Launch pricing — Azure certification prep for €19,95 €29,95

← All DP-900 topicsDP-900 · Analytics Workloads on Azure

DP-900 Real-Time and Streaming Analytics Practice Questions

Streaming questions start with the difference between batch and stream processing, and architectures that combine both such as lambda and kappa. You should know where events come from (Azure Event Hubs, IoT Hub) and how they are processed.

For processing, know Azure Stream Analytics and its windowing functions (tumbling, hopping, sliding), Spark Structured Streaming, and Microsoft Fabric Real-Time Intelligence with eventstreams and KQL.

Below are 5 free questions on this topic. Try each one before you open the answer.

Question 1

An architecture uses a batch processing path for historical analysis and a separate stream processing path for real-time insights, both feeding results into a serving layer. What is this architectural pattern called?

  • A.Kappa architecture
  • B.Lambda architecture
  • C.Medallion architecture
  • D.Hub-and-spoke architecture
chevron_rightShow answer and explanation

Correct answer: B. Lambda architecture

Lambda architecture combines a batch layer (processes all historical data for accuracy) and a speed layer (processes incoming streaming data for low latency) whose results are merged in a serving layer. This allows both accurate historical queries and near-real-time views. Kappa architecture simplifies this by using only a streaming layer for both historical reprocessing and real-time data, eliminating the separate batch path.

Question 2

A company needs to analyze a stream of IoT sensor events in real time and trigger an alert when the average temperature across all sensors exceeds a threshold over a 5-minute window. Which Azure service is best suited for this?

  • A.Azure Data Factory
  • B.Azure Synapse dedicated SQL pool
  • C.Azure Stream Analytics
  • D.Azure Table Storage
chevron_rightShow answer and explanation

Correct answer: C. Azure Stream Analytics

Azure Stream Analytics is a real-time event processing service that can ingest data from Azure Event Hubs, IoT Hub, or Blob Storage, apply SQL-like queries with windowing functions (tumbling, hopping, sliding windows), and output results to various sinks such as Power BI, Cosmos DB, or SQL Database. It is specifically designed for time-series stream processing scenarios like IoT monitoring and alerting.

Question 3

An IoT platform collects temperature sensor readings from 100,000 devices every second. A dashboard must show the average temperature per region in the last 5 minutes, updated every 30 seconds. The raw sensor data does not need to be stored long-term. Which Azure service is the best fit for computing these real-time aggregations?

  • A.Azure Data Factory with a tumbling window trigger
  • B.Azure Synapse Analytics dedicated SQL pool
  • C.Azure Stream Analytics
  • D.Azure Databricks batch job scheduled every 30 seconds
chevron_rightShow answer and explanation

Correct answer: C. Azure Stream Analytics

Azure Stream Analytics is purpose-built for real-time stream processing with SQL-like queries over sliding, tumbling, and hopping time windows. A tumbling window query can group sensor readings by region over the last 5 minutes and output aggregated results every 30 seconds to Power BI or another sink. It requires no long-term storage of raw data. Azure Data Factory is a batch ETL orchestrator — its tumbling window trigger has a minimum interval of minutes and is not designed for second-level streaming. Synapse dedicated SQL pools require data to be loaded before querying. Databricks can do streaming but is significantly more complex and expensive to set up for this simple aggregation scenario.

Question 4

Determine whether each statement about batch processing versus stream processing is correct.

  1. 1.Batch processing typically has higher latency than stream processing.
  2. 2.Stream processing can access the entire historical dataset at any point during execution.
  3. 3.Stream processing is well suited for simple aggregations such as rolling averages over time windows.
chevron_rightShow answer and explanation
  1. Statement 1: Yes
  2. Statement 2: No
  3. Statement 3: Yes

Statement 1 is correct: batch processing latency is measured in hours because data must be collected and processed in groups, whereas stream processing latency is typically seconds or milliseconds. Statement 2 is incorrect: stream processing typically only has access to the most recent data or data within a rolling time window — it does not have access to full historical datasets, which is a characteristic of batch processing. Statement 3 is correct: stream processing excels at simple real-time computations like rolling averages, counts over time windows, and threshold monitoring.

Question 5

Determine whether each statement about Spark Structured Streaming is correct.

  1. 1.Spark Structured Streaming uses a dataframe-based API to process perpetual streams of data.
  2. 2.Spark Structured Streaming can only be used in Azure Databricks and cannot run on Microsoft Fabric.
  3. 3.Delta Lake can be used as both a streaming source and a streaming sink in a Spark Structured Streaming solution.
chevron_rightShow answer and explanation
  1. Statement 1: Yes
  2. Statement 2: No
  3. Statement 3: Yes

Statement 1 is correct: Spark Structured Streaming extends Spark's dataframe API to handle unbounded, continuously arriving data as a 'boundless' dataframe processed with perpetual queries. Statement 2 is incorrect: Spark Structured Streaming runs on both Azure Databricks and Microsoft Fabric — both services include Spark runtimes that support Structured Streaming. Statement 3 is correct: Delta Lake tables can act as a streaming source (for downstream queries) or as a streaming sink (to receive and persist stream output), making it a versatile component in Spark-based streaming architectures.

These are 5 of 17 Real-Time and Streaming Analytics questions in the full bank

Practice all 400 DP-900 questions in exam format, including drag-and-drop and multi-select questions, with a mock exam and progress tracking.