby state, by issuing the PXM45 CLI command, dspcds.

See Possible Solutions for "Active Card—VXSM card did not become active".

Active Card—
VXSM card did not become active.

VXSM card was inserted into the wrong slot. Slot 7, 8, 15 and 16 are reserved for other service modules.

Remove the VXSM card. Verify that the pins on the backplane are not bent and the power key is intact. Attempt to insert the VXSM card into a slot that is not reserved.

PXM45 and VXSM run time images are not compatible

Check the latest Cisco MGX8850 Release Notes and make sure that the PXM45 and VXSM images are current and are compatible. If they are not, update the VXSM boot image and runtime image by issuing the following PXM45 CLI commands:

1. burnboot

2. setrev

VXSM runtime image is not available from the PXM45 hard drive or is corrupted.

Download the correct VXSM image to PXM45-HD. Use the PXM45 dsprevs CLI command in order to verify that the image file name and size are correct. Then, issue the clrsmcnf command and the setrev command to clear and reload the correct image.

The setrev command was not executed for the VXSM card.

Issue the PXM45 CLI command, setrev.

The VXSM card is inserted in a slot that was previously configured for a different card type and the slot is still in the Reserved state.

Insert the VXSM card into a different slot or issue the clrsmcnf CLI command.

Alarm—
PVC is in alarm.

The OC3 cable is bad, does not exist, or is not connected properly.

1. Check for alarms on the AXSM line associated with the VXSM PVC using the dsplns and dsplnalm CLI commands.

2. On the AXSM, issue the addlnloop command with the option for local loopback set.

3. Check whether the alarm is cleared using the dspcons command.

4. If the alarm is cleared, check the TX/RX connections between the AXSM and the router's ATM interface.

5. If the TX/RX connections are correct, replace the OC3 cable.

The router's ATM interface is not configured.

On the router, issue the IOS command, show run, to verify whether the VPI/VCI and traffic parameters are correct. If necessary, reconfigure the VPI/VCI and traffic parameters.

The PVC has been added on the VXSM, but not on the AXSM, or vice versa.

Verify that the PVC has been added on both cards by issuing the dspcons command on both the VXSM and the AXSM. If necessary, add the connection, using the addcon command.

MGC-Initiated Commands Ignored

An association does not exist between the MGC and the VXSM.

Display the H.248 association state using the dsph248state CLI command (or SNMP equivalent) and specifying the GatewayLinkID.

VXSM is Rejecting Call Attempts

   

TDM/SCN termination does not exist.

Display the VIFs (Voice Interfaces) using the dspvifs CLI command (or SNMP equivalent). If the VIF corresponding to the TDM/SCN termination does not exist, it must be added using the addvif CLI command (or SNMP equivalent).

H.248 packages not associated with VIF.

Display the VIFs (Voice Interfaces) using the dspvifs CLI command (or SNMP equivalent). Verify that the appropriate H.248 packages have been added corresponding to the VIFs. H.248 packages can be added using cnfvifterm CLI command (or SNMP equivalent).

   

Inadequate resources (CPU, memory, message queues...).

Display the system resources using the following CLI commands,

dspconcacs

dspcrr

dspcpursc

dspmemrsc

dsprpcrsc

dspmsgqrsc

A H.248 Add TDM termination command from the CA may be rejected if the TDM termination is already associated with a (non-null) context.

The CA should attempt to clean up the orphaned TDM termination.

In the event that the CA is not able to subtract the orphaned termination, the cnfh248oos CLI command (or SNMP equivalent) can be used to forcefully subtract ALL of the terminations (and calls) on the VXSM. This should be used as a second-to-last resort.

In the event that the cnfh248oos CLI command won't subtract the termination, the card will need to be reset.

The CA port has not been provisioned on the VXSM.

The CA port can be provisioned using the addmgcgrpmgc CLI command (or SNMP equivalent).

Echo—
Echo is present in voice call.

Echo cancellation (ECAN) is either not enabled or is not configured properly.

Ensure that the ECAN feature is enabled.

Fail—
dspcds command output on the PXM shows VXSM card as failed.

VXSM card is in the initial boot process.

Use the PXM45 CLI dspcds command in order to monitor the boot process and card bring-up. Allow time for the boot process to complete.

VXSM boot flash or configuration file is corrupted.

Capture the output of the PXM45 dsplog -sl CLI command.Capture the console output. Issue the PXM45 CLI command clrsmcnf with the option all to clear the configuration for the slot. If the problem persists, contact Cisco Systems, Inc.

VXSM slot is configured for another service module

Physically reset the VXSM card or move it to another slot. Issue the PXM45 CLI command, clrsmcnf with the option all to clear configuration for the slot.

Firmware—
Firmware does not download to VXSM card.

VXSM card is not present or is not seated properly.

Make sure that the VXSM card is seated properly in the slot so that the top and bottom portions of card are making electrical contact with back plane.

VSXM card is not in the active state or the standby state.

Verify that the VXSM card is in either the active or standby state, by issuing the PXM45 dspcds CLI command. Check for a card state of Boot/Init or Failed.

Perform the PXM45 dsprevs CLI command to verify that the proper firmware image resides on the PXM.

See Possible Solutions for "Active Card—VXSM card did not become active".

