DP-900 Describe core data concepts Practice Test 18 — 15 Questions

Question 1

Your organization has a data warehouse in Azure Synapse Analytics. You need to load data from Azure Blob Storage daily, transforming it using a data flow. Which Azure service should you use for the ETL process?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Azure Data Factory. Azure Data Factory (ADF) is the correct choice because it provides native integration with Azure Synapse Analytics and Azure Blob Storage, and it includes a visual data flow designer for transforming data without writing code. ADF's mapping data flows execute at scale on Spark clusters, making it ideal for daily ETL workloads that require both ingestion and transformation.

Answer

Azure Databricks

Answer

Azure Logic Apps

Answer

Azure Synapse Pipelines

Question 2

Which TWO Azure services are appropriate for real-time data ingestion from IoT devices?

Accepted Answer

Azure IoT Hub. Azure IoT Hub (A) is correct because it is a fully managed, cloud-hosted service designed specifically for bidirectional communication with IoT devices, supporting device-to-cloud telemetry ingestion over protocols like MQTT, AMQP, and HTTPS at scale. Azure Event Hubs (E) is correct because it is a big-data streaming platform and event ingestion service capable of receiving millions of events per second in real time, commonly used as the ingestion endpoint for IoT telemetry pipelines. Azure Data Factory (B) is a batch-oriented data integration and orchestration service, not a real-time ingestion endpoint for devices. Azure SQL Database (C) is a relational database for storing and querying structured data, not a device ingestion service. Azure Blob Storage (D) is object storage for persisting data, not a real-time ingestion mechanism for IoT devices.

Answer

Azure Data Factory

Answer

Azure SQL Database

Answer

Azure Blob Storage

Question 3

Which THREE are valid use cases for Azure Cosmos DB?

Accepted Answer

Storing IoT telemetry data with low latency. Azure Cosmos DB is a globally distributed, multi-model NoSQL database designed for low-latency, high-throughput workloads, making option A correct: storing IoT telemetry data benefits from its single-digit-millisecond reads/writes and elastic scale across regions. Option C is correct because session state management for web applications fits Cosmos DB's key-value/document model with a TTL (time-to-live) feature that automatically expires session data, plus low-latency access from any region. Option D is correct because personalization and recommendation engines rely on fast lookups of user profiles and interaction data, which Cosmos DB supports through its multi-model APIs and global distribution. Option B is not valid because relational OLTP with complex joins requires a relational engine with strong schema enforcement and JOIN semantics, which Cosmos DB does not provide as a core capability. Option E is not valid because storing large files like images and videos is better suited to Azure Blob Storage, since Cosmos DB is optimized for small, structured JSON documents rather than large binary objects.

Answer

Relational OLTP with complex joins

Answer

Storing large files like images and videos

Question 4

Refer to the exhibit. You are deploying an Azure Storage account. The JSON snippet represents a template parameter. What does the 'isHnsEnabled' property enable?

Accepted Answer

Hierarchical namespace for the storage account. The 'isHnsEnabled' property enables the hierarchical namespace for the storage account, which is a core feature of Azure Data Lake Storage Gen2. When set to true, it allows the storage account to organize blobs into a directory hierarchy, enabling POSIX-like access control lists (ACLs) and file system semantics. This is essential for big data analytics workloads that require a file system structure rather than a flat blob storage model.

Answer

Blob versioning

Answer

Soft delete for blobs

Answer

Geo-redundant storage

Question 5

Your company runs a global e-commerce platform that generates over 5 TB of clickstream data daily. The data is currently stored as raw CSV files in Azure Blob Storage. The data engineering team needs to transform this data into a star schema for business intelligence reporting. They want to use a serverless, code-first approach where they can write Python or SQL transformations. The transformed data should be stored in a format that optimizes query performance for Power BI. You also need to ensure that the solution can handle variable data volumes without manual scaling. Which Azure service should you use for the transformation?

Accepted Answer

Azure Databricks. Azure Databricks is the correct choice because it provides a serverless, code-first environment where data engineers can write Python or SQL transformations using Apache Spark. It can handle variable data volumes without manual scaling, and it can output transformed data in optimized formats like Parquet, which significantly improves query performance for Power BI. This aligns perfectly with the requirement for a serverless, code-first approach and star schema transformation.

