If you do not currently use application or user-oriented scratch pools based on job names and data set names, use DFSMSrmm system-based pooling or a general, installation-wide scratch pool. If you currently use application or user-oriented scratch pools that are based on job names and data set names, use SMS ACS routines to assign storage groups. You can use storage groups to pool volumes based on job names and data set names as described in Using SMS tape storage groups for DFSMSrmm scratch pooling. You can also implement DFSMSrmm pooling using DFSMSrmm VLPOOL options.

If you are already using some form of scratch pooling based on information such as job name and data set name, review your current use of these pools in order to plan or change your implementation under DFSMSrmm.