You can back up a data set during job processing by invoking DFSMShsm to perform inline backup.

DFSMShsm provides two distinct methods to perform inline backup. Each has its benefits over the other. The first method uses the HBACKDS command in a TSO terminal monitor program (TMP) batch job step. This allows for data set filtering using % and * characters for masking, and provides an option to only perform the backup if the data set was modified since the last backup was taken by DFSMShsm. The second method is to use the ARCINBAK program in a batch job step. Each data set is identified by a unique DD statement and supports the use of JCL name substitution and relative data flow diagram names. The second method can be used while the job has exclusive use of the data set, for example, if it was created by an earlier job step.

You can request inline backup copies by placing the appropriate statements in a job step. The copies are created by DFSMShsm incremental backup during processing. Inline backup writes the backup version on a ML1 DASD volume to minimize unplanned tape mounts. The backup versions can be moved from ML1 to incremental backup volumes automatically during incremental backup processing, or by command.