Answer

Azure Stream Analytics

Answer

Azure Synapse Serverless SQL

Answer

Azure Data Factory

Question 6

A marketing team needs to analyze customer sentiment from social media posts in real time. The solution must ingest a stream of tweets, perform sentiment analysis using a pre-built AI model, and store the results in a dashboard for immediate visualization. The team has limited coding experience and prefers a low-code/no-code approach. Which combination of Azure services should you recommend?

Accepted Answer

Azure Event Hubs, Azure Stream Analytics, and Power BI. Azure Event Hubs ingests the real-time tweet stream, Azure Stream Analytics performs sentiment analysis using its built-in machine learning functions (a low-code/no-code approach), and Power BI provides the dashboard for immediate visualization. This combination meets the real-time, low-code requirement without custom coding.

Answer

Azure Event Hubs, Azure Functions, and Azure Cosmos DB

Answer

Azure IoT Hub, Azure Data Factory, and Power BI

Answer

Azure Event Hubs, Azure HDInsight, and Power BI

Question 7

Your company has a data lake in Azure Data Lake Storage Gen2 containing terabytes of parquet files. Data scientists need to explore and prepare this data using Python and SQL. They want to use a collaborative notebook environment that integrates with Git for version control. The solution should automatically scale compute resources based on workload demand and minimize management overhead. Which Azure service should you use?

Accepted Answer

Azure Databricks. Azure Databricks is the correct choice because it provides a collaborative notebook environment that natively supports Python and SQL, integrates with Git for version control, and offers auto-scaling clusters that dynamically adjust compute resources based on workload demand. It is purpose-built for big data analytics and data preparation on data lakes, minimizing management overhead through its serverless and managed Spark infrastructure.

Answer

Azure Machine Learning studio

Answer

Azure Data Studio

Answer

Azure Synapse Studio

Question 8

You are designing a data pipeline for a social media analytics platform. The pipeline needs to ingest posts from multiple sources (Twitter, Facebook) in real time, transform the data by adding sentiment scores, and store the results in a data store for later analysis. The transformation logic is simple and can be expressed as a SQL query. You want to minimize coding effort. Which Azure service should you use for the transformation step?

Accepted Answer

Azure Stream Analytics. Azure Stream Analytics is the correct choice because it is designed for real-time data processing with SQL-like query language, allowing you to transform streaming data (e.g., from Twitter and Facebook) by adding sentiment scores using simple SQL expressions without writing custom code. It integrates natively with Azure Event Hubs or IoT Hub for ingestion and outputs to Azure SQL Database, Cosmos DB, or Blob Storage for analysis, minimizing coding effort.

Answer

Azure Data Factory

Answer

Azure Databricks

Answer

Azure Functions

Question 9

A company is designing a data solution for a retail application. The solution must support real-time analytics on streaming sales data, and also provide historical reports for business intelligence. Which TWO data processing models should be combined to meet these requirements?

Accepted Answer

Batch processing. Batch processing (B) is correct because it is used to process large volumes of historical sales data at scheduled intervals, enabling the generation of comprehensive business intelligence reports. Stream processing (E) is correct because it handles real-time data ingestion and analytics on streaming sales data, allowing the application to react instantly to sales events. Combining these two models (often called a Lambda architecture) meets both the real-time and historical reporting requirements.

Answer

Distributed processing

Answer

Data lake storage

Answer

Transactional database

Question 10

Your company, Contoso Ltd., operates a global e-commerce platform. The data engineering team ingests over 10 TB of raw clickstream data daily into Azure Data Lake Storage Gen2. The data is partitioned by date and hour. Business analysts need to query this data using Azure Synapse Serverless SQL to generate daily sales reports. However, the reports are taking over 30 minutes to run, and the team needs to improve query performance without moving data to a dedicated SQL pool. You are asked to recommend a solution. Which action should you take?

Accepted Answer

Create external tables using a partition elimination strategy and ensure the data is partitioned by date.. Azure Synapse Serverless SQL can use external tables with partition elimination to skip irrelevant partitions (e.g., date/hour folders) during query execution. This reduces the amount of data scanned, directly improving query performance without moving data. Partition elimination works by filtering on the partition column (e.g., date) in the WHERE clause, allowing the query engine to read only the necessary files.

Answer

Convert the data from JSON to Parquet format and apply Snappy compression.

Answer

Use Azure Data Factory to copy the data into Azure SQL Database and create indexes.

Answer

Create a dedicated SQL pool and distribute the data across 60 distributions.