VXSM card or MGX slot is defective.

Attempt to insert the VXSM card into another slot. If the problem persists, contact Cisco Systems, Inc.

LEDs—
All VXSM front panel LEDs are off.

Card is not seated properly in the slot.

Make sure that the VXSM card is seated properly in the slot so that the top and bottom portions of card are making electrical contact with back plane.

Mismatch—
Front card, back card mismatch.

Configuration mismatch

Use the dspcds, dspcd and dsplog CLI commands to identify the configuration mismatch.

Mismatch—
T1/E1 mismatch.

Configuration mismatch

Issue the PXM45 dspcds and dspsmcnf CLI commands to identify a configuration mismatch.

After the slot is identified, issue the PXM45 dsplog CLI command to show the card mismatch log entry.

Ping—
Card does not respond to ping from router.

PVC connection is not established between VXSM and AXSM/RPM-XF.

Verifying that the physical connection between the AXSM and the router's ATM interface is properly connected.

Verify that the LEDs for the ATM interface and the VXSM interface are green.

Verify that the TX/RX cables are not crossed.

PVC connection is in fail or alarm status.

Verify the alarm status by issuing the PXM45 dspcons or dspalms commands.

The router's ATM interface is not configured.

From the router, issue the IOS show atm pvc and show run commands to verify that the ATM PVC status is up and that the VPI/VCI, traffic parameters and the IP address are configured properly.

The IP address is incorrect.

From VXSM and AXSM, issue the PXM45 CLI command dspcons to verify that the VPI/VCI for the PVC between the VXSM and AXSM is correct.

Issue the PXM45 dspconips CLI command and the IOS route show command to verify that the IP address and netmask for the PVC are correct.

An IP address has not been provisioned corresponding to the AAL5 bearer PVC.

Issue the VXSM addipcon CLI command.

Resets—
Card constantly reboots-becomes active and then resets again.

The MGX switch is fully loaded and the power is reaching its shut off threshold.

Issue the PXM45 dspenvalms and dspndalms CLI commands to see if the power and temperature for the shelf and the card are within acceptable limits. If the problem persists, contact Cisco Systems, Inc.

VXSM card is getting too hot due to poor air circulation

The MGX fans are not functioning.

Resets—
Card resets occasionally.

Too many calls are made while a debugging command is enabled.

Minimize the number of calls when debugging is enabled, reduce the trace level in order to minimize the number of trace messages or turn off debugging altogether.

Slot—
VXSM card is in place, but the dspcds command on the PXM shows that the slot is empty.

The pins on the back plane are bent or a power key is missing.

Remove the VXSM from the slot and observe the back panel to make sure that the pins on that slot are not bent and that the power key (the orange or yellow plastic cap in the center of the slot) exists.

Insert the VXSM card into another slot.

If the problem persists, contact Cisco Systems.


Obtaining Information on Current Voice Calls—H.248

With H.248, calls are modeled using a context. A context contains one or more terminations. In Release 5, a context can contain at most two terminations. In this release, there are two types of terminations: TDM (also referred to as SCN or Switched Circuit Network) and RTP.

A basic VoIP call leg is modeled as a context containing one TDM termination and one RTP termination.

Figure 8-1 VoIP Call Leg Model

.

A TDM hairpin call is modeled as a context containing two TDM terminations.

Figure 8-2 TDM Hairpin Model

Table 8-3 lists commands for obtaining information about current voice calls on the VXSM.

Table 8-3 Displaying Current Voice Calls on the VXSM 

Step
CLI Commands

1. Display all of the H.248 terminations on the VXSM card.

dsph248calls <GatewayLinkId>

2. Display the status of all the H.248 contexts on the VXSM card.

dsph248status -cntxs

3. Display the status of a particular H.248 context on the VXSM card.

dsph248status -cntx <context ID>

4. Display the status of a particular H.248 TDM/SCN termination on the VXSM card.

dsph248status -term 1 1 <term_id>

dsph248status -term 1 2 <term_name>

dsph248status -term 1 3 <oc3> <ds1> <ds0>

5. Display the status of a particular H.248 RTP termination on the VXSM card.

dsph248status -term 3 1 <term_id>

dsph248status -term 3 2 <term_name>

6. Display information pertaining to all of the H.248 associations on the VXSM card.

dsph248status -assoc

7. Display information pertaining to a particular H.248 termination on the VXSM card.

dsph248call -term

8. Display counters corresponding to all of the H.248 physical (TDM/SCN) terminations on the VXSM card.

dsph248cnts -phyterm

9. Display H.248 command counters.

dsph248cnt -cmd <association>

10. Display counters pertaining to a particular H.248 physical (TDM/SCN) termination on the VXSM card.

dsph248cnt -phyterm 1 <bay.line.path.vtg/ds3.vt/ds1:ds0>

11. Display information regarding all of the H.248 contexts.

dsph248cnt -cntx <association>

12. Display counters pertaining to all H.248 ephemeral (RTPยบ) terminations on the VXSM card.

dsph248cnt -ephterm

13. Display counters pertaining to a particular H.248 association.

dsph248cnt -assoc <association>


Examples

The dsph248calls CLI command displays all of the H.248 terminations.

