TRAILCHARSETEBCDIC to cause character data to be written to the trail file in the local EBCDIC code page of the DB2 subsystem from which data is to be captured.
Specification of this parameter causes all character data to be written to the trail file in the EBCDIC code page of the job in which Extract is running.
Specification of this parameter on a single-byte DB2 z/OS subsystem causes character data from non-Unicode tables to be written to the trail file in the installed EBCDIC single-byte CCSID. Data from ASCII tables is converted to this EBCDIC CCSID.
Specification of this parameter on a multi-byte DB2 z/OS subsystem causes Extract to process only EBCDIC and Unicode tables. Extract abends with an error if it encounters ASCII tables. Data from EBCDIC tables is written to the trail file in the installed EBCDIC mixed CCSID.
TRAILCHARSETEBCDIC is required if the target is a multi-byte system. To replicate both ASCII and EBCDIC tables to a multi-byte DB2 z/OS target, process each character set with a different Oracle GoldenGate processing stream: one Extract, data pump, and Replicat for the ASCII tables and another Extract, data pump, and Replicat for the EBCDIC tables.