This topic provides a high level architecture view of the Asynchronous Processing framework.
The Asynchronous Processing framework logic is associated with a Java™ context object, AsynchronoousWorkBObj. This context object identifies transient data objects and the asynchronous logic that each data object needs to run.
To initiate asynchronous processing, the asynchronous work producer instantiates and sends the required asynchronous work instance to the messaging middleware. It then can continue the rest of the required actions without having to wait for the outcome of the asynchronous work.
AsynchronousWorkMDB is a message driven bean that receives and processes the asynchronous work messages. The following figure shows a high level communication diagram for this message driven bean.
The Asynchronous Processing framework enables InfoSphere® MDM Server to complete any required asynchronous actions, as coded and customized by developers. The framework also supports invocation retries if an exception is thrown by the AsynchronousWorkProcessor.execute() method.
You can configure the retry behavior with the following elements in the Configuration and Management component:
- /IBM/DWLCommonServices/AsynchronousWork/maxRetryLimit
- The maximum number of retries when an error occurs. The default value is 3.
- /IBM/DWLCommonServices/AsynchronousWork/RollbackMessageOnFailure/enabled
- Determines whether the AsynchronousWorkMDB:
- Returns erroneous messages to the messaging middleware after reaching the maximum retry limit (true); or
- Treats failed messages as consumed, and does not return them to the messaging middleware (false).
Tip: This configuration item is useful if you want the messaging middleware to handle failed messages and retries. For example, you can use this configuration item to forward failed messages to a dead queue so that you can investigate them later.
