Practice Exams:

View All

Microsoft Fabric Analytics Engineer DP-600 Enterprise Semantic Model Performance And Direct Lake Practice Test

Skill 3.2 – 67 original questions

This Microsoft DP-600 practice test focuses on enterprise semantic model performance and direct lake through original scenario-based questions aligned to the active DP-600 skills measured as of July 21, 2026. Use the complete ExamSnap DP-600 collection for broader practice across Microsoft Fabric analytics lifecycle, data preparation, querying, and semantic modeling. For broader exam preparation, review the Microsoft DP-600 Exam Dumps page.

Instructions: Select the best answer for each question. Review the explanation after answering; each option includes a reason it is or is not the strongest choice for the scenario.

Question 1

For the marketing semantic model, Fourth Coffee has documented a business requirement to identify which report visual is issuing the slowest query before redesigning the page. The data governance group must meet it in a way where the design should minimize duplicated data. What is the best choice? Existing users should keep their current access.

Use incremental refresh and define a shorter refresh window than the historical retention window
Use the appropriate DAX window function with an explicit ordering definition
Use DAX Studio or equivalent server-timing diagnostics to profile the measure
Use Power BI Performance Analyzer to capture and compare visual query durations
Configure the semantic model to use large semantic model storage format

Correct answer: D

Why: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This directly matches the stated requirement.

Option review:

A: The policy can retain a long history while refreshing only the recent period that is still subject to change. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

B: DAX window functions support ordered row-window calculations without manually reconstructing the entire sequence logic. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

C: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

D: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This directly matches the stated requirement.

E: The large storage format supports enterprise-scale models and related management operations on supported capacity. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

Learning point: Use Power BI Performance Analyzer to capture and compare visual query durations

Question 2

A governance review of Wide World Importers’s supply-chain lakehouse asks for evidence that the solution can reduce the cost of a report page that renders many high-cardinality visuals at once. Because least privilege must be preserved, which action should be approved? Existing users should keep their current access.

Reduce unnecessary visuals or interactions and lower the amount of data each visual requests
Use the appropriate DAX window function with an explicit ordering definition
Model a fact table surrounded by dimension tables with clear one-to-many relationships
Choose Direct Lake on OneLake
Use incremental refresh and define a shorter refresh window than the historical retention window

Correct answer: A

Why: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This directly matches the stated requirement.

Option review:

A: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This directly matches the stated requirement.

C: A star schema gives the semantic model a predictable analytical structure and simplifies filter propagation. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

D: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This can be valid for ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

E: The policy can retain a long history while refreshing only the recent period that is still subject to change. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

Learning point: Reduce unnecessary visuals or interactions and lower the amount of data each visual requests

Question 3

Before expanding the IoT telemetry solution, the finance analytics squad at Northwind Traders must improve model query efficiency by replacing a snowflaked reporting path with a simpler analytical relationship pattern. The rollout plan says that the rollout must support controlled validation. Which option most directly addresses the requirement? Existing users should keep their current access.

Create a field parameter
Choose Direct Lake on the SQL analytics endpoint
Optimize the model toward a clear star schema and efficient relationships
Create both relationships and use an inactive relationship for the secondary date when appropriate
Configure an incremental refresh policy with RangeStart and RangeEnd parameters

Correct answer: C

Why: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This directly matches the stated requirement.

Option review:

A: Field parameters let consumers dynamically choose which modeled fields a visual uses. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

B: Direct Lake on SQL uses the SQL analytics endpoint for metadata and can fall back to DirectQuery in supported scenarios. This can be valid for ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

C: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This directly matches the stated requirement.

D: Role-playing date scenarios commonly use one active relationship and an inactive relationship activated explicitly in DAX when needed. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

E: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

Learning point: Optimize the model toward a clear star schema and efficient relationships

Question 4

Trey Research is redesigning its finance reporting platform. The analytics engineering team must identify which report visual is issuing the slowest query before redesigning the page. In addition, the team wants predictable performance and behavior. Which action is the best fit? No unrelated workspace or model permissions should be changed.

Use DAX Studio or equivalent server-timing diagnostics to profile the measure
Introduce a bridge table at the customer-segment membership grain
Use a dynamic format string
Create a calculation group
Use Power BI Performance Analyzer to capture and compare visual query durations

Correct answer: E

Why: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This directly matches the stated requirement.

Option review:

A: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

B: A bridge table resolves a many-to-many business relationship into controlled one-to-many relationships for filtering and measures. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

C: Dynamic format strings change presentation without converting the measure result to text. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

D: Calculation groups centralize reusable calculation logic and reduce measure proliferation. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

E: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This directly matches the stated requirement.

Learning point: Use Power BI Performance Analyzer to capture and compare visual query durations

Question 5

During a design review for Alpine Ski House’s retail performance dashboard, one requirement is non-negotiable: reduce the cost of a report page that renders many high-cardinality visuals at once. Because the choice should use a native Fabric capability, what should the operations data team implement? No unrelated workspace or model permissions should be changed.

Configure the semantic model to use large semantic model storage format
Reduce unnecessary visuals or interactions and lower the amount of data each visual requests
Use Direct Lake storage mode
Apply the RangeStart and RangeEnd filters in a foldable transformation
Use Dual storage mode for the compatible dimension

Correct answer: B

Why: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This directly matches the stated requirement.

Option review:

A: The large storage format supports enterprise-scale models and related management operations on supported capacity. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

B: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This directly matches the stated requirement.

C: Direct Lake reads OneLake-backed Delta data for semantic-model queries without a conventional import copy. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

D: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

E: Dual can act as Import or DirectQuery depending on the query path and can improve composite-model dimension behavior. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

Learning point: Reduce unnecessary visuals or interactions and lower the amount of data each visual requests

Question 6

The data governance group at Contoso is preparing the next release of its marketing semantic model. They need to improve model query efficiency by replacing a snowflaked reporting path with a simpler analytical relationship pattern; the design should minimize duplicated data. Which choice most directly satisfies the requirement? No unrelated workspace or model permissions should be changed.

Use Import storage mode
Configure an incremental refresh policy with RangeStart and RangeEnd parameters
Use a semantic-model star schema with dimensions filtering the fact table
Use DAX Studio or equivalent server-timing diagnostics to profile the measure
Optimize the model toward a clear star schema and efficient relationships

Correct answer: E

Why: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This directly matches the stated requirement.

Option review:

A: Import loads data into the VertiPaq in-memory engine and generally provides the fastest query performance when refresh latency is acceptable. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

B: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

C: Dimension-to-fact filtering aligns the model with common analytical patterns and reduces relationship ambiguity. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

D: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

E: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This directly matches the stated requirement.

