The Termination Status report summarizes the completion status of the selected job types running on a scheduler during a certain time period. It shows the number of jobs ending in Failure, Terminated, or Success status, for each scheduler and day, and optionally shows details of Failure and Terminated runs. Such reports are very useful, but can affect performance of AAI if too many/large selection is being run at the same time.
This article describes what parameters can be modified/looked at in order to control the amount of generating termination status reports.
Release : 5.X, 6.X
Component : AUTOMIC AUTOMATION INTELLIGENCE ENGINE
All these values can be found in the Params section of the Configuration Tool, and 'Show less commonly configured' has to be selected. Parameters are listed below:
reportMaxParrallelGenerators - Maximum number of report generators that can run simultaneously. Default: 5
reportMaxParallelPDFWriters - Maximum number of PDF writers that can run simultaneously. Default: 1
reportMaxParallelWriters - Maximum number of Non-PDF writers that can run simultaneously. Default: 5.
All the default values are recommended values by Broadcom. For reportMaxParrallelGenerators and reportMaxParallelWriters can be limited to 1 or 2 if the system is big, and performance issues were experienced in the past due to termination status reports generation.
Keep in mind that the selection of statuses as well as dates when running termination report can also be an issue if a big amount of data is selected. For example, it is not recommended to put something like thousands of days for 1 report if it is a big system. Please see the screenshot below:
All unnecessary statuses should not be selected. In this example, 'Success' not selected (which is optional, just be aware it can be pulling too much data if a big time frame is selected for past runs). With such selection, there will be 50,000 days worth of data for termination status report. The more days put in, the more data report would try to pull out of the database, and that means long running query parsing through enormous amount of data, that can cause performance issues on AAI side.