You are designing a real-time system for a ride hailing app that identifies areas with high demand for rides to effectively reroute available drivers to meet the demand. The system ingests data from…

Question

Options

• AGroup the data by using a tumbling window in a Dataflow pipeline, and write the aggregated data to Memorystore.
• BGroup the data by using a hopping window in a Dataflow pipeline, and write the aggregated data to Memorystore.
• CGroup the data by using a session window in a Dataflow pipeline, and write the aggregated data to BigQuery.
• DGroup the data by using a hopping window in a Dataflow pipeline, and write the aggregated data to BigQuery.

• A
  9% (2)
• B
  77% (17)
• C
  5% (1)
• D
  9% (2)

Explanation

Option B is correct because the requirement calls for a 30-second aggregation window recalculated every 2 seconds - this is the definition of a hopping (sliding) window: a fixed-size window that advances by a smaller hop interval, producing overlapping results. Memorystore (managed Redis) is the right sink because the requirement explicitly demands low-latency storage for real-time dashboard visualization, which is Redis's core strength.

A is wrong because a tumbling window is non-overlapping and fixed - a 30-second tumbling window would only produce output every 30 seconds, not every 2 seconds as required.

C is wrong on two counts: session windows are activity-gap-based (used for modeling user sessions, not fixed-time aggregations), and BigQuery is a data warehouse optimized for analytical queries, not sub-second dashboard reads.

D is wrong because while the hopping window is correct, BigQuery's ingestion and query latency makes it unsuitable for the low-latency visualization requirement - Memorystore is the appropriate low-latency store here.

Memory tip: Think of it as "hop = overlap." When the output frequency (every 2s) is shorter than the window size (30s), you need a hopping window. And whenever you see "low-latency" + "real-time dashboard" on GCP, that's a signal pointing to Memorystore, not BigQuery.

No community discussion yet for this question.