[Task]: Add UseDataStreamForBatch option to the Flink runner. #25740
Description
What needs to happen?
This is the second task for migrating Flink batch job execution from DataSet to DataStream API. We will add a new option of UseDataStreamForBatch to the Flink runner. When it is set true, the Flink runner will use DataStream to execute the batch jobs, otherwise the existing DataSet API will be used.
Issue Priority
Priority: 2 (default / most normal work should be filed as P2)
Issue Components
- Component: Python SDK
- Component: Java SDK
- Component: Go SDK
- Component: Typescript SDK
- Component: IO connector
- Component: Beam examples
- Component: Beam playground
- Component: Beam katas
- Component: Website
- Component: Spark Runner
- Component: Flink Runner
- Component: Samza Runner
- Component: Twister2 Runner
- Component: Hazelcast Jet Runner
- Component: Google Cloud Dataflow Runner