Setting up log files for analysis

To use the Voice Trace Analyzer, you must first set up log files for analysis on your IBM WebSphere® Voice Server. To do so, you should enable the following trace log settings, depending on the version you are using.

Setting up IBM WebSphere Voice Server logging on WebSphere Application Server 6.1.x

Setting up IBM WebSphere Voice Server logging on WebSphere Application Server 5.1.x

Setting up IBM WebSphere Voice Server logging on WebSphere Application Server 6.1.x

Table 1. Trace log settings
Voice Trace Analyzer Function Trace Required to Enable
Trace Log Analysis and Statistics (REQUIRED FOR ALL FUNCTIONS) IBM.WVS.COMMON.RTSPBridge=finest
Analyze or Listen to Audio Recordings:
  • IBM.WVS.COMMON.MediaConv stores ASR (un-endpointed) and TTS audio for the entire session as processed by the server.
  • IBM.WVS.ASR.ASRAPI stores the un-endpointed audio as is it received by the ASR engine, before speech recognition is completed.
  • IBM.WVS.ASR.ASRAUD stores the endpointed audio after recognition, clipped to the recognized speech.
  • IBM.WVS.COMMON.MediaConv=fine
  • IBM.WVS.ASR.ASRAPI=finer
  • IBM.WVS.ASR.ASRAUD=finer
Recognition Accuracy Computation (storage of grammar information) IBM.WVS.ASR.ASRBEAN=fine:IBM.WVS.ASR.ASRAPI=finer
To enable trace on your WebSphere Voice Server, follow these steps:
  1. Log on to the WebSphere Application Server administrative console.
  2. Click Troubleshooting > Logs and Trace in the console navigation tree. Click on the desired server name and click server1 and then click Change Log Detail Levels.
  3. Select the Configuration tab.
  4. Configure the trace output and trace output format if a change from the existing one is desired.
  5. Select the Runtime tab.
  6. Select Save runtime changes to configuration as well if you want to write your changes back to the server configuration
  7. Click Apply.
To enable IBM WebSphere Voice Server component tracing, follow these steps:
  1. Under Additional Properties, click Change Log Detail Levels. The Change Log Detail Levels window lists the components.
  2. Select IBM.WVS.ASR to expand a list of basic WebSphere Voice Server Asynchronous Speech Recognition components.
  3. (Optional) For capturing endpointed ASR audio, click IBM.WVS.ASR.ASRAUD and select Messages and Trace Levels > finer.
  4. (Optional) For capturing un-endpointed ASR audio or calculating recognition accuracy, click IBM.WVS.ASR.ASRAPI and select Messages and Trace Levels > finer.
  5. (Optional) For calculating recognition accuracy, click IBM.WVS.ASR.ASRBEAN and select Messages and Trace Levels > fine.
  6. Select IBM.WVS.COMMON to expand a list of basic WebSphere Voice Server components.
  7. (Optional) For capturing un-endpointed ASR and TTS engine audio, click IBM.WVS.COMMON.MediaConv and select Messages and Trace Levels > fine.
  8. For IBM.WVS.COMMON.RTSPBridge, click IBM.WVS.COMMON.RTSPBridge and select Messages and Trace Levels > finest.
  9. Click Apply > Close to exit the window and save your changes.
    The following entry is displayed in the Components section:
    *=info: IBM.WVS.ASR.ASRAPI=finer: IBM.WVS.ASR.ASRBEAN=fine: 
IBM.WVS.ASR.ASRAUD=finer: IBM.WVS.COMMON.RTSPBridge=finest
  10. Scroll to the bottom of the Configuration window and click Apply.
  11. To save configuration changes, click Save from the menu.
  12. Click Save to save the master configuration.

WebSphere Voice Server logs its traces to a trace log located at <was_root>/profiles/AppSrv01/logs/server1 where was_root is the directory where WebSphere Application Server is installed. The trace log shows trace events, such as function entry and exits, component events, and debug information.

After the tracing is applied, the trace entries are written when the WebSphere Voice Server processes an MRCP request. You should then run the WebSphere Application Server Collector tool to collect the appropriate WebSphere Voice Server trace data.
Note: For detailed information on enabling and monitoring trace, see the IBM WebSphere Voice Server for Multiplatforms Information Center.

Setting up IBM WebSphere Voice Server logging on WebSphere Application Server 5.1.x

Table 2. Trace log settings
Voice Trace Analyzer Function Trace Required to Enable
Trace Log Analysis and Statistics (REQUIRED FOR ALL FUNCTIONS) RTSPBridge=all=enabled
Analyze or Listen to Audio Recordings:
  • MediaConv stores ASR (un-endpointed) and TTS audio for the entire session as processed by the server.
  • ASRAPI stores the un-endpointed audio as is it received by the ASR engine, before speech recognition is completed.
  • ASRAUD stores the endpointed audio after recognition, clipped to the recognized speech.
  • MediaConv=event=enabled
  • ASRAPI=entryExit=enabled
  • ASRAUD=entryExit=enabled
Recognition Accuracy Computation (storage of grammar information) ASRBEAN=event=enabled:ASRAPI=entryExit=enabled
To enable trace on your WebSphere Voice Server, follow these steps:
  1. Log on to the WebSphere Application Server administrative console.
  2. Click Troubleshooting > Logs and Tracing in the console navigation tree. Click on the desired server name and then click Diagnostic Trace.
  3. Select the Runtime tab.
  4. Select the Save Trace check box if you want to write your changes back to the server configuration.
  5. Change the existing trace state by changing the trace specification to the desired state.
  6. Configure the trace output if a change from the existing one is desired.
  7. Click Apply.
To enable IBM WebSphere Voice Server component tracing, follow these steps:
  1. Click Modify under the Trace Specification section of the Runtime tab. A window is displayed listing components.
  2. Select the Save Trace check box if you want to write your changes to the server configuration.
  3. Select IBM WVS ASR to expand a list of basic WebSphere Voice Server Asynchronous Speech Recognition components.
  4. (Optional) For capturing endpointed ASR audio, select entry/exit from the ASRAUD pop-up menu.
  5. (Optional) For capturing un-endpointed ASR audio or calculating recognition accuracy, select entry/exit from the ASRAPI pop-up menu.
  6. (Optional) For calculating recognition accuracy, select event from the ASRBEAN pop-up menu.
  7. Select IBM WVS to expand a list of basic WebSphere Voice Server components.
  8. (Optional) For capturing un-endpointed ASR and TTS engine audio, select event from the MediaConv pop-up menu.
  9. For RTSPBridge, click RTSPBridge and then select all enabled from the pop-up menu.
  10. Click Apply > Close to exit the window and save your changes.
    The following entry is displayed in the Trace Specification section:
    ASRAPI=entryExit=enabled:ASRBEAN=event=enabled:RTSPBridge=all=enabled
  11. Scroll to the bottom of the Configuration window and click Apply.
  12. To save configuration changes, click Save from the menu.
  13. Click Save to save the master configuration.

WebSphere Voice Server logs its traces to a trace log located at <was_root>/logs/server1/ where was_root is the directory where WebSphere Application Server is installed. The trace log shows trace events, such as function entry and exits, component events, and debug information.

After the tracing is applied, the trace entries are written when the WebSphere Voice Server processes an MRCP request. You should then run the WebSphere Application Server Collector tool to collect the appropriate WebSphere Voice Server trace data.
Note: For detailed information on enabling and monitoring trace, see the IBM WebSphere Voice Server for Multiplatforms V5.1.x Information Center.
Parent topic: Voice Trace Analyzer