1
Microsoft Data
Fundamentals (DP-900)
— Advanced Practice
Exam V2.0 150 Multiple-
Choice Questions with
Answers and Rationales
a well detailed one 2025
/ 2026 written and
graded A+ upgraded
, 2
DOMAIN 1: CORE DATA CONCEPTS (15–20%)
Question 1
A multinational corporation is building a centralized data platform. The platform must store raw
log files from web servers, sensor data from manufacturing equipment, and structured sales
records from an ERP system, all within a single storage account. Which Azure storage service
should serve as the foundation for this platform?
A. Azure Blob Storage
B. Azure SQL Database
C. Azure Cosmos DB
D. Azure Table Storage
- detailed answer 100% correct :- A — Azure Blob Storage
Rationale: Azure Blob Storage is designed to store massive amounts of unstructured data,
including log files, sensor data, and any type of file. It can serve as a data lake foundation for
diverse data types. SQL Database is for relational data only (B). Cosmos DB is for NoSQL data (C).
Table Storage is for key-value data (D).
Question 2
A company's database contains a Customer table with a CustomerID column that uniquely
identifies each customer. This column is used as a foreign key in the Orders table. What property
must the CustomerID column in the Customer table have?
A. It must allow NULL values
B. It must be a primary key
, 3
C. It must be an indexed column
D. It must be a composite key
- detailed answer 100% correct :- B — It must be a primary key
Rationale: For a column to be referenced as a foreign key in another table, it must be a primary
key (or have a unique constraint) in its own table. Primary keys uniquely identify each row.
Foreign keys cannot reference NULL values (A). While indexing is common, it's not a
requirement (C). Composite keys are possible but not required (D).
Question 3
A retail company processes millions of transactions daily. Each transaction must be recorded
immediately, and inventory levels must be updated in real-time to prevent overselling. Which
type of data processing workload best describes this scenario?
A. Batch processing
B. Stream processing
C. Online Transaction Processing (OLTP)
D. Online Analytical Processing (OLAP)
- detailed answer 100% correct :- C — Online Transaction Processing (OLTP)
Rationale: OLTP systems handle high-volume, short-duration transactions in real-time, ensuring
immediate data consistency. This scenario describes a classic OLTP workload. Batch processing
would introduce unacceptable latency (A). Stream processing is for continuous data flows but
OLTP is more specific to transactional databases (B). OLAP is for analytical queries, not
operational transactions (D).
Question 4
An organization needs to store video surveillance footage, audio recordings of customer calls,
and scanned document images for compliance purposes. How should these data types be
classified collectively?
A. Structured data
B. Semi-structured data
C. Unstructured data
D. Tabular data
, 4
- detailed answer 100% correct :- C — Unstructured data
Rationale: Video, audio, and scanned images lack a predefined data model or schema, making
them classic examples of unstructured data. Structured data fits a rigid schema like tables (A).
Semi-structured data has some organization like JSON or XML (B). Tabular data is a form of
structured data (D).
Question 5
A database administrator is designing a new database and wants to eliminate duplicate
customer address data that is currently stored in multiple tables. Which process should be
applied to achieve this?
A. Data denormalization
B. Data normalization
C. Data replication
D. Data partitioning
- detailed answer 100% correct :- B — Data normalization
Rationale: Normalization is the process of organizing data to reduce redundancy and improve
data integrity by structuring tables to minimize duplicate data. Denormalization introduces
redundancy for performance (A). Replication copies data (C). Partitioning splits data for
scalability (D).
Question 6
Which statement accurately describes the role of a Data Engineer compared to a Data Analyst?
A. Data Engineers create dashboards; Data Analysts build data pipelines
B. Data Engineers build and maintain data pipelines; Data Analysts create insights and reports
C. Data Engineers manage database security; Data Analysts build machine learning models
D. There is no meaningful difference between the roles
- detailed answer 100% correct :- B — Data Engineers build and maintain data
pipelines; Data Analysts create insights and reports
Rationale: Data engineers design, build, and maintain data pipelines and infrastructure. Data
analysts focus on analyzing data, creating reports, and generating business insights. The roles
are distinct and complementary.
