IBM Z and LinuxONE Software

Software

Software

In order to compete you have to be trusted. System software delivers high-performing system and database management, faster insights, secure cloud and agile DevOps, plus the reliability, flexibility and protection you expect from the mainframe.

 View Only

PSF For z/OS AFRREBLK Error

  • 1.  PSF For z/OS AFRREBLK Error

    Posted Mon March 31, 2025 11:13 AM

    Hi,

    The re-block program is generating the following message:

    Here's the first 3 bytes of the AFP input record = 5A4040
    The 1st byte is the AFP record identifier        = 5A
    The 2nd and 3rd bytes is the length of the
     data that follows the AFP record identifier     =   4040
                             In decimal this value is:  16448
    And don't forget the 4-byte RDW that must
    be included in the AFP record length.

        Let's calculate the TOTAL INPUT RECORD LENGTH.....
                            RDW (Record Descriptor Word)   :     4
                            AFP Input Record               : 16449
                            GRAND TOTAL INPUT RECORD LENGTH: 16453

    MAXIMUM RECORD LENGTH ALLOCATED FOR THE OUTPUT DATA SET: 8209

    There's your problem. The output record length is less than
                          the length of the AFP record.


    Total length of the AFP input record            = 16453
    Maximum length allowed for the output data set  = 8209

    Input DSN = TSTBT.XXXX.WORK.CCL1
    AFP record length = 16453 (TSTBT.XXXX.WORK.CCL1)

    Output DSN = TSTBT.XXXX.WORK.CCL1.RBLK
    LRECL   = 8209 (TSTBT.XXXX.WORK.CCL1.RBLK)
    BLKSIZE = 32760 (TSTBT.XXXX.WORK.CCL1.RBLK)

    Profile DSN = X824254.AFRREBLK.PROFILE
    Output LRECL   = 32756 (X824254.AFRREBLK.PROFILE)
    Output BLKSIZE = 32760 (X824254.AFRREBLK.PROFILE)


    For more information about the AFRREBLK utility,
    look in the 'PSF for z/OS User's Guide'
    and view Chapter: 'Using the AFP Reblocking Program'

    The invalid data was identified using the last records generated record on the output file but was still a guess.  The information provided in the output did not seem to help.  This was a production problem and was overcome by re-generating the input file from the source on a Windows platform - it took almost half a day.

    Is there any method to identify which are the offending records and remove them?

    Thank you.



    ------------------------------
    Girish Sajja
    ------------------------------