The Case Client In-basket can be used to view, open and work with BAW and FileNet P8 work items. Some users prefer to have access to large numbers of work items at a time, e.g. for sorting or filtering relevant items.
Depending on the number of items in the system and the number of case properties, this can lead to unsatisfactory performance, especially with complex sort and filter operations.
This blog describes the relevant settings to enable higher number of work items while maintaining satisfactory end user performance for in-baskets.
You can control the number of work items displayed on a single page. The default page size is 50 but this can be increased by changing the "Number of work items in In-basket page" option in the IBM Business Automation Workflow Case client plug-in:
Saved search count limit
For work items which are implemented as BAW work items (i.e. "Processes"), the in-basket uses the same API as the Process Portal. This implies that search optimization features used there apply to Case client in-baskets as well.
One optimization is the number of work items that are returned as the result of a query. This is controlled by the process-search-engine-count-optimization property. The default is 500. If you increase this value, you will get a more work items returned in a single query and a more accurate count of items available but at the cost of decreased performance.
Saved search optimization (pivot tables)
There is a BAW feature to optimize the performance of saved searches which applies to BAW work items displayed in in-baskets as well. This is called "Saved search acceleration tools" or "pivot tables". The BAW documentation describes how to execute these tools: Saved search acceleration tools
Impact of search optimization
This diagram show API performance for a Case client role based in-basket on IBM BAW 22.x (on premise). The total number of case work items is 10,000, the page size is set to 100, process-search-engine-count-optimization is set to 10,000 and different filter combinations are employed.
Note that even with no work item filters applied, performance increases dramatically after having run the Saved search acceleration tools