shelf.1.VXSM.a > dsph248calls

TermId TermType Media Codec Vad Ecan

------ -------- -------- ----- ----- ----

1 SCN 'Voice' 'G.711 U' ----- 'Enabled'

8067 RTP 'Voice' 'G.711 U' 'Enabled' -----

2 SCN 'Voice' 'G.711 U' ----- 'Enabled'

8068 RTP 'Voice' 'G.711 U' 'Enabled' -----

The dsph248status -cntxs CLI command displays the status of all H.248 contexts.

shelf.1.VXSM.a > dsph248status -cntxs

=======================================

Status for all contexts

=======================================

Number of active contexts:2

List of contexts

Status of context 4:

Creation date and time:02/03/2004, 17:31:21

Number of terminations in context:2

List of termination IDs:

8068 2

Status of context 3:

Creation date and time:02/03/2004, 17:31:21

Number of terminations in context:2

List of termination IDs:

8067 1

The dsph248status -cntx CLI command displays the status of a particular H.248 context.

shelf.1.VXSM.a > dsph248status -cntx 1

=======================================

Status of context 1

=======================================

Creation date and time: 07/04/2003, 13:55:45

Number of terminations in context: 2

List of terminations:

term_id: 8065, term_type: MG_TERM_TYPE_PDN_IP, term_name: RTP8065

term_id: 3, term_type: MG_TERM_TYPE_SCN, term_name: ds/1/1/3

The dsph248status -term CLI command displays the status of a particular H.248 termination.

shelf.1.VXSM.a > dsph248status -term 1 1 1

==========================================

Status of termination with ID 1

==========================================

Termination type:MG_TERM_TYPE_SCN

Termination name:ds/1/1/1

Termination state:MG_TERM_STATE_INSERVICE

Termination test flag:False

Termination context id:3

Id of profile on termination:0

Number of streams on termination:1

Termination about to be deleted:False

Next state of termination:MG_NEXT_STATE_INSERVICE

Last modified/updated:02/03/2004, 17:31:21

NOTE:This timestamp is not preserved across switchovers.

It is independently generated on each card.

Supported packages:

H248_GENERIC_PKG H248_TDMC_PKG

The dsph248call -term CLI command displays information pertaining to a particular H.248 terminations:

shelf.1.VXSM.a > dsph248call -term 1

Term Id :1

Term Type :SCN

Term Name :ds/1/1/1

connMode :4 ('Send-recv')

loopBackType :1 ('None')

RTP Encapsulation Parameters ...

ssrc :424285698 (0x194a1602)

RTCP Enable :TRUE

dspCodec :2 ('G.711 U')

packetPeriod :'20 ms'

vadMode :2 ('Enabled')

vadThreshold :-38

vadHangoverTime :250 ms

ecan enabled :'Enabled'

ecanTail :128 ms

ecanFlags :0x0

Type <CR> to continue, Q<CR> to stop:

upspeedCodec :2 ('G.711 U')

jitterMode :2 ('Adaptive')

jitterMinDelay :5 ms

jitterMaxDelay :100 ms

jitterNomDelay :30 ms

jitterFaxNomDelay :0 ms

txGain :0

rxGain :0

nsePayloadType :100 (0x64)

nteTxPayloadType :0 (0x0)

nteRxPayloadType :0 (0x0)

profileType :0

profileNum :0

hsRedCount :0

lsRedCount :0

toneDetect :0x3c32

toneDetectBitmap :0x0

Type <CR> to continue, Q<CR> to stop:

digitDetect :1 ('DTMF')

digitRelayMethod :1 ('Send as Voice')

icsEnable :0 (FALSE)

nx64FramePattern :0 (CRML_PATTERN_NONE)

trunkMode :0 (CRML_MODE_NONE)

nx64FrameFlagCnt :0

dtmfTransport :0 (FALSE)

trunkCond :0 (FALSE)

The dsph248cnts -phyterm CLI command displays counters corresponding to all of the H.248 physical (TDM/SCN) terminations on the VXSM card:

shelf.1.VXSM.a > dsph248cnts -phyterm

=================================================================

All Gateway Physical Terminations Statistic

DS1:DS0 Termination Termination Num of Num of Num of OOS Num of OOS

line Id Name Add Failure from MGC from OAM

============ =========== ============ ====== ======= ==========

1.1.1.2.3:1 361 DS/1/16/1 1 0 0 0

1.1.1.2.3:2 362 DS/1/16/2 0 0 0 0

......

1.1.1.2.3:23 383 DS/1/16/23 0 0 0 0

1.1.1.2.3:24 384 DS/1/16/24 1 0 0 0

The dsph248cnt -ephterm CLI command displays counters corresponding to all of the H.248 ephemeral (such as RTP) terminations on the VXSM card:

shelf.1.VXSM.a > dsph248cnt -ephterm

=========================================================

Gateway Ephemeral Terminations Statistic

=========================================================

Num of Add Commands : 2

Num of Commands Failure : 0

Obtaining Information on Current Voice Calls—xGCP

The Current Voice xGCP call (xGCP) model is based upon endpoints and connections. Connections have modes. Supported modes include Receive-Only, Send/Receive, Inactive, Loopback, and Continuity Test. The connection model is shown in Figure 8-3.