Learning point: Optimize the model toward a clear star schema and efficient relationships

Question 7

A production readiness review at Wingtip Toys found a gap in the supply-chain lakehouse. The remediation must identify which report visual is issuing the slowest query before redesigning the page, and least privilege must be preserved. What is the most appropriate action? The team will validate the change first in a nonproduction environment.

Apply the RangeStart and RangeEnd filters in a foldable transformation
Configure the semantic model to use large semantic model storage format
Use incremental refresh and define a shorter refresh window than the historical retention window
Use Power BI Performance Analyzer to capture and compare visual query durations
Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback

Correct answer: D

Why: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This directly matches the stated requirement.

Option review:

A: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

B: The large storage format supports enterprise-scale models and related management operations on supported capacity. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

C: The policy can retain a long history while refreshing only the recent period that is still subject to change. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

D: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This directly matches the stated requirement.

E: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This can be valid for ‘Configure Direct Lake, including default fallback and refresh behavior’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

Learning point: Use Power BI Performance Analyzer to capture and compare visual query durations

Question 8

For a new phase of the IoT telemetry solution, Proseware asks the finance analytics squad to reduce the cost of a report page that renders many high-cardinality visuals at once. The architecture decision record also states that the rollout must support controlled validation. Which approach should be selected? The team will validate the change first in a nonproduction environment.

Reduce unnecessary visuals or interactions and lower the amount of data each visual requests
Introduce a bridge table at the customer-segment membership grain
Store the intermediate expression in a DAX VAR and reference the variable
Configure an incremental refresh policy with RangeStart and RangeEnd parameters
Build a composite model that uses the appropriate storage mode for each table

Correct answer: A

Why: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This directly matches the stated requirement.

Option review:

A: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This directly matches the stated requirement.

C: Variables reduce repeated expressions and make complex DAX easier to read and maintain. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

D: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

E: Composite models can combine tables that use different storage modes so each part of the model matches its latency and performance needs. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

Learning point: Reduce unnecessary visuals or interactions and lower the amount of data each visual requests

Question 9

Blue Yonder Airlines is standardizing how the finance reporting platform is managed. The immediate goal is to improve model query efficiency by replacing a snowflaked reporting path with a simpler analytical relationship pattern. Given that the team wants predictable performance and behavior, which option should the analytics engineering team choose? The team will validate the change first in a nonproduction environment.

Configure an incremental refresh policy with RangeStart and RangeEnd parameters
Choose Direct Lake on the SQL analytics endpoint
Create a composite model over the remote semantic model and the local data
Optimize the model toward a clear star schema and efficient relationships
Compute the expression once in a DAX variable and reuse it

Correct answer: D

Why: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This directly matches the stated requirement.

Option review:

A: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

B: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This can be valid for ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

C: Composite models can extend a remote semantic model with additional local model objects when supported. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

D: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This directly matches the stated requirement.

E: Variables can reduce repeated evaluation and also make the measure easier to optimize. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

Learning point: Optimize the model toward a clear star schema and efficient relationships

Question 10

An internal audit of Fabrikam’s retail performance dashboard identifies this requirement: identify which report visual is issuing the slowest query before redesigning the page. The operations data team also notes that the choice should use a native Fabric capability. What should they do? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Compute the expression once in a DAX variable and reuse it
Create both relationships and use an inactive relationship for the secondary date when appropriate
Use Power BI Performance Analyzer to capture and compare visual query durations
Use Dual storage mode for the compatible dimension
Create a composite model over the remote semantic model and the local data

Correct answer: C

Why: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This directly matches the stated requirement.

Option review:

A: Variables can reduce repeated evaluation and also make the measure easier to optimize. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

B: Role-playing date scenarios commonly use one active relationship and an inactive relationship activated explicitly in DAX when needed. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

C: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This directly matches the stated requirement.

D: Dual can act as Import or DirectQuery depending on the query path and can improve composite-model dimension behavior. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

E: Composite models can extend a remote semantic model with additional local model objects when supported. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

Learning point: Use Power BI Performance Analyzer to capture and compare visual query durations

Question 11

The marketing semantic model at Litware is moving from proof of concept to production. Before rollout, the data governance group must reduce the cost of a report page that renders many high-cardinality visuals at once, while ensuring that the design should minimize duplicated data. Which action best meets both needs? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Use incremental refresh and define a shorter refresh window than the historical retention window
Use a semantic-model star schema with dimensions filtering the fact table
Use Direct Lake storage mode
Apply the RangeStart and RangeEnd filters in a foldable transformation
Reduce unnecessary visuals or interactions and lower the amount of data each visual requests

Correct answer: E

Why: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This directly matches the stated requirement.

Option review:

B: Dimension-to-fact filtering aligns the model with common analytical patterns and reduces relationship ambiguity. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

E: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This directly matches the stated requirement.

Learning point: Reduce unnecessary visuals or interactions and lower the amount of data each visual requests

Question 12

Woodgrove Bank has a change request for the supply-chain lakehouse: improve model query efficiency by replacing a snowflaked reporting path with a simpler analytical relationship pattern. The enterprise reporting group wants a solution where least privilege must be preserved. Which implementation is most suitable? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Use a semantic-model star schema with dimensions filtering the fact table
Configure an incremental refresh policy with RangeStart and RangeEnd parameters
Use the supported Direct Lake refresh or framing operation
Optimize the model toward a clear star schema and efficient relationships
Use Dual storage mode for the compatible dimension

Correct answer: D

Why: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This directly matches the stated requirement.

Option review:

A: Dimension-to-fact filtering aligns the model with common analytical patterns and reduces relationship ambiguity. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

C: Direct Lake refresh or framing updates the model view of OneLake data rather than loading a conventional imported copy. This can be valid for ‘Configure Direct Lake, including default fallback and refresh behavior’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

D: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This directly matches the stated requirement.

Learning point: Optimize the model toward a clear star schema and efficient relationships

Question 13

A solution architect reviewing Coho Winery’s IoT telemetry solution asks the finance analytics squad to identify which report visual is issuing the slowest query before redesigning the page. Since the rollout must support controlled validation, which recommendation is strongest? The design decision will be reviewed by both data engineering and BI owners.

Use Dual storage mode for the compatible dimension
Use Import storage mode
Model a fact table surrounded by dimension tables with clear one-to-many relationships
Use DirectQuery storage mode
Use Power BI Performance Analyzer to capture and compare visual query durations

Correct answer: E

Why: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This directly matches the stated requirement.

