Prepare for the DP-900 Microsoft Azure Data Fundamentals Exam

Free DP-900 Practice Questions (With Answers)

5 real questions from the DP-900 bank, with the full explanation for each answer. No sign-up needed to read them.

1. Question 1Relational Data on Azure [Views]

   A data analyst at a healthcare company needs to let department managers see only patient data for their own department, without creating separate copies of the data. Which database object would best serve this purpose?

   • A.A stored procedure that filters by department
   • B.A separate copy of the table for each department
   • C.An index on the department column
   • D.A view that filters the patient table by department
   Show answer and explanation

   Correct answer: D

   Correct. A view is a virtual table based on a SELECT query—it doesn't store data but presents a filtered or shaped subset. This lets each manager query 'their' view as if it were a table, while the underlying data remains in one place.

   Why the other options are wrong

   A. Incorrect. Stored procedures execute logic but aren't queryable like tables. Managers would need to run the procedure rather than query it directly in their reporting tools.

   B. Incorrect. Copying data creates maintenance nightmares—updates must happen in every copy. Views solve this by providing different 'windows' into the same underlying data.

   C. Incorrect. An index speeds up queries but doesn't restrict what data users can see. It's a performance tool, not a security or presentation tool.

2. Question 2Non-Relational Data on Azure [Cosmos DB Consistency Trade-offs]

   In Cosmos DB, what is the trade-off when choosing Eventual consistency over Strong consistency?

   • A.Higher storage costs, because every replica keeps its own separate extra copy
   • B.Stronger data validation, enforced by the service at the row level
   • C.Better query performance, but a noticeably higher Request Unit charge on every read
   • D.Lower latency and higher availability, but reads might return stale data
   Show answer and explanation

   Correct answer: D

   Correct. Eventual consistency provides the lowest latency and highest availability because reads don't wait for writes to propagate. The trade-off is that you might read data that's slightly out of date.

   Why the other options are wrong

   A. Incorrect. Consistency levels affect read behavior (freshness and latency), not storage costs. All data is stored the same way regardless of which consistency level you choose.

   B. Incorrect. Consistency levels control the trade-off between read latency and data freshness—not data validation or format enforcement. Choosing Eventual vs Strong consistency affects when replicas reflect the latest write.

   C. Incorrect. Cost runs the other way: strong and bounded-staleness reads consume MORE Request Units than eventual reads. The trade-off eventual consistency actually makes is freshness — lower latency in exchange for possibly stale data.

3. Question 3Core Data Concepts [Data Types]

   A retail company stores product information in a database with columns for ProductID, Name, Price, and Category. Each product has the same set of attributes. What type of data is this?

   • A.Unstructured data
   • B.Semi-structured data
   • C.Binary data
   • D.Structured data
   Show answer and explanation

   Correct answer: D

   Correct. The fundamental characteristic of structured data is that it adheres to a predefined schema with fixed columns and data types. Data organized in tables with rows and columns, where each row has the same attributes, is the definition of structured data.

   Why the other options are wrong

   A. Incorrect. Unstructured data has no predefined schema or organization that a database can interpret, such as images or videos. This data has a clear tabular structure.

   B. Incorrect. Semi-structured data contains tags or markers to identify elements but allows flexible schemas where records can have different fields. This data has a fixed, rigid schema.

   C. Incorrect. Binary data refers to files like images or executables that contain non-text content. This product data is text-based tabular information.

4. Question 4Core Data Concepts [Data Roles]

   A team member creates Power BI reports and dashboards to help executives understand sales performance. Which role does this describe?

   • A.Database Administrator
   • B.Data Engineer
   • C.Data Analyst
   • D.Solutions Architect
   Show answer and explanation

   Correct answer: C

   Correct. Data Analysts focus on insights and visualization. Creating reports and dashboards to communicate data findings to stakeholders is the fundamental responsibility of this role.

   Why the other options are wrong

   A. Incorrect. DBAs manage database infrastructure, not visualization and reporting.

   B. Incorrect. Data Engineers build data pipelines, not business reports and dashboards.

   D. Incorrect. Solutions Architects design overall system architecture, not day-to-day reporting.

5. Question 5Relational Data on Azure [SQL Commands]

   Which SQL command would you use to add a new row to an existing table?

   • A.CREATE
   • B.UPDATE
   • C.ALTER
   • D.INSERT
   Show answer and explanation

   Correct answer: D

   Correct. INSERT is a DML (Data Manipulation Language) command that adds new rows to a table. This is the fundamental command for adding data to relational databases.

   Why the other options are wrong

   A. Incorrect. CREATE is DDL (Data Definition Language) used to create new tables or databases, not to add data to existing tables.

   B. Incorrect. UPDATE modifies existing rows. To add a new row, you need a different command.

   C. Incorrect. ALTER is DDL used to modify table structure (add columns, change data types), not to add data rows.