Figure 8-3 xGCP Connection Model

To obtain information regarding the status of TGCP activities. Use the PXM and VXSM commands as follows. The PXM CLI commands in Table 8-4 display the list of events and errors logged by the active and standby VXSM cards, starting with the most recent event or error.

Table 8-4 PXM CLI Commands for Troubleshooting VXSM 

Command
Displays

dsperr -sl <primary slot>

Error for a primary VXSM.

dsperr -sl <secondary slot>

Error for a secondary VXSM.

dsplog -sl <primary slot>

Log for a primary VXSM.

dsplog -sl <secondary slot>

Log for a secondary VXSM.


Use the VXSM CLI commands in Table 8-5 to monitor the basic operation of the currently-active VXSM card and to check if calls are being processed after switchover:

Table 8-5 VXSM CLI Commands for Troubleshooting VXSM 

Command
Description

dspxgcpcnts

Displays xgcp command counts

dspxgcpdetailcnts

Displays xgcp details

dspxgcpendpts

Displays xgcp endpoints

dspxgcpendptcons

Displays xgcp endpoint connections


dspxgcpcnts—Displays a static summary of the number of TGCP commands that failed/succeeded on the ACTV card at that moment; executing this command a few times will give the user an idea whether commands are being sent to the VXSM and whether VXSM is processing those commands.

dspxgcpdetailcnts—Displays number of times a TGCP command is received/sent/retransmitted.

dspxgcpendpts <endpt-name>—Displays all the endpt state if connection is present on that endpoint.

dspxgcpendpts *—Can be used to display all.

-dspxgcpendptcons <endpt-name>—Displays the connection ID. Call ID of the connection on that endpoint.

Bearer Tracing Feature

The VXSM card includes a useful troubleshooting tool that has the ability to collect fundamental information about the content of bearer streams and to transmit the information in real-time to an external file server for offline analysis. This feature, known as Bearer Tracing, involves the VXSM card, the Media Gateway Controller for VoIP applications, and a file server to receive the bearer traces (Figure 8-4).

Figure 8-4 Bearer Tracing Major Elements

Within the VXSM card, the Bearer Tracing feature is performed in the voice module (Figure 8-5).

Figure 8-5 VXSM Bearer Tracing High-Level Function

The Voice Module contains a number of probes on the card that are situated at various points in the path of the bearer stream through the card. There are 10 probes (Figure 8-6).

Figure 8-6 Bearer Trace Probe Locations


Note The numbering used to identify the probes in the diagram above is the same numbering that the VXSM Bearer Tracing command use to identify the probes.


The probes are:

1 = PCM Input from TDM network

2 = PCM Output to the TDM network

3 = Input from the packet network

4 = Output to the packet network

5 = Output from echo canceller (Ecan/Sout)

6 = Voice Playout Unit - Jitter buffer events

7 = T.38 information at the Decoder—level 1

8 = T.38 information at the Decoder—level 2

9 = DIM Message Trace to and from DSP

10 = Voice Playout Unit—segments

Bearer Trace Operation


Step 1 To be able to collect bearer traces on a endpoint, VXSM must:

a. Have an active bearer channel

In VoIP—A call must be set up
or

In AAL2—An AAL2 nailed up connection must be added by adding a CID

And

b. Be connected to a computer where the trace files are to be stored (referred to as a fileserver, server, or network).


Note If a call agent is used, the same machine can be a call agent and a file server at the same time.


Step 2 To start collecting bearer traces:

a. If bearer tracing is set up and activated on an endpoint that has an active bearer channel, tracing starts immediately, otherwise, it starts as soon as a call (VoIP) or a connection (AAL2) is activated on the endpoint.

b. If bearer tracing is set up and not activated on an endpoint, tracing must be initiated manually even if there is an active call/connection on the endpoint.

Step 3 To stop collecting bearer traces:

a. End the active call/connection on the endpoint, or

b. Manually stop the trace


Note As long as the trace is not stopped, VXSM continues to send trace samples to the file server


Step 4 VXSM generates a trace filename each time a trace is started and active on an endpoint. The filename is based on the probe type, the endpoint index and on specific user input that constitutes the file signature.


Note The file signature could be set to remain the same from one trace session to another or, in some cases; it could be set to be specific to each session (see Trace Files).


If the connection to the file server cannot be established, bearer tracing is deactivated automatically. To reactivate bearer tracing:

a. Fix the network issues that prevent the network connection to be established.

And

b. Restart the call on the endpoint, or stop then start the trace manually.

Step 5 All trace probes can be activated on the same endpoint. Each probe initiates its own trace stream to the network and writes its data to its own file except for VPU traces where VPU event traces and VPU segment traces are written to the same file.

Step 6 The level 1 and level 2 T.38 probes (7 and 8) cannot be activated at the same time. These probes also need PCM trace to be on.

Step 7 The total number of probes activated at the same time is limited, but the amount of data transferred per probe is not limited.


