Overview
The Spark Connector supports streaming mode, which uses Spark Structured Streaming to process data as soon as it's available instead of waiting for a time interval to pass. Spark Structured Streaming is a data-stream-processing engine that you can access by using the Dataset or DataFrame API.
Important
Apache Spark contains two different stream-processing engines:
Spark Streaming with DStreams, now an unsupported legacy engine
This guide pertains only to Spark Structured Streaming.
The following sections show you how to use the Spark Connector to read data from MongoDB and write data to MongoDB in streaming mode:
Tip
Apache Spark Documentation
To learn more about using Spark to process streams of data, see the Spark Programming Guide.