Option review:

A: Dual can act as Import or DirectQuery depending on the query path and can improve composite-model dimension behavior. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

B: Import loads data into the VertiPaq in-memory engine and generally provides the fastest query performance when refresh latency is acceptable. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

D: DirectQuery leaves data in the source and sends queries at report time, trading source dependency for fresher access without a full import. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

E: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This directly matches the stated requirement.

Learning point: Use Power BI Performance Analyzer to capture and compare visual query durations

Question 14

The next sprint for Adventure Works’s finance reporting platform includes a task to reduce the cost of a report page that renders many high-cardinality visuals at once. The acceptance criteria add that the team wants predictable performance and behavior. Which Fabric or Power BI action is appropriate? The design decision will be reviewed by both data engineering and BI owners.

Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback
Reduce unnecessary visuals or interactions and lower the amount of data each visual requests
Introduce a bridge table at the customer-segment membership grain
Apply the RangeStart and RangeEnd filters in a foldable transformation
Use a semantic-model star schema with dimensions filtering the fact table

Correct answer: B

Why: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This directly matches the stated requirement.

Option review:

A: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This can be valid for ‘Configure Direct Lake, including default fallback and refresh behavior’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

B: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This directly matches the stated requirement.

C: A bridge table resolves a many-to-many business relationship into controlled one-to-many relationships for filtering and measures. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

E: Dimension-to-fact filtering aligns the model with common analytical patterns and reduces relationship ambiguity. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Implement performance improvements in queries and report visuals’.

Learning point: Reduce unnecessary visuals or interactions and lower the amount of data each visual requests

Question 15

For the supply-chain lakehouse, Trey Research has documented a business requirement to find the storage-engine and formula-engine work behind a slow measure. The data governance group must meet it in a way where the implementation should reduce operational overhead. What is the best choice? Existing users should keep their current access.

Create both relationships and use an inactive relationship for the secondary date when appropriate
Use DAX Studio or equivalent server-timing diagnostics to profile the measure
Use DirectQuery storage mode
Use incremental refresh and define a shorter refresh window than the historical retention window
Configure the semantic model to use large semantic model storage format

Correct answer: B

Why: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This directly matches the stated requirement.

Option review:

A: Role-playing date scenarios commonly use one active relationship and an inactive relationship activated explicitly in DAX when needed. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

B: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This directly matches the stated requirement.

C: DirectQuery leaves data in the source and sends queries at report time, trading source dependency for fresher access without a full import. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

D: The policy can retain a long history while refreshing only the recent period that is still subject to change. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

Learning point: Use DAX Studio or equivalent server-timing diagnostics to profile the measure

Question 16

A governance review of Alpine Ski House’s IoT telemetry solution asks for evidence that the solution can avoid recalculating the same expensive expression multiple times inside one measure. Because the solution must remain maintainable as usage grows, which action should be approved? Existing users should keep their current access.

Introduce a bridge table at the customer-segment membership grain
Use the supported Direct Lake refresh or framing operation
Build a composite model that uses the appropriate storage mode for each table
Use the appropriate DAX window function with an explicit ordering definition
Compute the expression once in a DAX variable and reuse it

Correct answer: E

Why: Variables can reduce repeated evaluation and also make the measure easier to optimize. This directly matches the stated requirement.

Option review:

A: A bridge table resolves a many-to-many business relationship into controlled one-to-many relationships for filtering and measures. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

B: Direct Lake refresh or framing updates the model view of OneLake data rather than loading a conventional imported copy. This can be valid for ‘Configure Direct Lake, including default fallback and refresh behavior’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

C: Composite models can combine tables that use different storage modes so each part of the model matches its latency and performance needs. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

D: DAX window functions support ordered row-window calculations without manually reconstructing the entire sequence logic. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

E: Variables can reduce repeated evaluation and also make the measure easier to optimize. This directly matches the stated requirement.

Learning point: Compute the expression once in a DAX variable and reuse it

Question 17

Before expanding the finance reporting platform, the finance analytics squad at Contoso must replace unnecessary row-by-row filtering over a large fact table with a simpler filter that the engine can optimize. The rollout plan says that the team wants the smallest administrative blast radius. Which option most directly addresses the requirement? Existing users should keep their current access.

Create both relationships and use an inactive relationship for the secondary date when appropriate
Use direct column filters or more efficient set-based DAX where semantics allow
Create a composite model over the remote semantic model and the local data
Use Power BI Performance Analyzer to capture and compare visual query durations
Choose Direct Lake on the SQL analytics endpoint

Correct answer: B

Why: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This directly matches the stated requirement.

Option review:

B: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This directly matches the stated requirement.

D: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

E: Direct Lake on SQL uses the SQL analytics endpoint for metadata and can fall back to DirectQuery in supported scenarios. This can be valid for ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

Learning point: Use direct column filters or more efficient set-based DAX where semantics allow

Question 18

Wingtip Toys is redesigning its retail performance dashboard. The analytics engineering team must find the storage-engine and formula-engine work behind a slow measure. In addition, the team wants to keep governance centralized. Which action is the best fit? No unrelated workspace or model permissions should be changed.

Use DAX Studio or equivalent server-timing diagnostics to profile the measure
Use incremental refresh and define a shorter refresh window than the historical retention window
Apply the RangeStart and RangeEnd filters in a foldable transformation
Use a supported composite model with Direct Lake on OneLake plus Import
Create a composite model over the remote semantic model and the local data

Correct answer: A

Why: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This directly matches the stated requirement.

Option review:

A: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This directly matches the stated requirement.

B: The policy can retain a long history while refreshing only the recent period that is still subject to change. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

C: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

D: Direct Lake on OneLake supports composite modeling with other compatible storage modes in supported scenarios. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

Learning point: Use DAX Studio or equivalent server-timing diagnostics to profile the measure

Question 19

During a design review for Proseware’s marketing semantic model, one requirement is non-negotiable: avoid recalculating the same expensive expression multiple times inside one measure. Because the approach should remain understandable to future maintainers, what should the operations data team implement? No unrelated workspace or model permissions should be changed.

Build a composite model that uses the appropriate storage mode for each table
Use a supported composite model with Direct Lake on OneLake plus Import
Compute the expression once in a DAX variable and reuse it
Reduce unnecessary visuals or interactions and lower the amount of data each visual requests
Choose Direct Lake on OneLake

Correct answer: C

Why: Variables can reduce repeated evaluation and also make the measure easier to optimize. This directly matches the stated requirement.

Option review:

A: Composite models can combine tables that use different storage modes so each part of the model matches its latency and performance needs. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

B: Direct Lake on OneLake supports composite modeling with other compatible storage modes in supported scenarios. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

C: Variables can reduce repeated evaluation and also make the measure easier to optimize. This directly matches the stated requirement.

