AWS BluAge L3 Workshop - Execution - Batch Testing - COMBTRAN(SORT) - Issue

0

Batch Testing jobs flow: TRANBKP -> POSTTRAN -> INTCALC -> TRANBKP(2) -> COMBTRAN (Issue here)-> CREASTMT.

POSTTRAN - Written 262 transaction records INTCALC - Interest calculation - Systemtran written 50 records

Job - COMBTRAN

//STEP05R EXEC PGM=SORT
//SORTIN DD DISP=SHR,
// DSN=AWS.M2.CARDDEMO.TRANSACT.BKUP(0)
// DD DISP=SHR, // DSN=AWS.M2.CARDDEMO.SYSTRAN(0) //SYSIN DD *
SORT FIELDS=(1,16,CH,A)
/*
//SYSOUT DD SYSOUT=*
//SORTOUT DD DISP=(NEW,CATLG,DELETE),
// UNIT=SYSDA,
// DCB=(*.SORTIN),
// SPACE=(CYL,(1,1),RLSE),
// DSN=AWS.M2.CARDDEMO.TRANSACT.COMBINED(+1)

I am getting SORT output with only 263 records instead of 312 (262 + 50) Out of 263 records, all 262 records from AWS.M2.CARDDEMO.TRANSACT.BKUP(0) but only ONE record from AWS.M2.CARDDEMO.SYSTRAN(0) instead of 50.

Code changes in COMBTRAN job is, Adding "recordSize=350". Verified groovy script and matching with document.

Any reason to get only one record from Systran(0) file?

asked a month ago99 views
1 Answer
0

The most likely reason you are only getting one record from the Systran(0) file instead of 50 is that the sort key value for those records may not be unique.

When SORT operation is performed, it orders the records lexicographically based on the sort key value. If multiple records have the same sort key value, only one of those duplicates will be returned.

A few things you can check:

Verify that the sort key field contains a unique value for each record in the Systran(0) file. If any records have the same sort key, only one of the duplicates will be returned.

Check if the sort key type is appropriate - dates/numbers stored as strings may not sort correctly. The sort key should be of a type that provides unique ordering like a number or date.

Try running the SORT without the "recordSize=350" parameter to see if it returns more records from Systran(0) file. That parameter could be influencing the number of records returned.

Check AWS documentation for latest limits or behaviors of SORT operation that could impact the number of returned records.

profile picture
EXPERT
answered a month ago
  • Verified SYSTRAN(0) records and all 50 are unique transaction ids yyyymmdd000000xx where xx is number 01 to 50. So duplicate is not an issue. However I think Normal JCL SORT many not remove duplicates without explicitly specifying SUM FILEDS= NONE. On recordSize parameter. You mentioned remove but as per bluAge workshop documents, explicityly mentioned to include recordSize. Prior to this jobs are having recordSize parameters. So any other possibilities to get only one record in SORTOUT?

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions