Configuring Datastores for Oracle

Use these steps to configure datastore databases in Oracle.
EBCDIC data: By default, Oracle does not support individual table columns to be defined with a specific collation sequence. This means that any alphanumeric key fields defined for EBCDIC files cannot be explicitly collated using an EBCDIC collating sequence. Instead, MFDBFH defines these fields as binary data, resulting in binary collation be used. For the most part, binary collation produces the expected sequence order for EBCDIC keys.

However, if true EBCDIC collation is required, the Oracle database hosting the MFDBFH datastore must have been created with, or upgraded specifying, the MAX_STRING_SIZE configuration parameter set to a value of EXTENDED. By default, Oracle databases are created with MAX_STRING_SIZE=STANDARD. When MAX_STRING_SIZE=EXTENDED is set for the database, MFDBFH will define each alphanumeric key column of an EBCDIC file to use EBCDIC collation. For MFDBFH to recognise a key as being alphanumeric, the COBOL program creating the file must either have been compiled using the IXNUMKEY directive, or key types for the file must have been pre-defined using the dbfhadmin -keytypes -add command.

  1. Create Oracle Datastores Using Script Files
    Create database script files that can be run ahead-of-time to create Oracle datastores.
  2. Update the Configuration File for Oracle Datastores
    Add the database entries and connection details for your datastore databases in the MFDBFH configuration file.
  3. Upload Files to an Oracle Datastore
    Use the dbfhdeploy command line utility to upload data files to the datastores.
  4. Create an XA Resource
    If you are handling transactional data, you are required to create an XA resource for the datastore that houses the transactional data files.
Parent topic: Using Oracle