User Guide : Best Practices : Message Pattern Integrations : RDBMS Considerations
 
Share this page             
RDBMS Considerations
Application integrations generally involve API endpoints. However, you will often have the need to integrate with relational database systems. See Bulk or Batch Pattern Integrations as needed. Also, initial data loads and subsequent data refresh can often involve large batch sizes.
The following points are useful when integrating applications to or from a relational database:
Transformation Maps:
When an application is the source, use timestamps as part of the query to limit data sizes to the most current changed data
For an initial load or complete refresh of the database, consider the source objects as a large file and use a query to split the dataset into multiple smaller batch sizes.
When the database is the source, consider the max batch size of the application and write a query to conform to it.
Processes:
Most database servers can be configured to send changed data to a queue. Use the appropriate Queue component in Process editor to pass data to/from a Transformation Map (that is, Oracle AQ, Microsoft MQ, WebSphere MQ, and so on.)
Use the Studio API invoker step with any Web Service API and pass it to/from a Transformation Map to leverage one of the RDBMS connectors through a djMessage object variable.