AZ-900 Describe Azure architecture and services Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Azure Stream Analytics

  Why it's wrong here

  Azure Stream Analytics is purpose-built for real-time, in-memory processing of streaming data from sources like Azure Event Hubs and IoT Hub. It is not an ETL/ELT service for batch data movement because it does not offer connectors for moving static data between 90+ stores, nor does it support scheduled pipeline orchestration. Its job is to run continuous queries against live data streams and emit results to outputs, which is fundamentally different from the batch-oriented orchestration that Data Factory provides.

• ✓

  Azure Data Factory

  Why this is correct

  Azure Data Factory is the canonical ETL/ELT orchestration service in Azure. It provides over 90 native connectors to on-premises and cloud data sources, enabling you to design, schedule, and monitor data movement and transformation pipelines visually or via code. Data Factory supports both code-free mapping data flows and hand-written activities that can invoke compute engines like Azure Databricks and HDInsight, making it the dedicated service for orchestrating end-to-end data integration.

• ✗

  Azure Databricks

  Why it's wrong here

  Azure Databricks is an Apache Spark-based analytics platform optimized for big data engineering and machine learning. While it absolutely can perform ETL transformations within its scalable notebooks and clusters, it is fundamentally a compute and processing engine rather than a data orchestration service; it lacks the prebuilt connectors to 90+ sources and the pipeline scheduling/control-flow capabilities of Data Factory. In practice, Data Factory often orchestrates Databricks jobs, not the other way around.

• ✗

  Azure HDInsight

  Why it's wrong here

  Azure HDInsight is a managed cloud service that deploys open-source big data frameworks such as Hadoop, Spark, Hive, and Kafka on dedicated clusters. It offers the raw compute power to execute large-scale batch transformations, but it does not provide a native, productized ETL/ELT pipeline service with built-in connectors across diverse data stores. You would use HDInsight as a compute target within a Data Factory pipeline to run Hive or Spark jobs, not as the service that moves and orchestrates data between sources and destinations.