Describe an Analytics Workload on Azure (DP-900) Questions

Question 1

A team is new to Azure Databricks and wants to understand how data is stored when they create Delta Lake tables in their workspace. They notice that queries run faster on Delta tables compared to standard Parquet files. What is the primary storage format that Delta Lake uses underneath to achieve this performance?

Delta Lake stores data as Avro files combined with a metadata catalog managed by the Databricks Runtime Delta Lake stores data as optimized CSV files with an indexing mechanism built into Azure Blob Storage Delta Lake stores data as ORC files with a caching layer that integrates with the Azure Data Lake filesystem Delta Lake stores data as Parquet files with an additional transaction log layer on top

Correct Answer: Delta Lake stores data as Parquet files with an additional transaction log layer on top

Delta Lake is built on top of Apache Parquet as its underlying storage format, augmented by a transaction log (also known as the Delta Log) that enables ACID transactions, schema enforcement, time travel, and optimized query performance.

CSV files lack the columnar storage and compression benefits needed for high-performance analytics. Avro is a row-based format primarily used for streaming and serialization, not as the foundation for Delta Lake. ORC is a columnar format used mainly in the Hadoop ecosystem and is not the format underlying Delta Lake.

The key concept tested is understanding Delta Lake's architecture, specifically that it uses Parquet files combined with a transaction log to deliver reliability and performance.

Question 2

Which operational capability provides the strongest justification for selecting Azure Stream Analytics in a modern architecture?

The capability to execute deep machine learning model training over massive historical event datasets. The capability to orchestrate routine daily batch transformations across diverse hybrid cloud integration platforms. The capability to process continuous event feeds and compute metrics across sequential temporal windows. The capability to continuously archive structured relational records into multiple hierarchical warehouse configurations.

Correct Answer: The capability to process continuous event feeds and compute metrics across sequential temporal windows.

Azure Stream Analytics is specifically designed to ingest real-time data feeds and process information across continuous time-based windows.
Training large machine learning models requires compute services like Azure Machine Learning rather than streaming engines. Orchestrating scheduled batch transformations is a task suited for Azure Data Factory. Archiving structured relational records into data warehouses is handled by standard ETL pipelines, not real-time event processors.
The key concept tested is the core use case of Azure Stream Analytics for real-time temporal window processing.

Question 3

What type of processing model does Azure Stream Analytics use to handle incoming data?

Real-time stream processing that analyzes data continuously as it arrives On-demand processing that triggers analysis only when a user submits a manual query request Micro-batch processing that groups events into small periodic batches for analysis Batch processing that collects data into scheduled intervals before analysis

Correct Answer: Real-time stream processing that analyzes data continuously as it arrives

Azure Stream Analytics is specifically designed as a real-time stream processing engine that continuously analyzes and processes data as it arrives from sources like IoT devices, event hubs, and other streaming inputs.

Batch processing collects and stores data before processing it at scheduled intervals, which is the opposite of stream processing. Micro-batch processing, used by services like Spark Structured Streaming, buffers events into small batches rather than processing them truly continuously. On-demand processing requires manual user intervention, whereas Stream Analytics runs continuously and automatically.

The key concept tested is understanding Azure Stream Analytics as a real-time stream processing service within the Azure data platform.

Describe an Analytics Workload on Azure

DP-900 - Describe an Analytics Workload on Azure Example Questions

Question 1

Question 2

Question 3

Master Data Fundamentals on Azure