Has anyone tried to migrate a file from Adabas in Windows to a mainframe Adabas database. We frequently migrate in the other direction but I cannot get a migration up to a mainframe to work.
I have tried all of the available options in the Decompress and then tried using and not using the “cvt-fmt” program before the FTP.
The record gets up to the mainframe in a readable format but the record is not blocked. I have tried FB and VB format. If defined as variable block the record has the 2 byte variable RDW count rather than the normal 4 bytes on a mainframe record.
These records are variable length because of the MUs and PEs in the record.
SAG has advised to use the RDW option in the decompress and NOT to use the "cvt-fmt’ conversion however this does not work.
By “not blocked” I meant that the record gets written as though the blocking were being ignored - the second record follows on immediately after the first record and then wraps around at the end of the physical blocksize. It is the same whether it is FB or VB. So if the blocking factor is 2000 and the first record is 1800 you get the first record and 200 bytes of the second record on the first line.
To answer my own question, we finally worked out how to migrate data to the mainframe.
The only way to force the records to block correctly is to define the maximum occurrences for each of the MUs and PEs. Using the RDW parameter in the Decompress did not work because it includes 2 length bytes at the beginning of each record - we used the ASCII parameter. The fixed blocking factor is then set at the maximum record length (including max. MUs and PEs).