DP-900 Describe an Analytics Workload Practice Questions

Free DP-900 Describe an Analytics Workload practice test questions

8 questions from this domain with answers and explanations - different from the samples on the main DP-900 page. Sign up free to practice the full set.

1. Question 1Describe an Analytics Workload

   A large retailer wants to consolidate data from its point-of-sale systems, online store, and supply chain into a single repository optimized for business intelligence reporting. The data should be organized using fact and dimension tables. Which Azure service is best suited for building this solution?

   • AAzure Blob Storage
   • BAzure Synapse AnalyticsCorrect
   • CAzure SQL Database
   • DAzure Cosmos DB
   ✓ Correct answer: B

   Azure Synapse Analytics is specifically designed as a data warehouse platform that excels at consolidating data from multiple sources and organizing it using dimensional modeling (fact and dimension tables) for business intelligence. It combines the analytical power of Synapse SQL pools (for data warehousing) with Spark capabilities, making it ideal for building enterprise BI solutions with large-scale data consolidation.

   Why the other options are wrong

   • AAzure Blob Storage is incorrect because it is object storage, not a data warehouse platform, and does not provide query capabilities or schema organization for BI reporting.
   • CAzure SQL Database is incorrect because while it is a relational database, it is optimized for OLTP transactional workloads rather than large-scale analytics and dimensional modeling.
   • DAzure Cosmos DB is incorrect because it is a NoSQL database designed for operational workloads, not BI reporting with fact and dimension table schemas.
2. Question 2Describe an Analytics Workload

   In Microsoft Fabric, which workload is designed to analyze streaming and event-based data as it arrives?

   • AData Warehouse
   • BReal-Time IntelligenceCorrect
   • CData Factory
   • DPower BI paginated reports
   ✓ Correct answer: B

   Real-Time Intelligence is the Microsoft Fabric workload built for ingesting, querying, and analyzing high-volume streaming and event data with low latency, and for triggering actions on it. Within Fabric it plays the role that Event Hubs and Stream Analytics play in a classic Azure streaming pipeline. Data Warehouse serves large-scale historical SQL analytics, Data Factory handles batch data integration, and paginated reports are a Power BI format - none of these is the streaming-focused workload described.

   Why the other options are wrong

   • AData Warehouse is for historical, batch relational analytics, not continuously arriving event data.
   • CData Factory orchestrates batch data movement and transformation, not real-time event analysis.
   • DPaginated reports are a static reporting format in Power BI, not a real-time analytics workload.
3. Question 3Describe an Analytics WorkloadSelect all that apply

   Datum Corporation is building an analytics solution. Which of the following are components typically found in a modern data analytics pipeline? Select all that apply.

   • AData transformation and processingCorrect
   • BData ingestion from source systemsCorrect
   • CData encryption at the source application level
   • DData visualization and reportingCorrect
   ✓ Correct answer: A, B, D

   A modern data analytics pipeline consists of several connected stages that move data from source systems through transformation and analysis to presentation. Data ingestion retrieves raw data from various source systems (databases, APIs, files, streaming sources). Data transformation and processing prepares and enriches this data by cleaning, validating, aggregating, and reshaping it into analytical structures. Data visualization and reporting presents the processed data in accessible formats (dashboards, reports, charts) that enable business users to extract insights and make decisions. These three components form the core flow of analytical data processing.

   Why the other options are wrong

   • CData encryption at the source application level is incorrect because while security is important in analytics solutions, encryption at the source application level is not a component of the analytics pipeline itself-it is an infrastructure and security concern that may be applied across the pipeline. This option describes a security practice rather than a pipeline functional component.
4. Question 4Describe an Analytics Workload

   A data warehouse at Humongous Insurance uses a dimension table that slowly changes over time. For example, when a customer changes their address, both the old and new addresses need to be retained for historical accuracy. What is this design pattern called?

   • ARapidly changing dimension
   • BSlowly changing dimension (SCD)Correct
   • CDegenerate dimension
   • DConformed dimension
   ✓ Correct answer: B

   A Slowly Changing Dimension (SCD) is a data warehouse design pattern that handles the evolution of dimension attributes over time while preserving historical accuracy. When attribute changes occur (such as a customer address change), SCD techniques maintain both the previous and new values, allowing analysts to report results as they appeared at different points in time. This approach is critical for historical analysis, audit trails, and accurate period-over-period comparisons in analytical systems. Multiple SCD types exist (Type 1 overwrites, Type 2 adds history rows, Type 3 maintains current and previous values), but all address the requirement to manage temporal changes in dimensional data.

   Why the other options are wrong

   • ARapidly changing dimension is incorrect because it refers to dimensions with frequent attribute changes that can impact star schema performance due to high-volume updates.
   • CDegenerate dimension is incorrect because this is a dimension that contains only an identifier without descriptive attributes, used for transaction-level detail.
   • DConformed dimension is incorrect because this term refers to dimensions that are reused across multiple fact tables with identical structure, not to handling temporal changes.
5. Question 5Describe an Analytics Workload

   Which Azure service provides a fully managed Apache Spark environment for big data processing and analytics?