D: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

E: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This can be valid for ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

Learning point: Compute the expression once in a DAX variable and reuse it

Question 20

The data governance group at Blue Yonder Airlines is preparing the next release of its supply-chain lakehouse. They need to replace unnecessary row-by-row filtering over a large fact table with a simpler filter that the engine can optimize; the implementation should reduce operational overhead. Which choice most directly satisfies the requirement? No unrelated workspace or model permissions should be changed.

Optimize the model toward a clear star schema and efficient relationships
Configure the semantic model to use large semantic model storage format
Use a supported composite model with Direct Lake on OneLake plus Import
Use Direct Lake storage mode
Use direct column filters or more efficient set-based DAX where semantics allow

Correct answer: E

Why: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This directly matches the stated requirement.

Option review:

A: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

C: Direct Lake on OneLake supports composite modeling with other compatible storage modes in supported scenarios. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

D: Direct Lake reads OneLake-backed Delta data for semantic-model queries without a conventional import copy. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

E: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This directly matches the stated requirement.

Learning point: Use direct column filters or more efficient set-based DAX where semantics allow

Question 21

A production readiness review at Fabrikam found a gap in the IoT telemetry solution. The remediation must find the storage-engine and formula-engine work behind a slow measure, and the solution must remain maintainable as usage grows. What is the most appropriate action? The team will validate the change first in a nonproduction environment.

Use incremental refresh and define a shorter refresh window than the historical retention window
Store the intermediate expression in a DAX VAR and reference the variable
Use DAX Studio or equivalent server-timing diagnostics to profile the measure
Use an iterator such as SUMX over the required table expression
Configure an incremental refresh policy with RangeStart and RangeEnd parameters

Correct answer: C

Why: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This directly matches the stated requirement.

Option review:

B: Variables reduce repeated expressions and make complex DAX easier to read and maintain. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

C: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This directly matches the stated requirement.

D: Iterator functions evaluate an expression for each row and then aggregate the results. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

Learning point: Use DAX Studio or equivalent server-timing diagnostics to profile the measure

Question 22

For a new phase of the finance reporting platform, Litware asks the finance analytics squad to avoid recalculating the same expensive expression multiple times inside one measure. The architecture decision record also states that the team wants the smallest administrative blast radius. Which approach should be selected? The team will validate the change first in a nonproduction environment.

Choose Direct Lake on OneLake
Use a dynamic format string
Use Power BI Performance Analyzer to capture and compare visual query durations
Use the appropriate DAX window function with an explicit ordering definition
Compute the expression once in a DAX variable and reuse it

Correct answer: E

Why: Variables can reduce repeated evaluation and also make the measure easier to optimize. This directly matches the stated requirement.

Option review:

A: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This can be valid for ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

B: Dynamic format strings change presentation without converting the measure result to text. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

C: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

E: Variables can reduce repeated evaluation and also make the measure easier to optimize. This directly matches the stated requirement.

Learning point: Compute the expression once in a DAX variable and reuse it

Question 23

Woodgrove Bank is standardizing how the retail performance dashboard is managed. The immediate goal is to replace unnecessary row-by-row filtering over a large fact table with a simpler filter that the engine can optimize. Given that the team wants to keep governance centralized, which option should the analytics engineering team choose? The team will validate the change first in a nonproduction environment.

Use Import storage mode
Use incremental refresh and define a shorter refresh window than the historical retention window
Use direct column filters or more efficient set-based DAX where semantics allow
Use Direct Lake storage mode
Configure the model for DirectLakeOnly behavior where supported

Correct answer: C

Why: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This directly matches the stated requirement.

Option review:

C: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This directly matches the stated requirement.

E: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This can be valid for ‘Configure Direct Lake, including default fallback and refresh behavior’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

Learning point: Use direct column filters or more efficient set-based DAX where semantics allow

Question 24

An internal audit of Coho Winery’s marketing semantic model identifies this requirement: find the storage-engine and formula-engine work behind a slow measure. The operations data team also notes that the approach should remain understandable to future maintainers. What should they do? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Optimize the model toward a clear star schema and efficient relationships
Use an iterator such as SUMX over the required table expression
Use a dynamic format string
Use DAX Studio or equivalent server-timing diagnostics to profile the measure
Use CALCULATE with an appropriate FILTER table expression

Correct answer: D

Why: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This directly matches the stated requirement.

Option review:

B: Iterator functions evaluate an expression for each row and then aggregate the results. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

D: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This directly matches the stated requirement.

E: FILTER can construct the row set while CALCULATE evaluates the measure under the resulting filter context. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

Learning point: Use DAX Studio or equivalent server-timing diagnostics to profile the measure

Question 25

The supply-chain lakehouse at Adventure Works is moving from proof of concept to production. Before rollout, the data governance group must avoid recalculating the same expensive expression multiple times inside one measure, while ensuring that the implementation should reduce operational overhead. Which action best meets both needs? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Optimize the model toward a clear star schema and efficient relationships
Create a composite model over the remote semantic model and the local data
Use Dual storage mode for the compatible dimension
Compute the expression once in a DAX variable and reuse it
Use a dynamic format string

Correct answer: D

Why: Variables can reduce repeated evaluation and also make the measure easier to optimize. This directly matches the stated requirement.

Option review:

B: Composite models can extend a remote semantic model with additional local model objects when supported. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

C: Dual can act as Import or DirectQuery depending on the query path and can improve composite-model dimension behavior. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

D: Variables can reduce repeated evaluation and also make the measure easier to optimize. This directly matches the stated requirement.

E: Dynamic format strings change presentation without converting the measure result to text. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

Learning point: Compute the expression once in a DAX variable and reuse it

Question 26

Tailspin Toys has a change request for the IoT telemetry solution: replace unnecessary row-by-row filtering over a large fact table with a simpler filter that the engine can optimize. The enterprise reporting group wants a solution where the solution must remain maintainable as usage grows. Which implementation is most suitable? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Configure the semantic model to use large semantic model storage format
Create a calculation group
Optimize the model toward a clear star schema and efficient relationships
Use direct column filters or more efficient set-based DAX where semantics allow
Reduce unnecessary visuals or interactions and lower the amount of data each visual requests

Correct answer: D

Why: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This directly matches the stated requirement.

Option review:

B: Calculation groups centralize reusable calculation logic and reduce measure proliferation. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

C: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

D: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This directly matches the stated requirement.

