Develop a Stream Processing Solution DP-203 Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   In Spark Structured Streaming, what is the primary purpose of checkpoints?

   • ATo store the state and progress of the streaming query for fault tolerance
   • BTo improve the performance of the query
   • CTo validate the schema of incoming data
   • DTo encrypt the output data
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   When optimizing a data pipeline for analytical purposes, what is a common strategy?

   • AUsing row-based storage formats like JSON
   • BUsing columnar storage formats like Parquet
   • CAvoiding indexing
   • DStoring data in a single large file
   Reveal Solution Discussion Select an answer first
4. 23foundation · easy

   Which feature in Azure Stream Analytics allows you to handle schema drift by automatically adding new fields to the output?

   • AEvent ordering
   • BPartitioning
   • CCheckpointing
   • DSchema drift handling
   Reveal Solution Discussion Select an answer first
5. 24foundation · easy

   In Azure Stream Analytics, what happens when a job is stopped and then restarted?

   • AThe job resumes from the last checkpoint if output is configured with exactly-once semantics
   • BThe job automatically scales up to handle the backlog
   • CThe job loses all state and must be reconfigured
   • DThe job always starts from the beginning of the input stream
   Reveal Solution Discussion Select an answer first
6. 25application · medium

   A telemetry company ingests IoT sensor data into an Event Hubs namespace with 32 partitions. They need to compute a rolling 5-minute average of temperature readings per device and output results to a Power BI dashboard. During peak hours, the ingestion rate triples, causing the Stream Analytics job to fall behind. The company wants to maintain low latency for the dashboard while minimizing cost. What should they do?

   • AIncrease the number of Stream Analytics streaming units (SUs) and enable autoscaling on the job.
   • BChange the query to use a hopping window instead of a tumbling window.
   • CAdd a second Event Hubs namespace and split the incoming traffic across both namespaces.
   • DIncrease the partition count of the existing Event Hubs namespace to 64.
   Reveal Solution Discussion Select an answer first