Question 11

Your company is implementing a data governance solution using Microsoft Purview. The data catalog must automatically scan and classify sensitive data in Azure SQL Database, Azure Synapse Analytics, and Amazon S3. The company uses Microsoft Entra ID for identity management. You need to ensure that the Purview managed identity can authenticate to these data sources. Which authentication method should you configure for the Amazon S3 connection?

Accepted Answer

AWS IAM authentication. Amazon S3 is an external cloud storage service that does not support Microsoft Entra ID, SQL Authentication, or Windows Authentication. To authenticate Purview's managed identity to S3, you must configure AWS IAM authentication, which allows Purview to assume an IAM role with permissions to read the S3 bucket metadata and data for scanning and classification.

Answer

SQL Authentication

Answer

Windows Authentication

Answer

Microsoft Entra ID authentication

Question 12

Your organization uses Azure Cosmos DB for a real-time inventory application. The data includes a container with items that have a `category` property. The operations team frequently queries for all items in a specific category. To optimize query performance and minimize request unit (RU) consumption, you decide to implement a materialized view. Which Azure Cosmos DB feature should you use to achieve this?

Accepted Answer

Materialized views (preview). Azure Cosmos DB's materialized views (preview) feature allows you to pre-join, aggregate, and transform data from a source container into a separate container optimized for specific query patterns, such as filtering by `category`. This reduces RU consumption by avoiding full scans or expensive cross-partition queries, as the view is pre-computed and indexed according to the target query.

Answer

Partition key design

Answer

Change feed

Answer

Composite indexes

Question 13

A university registrar's office needs to store student enrollment records where every student has exactly one student ID, one full name, one declared major, and one enrollment date. The registrar requires that no enrollment record can exist without a valid student ID, and queries always filter by major. Which data model characteristic best describes this requirement?

Accepted Answer

A relational model with a defined schema and a primary key on student ID. The registrar's data has a fixed shape, a required unique identifier, and consistent filtering, which are hallmarks of a relational model. A declared schema enforces that every enrollment record carries a valid student ID and the expected attributes, while indexed columns support efficient filtering by major. The other models either defer structure or optimize for relationship traversal, neither of which matches these requirements.

Answer

A graph model where students are nodes connected by enrollment edges to major nodes

Answer

A key-value model where each student is stored as an opaque blob without defined attributes

Answer

A schema-on-read model where structure is applied only when data is queried

Question 14

A media company stores metadata about its video assets. Each asset has a fixed set of fields: AssetID, Title, DurationSeconds, and UploadDate. Analysts run daily reports that sum DurationSeconds grouped by UploadDate. The company wants the storage engine to enforce that DurationSeconds is numeric and that each AssetID is unique. Which characteristic of the data most directly justifies choosing a relational database over a non-relational store?

Accepted Answer

The metadata fields are identical for every asset and require type and uniqueness enforcement. The uniform field set combined with explicit requirements for numeric typing and a unique AssetID points to a relational model, where such constraints are declared in the schema and enforced by the engine. Non-relational stores typically leave typing and uniqueness to application code, which weakens guarantees. Aggregation and client type do not distinguish the models here, so the structural enforcement requirement is the decisive factor.

Answer

The metadata will be consumed by a web application rather than a desktop client

Answer

The reports aggregate numeric values, which only relational engines can compute

Answer

The data volume is expected to grow quickly over time

Question 15

A regional utility is designing a data platform. Engineers will capture continuous readings from smart meters that arrive as a stream and must be analyzed within seconds for anomaly detection. Separately, billing analysts need to run complex queries joining customer contracts with historical usage, and those queries must always return consistent results even if several billing tables are updated in the same operation. Which two data processing approaches are most appropriate for these requirements? (Choose two.)

Accepted Answer

A relational transactional workload for the billing joins, ensuring consistent results across multi-table updates. The two requirements have distinct characteristics: continuous meter readings needing seconds-level analysis map to stream processing, while billing joins across multiple tables with guaranteed consistency map to a relational transactional workload. Stream processing minimizes latency, and ACID transactions ensure that multi-table updates either commit fully or roll back, keeping query results consistent. The remaining approaches either add latency, defer structure, or lack transactional guarantees.

Answer

A schema-on-read data lake for the billing joins to defer structure until query time

Answer

A key-value cache for the billing joins to speed up repeated lookups of contract rows

Answer

Batch processing for the smart meter readings to aggregate them once per day