The job may run successfully if a larger REGION can be specified. If this is impossible, or if the job fails with larger REGIONs, submit a multistep job, breaking up the user data sets into portions as indicated by the messages that follow this one. If possible, isolate the USRDSN USEROPTS into their own steps. If a substantial number of RACGRP USEROPTS are specified in the JCL, isolate these into their own jobsteps also. Steps in which only USRDSNs or RACGRPS are specified should specify FUNCTION USRDSN or FUNCTION RACGRP to avoid the default (FUNCTION ALL) processing.