E: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

Learning point: Use direct column filters or more efficient set-based DAX where semantics allow

Question 27

A solution architect reviewing Fourth Coffee’s finance reporting platform asks the finance analytics squad to find the storage-engine and formula-engine work behind a slow measure. Since the team wants the smallest administrative blast radius, which recommendation is strongest? The design decision will be reviewed by both data engineering and BI owners.

Use DAX Studio or equivalent server-timing diagnostics to profile the measure
Configure an incremental refresh policy with RangeStart and RangeEnd parameters
Use the supported Direct Lake refresh or framing operation
Choose Direct Lake on OneLake
Use DirectQuery storage mode

Correct answer: A

Why: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This directly matches the stated requirement.

Option review:

A: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This directly matches the stated requirement.

E: DirectQuery leaves data in the source and sends queries at report time, trading source dependency for fresher access without a full import. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

Learning point: Use DAX Studio or equivalent server-timing diagnostics to profile the measure

Question 28

The next sprint for Wide World Importers’s retail performance dashboard includes a task to avoid recalculating the same expensive expression multiple times inside one measure. The acceptance criteria add that the team wants to keep governance centralized. Which Fabric or Power BI action is appropriate? The design decision will be reviewed by both data engineering and BI owners.

Compute the expression once in a DAX variable and reuse it
Create a composite model over the remote semantic model and the local data
Create both relationships and use an inactive relationship for the secondary date when appropriate
Introduce a bridge table at the customer-segment membership grain
Use an iterator such as SUMX over the required table expression

Correct answer: A

Why: Variables can reduce repeated evaluation and also make the measure easier to optimize. This directly matches the stated requirement.

Option review:

A: Variables can reduce repeated evaluation and also make the measure easier to optimize. This directly matches the stated requirement.

C: Role-playing date scenarios commonly use one active relationship and an inactive relationship activated explicitly in DAX when needed. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

D: A bridge table resolves a many-to-many business relationship into controlled one-to-many relationships for filtering and measures. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

E: Iterator functions evaluate an expression for each row and then aggregate the results. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Improve DAX performance’.

Learning point: Compute the expression once in a DAX variable and reuse it

Question 29

For the IoT telemetry solution, Wingtip Toys has documented a business requirement to prevent a Direct Lake on SQL model from silently switching to DirectQuery when a query cannot run in Direct Lake. The data governance group must meet it in a way where the solution should avoid unnecessary custom code. What is the best choice? Existing users should keep their current access.

Configure the model for DirectLakeOnly behavior where supported
Build a composite model that uses the appropriate storage mode for each table
Use incremental refresh and define a shorter refresh window than the historical retention window
Use a semantic-model star schema with dimensions filtering the fact table
Introduce a bridge table at the customer-segment membership grain

Correct answer: A

Why: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This directly matches the stated requirement.

Option review:

A: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This directly matches the stated requirement.

B: Composite models can combine tables that use different storage modes so each part of the model matches its latency and performance needs. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

C: The policy can retain a long history while refreshing only the recent period that is still subject to change. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

D: Dimension-to-fact filtering aligns the model with common analytical patterns and reduces relationship ambiguity. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

E: A bridge table resolves a many-to-many business relationship into controlled one-to-many relationships for filtering and measures. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

Learning point: Configure the model for DirectLakeOnly behavior where supported

Question 30

A governance review of Proseware’s finance reporting platform asks for evidence that the solution can make a Direct Lake model recognize newly committed Delta files and metadata changes without performing a traditional Import refresh. Because the design must preserve a clear development lifecycle, which action should be approved? Existing users should keep their current access.

Model a fact table surrounded by dimension tables with clear one-to-many relationships
Choose Direct Lake on the SQL analytics endpoint
Use a semantic-model star schema with dimensions filtering the fact table
Use the supported Direct Lake refresh or framing operation
Compute the expression once in a DAX variable and reuse it

Correct answer: D

Why: Direct Lake refresh or framing updates the model view of OneLake data rather than loading a conventional imported copy. This directly matches the stated requirement.

Option review:

A: A star schema gives the semantic model a predictable analytical structure and simplifies filter propagation. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

B: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This can be valid for ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

C: Dimension-to-fact filtering aligns the model with common analytical patterns and reduces relationship ambiguity. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

D: Direct Lake refresh or framing updates the model view of OneLake data rather than loading a conventional imported copy. This directly matches the stated requirement.

E: Variables can reduce repeated evaluation and also make the measure easier to optimize. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

Learning point: Use the supported Direct Lake refresh or framing operation

Question 31

Before expanding the retail performance dashboard, the finance analytics squad at Blue Yonder Airlines must understand why one Direct Lake on SQL query is suddenly slower after model changes. The rollout plan says that the team must avoid granting broader access than required. Which option most directly addresses the requirement? Existing users should keep their current access.

Use direct column filters or more efficient set-based DAX where semantics allow
Choose Direct Lake on the SQL analytics endpoint
Use Power BI Performance Analyzer to capture and compare visual query durations
Build a composite model that uses the appropriate storage mode for each table
Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback

Correct answer: E

Why: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This directly matches the stated requirement.

Option review:

A: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

B: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This can be valid for ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

D: Composite models can combine tables that use different storage modes so each part of the model matches its latency and performance needs. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

E: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This directly matches the stated requirement.

Learning point: Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback

Question 32

Fabrikam is redesigning its marketing semantic model. The analytics engineering team must prevent a Direct Lake on SQL model from silently switching to DirectQuery when a query cannot run in Direct Lake. In addition, the change must be easy to audit later. Which action is the best fit? No unrelated workspace or model permissions should be changed.

Apply the RangeStart and RangeEnd filters in a foldable transformation
Create a composite model over the remote semantic model and the local data
Configure the model for DirectLakeOnly behavior where supported
Build a composite model that uses the appropriate storage mode for each table
Use Dual storage mode for the compatible dimension

Correct answer: C

Why: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This directly matches the stated requirement.

Option review:

A: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

C: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This directly matches the stated requirement.

E: Dual can act as Import or DirectQuery depending on the query path and can improve composite-model dimension behavior. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

Learning point: Configure the model for DirectLakeOnly behavior where supported

Question 33

During a design review for Litware’s supply-chain lakehouse, one requirement is non-negotiable: make a Direct Lake model recognize newly committed Delta files and metadata changes without performing a traditional Import refresh. Because the implementation should be easy to troubleshoot, what should the operations data team implement? No unrelated workspace or model permissions should be changed.

Optimize the model toward a clear star schema and efficient relationships
Create a composite model over the remote semantic model and the local data
Compute the expression once in a DAX variable and reuse it
Use the supported Direct Lake refresh or framing operation
Use incremental refresh and define a shorter refresh window than the historical retention window

Correct answer: D

Why: Direct Lake refresh or framing updates the model view of OneLake data rather than loading a conventional imported copy. This directly matches the stated requirement.

