The Bulk Processing framework provides the ability for data stewards to invoke a bulk set of InfoSphere® MDM Server transactions to process entities that meet the desired criteria.
Te Bulk Processing framework is an extension to the Batch Controller framework in which the registered bulk processes are picked up and, based on user criteria, can invoke a bulk set of transactions. These bulk processes are also managed using the InfoSphere MDM Server Task Management feature.
The following diagram shows the architecture of the Bulk Processing framework:
The Bulk Processing Controller performs the following tasks:
- Reads the new bulk processes from task management tables.
- Updates the status of the bulk processing tasks.
- Looks up the entities that satisfy the bulk processing criteria.
- Constructs an InfoSphere MDM Server request XML for each searched entity.
- Submits the request XML for InfoSphere MDM Serverfor each entity.
- Receives the response XML from InfoSphere MDM Serverfor each entity.
- Updates the bulk process task status for each entity.
The Bulk Processing Controller is essentially an extension to the
standard Batch Controller. The following Bulk Controller reader and
writer queue classes must be configured:
- TaskBasedEnitiySearcherQueue – Overrides the out of the box FileReaderQueue class. This class reads the evergreen task details, searches the entities based on the provided criteria, and constructs as many InfoSphere MDM Server transaction requests as required to complete the entire evergreen process. The constructed transaction requests will be later submitted to InfoSphere MDM Server by the batch controller’s submitter.
- TaskBasedWriterQueue – Overrides the out of the box FileWriterQueue class. This class logs the returned response from InfoSphere MDM Server and refreshes the task's progress status repeatedly after a configured period of time.
