Mainframe, MVS and zOS Discussion

Hi,

There was an abend in one of the Jobs in production yesterday. Upon investigation, it was found that there was a data field in the file which came "empty" in the last record of the file. When the record was removed the job ran successfully.

To this one of our senior guy said, the record was not corrupt, instead the file missed its 'end-of-marker' and so failed. As this file comes from outside to mainframe, it should be reallocated and data should be cut pasted rather than running the abended Job by removing the last record. We did not hear about this as the job executed successfully after removing the record. But today we got to know that "empty date" is a usual condition and had happened in past many time and it did not abend the job.

Question is, what is "end of file marker" for a sequential dataset in mainframes?

What medium are we talking about?

Akatsukami wrote: ↑Thu Dec 13, 2018 11:27 pmWhat medium are we talking about?

File is on mainframe. Sent from a windows server, using FTP or MFT.

MEDIUM - Tape, floppy disk, DASD, paper...

To this one of our senior guy said, the record was not corrupt, instead the file missed its 'end-of-marker' and so failed.
.....
We did not hear about this as the job executed successfully after removing the record.

THE RECORD WAS CORRUPTED

a missing - as Your idiot senior says - End of File marker.
( whatever is the meaning )
would have caused a completely different error

and, if You were able to delete the offending record
the file structure was not damaged