Understanding the storage group construct

For data sets, the storage group construct simplifies the task of administering external data. By putting a number of homogeneous data sets into one storage group, they can be viewed as one entity.

An Object storage group is composed of a set of volumes. Each installation develops Object storage groups according to its individual needs. Storage groups can be used to segregate different types of data (such as production versus development) and aggregate like types of data.

By separating the physical volumes from the service level (as defined by the storage class construct), Object storage groups also allow installations to change the physical aspects of storage without affecting the logical requirements of data access. For example, a new volume or device can be added to the storage group without affecting end-user routines.

In an OAM environment, Object storage groups allow the storage administrator to define an object storage hierarchy. The object storage hierarchy classifies storage areas according to location and, therefore, according to retrieval response time. Each object storage hierarchy must contain an object directory, containing control information about each object. Additionally, the hierarchy can have:

DB2 object storage tables on disk
A file system (NFS or z/FS)
Optical volumes that are associated with optical libraries (real or pseudo), and operator-accessible optical disk drives
Tape volumes that are associated with tape libraries or stand-alone tape drives

During an object’s lifetime, it can move from one OAM storage hierarchy level (storage location) to another, ascending or descending depending on its performance objectives.

Related reading: For more information on storage groups for data sets and objects, see z/OS DFSMSdfp Storage Administration.