A batch analysis of a Migration Strategy fails with an SB37 space abend on ddname DDLFILET. As this dataset is dynamically allocated, how do I resolve this space problem?

Document ID : KB000054659
Last Modified Date : 14/02/2018
Show Technical Document Details

Description:

When performing an analysis of a Migration Strategy in batch, the job fails with an SB37 on ddname DDLFILET.

The job output contains the following messages:

 RMA365E ************************************************************** 
 RMA365E *                                                            * 
 RMA365E *        RC/MIGRATOR ANALYSIS ABEND HAS OCCURED              * 
 RMA365E *      RMA@ESTA GENERAL PURPOSE ESTAE EXIT ENTERED           * 
 RMA365E *                                                            * 
 RMA365E *        ABEND CODE WAS:SB37.  REASON CODE WAS: 00000004     * 
 RMA365E *                                                            * 
 RMA365E *   ABEND OCCURED IN NON-MIGRATOR MODULE: IFG0199B           * 
 RMA365E *   MODULE LOAD ADDRESS: 00E5B000                            * 
 RMA365E *   MODULE ENTRY ADDRESS: 00E5B000                           * 
 RMA365E *   MODULE LENGTH: 000021F0                                  *  
 RMA365E *      RMA@ESTA GENERAL PURPOSE ESTAE EXIT EXITING           * 
 RMA365E *                                                            * 
 RMA365E ************************************************************** 

The job completes with a return code of SB37.

The DDLFILET dataset is being dynamically allocated because parameter PDSSHR has been set to YES in the MIGRATOR PARMLIB member.

 PDSSHR  (YES)               /* OUTPUT PDS USAGE.  SPECIFY    */ 
                             /* YES IF THE ANALYSIS PDS WILL  */ 
                             /* BE SHARED BY MULTIPLE USERS.  */

The DDLFILET dataset is dynamically allocated with a space allocation of CYL(1,10), which on 3390 packs allows for a maximum of approximately 1 million rows.

Solution:

As the space allocation of CYL(1,10) is not enough to perform the failing batch analysis, it will be necessary to perform this analysis with PDSSHR set to NO, so that the DDLFILET dataset is not dynamically allocated.

Edit the MIGRATOR PARMLIB member, changing the PDSSHR parameter to NO.

 PDSSHR  (NO)                /* OUTPUT PDS USAGE.  SPECIFY    */ 
                             /* YES IF THE ANALYSIS PDS WILL  */ 
                             /* BE SHARED BY MULTIPLE USERS.  */

Then rerun the failed batch analysis job/step.
The DDLFILET dataset will now not be dynamically allocated.
After the job has completed successfully then the PDSSHR parameter can be reset back to YES if required.