Control data sets are used to track and control application programs within the hierarchy. SCLM stores accounting and audit information in VSAM data sets whose names are defined in the project definition. VSAM data sets consist of VSAM clusters. A VSAM cluster is a named structure consisting of a group of related components. While it is not required that the first qualifier of VSAM data sets match the project name, it makes project maintenance easier. There are six types of VSAM data sets that can be associated with a project.

Primary Accounting

The accounting data set contains information about the software components in the project including statistics, dependency information and build maps (information about the last build of the member). At least one accounting data set is required for a project.

Secondary Accounting

The secondary accounting data set is a backup of the information in the accounting data set.

Export Accounting

The export accounting data set contains accounting information that has been exported from the accounting data set.

Primary Audit Control

The audit control data set contains audit information about changes to the software components in the project for groups that have auditing enabled.

Secondary Audit Control

The secondary audit control data set is a backup of the information in the audit control data set.

Cross-dependency data set

The Cross-dependency data set contains information used by the Where-used function.

Most projects start out with one VSAM data set, the primary accounting data set. Additional data sets can be added as the project evolves and more advanced SCLM capabilities are needed. Additional VSAM data sets are required for Import, Export, Auditing, automatic backup of accounting data and multiple control data set support. In some cases, it is desirable to use multiple VSAM data sets instead of one or two. If this is the case, see Splitting project VSAM data sets for additional information.

The VSAM data sets cannot be shared between systems in a sysplex under any other condition. Accessing any of the VSAM data sets from multiple systems when VSAM RLS is not available can result in the corruption of data, system errors, or other integrity problems. To avoid these problems, the project manager must allocate VSAM data sets so that they cannot be accessed from multiple systems. What this means is you should ensure that the VSAM data sets are allocated on DASD, or catalogs, not available to other systems in the sysplex.

Except when using the RLS, the share options required on the VSAM file allocations are 4,3. This means that the data set can be fully shared by any number of users in a single system, and VSAM refreshes the data and index components buffer pools for direct processing, to guarantee the coherency of the data in the buffer pool. Coherency, in this case, means that SCLM gets the most updated contents of the requested record.

All VSAM data sets should be REPROed periodically using the IDCAMS reproduction utility. This will reduce fragmentation and optimize the performance of your VSAM data sets.