Note a. Ensure that there is enough free space on the file server to store the trace files generated by VXSM. For an estimation of the required space per trace, see the "Per Trace Bandwidth and File Size Requirements" section.

b. If the network connection experiences problems while bearer tracing is writing trace data into the trace file, bearer tracing may be deactivated automatically. To reactivate bearer tracing:

1. Fix the network issues that are preventing the network connection to be established
2. Restart the call on the endpoint or stop then start the trace manually



Bearer Trace Types

Bearer Traces are transmitted to the server as files with file names that indicate the probe type, the endpoint being probed and, optionally, a timestamp. Files are assigned names by the Bearer Tracing feature that indicate these items. For more details on file naming see Trace Files.

PCM Traces

There are 2 PCM traces, pcmin and pcmout, to and from the voice network, respectively. In CLI, pcmin Probe 1 collects pcmin and probe 2 collects pcmout. The pcmin trace files generated by VXSM have the pcmin prefix and.pcm suffix, pcmout trace file have the pcmout prefix and the .pcm suffix.

The format for a PCM trace is either 8-bit PCM A-law or 8-bit PCM mu-law, depending on the physical interface being used. VXSM currently ties the PCM encoding to the physical interface.

Echo Canceller Traces—Sout

A sout trace is collected from the output of the echo canceller and is collected by probe 5.

The format of the sout stream is a big-endian signed 16-bit linear format. The value 32757 represents full scale. The sout trace file generated by VXSM has the sout prefix and .pcm suffix

The sout stream is synchronized to the PCM In and PCM Out signals.

The sout tracing cannot be enabled if the Ecan is not enabled on the given connection

Packet Traces

The packet traces are collected at the input and output of the packet IO interface.

The current implementation focuses on RTP traces. The packet trace is in raw IP pcap format that is readable by packet sniffers such as Ethereal.

Voice Playout Unit Traces

There are two type of voice playout unit (VPU) traces:

VPU event on probe 6—VPU Event trace (vpevt) reports jitter-related events when they happen.

VPU segment on probe 10—The Voice Playout Unit (jitter buffer) segment trace (vpseg) reports jitter-related data every segment (that is, every 5 ms).

Although there are 2 types of VPU traces, only one file is created on the server for both traces. The vpu file generated by VXSM has the vpu prefix and the vpu suffix.

T.38 Traces

When a fax call is active using T.38 fax relay, traces can be gathered from the Fax module within the DSP. The user can select to gather 2 levels of traces:

Level 1—Basic events to decipher T.30 and other signaling progress

Level 2—Proprietary Telogy format

When enabling the tracing functionality for T.38 the user can chose to monitor any DS0 in a DS1 or a specific DS0 in a DS1. When monitoring any DS0 in a DS1 the first fax relay session activated on the monitored span will have tracing enabled. User can choose to monitor any DS0 in a DS1 by choosing a special DS0 attribute in the addbearertraceendpt CLI.

A total of 10 simultaneous DS0 can be monitored at any one time. In addition to this, a maximum number of 10 trace sessions can be activated simultaneously.

The file containing the T38 traces will have a header that provides some information about the fax relay session. It contains ds1 and ds0 information, tcid/dsp information, and the timestamp indicating when the file was created.

In CLI, the T38 level l trace 1is activated by probe 8 and T38 level 2 trace is activated by probe 7.

The file created on the server for T38 level 1 trace has the suffix DBG1.

The file created on the server for T38 level 2 trace has the suffix DBG2.

Both level 1 and level 2 traces have the prefix FaxRelayTrace.

After the filename base is entered by the user, the filename is also appended to the endpoint prefix, followed by the DS0 number.

For example, in the case of FTP where the timestamp is used as a filename base, the filename for a level 1 trace on endpoint 2 ds0 1 is: FaxRelayTrace_022306172604_2_1.DBG1. See Table 8-9 for details.

MSG Traces

Probe 9 collects MSG traces. The message trace file generated by VXSM has the msg prefix and the msg suffix.

Message traces contain the information within DIM traces. MSG Traces are asynchronous and depend on whether or not the activity on the card generates DIM traces.

Bearer Tracing Connectivity

External Interface

VXSM can be connected to the file server either through the control PVC or through the VXSM Ethernet port.

If the control PVC is used, the provisioning of the control PCV needs to include the bandwidth required for the transfer of trace data to the network.

Per Trace Bandwidth and File Size Requirements

Bearer tracing requires that enough bandwidth be available to accommodate the bearer trace traffic data and the transport protocol packet and control overhead. Table 8-6 and Table 8-7 below show the estimate per-trace bandwidth requirements in the case of FTP and TFTP. The file size of the accumulated bearer trace data on the server is also provided so that enough space is set aside for the trace files to be stored in the FTP server. The file size of the trace files is limited only by the free space on the file server. To make sure that trace files have no loss of data, provide enough space on the file server for the trace files.

Table 8-6 FTP Bandwidth and File Sizes

Probe Type

AAL5 Bandwidth per Trace in cells/s

Ethernet Bandwith in Bytes/s

File Size for 5 min of tracing in Bytes

PCM in, PCM out, Sout

200

9000

2400000

Pkt in, Pkt out

250

11000

3600000

VPU segment

100

3000

920000

Sout

400

18000

4800000


Table 8-7 FTP Bandwidth and File Sizes

Probe Type

AAL5 Bandwidth per Trace in cells/s

Ethernet Bandwith in Bytes/s

File Size for 5 min of tracing in Bytes

PCM in, PCM out, Sout

210

9500

2400000

VPU Segment

90

4300

920000

Sout

420

19000

4800000


Bearer Tracing for AAL2 Connections

FTP and TFTP protocols run on an IP network. For Voice over IP, VXSM inherently uses an AAL5 control PVC to communicate with the MGC. The same control PVC can be used to communicate with the external FTP/TFTP server.

