I see two OUTFIL statements, so the job would create two different files. Each should have all of the input records, but SORTOUT would have the first 10 bytes of each record and ADDOUT would have the first 60 bytes of each record.
Are you saying that only 50000 of the 200000 input records were written to the output data sets? Or are you saying something else? Please clarify.
Due to security reasons and complexity (using many procs) i can't paste the JCL.
But syslog looks like this -
SORTOUT : DELETED = 0, REPORT = 0, DATA = 150000
0 SORTOUT : TOTAL IN = 150000, TOTAL OUT = 150000
0 addOUT : DELETED = 0, REPORT = 0, DATA = 150000
0 addOUT : TOTAL IN = 150000, TOTAL OUT = 150000
0 dateOUT : DELETED = 0, REPORT = 0, DATA = 150000
0 dateOUT : TOTAL IN = 150000, TOTAL OUT = 150000
I wanted to see your JCL and all of the messages to determine if my guess about what's going on was reasonable. My guess is that you're running into the system restriction described in the second bullet here:
If that's not it, then I need to see the input DD statement or statements (if concatenated) and the //SORTOUT DD and //ADDOUT DD statements. I also need to see ALL of the messages with their ICExxxs numbers. If you want to send all of that to me offline (firstname.lastname@example.org) rather than posting them here, that's fine. Just put "DFSORT" somewhere in your Subject to catch my attention.
Your references to System 1 and System 3 don't really mean anything to anyone else on this list. Since you said the link I provided was helpful, I assume that the problem was in fact some variation of the system restriction described there and you were able to do something to avoid that system restriction.