An ODBC-based load provides an extensive option for acquiring data from many sources. It will be slower than most other forms of loads but may still be a viable option depending on the volume of data being loaded.
An ODBC-based 'interactive load' when using the WhereScape RED tool will use the established ODBC connection to pull the data back to the local PC and then push the data to the data warehouse database via a sql ODBC statement.
A scheduler load will perform in the same way except that the data is loaded into the server that is running the scheduler and then pushed to the data warehouse database.
This is normally achieved via a 'During' load transformation such as TO_CHAR(source_date_column,'YYYY-MM-DD HH24:MI:SS')
The obvious disadvantage in an ODBC-based load is the two network transactions required and the overhead placed on the Scheduler server. The UNIX scheduler does not support ODBC-based loads, therefore a Windows scheduler must be available to handle ODBC-based loads.
A Windows scheduler must be available to handle ODBC-based loads.
The Properties screens for an ODBC-based load are the same as those of a database link load. Refer to the previous sections for details.