Control PVC is not inherent in an AAL2 configuration (for example, AAL2 trunking). To access the IP cloud in such a situation, either a control PVC or the Ethernet console need to be configured. Unless pinging the file server is possible, tracing does not operate.

IPSec and Bearer Tracing

The Bearer Tracing mechanism uses FTP/TFTP connections to transfer the traces to the server. The trace data are carried over the control PVC, which is also used to communicate with MGC. For many solutions, IPSec is used to secure the communication with the MGC over control PVC. In such cases, the FTP/TFTP traffic is bypassed by IPSec to transfer trace data to the bearer tracing server.

Bearer Trace CLI Requirements

Persistency

The operational state of bearer tracing is not saved in disk database. Hence, all the information is lost if the card resets. All the bearer tracing commands are non-MIB commands.

Backward Compatibility

Commands introduced in the PCM tracing functionality in the earlier VXSM releases are no longer functional. They are replaced by the new bearer tracing commands in VXSM Release 5.3.

Bearer Trace Configuration

After a bearer trace or a server profile is added, the only way to change its configuration is by deleting the trace or the server profile and adding it again with the new configuration.


Note For bearer tracing to occur, there must be a connection between VXSM and the file server. An FTP/TFTP server need to be running on the file server and access permission to the directory (and in the case of TFTP, to the files) must be granted.



Note Before you delete either configuration, make sure that the trace is stopped.


Considerations and Limitations

1. A VXSM card supports up to 32 simultaneous trace sessions.

2. A VXSM card supports the following max number of traces per probe type (Table 8-8):

Table 8-8 Trace Limitations

Probe Type
Maximum Number of Traces

Pcmin

5

Pcmout

5

Pkt in

5

Pkt out

5

Ecan/Sout

5

VPU

5

MSG

2


3. Only one trace per DSP core is possible, consequently, adding tracing on an endpoint will fail if the dsp core is found to have tracing enabled for a different endpoint.

4. Network latencies beyond 20 seconds have an impact on the quality of traces. To avoid loss of data during the trace collection, have the file server on the same LAN segment as VXSM.

5. Bearer tracing requires about 1 percent of CPU use. For tracing to operate properly, only enable the number of traces that the system CPU use can manage. Also, under a heavy call-rate, there might not be enough CPU bandwidth to accommodate the maximum number of traces.

6. In Packet Trace, to capture events such as DTMF or Tones in a trace file, stop bearer tracing before using the trace file in an offline processing.

7. FTP sessions time out after 30 seconds of inactivity so that dead network connections do not affect the system adversely.

8. All traces must be set up at the same time as there is no provision to set up each one individually.

9. We recommend that bearer traces and T.38 traces are run separately.

Configuring the Bearer Tracing Feature

Configuration Summary

The following steps provide a quick summary of the procedure for configuring the Bearer Tracing feature.


Step 1 In a typical network scenario, identify the ds0 that needs to be monitored for bearer trace collection. Use addbearertraceendpt to define the endpoint and the trace probes that need to be enabled.

Step 2 Use addbearertracesrvprof to add TFTP or FTP server profile.

Step 3 Use addbearertrace to tie the endpoint to the defined server profile.

Step 4 Make sure that the necessary files are created on the server with correct write permissions if TFTP transfer mode is to be used.

Step 5 Use bearertracestart to enable the tracing on the given endpoint if it is not already enabled using addbearertrace or simply start a call on the endpoint

Step 6 Use bearertracestop to stop the tracing on the given endpoint, or simply stop the call in a VOIP configuration or delete the connection in an AAL2 configuration


Detailed Configuration

To set up bearer tracing, use the following procedure.


Step 1 Set up a connection to an external server to receive the trace information. See Bearer Tracing Connectivity for connectivity details

Step 2 Use the addbearertraceendpt command create the TDM segment of a bearer trace session.

This command is used to create a session with a specified DS0 or DS1 endpoint for bearer tracing. See T.38 Traces for T.38 endpoints.

The format of this command is:

addbearertraceendpt <endptIndex> <LineNum> <ds0> -trace <probe>

For <endptIndex> enter a number in the range 1 to 32. This parameter uniquely identifies the bearer trace session.

For <LineNum>, enter line number of the endpoint in one of the following formats.

Interface Type Line Number Format Values

OC3/SDH bay.line.path.vtg.vt bay {1 - upper}
or bay.line.path.ds1 line (range=1..4)
path (range=1..3)
vtg (range=1..7)
vt (range=1..4)(ds1)
(range=1..3)(e1)
ds1 (range=1..28)

T1/E1 bay.line bay {1 - upper}
line (range=1..24)

T3 bay.line.ds1 bay {1 - upper}
line (range=1..6)
ds1 (range=1..28)

For <ds0>, enter:

A number in the range 1 to 24 or all for a T1 interface.

A number in the range 1 to 32 or all for an E1 interface.

