Performance of Import Process
To optimize the performance, the activities used in import are processed asynchronously. Data is inserted in the database in batches. For each activity in the workflow, the RecordperAsyncCall and isAsyncProcessing variables are defined. If isAsyncProcessing is set to true, asynchronous processing takes place.
In Asynchronous processing, the Import process populates a queue with batches of records. The batch size, or the number of records held by a batch, will be as defined in the RecordperAsyncCall variable.You can also define the pool size by using the Async_PoolSize variable. The pool size is the number of batches that can be processed at a time. For example, if Async_PoolSize is set to 2, a machine can process a maximum of two batches at a time. For more information on these variables, refer to TIBCO MDM Workflow Reference.
Machines involved in the import process can work on the records in each of these batches without waiting for another machine to finish work on a batch. Once they complete working on the batches, they pick up more batches from the queue.
For performance improvements during import, UpdateRecordState should be run using Event mode. This can be done by not passing a record list to this activity.