   • AAzure Cosmos DB with its globally distributed NoSQL engine
   • BAzure SQL Database with its managed relational engine
   • CAzure HDInsight or Azure Synapse Spark poolsCorrect
   • DAzure Data Factory with its pipeline orchestration engine
   ✓ Correct answer: C

   Both Azure HDInsight and Azure Synapse Spark pools provide fully managed Apache Spark environments for big-data processing and analytics, so the combined option correctly identifies where managed Spark runs on Azure. They let teams run distributed Spark jobs without provisioning and maintaining clusters by hand. Offering managed Spark compute is exactly what these services do.

   Why the other options are wrong

   • ACosmos DB is a NoSQL database, not a compute engine; managed Apache Spark for big-data processing is delivered by HDInsight or Synapse Spark pools.
   • BAzure SQL Database is a relational OLTP engine and does not run Spark; big-data Spark workloads run on HDInsight or Synapse Spark pools.
   • DData Factory orchestrates data movement but does not itself provide a managed Spark cluster; that is offered by HDInsight or Synapse Spark pools.
6. Question 6Describe an Analytics WorkloadSelect all that apply

   Which TWO Azure services are most appropriate for scheduled, large-scale BATCH data processing rather than real-time streaming? (Choose two.)

   • AAzure Data FactoryCorrect
   • BAzure Synapse AnalyticsCorrect
   • CAzure Event Hubs service
   • DAzure Stream Analytics
   ✓ Correct answer: A, B

   Scheduled, large-scale batch processing is the domain of Azure Data Factory, which orchestrates ETL and ELT pipelines, and Azure Synapse Analytics, which runs big-data and data-warehouse workloads over large historical datasets. These are designed to process data in bulk on a schedule. Azure Event Hubs and Azure Stream Analytics are built for the opposite pattern - ingesting and processing events continuously in real time with low latency. The choice comes down to bulk periodic processing (batch) versus continuous low-latency processing (streaming).

   Why the other options are wrong

   • CEvent Hubs is a real-time event ingestion front door, not a scheduled batch processing service.
   • DStream Analytics processes events continuously in real time, which is the opposite of batch processing.
7. Question 7Describe an Analytics WorkloadSelect all that apply

   A BI team wants a single cloud service that unifies large-scale data warehousing with big-data analytics in one workspace, and they need an author to build the data model and visuals on a Windows desktop before publishing. Which TWO components fit these roles? (Choose two.)

   • AAzure Synapse AnalyticsCorrect
   • BMicrosoft Power BI DesktopCorrect
   • CThe Power BI service (app.powerbi.com)
   • DAzure Analysis Services
   • EAzure Data Factory
   ✓ Correct answer: A, B

   Azure Synapse Analytics brings enterprise data warehousing and big-data analytics together in a single integrated workspace. Microsoft Power BI Desktop is the free Windows application where authors build the data model, queries, and visuals before publishing them to the service.

   Why the other options are wrong

   • CThe Power BI service is the browser-based platform for sharing and consuming published content, not the desktop authoring tool.
   • DAzure Analysis Services builds tabular semantic models but does not by itself unify warehousing with big-data analytics like Synapse.
   • EAzure Data Factory orchestrates data movement and transformation; it is neither the unified analytics workspace nor the desktop authoring tool.
8. Question 8Describe an Analytics Workload

   In an Azure Synapse dedicated SQL pool, what does its massively parallel processing (MPP) architecture do?

   • AIt divides one query across many compute nodes that run in parallelCorrect
   • BIt forces every query to run on a single node in strict order only
   • CIt caches the entire warehouse in memory on the analyst's own laptop
   • DIt turns query results into charts automatically without any coding
   ✓ Correct answer: A

   A dedicated SQL pool uses a massively parallel processing (MPP) engine that spreads rows across underlying distributions and assigns compute nodes to process them at the same time. A single large query is broken into smaller tasks that run in parallel, which is why MPP suits big analytical scans over billions of rows. This is the opposite of a traditional single-node database that processes a query on one CPU. MPP scales out across nodes, not just up.

   Why the other options are wrong

   • BSerial single-node execution is the model MPP avoids; MPP runs the work concurrently across many nodes.
   • CCaching a whole warehouse on a client device is not how MPP works; compute stays in the cloud across nodes.
   • DTurning results into visuals is the role of Power BI, not the MPP query engine inside the SQL pool.

How Describe an Analytics Workload is tested

This domain holds 204 of the 717 questions in the DP-900 bank, about 28%. The mix is 140 single-answer multiple choice, 35 multiple-response, 24 true/false, 3 hotspot, 1 yes/no scenario and 1 ordering, so it is worth practising the formats as well as the content.

Once you have a few attempts recorded, CertGrid scores every domain separately and points you at the weakest one, so you can drill Describe an Analytics Workload on its own rather than re-running full-length mocks.

Other DP-900 exam domains

• Describe Core Data Concepts199 questions
• Identify Considerations for Relational Data177 questions
• Describe Considerations for Non-Relational Data137 questions
• All DP-900 practice questions717 total
• Describe an Analytics Workload study notesKey concepts
• Microsoft practice examsAll Microsoft

DP-900 Describe an Analytics Workload FAQ