Some CICS® resource types can have resource names that are based on a prefix or a suffix concatenated with a counter or an address. These resource names can create millions of records that are stored in the CICS IA data space, VSAM files, and the DB2® database.
For example, if you could have a CICS TSQUEUE resource with a name based on QUEUEn, where: n is a 10-digit string that is controlled by a CICS counter.
If you have WRITEQ, READQ and, DELETEQ commands for this TSQUEUE resource in the program, CICS IA, will have three entries for each TSQUEUE name that is processed, starting with QUEUE0000000001 and ending with QUEUE9999999999. Based on this format, there are already 3 billion records for that single program. The number of records can increase substantially if the data for that program is collected in more than one region, or if the queue name is used by more than one program.
You can reduce the number of records to three if you create a "resource prefix list" table. This resource prefix list consists of a simple assembler CSECT where you can use prefixes, suffixes, or any pattern. Currently you can create these lists for CICS TSQUEUE names, CICS ENQ/DEQ names, and WebSphere® MQ queue names. You can configure the same table for each CICS region or you can configure individual ones dependent upon the content of that region.
When you first start collecting CICS IA data, use the CICS IA plug-in for CICS Explorer® to identify the resource names that match a pattern, and add these names to your resource prefix list. You must remove the duplicated data by deleting them from the VSAM file and the DB2 tables. A cleanup job is provided to remove duplicate resource names for CICS TS QUEUE names, CICS ENQUEUE, CICS DEQUEUE, and WebSphere MQ queue names.
For more information about removing duplicate resources, see Removing duplicate resources from the Collector files.
For more information about resource prefix lists, see Creating a resource prefix list.