Microsoft Data
Fundamentals (DP-900)
— Advanced Practice
Exam V2.0 150 Multiple-
Choice Questions with
Answers and Rationales
a well detailed one 2025
/ 2026 written and
graded A+ upgraded
, 2
DOMAIN 1: CORE DATA CONCEPTS (15–20%)
Question 1
A multinational corporation is building a centralized data platform. The platform must store raw
log files from web servers, sensor data from manufacturing equipment, and structured sales
records from an ERP system, all within a single storage account. Which Azure storage service
should serve as the foundation for this platform?
A. Azure Blob Storage
B. Azure SQL Database
C. Azure Cosmos DB
D. Azure Table Storage
- detailed answer 100% correct :- A — Azure Blob Storage
Rationale: Azure Blob Storage is designed to store massive amounts of unstructured data,
including log files, sensor data, and any type of file. It can serve as a data lake foundation for
diverse data types. SQL Database is for relational data only (B). Cosmos DB is for NoSQL data (C).
Table Storage is for key-value data (D).
Question 2
A company's database contains a Customer table with a CustomerID column that uniquely
identifies each customer. This column is used as a foreign key in the Orders table. What property
must the CustomerID column in the Customer table have?
A. It must allow NULL values
B. It must be a primary key
, 3
C. It must be an indexed column
D. It must be a composite key
- detailed answer 100% correct :- B — It must be a primary key
Rationale: For a column to be referenced as a foreign key in another table, it must be a primary
key (or have a unique constraint) in its own table. Primary keys uniquely identify each row.
Foreign keys cannot reference NULL values (A). While indexing is common, it's not a
requirement (C). Composite keys are possible but not required (D).
Question 3
A retail company processes millions of transactions daily. Each transaction must be recorded
immediately, and inventory levels must be updated in real-time to prevent overselling. Which
type of data processing workload best describes this scenario?
A. Batch processing
B. Stream processing
C. Online Transaction Processing (OLTP)
D. Online Analytical Processing (OLAP)
- detailed answer 100% correct :- C — Online Transaction Processing (OLTP)
Rationale: OLTP systems handle high-volume, short-duration transactions in real-time, ensuring
immediate data consistency. This scenario describes a classic OLTP workload. Batch processing
would introduce unacceptable latency (A). Stream processing is for continuous data flows but
OLTP is more specific to transactional databases (B). OLAP is for analytical queries, not
operational transactions (D).
Question 4
An organization needs to store video surveillance footage, audio recordings of customer calls,
and scanned document images for compliance purposes. How should these data types be
classified collectively?
A. Structured data
B. Semi-structured data
C. Unstructured data
D. Tabular data
, 4
- detailed answer 100% correct :- C — Unstructured data
Rationale: Video, audio, and scanned images lack a predefined data model or schema, making
them classic examples of unstructured data. Structured data fits a rigid schema like tables (A).
Semi-structured data has some organization like JSON or XML (B). Tabular data is a form of
structured data (D).
Question 5
A database administrator is designing a new database and wants to eliminate duplicate
customer address data that is currently stored in multiple tables. Which process should be
applied to achieve this?
A. Data denormalization
B. Data normalization
C. Data replication
D. Data partitioning
- detailed answer 100% correct :- B — Data normalization
Rationale: Normalization is the process of organizing data to reduce redundancy and improve
data integrity by structuring tables to minimize duplicate data. Denormalization introduces
redundancy for performance (A). Replication copies data (C). Partitioning splits data for
scalability (D).
Question 6
Which statement accurately describes the role of a Data Engineer compared to a Data Analyst?
A. Data Engineers create dashboards; Data Analysts build data pipelines
B. Data Engineers build and maintain data pipelines; Data Analysts create insights and reports
C. Data Engineers manage database security; Data Analysts build machine learning models
D. There is no meaningful difference between the roles
- detailed answer 100% correct :- B — Data Engineers build and maintain data
pipelines; Data Analysts create insights and reports
Rationale: Data engineers design, build, and maintain data pipelines and infrastructure. Data
analysts focus on analyzing data, creating reports, and generating business insights. The roles
are distinct and complementary.