Option review:

C: Variables can reduce repeated evaluation and also make the measure easier to optimize. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

D: Direct Lake refresh or framing updates the model view of OneLake data rather than loading a conventional imported copy. This directly matches the stated requirement.

E: The policy can retain a long history while refreshing only the recent period that is still subject to change. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

Learning point: Use the supported Direct Lake refresh or framing operation

Question 34

The data governance group at Woodgrove Bank is preparing the next release of its IoT telemetry solution. They need to understand why one Direct Lake on SQL query is suddenly slower after model changes; the solution should avoid unnecessary custom code. Which choice most directly satisfies the requirement? No unrelated workspace or model permissions should be changed.

Apply the RangeStart and RangeEnd filters in a foldable transformation
Use a semantic-model star schema with dimensions filtering the fact table
Use Direct Lake storage mode
Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback
Build a composite model that uses the appropriate storage mode for each table

Correct answer: D

Why: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This directly matches the stated requirement.

Option review:

A: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

B: Dimension-to-fact filtering aligns the model with common analytical patterns and reduces relationship ambiguity. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

C: Direct Lake reads OneLake-backed Delta data for semantic-model queries without a conventional import copy. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

D: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This directly matches the stated requirement.

E: Composite models can combine tables that use different storage modes so each part of the model matches its latency and performance needs. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

Learning point: Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback

Question 35

A production readiness review at Coho Winery found a gap in the finance reporting platform. The remediation must prevent a Direct Lake on SQL model from silently switching to DirectQuery when a query cannot run in Direct Lake, and the design must preserve a clear development lifecycle. What is the most appropriate action? The team will validate the change first in a nonproduction environment.

Use direct column filters or more efficient set-based DAX where semantics allow
Build a composite model that uses the appropriate storage mode for each table
Enable large semantic model storage format for the model
Configure the model for DirectLakeOnly behavior where supported
Choose Direct Lake on the SQL analytics endpoint

Correct answer: D

Why: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This directly matches the stated requirement.

Option review:

C: Large semantic model storage format is designed for supported capacities and models that need to exceed standard size limits. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

D: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This directly matches the stated requirement.

Learning point: Configure the model for DirectLakeOnly behavior where supported

Question 36

For a new phase of the retail performance dashboard, Adventure Works asks the finance analytics squad to make a Direct Lake model recognize newly committed Delta files and metadata changes without performing a traditional Import refresh. The architecture decision record also states that the team must avoid granting broader access than required. Which approach should be selected? The team will validate the change first in a nonproduction environment.

Choose Direct Lake on the SQL analytics endpoint
Use incremental refresh and define a shorter refresh window than the historical retention window
Use the supported Direct Lake refresh or framing operation
Enable large semantic model storage format for the model
Use Import storage mode

Correct answer: C

Why: Direct Lake refresh or framing updates the model view of OneLake data rather than loading a conventional imported copy. This directly matches the stated requirement.

Option review:

A: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This can be valid for ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

C: Direct Lake refresh or framing updates the model view of OneLake data rather than loading a conventional imported copy. This directly matches the stated requirement.

D: Large semantic model storage format is designed for supported capacities and models that need to exceed standard size limits. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

E: Import loads data into the VertiPaq in-memory engine and generally provides the fastest query performance when refresh latency is acceptable. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

Learning point: Use the supported Direct Lake refresh or framing operation

Question 37

Tailspin Toys is standardizing how the marketing semantic model is managed. The immediate goal is to understand why one Direct Lake on SQL query is suddenly slower after model changes. Given that the change must be easy to audit later, which option should the analytics engineering team choose? The team will validate the change first in a nonproduction environment.

Create a composite model over the remote semantic model and the local data
Create both relationships and use an inactive relationship for the secondary date when appropriate
Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback
Use an iterator such as SUMX over the required table expression
Enable large semantic model storage format for the model

Correct answer: C

Why: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This directly matches the stated requirement.

Option review:

A: Composite models can extend a remote semantic model with additional local model objects when supported. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

B: Role-playing date scenarios commonly use one active relationship and an inactive relationship activated explicitly in DAX when needed. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

C: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This directly matches the stated requirement.

E: Large semantic model storage format is designed for supported capacities and models that need to exceed standard size limits. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

Learning point: Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback

Question 38

An internal audit of Fourth Coffee’s supply-chain lakehouse identifies this requirement: prevent a Direct Lake on SQL model from silently switching to DirectQuery when a query cannot run in Direct Lake. The operations data team also notes that the implementation should be easy to troubleshoot. What should they do? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Use a semantic-model star schema with dimensions filtering the fact table
Configure the semantic model to use large semantic model storage format
Use a supported composite model with Direct Lake on OneLake plus Import
Configure the model for DirectLakeOnly behavior where supported
Create a field parameter

Correct answer: D

Why: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This directly matches the stated requirement.

Option review:

A: Dimension-to-fact filtering aligns the model with common analytical patterns and reduces relationship ambiguity. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

B: The large storage format supports enterprise-scale models and related management operations on supported capacity. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

D: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This directly matches the stated requirement.

E: Field parameters let consumers dynamically choose which modeled fields a visual uses. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

Learning point: Configure the model for DirectLakeOnly behavior where supported

Question 39

The IoT telemetry solution at Wide World Importers is moving from proof of concept to production. Before rollout, the data governance group must make a Direct Lake model recognize newly committed Delta files and metadata changes without performing a traditional Import refresh, while ensuring that the solution should avoid unnecessary custom code. Which action best meets both needs? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Use the appropriate DAX window function with an explicit ordering definition
Use the supported Direct Lake refresh or framing operation
Use an iterator such as SUMX over the required table expression
Use DirectQuery storage mode
Use Direct Lake storage mode

Correct answer: B

Why: Direct Lake refresh or framing updates the model view of OneLake data rather than loading a conventional imported copy. This directly matches the stated requirement.

Option review:

A: DAX window functions support ordered row-window calculations without manually reconstructing the entire sequence logic. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

B: Direct Lake refresh or framing updates the model view of OneLake data rather than loading a conventional imported copy. This directly matches the stated requirement.

