Ultimate access to all questions.
You have a streaming pipeline job that you wish to cancel, but you want to ensure all in-flight data is processed and written to the output. Which command in the Dataflow monitoring console should you use to achieve this?
Explanation:
Choosing the Drain option to stop your job instructs the Dataflow service to complete the job in its current state. This means the job will immediately cease accepting new data from input sources, but the Dataflow service will retain existing resources (like worker instances) to finish processing and writing any buffered data in your pipeline. For more details, refer to the Dataflow documentation.