For -trace<probe>, enter the probe number(s) to be used in the session. Enter either a single number or multiple numbers, for example, 1,3,5,7)

1 = PCM Input
2 = PCM Output
4 = Network Input
5 = Network Output
3 = Sout (echo canceller output)
6 = VPU - Events
7 = T38 Level 2
8 = T38 Level 1
9 = Message Events
10 = VPU - Segments


Note The line number all option can be used only when one of the probes 7 and 8 is chosen.


To verify the creation of the session, use the dspbearertraceendpt command. The format of this command is:

dspbearertraceendpt <endptIndex>

Step 3 Use the addbearertracesrvprof command to create the network segment of a bearer trace session.

The syntax of this command is:

addbearertracesrvprof <profIndex> <srvIP> <xferMode> <fileNameBase>[<uploadPath> <portNumber/login><password>]

For <profIndex>, enter a number in the range 1 to 10. This parameter uniquely identifies the bearer trace server profile.

For <srvIP>, enter the IP address of the server to receive the bearer trace reports. The IP address can be provided using either IP dot-notation (nnn.nnn.nnn.nnn) or a name (alpha.beta.com) with a proper DNS support.

For <xferMode>, specify the transfer mode for communicating with the server. Enter 0 for FTP, or 1 for TFTP.

For <fileNameBase>, specify a fileNameBase. For FTP, the user can also enter # to indicate the default fileNameBase is to be used (default is not allowed for TFTP).

VXSM generates filenames as follows (Table 8-9):

Table 8-9 Generated Bearer Trace Filenames

xferMode
fileNameBase
Filenames

1 = TFTP

FILENAMEBASE specified by user

PCM In - pcminFILENAMEBASE_EndPtID.pcm
PCM Out - pcmoutFILENAMEBASE_ EndPtID.pcm
Sout - soutFILENAMEBASE_ EndPtID.pcm
VPU - vpuFILENAMEBASE_ EndPtID.vpu
Network In - netwrxFILENAMEBASE_ EndPtID.pkt
Network Out - netwtxFILENAMEBASE_ EndPtID.pkt
T.38 level 1 - FaxRelayTrace_FILENAMEBASE_EndPtID_DS0.DBG1
T.38 level 2 - FaxRelayTrace_FILENAMEBASE_EndPtID_DS0.DBG2
MSG - msgFILENAMEBASE_EndPtID.msg (to and from DSP)

1 = TFTP

# for default

Not allowed

2 = FTP

FILENAMEBASE specified by user

PCM In - pcminFILENAMEBASE_EndPtID.pcm
PCM Out - pcmoutFILENAMEBASE_ EndPtID.pcm
Sout - soutFILENAMEBASE_ EndPtID.pcm
VPU - vpuFILENAMEBASE_ EndPtID.vpu
Network In - netwrxFILENAMEBASE_ EndPtID.pkt
Network Out - netwtxFILENAMEBASE_ EndPtID.pkt
T.38 level 1 - FaxRelayTrace_FILENAMEBASE_EndPtID_DS0.DBG1
T.38 level 2 - FaxRelayTrace_FILENAMEBASE_EndPtID_DS0.DBG2
MSG - msgFILENAMEBASE_EndPtID.msg (to and from DSP)

2 = FTP

# for default

PCM In - pcminMMDDYYHHMMSS_EndPtID.pcm
PCM Out - pcmoutMMDDYYHHMMSS_ EndPtID.pcm
Sout - soutMMDDYYHHMMSS_ EndPtID.pcm
VPU - vpuMMDDYYHHMMSS_ EndPtID.vpu
Network In - netwrxMMDDYYHHMMSS_ EndPtID.pkt
Network Out - netwtxMMDDYYHHMMSS_ EndPtID.pkt
T.38 level 1 - FaxRelayTrace_MMDDYYHHMMSS_EndPtID_DS0.DBG1
T.38 level 2 - FaxRelayTrace_MMDDYYHHMMSS_EndPtID_DS0.DBG2
MSG - msgMMDDYYHHMMSS_EndPtID.msg (to and from DSP)


For <uploadPath>, specify the path of the file in the server. For example. /pcm/tes


Note The server root directory is used if the path is not specified.


For <portNumber/login>, enter a TFTP port number if xferMode = 1 or the FTP login name If xferMode = 0.

The TFTP port number is a number in the range 1 to 255 with a default of 69.

For login, use the FTP login name, default is guest.

For <password>, use the FTP password, default is guest.

The creation of the profile can be verified by using the dspbearertracesrvprof command. The format of this command is:

dspbearertracesrvprof <profIndex>

Step 4 Use the addbearertrace command to link the TDM segment and the network segment of the bearer trace session. Optionally, this command can also start the bearer trace function.

The format of the addbearertrace command is:

addbearertrace <endptIndex> <profIndex> <active>

For <endptIndex>, enter the endpoint index number of the bearer trace session. A number in the range of 1 to 32.

For <profIndex>, enter the profile number of the bearer trace profile. A number in the range of 1 to10.

For <active>, specify 1 for enable or 0 for disable. If this parameter is omitted, the default is enable.


Note If the bearer tracing is enabled, tracing starts automatically whenever a call is added on the given bearer endpoint. If the bearer tracing is not enabled, then it must be explicitly started using the bearertracestart command (see also the corresponding bearertracestop command).


