DP-900 | 204 Questions and Answers
1. What three main types of workload can be found in a typical modern data warehouse?: - Streaming Data
- Batch Data
- Relational Data
2. A is a continuous flow of information, where contin- uous does not
necessarily mean regular or constant.: data stream
3. focuses on moving and transforming data at rest.: Batch
processing
4. This data is usually well organized and easy to understand. Data stored in relational databases is an example,
where table rows and columns represent entities and their attributes.: Structured Data
5. This data usually does not come from relational stores, since even if it could have some sort of internal
organization, it is not mandatory. Good examples are XML and JSON files.: Semi-structured Data
6. Data with no explicit data model falls in this category. Good examples include binary file formats (such as
PDF, Word, MP3, and MP4), emails, and tweets.: Unstructured Data
7. What type of analysis answers the question "What happened?": Descriptive Analysis
8. What type of analysis answers the question "Why did it happen?": Diagnos- tic Analysis
9. What type of analysis answers the question "What will happen?": Predictive Analysis
10.What type of analysis answers the question "How can we make it hap- pen?": Prescriptive Analysis
11.The two main kinds of workloads are and
.: extract-transform-load (ETL) extract-
load-transform (ELT)
12. is a traditional approach and has established best practices. It is more commonly found in on-
premises environments since it was around before cloud platforms. It is a process that involves a lot o data
movement, which is something you want to avoid on the cloud if possible due to its resource-intensive nature.:
ETL
13. seems similar to ETL at first glance but is better suited to big data scenarios since it leverages
the scalability and flexibility of MPP engines like Azure Synapse Analytics, Azure Databricks, or Azure
HDInsight.: ELT
14. is a cloud service that lets you implement, manage, and monitor a cluster for
Hadoop, Spark, HBase, Kafka, Store, Hive LLAP, and ML Service in an easy and effective way.: Azure HDInsight
, DP-900 | 204 Questions and Answers
15. is a cloud service from the creators of Apache Spark, combined with a great
integration with the Azure platform.: Azure Databricks
16. is the new name for Azure SQL Data Warehouse, but it extends it in many ways. It
aims to be the comprehensive analytics plat- form, from data ingestion to presentation, bringing together one-
click data exploration, robust pipelines, enterprise-grade database service, and report authoring.: Azure
Synapse Analytics
17.A displays attribute members on rows and measures on columns. A simple is
generally easy for users to understand, but it can quickly become difficult to read as the number of rows and
columns increases.: table
18.A is a more sophisticated table. It allows for attributes also on columns and can
auto-calculate subtotals.: matrix
19.Objects in which things about data should be captured and stored are called: .
A. tables
B. entities
C. rows
D. columns: B. entities
20.You need to process data that is generated continuously and near real-time responses are required. You
should use .
A. batch processing
B. scheduled data processing
C. buffering and processing
D. streaming data processing: D. streaming data processing
21.A. Extract, Transform, Load (ETL)
B. Extract, Load, Transform (ELT)
1. Optimize data privacy.
2. Provide support for Azure Data Lake.: 1 - A 2 - B
Extract, Transform, Load (ETL) is the correct approach when you need to filter sensitive data before loading the data into
analytical model. It is suitable for simple data models that do not require Azure Data Lake support. Extract, Load,
Transform (ELT) is the correct approach because it supports Azure Data Lake as the data store and manages large volum
of data.
, DP-900 | 204 Questions and Answers
22.The technique that provides recommended actions that you should take to achieve a goal or target is called
analytics.
A. descriptive
B. diagnostic
C. predictive
D. prescriptive: D. prescriptive
23.A. Tables
B. Indexes
C. Views
D. Keys
1. Create relationships.
2. Improve processing speed for data searches.
3. Store instances of entities as rows.
4. Display data from predefined queries.: 1 - D
2-B
3-A
4-C
24.The process of splitting an entity into more than one table to reduce data redundancy is called: .
A. deduplication
B. denormalization
C. normalization
D. optimization: C. normalization
25.Azure SQL Database is an example of -as-a-service.