C: Iterator functions evaluate an expression for each row and then aggregate the results. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

D: DirectQuery leaves data in the source and sends queries at report time, trading source dependency for fresher access without a full import. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

E: Direct Lake reads OneLake-backed Delta data for semantic-model queries without a conventional import copy. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

Learning point: Use the supported Direct Lake refresh or framing operation

Question 40

Northwind Traders has a change request for the finance reporting platform: understand why one Direct Lake on SQL query is suddenly slower after model changes. The enterprise reporting group wants a solution where the design must preserve a clear development lifecycle. Which implementation is most suitable? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Enable large semantic model storage format for the model
Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback
Compute the expression once in a DAX variable and reuse it
Use an iterator such as SUMX over the required table expression
Use CALCULATE with an appropriate FILTER table expression

Correct answer: B

Why: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This directly matches the stated requirement.

Option review:

A: Large semantic model storage format is designed for supported capacities and models that need to exceed standard size limits. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

B: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This directly matches the stated requirement.

Learning point: Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback

Question 41

A solution architect reviewing Trey Research’s retail performance dashboard asks the finance analytics squad to prevent a Direct Lake on SQL model from silently switching to DirectQuery when a query cannot run in Direct Lake. Since the team must avoid granting broader access than required, which recommendation is strongest? The design decision will be reviewed by both data engineering and BI owners.

Introduce a bridge table at the customer-segment membership grain
Create a composite model over the remote semantic model and the local data
Use Power BI Performance Analyzer to capture and compare visual query durations
Configure the model for DirectLakeOnly behavior where supported
Use a semantic-model star schema with dimensions filtering the fact table

Correct answer: D

Why: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This directly matches the stated requirement.

Option review:

D: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This directly matches the stated requirement.

E: Dimension-to-fact filtering aligns the model with common analytical patterns and reduces relationship ambiguity. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Configure Direct Lake, including default fallback and refresh behavior’.

Learning point: Configure the model for DirectLakeOnly behavior where supported

Question 42

For the executive reporting workspace, Coho Winery has documented a business requirement to require OneLake-native security behavior and a Direct Lake design that does not fall back to DirectQuery. The security analytics team must meet it in a way where the solution must remain maintainable as usage grows. What is the best choice? Existing users should keep their current access.

Choose Direct Lake on OneLake
Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback
Model a fact table surrounded by dimension tables with clear one-to-many relationships
Use DirectQuery storage mode
Create a field parameter

Correct answer: A

Why: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This directly matches the stated requirement.

Option review:

A: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This directly matches the stated requirement.

B: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This can be valid for ‘Configure Direct Lake, including default fallback and refresh behavior’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

C: A star schema gives the semantic model a predictable analytical structure and simplifies filter propagation. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

D: DirectQuery leaves data in the source and sends queries at report time, trading source dependency for fresher access without a full import. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

E: Field parameters let consumers dynamically choose which modeled fields a visual uses. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

Learning point: Choose Direct Lake on OneLake

Question 43

A governance review of Adventure Works’s customer 360 model asks for evidence that the solution can depend on security rules defined in the SQL analytics endpoint and need those delegated SQL rules to participate in access decisions. Because the team wants the smallest administrative blast radius, which action should be approved? Existing users should keep their current access.

Configure an incremental refresh policy with RangeStart and RangeEnd parameters
Apply the RangeStart and RangeEnd filters in a foldable transformation
Choose Direct Lake on the SQL analytics endpoint
Build a composite model that uses the appropriate storage mode for each table
Use a supported composite model with Direct Lake on OneLake plus Import

Correct answer: C

Why: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This directly matches the stated requirement.

Option review:

A: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

B: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

C: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This directly matches the stated requirement.

D: Composite models can combine tables that use different storage modes so each part of the model matches its latency and performance needs. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

E: Direct Lake on OneLake supports composite modeling with other compatible storage modes in supported scenarios. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

Learning point: Choose Direct Lake on the SQL analytics endpoint

Question 44

Before expanding the sales analytics solution, the customer insights team at Tailspin Toys must build the model from supported SQL analytics endpoint tables or views and accept possible DirectQuery fallback for unsupported Direct Lake cases. The rollout plan says that the team wants to keep governance centralized. Which option most directly addresses the requirement? Existing users should keep their current access.

Use Import storage mode
Optimize the model toward a clear star schema and efficient relationships
Use an iterator such as SUMX over the required table expression
Configure the semantic model to use large semantic model storage format
Choose Direct Lake on the SQL analytics endpoint

Correct answer: E

Why: Direct Lake on SQL uses the SQL analytics endpoint for metadata and can fall back to DirectQuery in supported scenarios. This directly matches the stated requirement.

Option review:

A: Import loads data into the VertiPaq in-memory engine and generally provides the fastest query performance when refresh latency is acceptable. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

B: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

C: Iterator functions evaluate an expression for each row and then aggregate the results. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

D: The large storage format supports enterprise-scale models and related management operations on supported capacity. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

E: Direct Lake on SQL uses the SQL analytics endpoint for metadata and can fall back to DirectQuery in supported scenarios. This directly matches the stated requirement.

Learning point: Choose Direct Lake on the SQL analytics endpoint

Question 45

Fourth Coffee is redesigning its risk analytics environment. The retail insights team must require OneLake-native security behavior and a Direct Lake design that does not fall back to DirectQuery. In addition, the approach should remain understandable to future maintainers. Which action is the best fit? No unrelated workspace or model permissions should be changed.

Reduce unnecessary visuals or interactions and lower the amount of data each visual requests
Use incremental refresh and define a shorter refresh window than the historical retention window
Model a fact table surrounded by dimension tables with clear one-to-many relationships
Use Direct Lake storage mode
Choose Direct Lake on OneLake

Correct answer: E

Why: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This directly matches the stated requirement.

Option review:

A: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

C: A star schema gives the semantic model a predictable analytical structure and simplifies filter propagation. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

E: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This directly matches the stated requirement.

Learning point: Choose Direct Lake on OneLake

Question 46

During a design review for Wide World Importers’s service-operations warehouse, one requirement is non-negotiable: depend on security rules defined in the SQL analytics endpoint and need those delegated SQL rules to participate in access decisions. Because the implementation should reduce operational overhead, what should the BI platform team implement? No unrelated workspace or model permissions should be changed.

Use DAX Studio or equivalent server-timing diagnostics to profile the measure
Use Dual storage mode for the compatible dimension
Choose Direct Lake on the SQL analytics endpoint
Use direct column filters or more efficient set-based DAX where semantics allow
Use the supported Direct Lake refresh or framing operation

Correct answer: C

Why: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This directly matches the stated requirement.

Option review:

A: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

B: Dual can act as Import or DirectQuery depending on the query path and can improve composite-model dimension behavior. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

C: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This directly matches the stated requirement.