The creation of the association can be verified by using the dspbearertrace command. The format of this command is:

dspbearertrace<endptIndex>


Bearer Trace Configuration Examples

Example 1

This sample procedure sets up the bearer tracing endpoint, the FTP/TFTP server, and a method of mapping them to each other before making and tracing a call. The call is on ds0=1 on ds1=1. Configure the bearer tracing endpoint and the servers:


Step 1 Configure the bearer trace endpoint with appropriate trace probes.

martler.2.VXSM.a > addbearertraceendpt 1 1.1.1.1.1 1 -trace 1,2,5

Step 2 Configure the bearer trace server with appropriate parameters.

martler.2.VXSM.a > addbearertracesrvprof 1 172.17.38.159 0 sarang/root password

Step 3 Map the bearer trace endpoint using the bearer trace server.

martler.2.VXSM.a > addbearertrace 1 1 1

Step 4 Verify that the mapping is successful.

martler.2.VXSM.a > dspbearertraceendpts

=============================================
         BEARER TRACE ENDPT PROFILES         
=============================================

   Index                   : 1
   line                    : 1.1.1.1.1
   lineNo                  : 1
   ds0No                   : 1
   bearerTraceSrvProf      : 0x1   <---- Server Profile is added
   bearerTraceProbes       : 0x13

Step 5 Verify that the parameters for bearer trace server are correct.

martler.2.VXSM.a > dspbearertracesrvprofs


=============================================
         BEARER TRACE SERVER PROFILES         
=============================================

Server Profile Index             : 1 
Bearer Trace Server IP           : 172.17.38.159
Server upload directory          : /
TFTP/FTP Mode                    : FTP
File Name Base                   : sarang
FTP server login                 : root
FTP server password              : password
---------------------------------------------

Step 6 Place a call on ds0=1 and ds1=1, and then delete it after some time. On the FTP server the following files are created.

-rw-r--r-- 1 root other 690480 Oct 16 12:17 pcminsarang_1.pcm
-rw-r--r-- 1 root other 690480 Oct 16 12:17 pcmoutsarang_1.pcm
-rw-r--r-- 1 root other 1380960 Oct 16 12:17 soutsarang_1.pcm


Example 2

The following example shows how to collect PCM traces for a call with both endpoints on the same card:


Step 1 Configure the bearer tracing endpoint for PCMin and PCMout traces.

addbearertraceendpt 1 1.1.1.1.1 1 -trace 1,2

addbearertraceendpt 2 1.1.1.1.1 2 -trace 1,2

Step 2 Add a Bearer Trace File Server.

addbearertracesrvprof 1 172.17.38.159 0 rxie /tftpboot root password

Step 3 Map the bearer endpt to the server profile.

addbearertrace 1 1 1

addbearertrace 2 1 1

When you make a call on endpoints 1 and 2, the PCM tracing starts automatically because of the last parameter in the addbearertrace command above.

However, If you use a different last parameter as in the following commands,

addbearertrace 1 1 0

addbearertrace 2 1 0

Tracing does not start automatically and the following commands are needed to start the tracing after the call is up:

bearertracestart 1

bearertracestart 2

Step 4 To stop bearer tracing at any time, use the following CLIs:

bearertracestop 1

bearertracestop 2


Trace Files

Server File Requirements

1. Before you enable bearer tracing, files corresponding to the trace files that will be generated by VXSM need to be created on the file server.

2. Write access to both the created traces files and to the directory where they reside must be set.

3. If a directory is specified when adding a server profile on VXSM, this directory is relative to the default TFTP directory specified in the TFTP config file (in UNIX, the default is set to /tftpboot in /etc/inet.conf.

FileNames

Bearer trace information is saved to the server as data chunks that will be stored in a file of which the name is generated by VXSM. The file naming scheme is slightly different depending upon whether TFTP or FTP are used as the transport protocol.

TFTP Filenames

Filenames used with TFTP consist of the following parts:

A probe type

A FILENAMEBASE

The endpoint index being probed

A file extension

For example, a trace for PCM In on a bearer trace endpoint of index 1 and with a Filename Base Mytrace yields a filename of:

pcminMytrace_1.pcm

Before enabling the bearer trace function, filenames corresponding to trace files must be created on the server. Further, VXSM must have write access to the trace files and the directory in which they reside. The path to the file is relative to the default TFTP directory (usually set as /tftpboot).


Note When using the FILENAMEBASE, subsequent traces for the same probe and same endpoint overwrite the previous file in the server. It is the responsibility of the server to take action to prevent this occurrence.


FTP Filenames

For systems using FTP, the use of a FILENAMEBASE is optional. If a FILENAMEBASE is specified, the same file naming scheme used for TFTP is used (see previous paragraphs). If a FILENAMEBASE is not specified, VXSM generates a filename having the following parts:

A probe type

A timestamp MMDDYYHHMMSS based upon a 24-hour clock)

The endpoint index being probed

A file extension

For example, a trace for PCM In on a bearer trace endpoint of index 1 yields a filename of:

pcmin111205104520_1.pcm


Note The filename method using timestamps automatically prevents the overwriting of previous trace files.