Key/non-key groups

You can further identify groups in the project hierarchy as key groups and non-key groups. Key groups are defined as the groups within a hierarchy that contain all the software components of the application under development. A key group is a group into which you move data during a promotion. A project can have as many key groups as you want as long as any hierarchical view has no more than 123 groups. The actual limiting factor is the MVS™ limit of 123 extents for a concatenated partitioned data set.

SCLM allows a project to specify transition groups between key groups. These groups are known as non-key groups. A non-key group is a group into which you copy (rather than move) data during a promotion. When you promote data in a hierarchy, SCLM does not purge data from a key group until it reaches the next key group. Therefore, in a project with non-key groups, SCLM temporarily duplicates data in the non-key groups and the next lower key group. Figure 1 illustrates the relationship between a key and a non-key group within a project hierarchy.

As the figure shows, two non-key groups (the STAGE layer) appear between the development groups (the DEV layer) and the test and integration group (the TEST layer.) Developers use the STAGE groups as an interim place into which they promote their work before it moves to the next layer.

Using non-key groups enables you to display the critical elements of the hierarchical structure on ISPF panels. Because ISPF panels allow you to display only four key groups at one time, it is difficult to display the highest group in the hierarchy when you have a complex project that contains many layers.