Previous Topic: Parallel ExtractNext Topic: Restrictions


FIRSTKEY= and LASTKEY= Keywords in Parallel Extract

It is important that the set of parallel extracts not miss any data rows, and it is your responsibility to edit the key values before executing the set of extract jobs. We strongly recommend that each set of extract jobs be executed through the DBUTLTY edit function and the output reviewed before execution. To do the edit, add the parameter EDIT as the first DBUTLTY function followed by your parallel extract control statements. Execute DBUTLTY and review the output to make certain that you are requesting what you expect.

If the parallel extract is being used to delete from the start or end of the key value range, the extract can accomplish this with one job step. A load of the output extract file would provide a data area loaded without the dropped rows.

When DBUTLTY echoes back the input parameters for the FIRSTKEY= and LASTKEY= keywords, it prints them as a three line set with the first line being printable characters, the second being the zone portion of the bytes, the third being the digit portion of the bytes. Following is an example:

 EXTRACT DBID=111,AREA=SAM,DDNAME=EXTRACT,SEQ=NATIVE
  FIRSTKEY=0512345,LASTKEY=0299

 FUNCTION=EXTRACT
    AREA=SAM
    DBID=00111
    DDNAME=EXTRACT
    SEQ=NATIVE
    FIRSTKEY=0512345
    . . .  .=FFFFFFF
    . . .  .=0512345
    LASTKEY=0299
    . .  . =FFFF
    . .  . =0299

The following is an example of parallel extract using three output datasets:

    EXTRACT DBID=790,AREA=A01,DDNAME=EXTRACT1,SEQ=NATIVE,
       FIRSTKEY=00,LASTKEY=043333

    EXTRACT DBID=790,AREA=A01,DDNAME=EXTRACT2,SEQ=NATIVE,
       FIRSTKEY=043334,LASTKEY=046666

    EXTRACT DBID=790,AREA=A01,DDNAME=EXTRACT3,SEQ=NATIVE,
       FIRSTKEY=046667,LASTKEY=00