Problem

Upgrading CFT from 3.1 to 3.2

Issue when an LRECL=80 dataset is concatenated with an LRECL=40 dataset

//CFTIN DD DSN=CFT.FILE.CMD(CFTCMD1),DISP=SHR
// DD DSN=CFT.FILE.REPPORT,
// DISP=(OLD,CATLG,CATLG)

Error: Open Error (Error no 21)

EDC5021I

Was working just fine in previous versions.

Resolution

It is possible to read files whose record length is < 80.

The issue is when PDS or PDS/E are concatenated.

Dataset concatenate with CFT 3.2.4 and higher:

• In case of PDSs concatenation, as specified in the IBM doc "the data set with the largest block size must appear first in the concatenation."
• It is not possible to concatenate PDSs from a different record length with RECFM=FB

Dataset concatenate with CFT 3.1.3 and below:

• It was possible to concatenate PDS(s) with the one with the smaller blksize in first position.
• It is possible to concatenate PDSs with different record length with RECFM=FB

Conclusion:

The new restrictions appeared since the use of the file opening routines in language "C" instead of ASM.

Transfer CFT documentation will be updated accordingly from 3.3.2 SP4.

IBM documentations:

https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.1.0/com.ibm.zos.v2r1.ieab600/iea3b6_Block_sizes_for_concatenated_data_sets.htm

https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.3.0/com.ibm.zos.v2r3.cbcpx01/concat.htm

The CFTIN reading has been enhanced from version 3.2 to support reading under USS, concatenation files with DD* (which was not possible before)