A.14.2 MONITORAVGLATENCYTHRESHOLD

Specifies whether the average latency of events between the start- and end-points of a path crosses a specified threshold.

You specify the start and end stages of the path, and whether it should start or end at the entry or exit points of each respective stage. If you specify the same stage for the start and end of the path, you can monitor the latency threshold of events flowing through a single stage.

Syntax

java wlevs.Admin 
    [ Connection Arguments ]
    [ User Credentials Arguments ]
    [ Common Arguments ] 
    MONITORAVGLATENCYTHRESHOLD -application application 
    -startStage startStage -startStagePoint stagePoint 
    -endStage endStage -endStagePoint stagePoint 
    -avgInterval avgInterval -timeUnit timeUnit -threshold threshold

Table A-32 MONITORAVGLATENCYTHRESHOLD Arguments

Argument	Definition
-application application	Specifies the name of the Oracle Stream Analytics application whose throughput and latency you want to monitor. To get the exact name of your application, you can: Use wlevs.Admin to query for the name (see Query for Application and Processor Names). Use the Oracle Stream Analytics Visualizer: Start the Oracle Stream Analytics Visualizer (see Using Visualizer for Oracle Stream Analytics). In the left pane, navigate to and expand the Applications node of the Oracle Stream Analytics instance to which the application is deployed. Each node under the Applications node is named with the exact application name. Look at the MANIFEST.MF file of the application; the application name is specified by the Bundle-SymbolicName header.
-startStage startStage	Specifies the name of the stage that starts the path for which you want to monitor the latency threshold. The stage is in the application specified by the -application option.
-startStagePoint startStagePoint	Specifies the specific starting point for monitoring the latency threshold of the specified start stage. You can start monitoring from the entry or exit point of the start stage. Valid values are entry and exit. Default value is entry.
-endStage endStage	Specifies the name of the stage that ends the path for which you want to monitor the latency threshold. The stage is in the application specified by the -application option.
-endStagePoint endStagePoint	Specifies the specific ending point for monitoring the latency threshold of the specified end stage. You can end monitoring from the entry or exit point of the end stage. Valid values are entry and exit. Default value is entry.
-avgInterval avgInterval	Specifies the average interval across which average the latency threshold is calculated. Default value is 100. Specify the units with the -timeUnit option; default is milliseconds.
-threshold threshold	Specifies the threshold value above which the metric event will be outputted at the end of every average interval. Default is 100. Specify the units with the -timeUnit option; default is milliseconds.

Example

The following example shows how to monitor the average latency threshold of events above 10 seconds average latency on the cqlprocessor stage, from entry point to exit point, of the helloworld application.

java wlevs.Admin 
     -url service:jmx:msarmi://localhost:9002/jndi/jmxconnector 
     -username wlevs -password wlevs 
     MONITORAVGLATENCY -application helloworld -startStage cqlprocessor
     -startStagePoint entry -endStage cqlprocessor -endStagePoint exit 
     -avgInterval 100 -timeUnit MILLISECONDS -threshold 100