Which of the following statements describes streaming with Spark as a model deployment strategy?

Question

Options

• AThe inference of batch processed records as soon as a trigger is hit
• BThe inference of all types of records in real-time
• CThe inference of batch processed records as soon as a Spark job is run
• DThe inference of incrementally processed records as soon as trigger is hit
• EThe inference of incrementally processed records as soon as a Spark job is run

• A
  2% (1)
• B
  6% (3)
• C
  10% (5)
• D
  2% (1)
• E
  80% (39)

Explanation

E correctly captures the two defining characteristics of Spark Structured Streaming as a deployment strategy: records are processed incrementally (micro-batches of new data since the last checkpoint, not a full dataset), and inference is triggered when the Spark job is run (the streaming application itself is submitted and executed as a job).

Why the distractors fail:

• A & C both say "batch processed records," which incorrectly implies the full dataset is reprocessed each time - that describes batch inference, not streaming.
• B is vague and technically inaccurate; it omits the incremental and job-execution mechanics entirely.
• D is the trickiest distractor - "incrementally processed" is correct, but "as soon as a trigger is hit" confuses the internal Spark Structured Streaming trigger (a micro-batch scheduling concept inside the job) with what initiates the deployment strategy. From a deployment perspective, it's the Spark job being run that kicks off inference.

Memory tip: Lock in the two keywords - Incremental (not full batch) + Spark Job (what you deploy and run). If a choice says "batch processed," eliminate it. If it says "trigger" without "Spark job," it's mixing internal streaming mechanics with deployment-level concepts.

No community discussion yet for this question.