How do I set Include results for individual XXX?s parameters? (NETIQKB37022)

The Knowledge Scripts that retrieve WebSphere performance data can be divided into several categories, based upon the components they are associated with.  These categories map to object types defined in the Operator Console TreeView.  For example, there are EJB scripts, and JDBC scripts, and so on.  In many cases, a Knowledge Script can be dropped either on a leaf node of the tree, or on a higher-level node.  For example, a J2C Connection Pool script can be dropped on either the J2C Connection Pools object, or on a particular J2C data source.  In cases such as this, if a script is dropped on a container (i.e., folder) node rather than a leaf node of the tree, you can choose to collect data (and trigger threshold events) for either the collection as a whole, or for each individual child of the collection (as well as the collection as a whole).

When you run a Knowledge Script against a container node of the tree, results are computed for each descendent (child or grandchild) of the container node, and these are used to compute an aggregate result for the container as a whole.  This aggregate data point is always included in the results.  However, the results for the child and grandchild nodes of the target node are included only if the corresponding "Include results for individual XXX's?" parameter is set to "y".

In some cases, a script can be dropped on any of three levels.  For example, a JDBC script can be dropped on the JDBC Connection Pools object, or a JDBC provider, or a JDBC data source.  In this case, if you drop the script on the top-level object (JDBC Connection Pools), you can choose to collect only the aggregate data (a single data point), or you can choose to have data aggregated at the JDBC provider level (an aggregated data point for the JDBC Collection Pools resource, plus individual aggregated data points for each provider), or you can choose to have data for each individual data source (as well as aggregates for the JDBC Connection Pools container as a whole, and optionally the aggregated results for each provider).

To provide a more explicit example, if you drop a JDBC Knowledge Script on the JDBC Connection Pools object in the tree, you can set the "Include results for individual XXX's?" parameters as follows:

1.         Include results for individual providers? y

Include results for individual data sources? y

This will produce results for each individual data source, as well as aggregated results for each provider, and a total aggregated result for the JDBC Connection Pools resource.

2.         Include results for individual providers? y

Include results for individual data sources? n

This will produce aggregated results for each provider, and a total aggregated result for the JDBC Connection Pools resource, but it will not produce results for the individual data sources.

3.         Include results for individual providers? n

Include results for individual data sources? y

This will produce results for each individual data source, as well as a total aggregated result for the JDBC Connection Pools resource, but it will not produce aggregated results for the individual providers.

4.         Include results for individual providers? n

Include results for individual data sources? n

This will produce only a single aggregated result for the JDBC Connection Pools resource as a whole.

Note that it's possible to set these parameters in such a way that the Knowledge Script will take no action.  For example, if you set "Include results for individual data sources?" to "n", and you run the script against a data source resource (rather than against a provider, or the top-level JDBC Connection Pools resource), you will not get any results.  In this case an event will be generated to inform you that the Knowledge Script performed no action.

Note also that if you run a Knowledge Script against a non-leaf node of the tree, you can choose to include results only for a selected set of descendent nodes, by deselecting the unwanted nodes in the tree shown in the Objects tab of the Knowledge Script.  If you do this, any aggregated values included in the results will be based upon only those resources which were left selected (i.e., only those nodes which were not deselected).