An OpenEdge database by default defines a sequence field in the schema as an INT64 data type. As a result, sequences migrated from a new OpenEdge database to a target Oracle DataServer automatically expand to support 64-bit data types.

The INT64 data type enables an Oracle DataServer to maintain OpenEdge compatibility with and provide 64-bit numeric capability for sequences when they are used to generate ROWIDS.In the Oracle DataServer, ROWIDS are supported by the 64-bitnative NUMBER data type. For details about NUMBER data type support, see Data types.

When generating sequences, remember that:

  • Sequences migrated from a new OpenEdge database expand to INT64.
  • Sequence functions return an INT64 data type, using the NUMBER data type.
  • An overflow condition can occur and subsequently an error message can appear on a client machine when a sequence generator is defined as a 64-bit value, but the value into which it is to be loaded is defined as a 32-bit data type (OpenEdge Release 10.1A or earlier).
  • When an Oracle database is the target database, and the sequence's upper limit is defined as the Unknown value (?) in the source database, the migration utility will specify the sequence's upper limit value as 922337203685775807. However, Oracle continues to generate sequence definitions with the NOMAXVALUE clause when the value for the upper limit is defined as the Unknown value (?) in the delta.df file. The value generated through this clause is the real upper limit of the sequence value in the OpenEdge client as distinguished from the upper limit defined through the INT64 data type.