Aggregating usage and discovery data

The Aggregator is a program that populates the Asset tables.

An audit table captures information of the Aggregator when this program is run, and these audit records can be browsed using an Analyzer report. As part of housekeeping, obsolete audit records from this program are deleted by the MDEL Deletion utility, together with audit records from other utilities.

These tables are accessed by Analyzer reports and batch jobs, and offer a higher level view of product discovery and usage at a version rather than a release or module level. The Aggregator also calculates product discovery and usage count totals that are used to speed up the Analyzer queries.

The Aggregator should be run as the final job in a batch run of Inquisitor Import and Usage Import jobs, and before the repository is accessed by the Analyzer. This ensures that asset reports and counts are synchronized with the latest collected discovery and usage details.

It is usual for asset level usage details to be kept longer than module usage details. This is controlled by the KEEPAGGR parameter of the Usage Deletion program, which has a longer default period than the KEEPDETAIL (or FIRSTDATE/LASTDATE) parameter. The aggregated entries are normally kept for the longer KEEPAGGR period and are usually only recalculated for periods where module usage details are available. Note that the KEEPAGGR parameter value determines the number of periods of Asset (aggregated entries) records to be retained when the COUNTUSAGEFULL=Y is set. See description of parameter COUNTUSAGEFULL in the TPARAM section.

The Aggregator run can be time consuming since it is run for the entire repository and periods rather than an individual SID (as in the case of the Inquisitor Import), or system and period (as in the case of the Usage Import). It is therefore more efficient to run the Aggregator only once following several imports of usage or discovery data. Ensure also that you do not keep usage details for longer than is necessary for your site, and that Usage Deletion is run on a regular basis.

An example of how to run the Aggregator can be seen in sample jobs that are generated by HZASCUST, such as HZASIQIM (Inquisitor Import) and HZASUIMP (Usage Import).

Note Db2® only:

The Aggregator program uses declared global temporary tablespace (DGTT) to process and summarize data in the repository database. And there are also extensive data sorts involved. For performance reasons, increase the sizes of the tablespaces belonging to the work file database and, also consider using different definitions for the Bufferpool (instead of the default BP0 or BP32K). In PARMLIB member HZASAGP1, you can change the storage group and index Bufferpool values for the DGTT (instead of the default BP0 or BP32K).