Explanation:
The correct choice is a side input, as it serves as an additional input for DoFn, enabling the inclusion of extra data during processing. Partitions divide a collection's elements into multiple outputs, watermarks signal that no earlier timestamped data will arrive, and custom windows are defined by WindowFn for data-driven gaps. For more details, visit Google Cloud Architecture.
Ultimate access to all questions.
No comments yet.
In the context of a Cloud Dataflow workflow, you've created a DoFn function but realize the PCollection lacks all necessary data for a computation. Which Apache Beam construct would you employ to supply extra input for each element processed in the PCollection?
A
Custom window
B
Watermark
C
Side input
D
Partition