You use a Google Dataproc cluster to run mappings on the Spark engine.

You can increase the performance of the mapping by running the mapping in Optimized Spark mode. When you use the Optimized Spark mode to read data, you can specify the number of partitions to use. You can specify whether to run the mapping in

Generic

or

Optimized

mode in the advanced read and write operation properties. Optimized Spark mode increases the mapping performance.

You can configure a SQL override to override the default SQL query used to extract data from the Google BigQuery source.

You can read or write data of the NUMERIC data type to Google BigQuery. The NUMERIC data type is an exact numeric value with 38 digits of precision and 9 decimal digits of scale. When you read or write the NUMERIC data type, the Data Integration Service maps the NUMERIC data type to the Decimal transformation data type and the allowed precision is up to 38 and scale upto 9.