Demonstration 2

This demonstrates the effect on elapse time and I/O activity when blocked records are processed.

The file is deliberately processed twice, one to write and the other to read, to ensure measurable results.

If the JCL is cut and pasted to your own system, the submitter's Userid will be substituted for &SYSUID which should minimise the need for changes.

The results will follow the second demonstration.

Programs IEFBR14, IEBDG and SORT are standard IBM supplied programs, and should exist on your system, even if an alternative to DFSORT is installed.

Demonstration 2 - Demonstrate the improvement on elapse time using blocked records.

//S0010    EXEC PGM=IEFBR14       /* MAKE JOB RE-RUNNABLE  */ 
//OUTDD    DD   DSN=&SYSUID..IBG.DEMO,DISP=(MOD,DELETE),      
//             SPACE=(TRK,0)                                  
//S0020    EXEC PGM=IEBDG         /* CREATE BASE DATA SET  */ 
//SYSPRINT DD   SYSOUT=*                                      
//OUTDD    DD   DSN=&SYSUID..IBG.DEMO,LRECL=60,RECFM=FB,      
//             DISP=(,CATLG),BLKSIZE=0,                       
//             SPACE=(60,(100,50),,,ROUND),AVGREC=K           
//SYSIN    DD   *                                             
  DSD OUTPUT=(OUTDD)                                          
  FD NAME=FIELD1,LENGTH=60,FORMAT=AL,ACTION=RP,STARTLOC=1     
  CREATE QUANTITY=500000,NAME=FIELD1                          
  END                                                         
//S0030    EXEC PGM=SORT          /* SORT THE DATA SET     */ 
//SYSOUT   DD   SYSOUT=*                                      
//SORTIN   DD   DSN=&SYSUID..IBG.DEMO,DISP=SHR                
//SORTOUT  DD   DSN=&SYSUID..IBG.DEMO,DISP=SHR                
//SYSIN    DD   *                                             
  SORT FIELDS=(1,8,CH,A)