Batch Processing in Analysis Services

In Microsoft SQL Server Analysis Services, you can process objects in batches. Using batch processing you can select the objects to be processed and control the processing order. Also, a batch can run as a series of stand-alone jobs or as a transaction in which the failure of one process causes a rollback of the complete batch.

Using batch processing, you can control how processing affects your online data. For example, when you fully process a dimension, any partition using that dimension is marked as unprocessed. As a result, cubes that contain the unprocessed partitions are unavailable for browsing. You can address this with a batch processing job by processing the dimensions together with the affected partitions. Running the batch processing job as a transaction makes sure that all objects included in the job remain available for queries until all processing is completed. As the transaction commits the changes, locks are put on the affected objects, and this makes the objects temporarily unavailable. For more information about processing options and settings, see Processing Options and Settings.

You can execute batch processing using Object Explorer in SQL Server Management Studio or Solution Explorer in Business Intelligence Development Studio. Also, you can run an XMLA script using the XMLA Query window in SQL Server Management Studio or as a scheduled task. For more information, see How to: Batch Process Analysis Services Objects and Scheduling Administrative Tasks Using SQL Server Agent.