WebSphere Business Space powered by WebSphere, Version 6.2.0


Returning results to WebSphere Business Modeler using tracking keys

If you want to return runtime values collected from WebSphere® Business Monitor to WebSphere Business Modeler, you require a measure with a tracking key for each value to be returned. Tracking keys are used to correlate the elements between the monitoring result and the process model.

Tracking keys are generated automatically when you specify monitored values to return using the Monitored Values tab in WebSphere Business Modeler, export the model, and then import the model into the Monitor action services. These tracking keys are based on internal IDs and are precise and accurate. They guarantee that a single match will be found for each tracking key.

If you want to create your own tracking keys for returning runtime results, you can base each key on a fully qualified path to a process element. This type of tracking key works only if there is no name change to the monitored element, but it can be created manually if you want to add support for returning values to WebSphere Business Modeler while working in the Monitor action services.

The name-based tracking key has the following structure:
	Process name/
	Process catalog name+/
	Process name/
	Sub-process or containing loop name*/
	Process element name/
	Attribute name	
where the plus sign (+) indicates that this step can occur one or more times, and the asterisk (*) indicates that this step can occur zero or more times, as required.

With this name-based type of tracking key, it is possible that more than one match might be found during processing. If more than one match is found, only one value in WebSphere Business Modeler is updated with the actual monitoring result. The value updated is the first match found by the processor.

The following example shows a tracking key that describes the processing time attribute of a task (My Task) contained by a subprocess (My Subprocess) in a process (My Process) in the default process catalog (Processes) in a project (My Project):
My Project/Processes/My Process/My Subprocess/My Task/processingTime
Similarly, the following example shows a tracking key that describes the output criterion probability of a decision contained by a subprocess in a process. In this case, there is a name segment for the decision node (My Decision) as well as for the name of the output criterion whose decision probability is being updated (Yes). Every named step in the hierarchy, starting with the project, is represented as a segment in the value path.
My Project/Processes/My Process/My Subprocess/My Decision/Yes/outputSetProbability
The following table shows the supported attribute names that can be used as part of the tracking key, as well as the constraints on each type of returned value.
Supported attribute names Data constraints Type
processingTime Values are in milliseconds. Double
startupCost Currency and distributions are not supported. Values are imported as specific values. Double
executionCost (Processing cost in WebSphere Business Modeler) Currency and distributions are not supported. Values are imported as specific values. Double
revenue Currency and distributions are not supported. Values are imported as specific values. Double
outputSetProbability (percentage on the decision branch in WebSphere Business Modeler) 30.5% is in the format of 30.50. Double

For instructions about returning the actual values after running WebSphere Business Monitor, see "Exporting values."




Support | Terms of use | Contact IBM
Copyright IBM Corporation 2007, 2008. All Rights Reserved.
This information center is powered by Eclipse technology. (http://www.eclipse.org)