Watch this video on YouTube
As the developer of a new Cloud Dataflow pipeline, you'd like to limit the processing resources used when testing a new pipeline. What parameter would you specify when executing your new Cloud Dataflow job?