A. platform
B. infrastructure
C. software
D. application: A. platform
26.A. Azure Data Studio
B. Azure Query editor
C. SQL Server Data Tools
, DP-900 | 204 Questions and Answers
1. Query data while working within a Visual Studio project.
2. Query data located in a non-Microsoft platform.
3. Query data from within the Azure portal: 1 - C
1. What three main types of workload can be found in a typical modern data warehouse?: - Streaming Data
- Batch Data
- Relational Data
2. A is a continuous flow of information, where contin- uous does not
necessarily mean regular or constant.: data stream
3. focuses on moving and transforming data at rest.: Batch
processing
4. This data is usually well organized and easy to understand. Data stored in relational databases is an example,
where table rows and columns represent entities and their attributes.: Structured Data
5. This data usually does not come from relational stores, since even if it could have some sort of internal
organization, it is not mandatory. Good examples are XML and JSON files.: Semi-structured Data
6. Data with no explicit data model falls in this category. Good examples include binary file formats (such as
PDF, Word, MP3, and MP4), emails, and tweets.: Unstructured Data
7. What type of analysis answers the question "What happened?": Descriptive Analysis
8. What type of analysis answers the question "Why did it happen?": Diagnos- tic Analysis
9. What type of analysis answers the question "What will happen?": Predictive Analysis
10.What type of analysis answers the question "How can we make it hap- pen?": Prescriptive Analysis
11.The two main kinds of workloads are and
.: extract-transform-load (ETL) extract-
load-transform (ELT)
12. is a traditional approach and has established best practices. It is more commonly found in on-
premises environments since it was around before cloud platforms. It is a process that involves a lot o data
movement, which is something you want to avoid on the cloud if possible due to its resource-intensive nature.:
ETL
13. seems similar to ETL at first glance but is better suited to big data scenarios since it leverages
the scalability and flexibility of MPP engines like Azure Synapse Analytics, Azure Databricks, or Azure
HDInsight.: ELT
14. is a cloud service that lets you implement, manage, and monitor a cluster for
Hadoop, Spark, HBase, Kafka, Store, Hive LLAP, and ML Service in an easy and effective way.: Azure HDInsight
, DP-900 | 204 Questions and Answers
15. is a cloud service from the creators of Apache Spark, combined with a great
integration with the Azure platform.: Azure Databricks
16. is the new name for Azure SQL Data Warehouse, but it extends it in many ways. It
aims to be the comprehensive analytics plat- form, from data ingestion to presentation, bringing together one-
click data exploration, robust pipelines, enterprise-grade database service, and report authoring.: Azure
Synapse Analytics
17.A displays attribute members on rows and measures on columns. A simple is
generally easy for users to understand, but it can quickly become difficult to read as the number of rows and
columns increases.: table
18.A is a more sophisticated table. It allows for attributes also on columns and can
auto-calculate subtotals.: matrix
19.Objects in which things about data should be captured and stored are called: .
A. tables
B. entities
C. rows
D. columns: B. entities
20.You need to process data that is generated continuously and near real-time responses are required. You
should use .
A. batch processing
B. scheduled data processing
C. buffering and processing
D. streaming data processing: D. streaming data processing
21.A. Extract, Transform, Load (ETL)
B. Extract, Load, Transform (ELT)
1. Optimize data privacy.
2. Provide support for Azure Data Lake.: 1 - A 2 - B
Extract, Transform, Load (ETL) is the correct approach when you need to filter sensitive data before loading the data into
analytical model. It is suitable for simple data models that do not require Azure Data Lake support. Extract, Load,
Transform (ELT) is the correct approach because it supports Azure Data Lake as the data store and manages large volum
of data.
, DP-900 | 204 Questions and Answers
22.The technique that provides recommended actions that you should take to achieve a goal or target is called
analytics.
A. descriptive
B. diagnostic
C. predictive
D. prescriptive: D. prescriptive
23.A. Tables
B. Indexes
C. Views
D. Keys
1. Create relationships.
2. Improve processing speed for data searches.
3. Store instances of entities as rows.
4. Display data from predefined queries.: 1 - D
2-B
3-A
4-C
24.The process of splitting an entity into more than one table to reduce data redundancy is called: .
A. deduplication
B. denormalization
C. normalization
D. optimization: C. normalization
25.Azure SQL Database is an example of -as-a-service.
A. platform
B. infrastructure
C. software
D. application: A. platform
26.A. Azure Data Studio
B. Azure Query editor
C. SQL Server Data Tools
, DP-900 | 204 Questions and Answers
1. Query data while working within a Visual Studio project.
2. Query data located in a non-Microsoft platform.
3. Query data from within the Azure portal: 1 - C