D: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

E: Direct Lake refresh or framing updates the model view of OneLake data rather than loading a conventional imported copy. This can be valid for ‘Configure Direct Lake, including default fallback and refresh behavior’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

Learning point: Choose Direct Lake on the SQL analytics endpoint

Question 47

The security analytics team at Northwind Traders is preparing the next release of its executive reporting workspace. They need to build the model from supported SQL analytics endpoint tables or views and accept possible DirectQuery fallback for unsupported Direct Lake cases; the solution must remain maintainable as usage grows. Which choice most directly satisfies the requirement? No unrelated workspace or model permissions should be changed.

Compute the expression once in a DAX variable and reuse it
Choose Direct Lake on the SQL analytics endpoint
Optimize the model toward a clear star schema and efficient relationships
Configure the semantic model to use large semantic model storage format
Use Power BI Performance Analyzer to capture and compare visual query durations

Correct answer: B

Why: Direct Lake on SQL uses the SQL analytics endpoint for metadata and can fall back to DirectQuery in supported scenarios. This directly matches the stated requirement.

Option review:

A: Variables can reduce repeated evaluation and also make the measure easier to optimize. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

B: Direct Lake on SQL uses the SQL analytics endpoint for metadata and can fall back to DirectQuery in supported scenarios. This directly matches the stated requirement.

E: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

Learning point: Choose Direct Lake on the SQL analytics endpoint

Question 48

A production readiness review at Trey Research found a gap in the customer 360 model. The remediation must require OneLake-native security behavior and a Direct Lake design that does not fall back to DirectQuery, and the team wants the smallest administrative blast radius. What is the most appropriate action? The team will validate the change first in a nonproduction environment.

Choose Direct Lake on OneLake
Use Dual storage mode for the compatible dimension
Configure an incremental refresh policy with RangeStart and RangeEnd parameters
Enable large semantic model storage format for the model
Use the supported Direct Lake refresh or framing operation

Correct answer: A

Why: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This directly matches the stated requirement.

Option review:

A: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This directly matches the stated requirement.

C: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This can be valid for ‘Implement incremental refresh for semantic models’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

D: Large semantic model storage format is designed for supported capacities and models that need to exceed standard size limits. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

Learning point: Choose Direct Lake on OneLake

Question 49

For a new phase of the sales analytics solution, Alpine Ski House asks the customer insights team to depend on security rules defined in the SQL analytics endpoint and need those delegated SQL rules to participate in access decisions. The architecture decision record also states that the team wants to keep governance centralized. Which approach should be selected? The team will validate the change first in a nonproduction environment.

Configure the model for DirectLakeOnly behavior where supported
Configure the semantic model to use large semantic model storage format
Use DirectQuery storage mode
Choose Direct Lake on the SQL analytics endpoint
Optimize the model toward a clear star schema and efficient relationships

Correct answer: D

Why: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This directly matches the stated requirement.

Option review:

A: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This can be valid for ‘Configure Direct Lake, including default fallback and refresh behavior’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

B: The large storage format supports enterprise-scale models and related management operations on supported capacity. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

D: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This directly matches the stated requirement.

E: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

Learning point: Choose Direct Lake on the SQL analytics endpoint

Question 50

Contoso is standardizing how the risk analytics environment is managed. The immediate goal is to build the model from supported SQL analytics endpoint tables or views and accept possible DirectQuery fallback for unsupported Direct Lake cases. Given that the approach should remain understandable to future maintainers, which option should the retail insights team choose? The team will validate the change first in a nonproduction environment.

Create a calculation group
Choose Direct Lake on the SQL analytics endpoint
Use direct column filters or more efficient set-based DAX where semantics allow
Enable large semantic model storage format for the model
Use a dynamic format string

Correct answer: B

Why: Direct Lake on SQL uses the SQL analytics endpoint for metadata and can fall back to DirectQuery in supported scenarios. This directly matches the stated requirement.

Option review:

A: Calculation groups centralize reusable calculation logic and reduce measure proliferation. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

B: Direct Lake on SQL uses the SQL analytics endpoint for metadata and can fall back to DirectQuery in supported scenarios. This directly matches the stated requirement.

C: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

D: Large semantic model storage format is designed for supported capacities and models that need to exceed standard size limits. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

Learning point: Choose Direct Lake on the SQL analytics endpoint

Question 51

An internal audit of Wingtip Toys’s service-operations warehouse identifies this requirement: require OneLake-native security behavior and a Direct Lake design that does not fall back to DirectQuery. The BI platform team also notes that the implementation should reduce operational overhead. What should they do? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Use Direct Lake storage mode
Use Dual storage mode for the compatible dimension
Use a supported composite model with Direct Lake on OneLake plus Import
Choose Direct Lake on OneLake
Model a fact table surrounded by dimension tables with clear one-to-many relationships

Correct answer: D

Why: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This directly matches the stated requirement.

Option review:

A: Direct Lake reads OneLake-backed Delta data for semantic-model queries without a conventional import copy. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

D: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This directly matches the stated requirement.

E: A star schema gives the semantic model a predictable analytical structure and simplifies filter propagation. This can be valid for ‘Implement a star schema for a semantic model’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

Learning point: Choose Direct Lake on OneLake

Question 52

The executive reporting workspace at Proseware is moving from proof of concept to production. Before rollout, the security analytics team must depend on security rules defined in the SQL analytics endpoint and need those delegated SQL rules to participate in access decisions, while ensuring that the solution must remain maintainable as usage grows. Which action best meets both needs? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Optimize the model toward a clear star schema and efficient relationships
Build a composite model that uses the appropriate storage mode for each table
Use a supported composite model with Direct Lake on OneLake plus Import
Create both relationships and use an inactive relationship for the secondary date when appropriate
Choose Direct Lake on the SQL analytics endpoint

Correct answer: E

Why: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This directly matches the stated requirement.

Option review:

B: Composite models can combine tables that use different storage modes so each part of the model matches its latency and performance needs. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

D: Role-playing date scenarios commonly use one active relationship and an inactive relationship activated explicitly in DAX when needed. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

E: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This directly matches the stated requirement.

Learning point: Choose Direct Lake on the SQL analytics endpoint

Question 53

Blue Yonder Airlines has a change request for the customer 360 model: build the model from supported SQL analytics endpoint tables or views and accept possible DirectQuery fallback for unsupported Direct Lake cases. The Fabric center of excellence wants a solution where the team wants the smallest administrative blast radius. Which implementation is most suitable? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Create both relationships and use an inactive relationship for the secondary date when appropriate
Use CALCULATE with an appropriate FILTER table expression
Build a composite model that uses the appropriate storage mode for each table
Choose Direct Lake on the SQL analytics endpoint
Create a calculation group

Correct answer: D

Why: Direct Lake on SQL uses the SQL analytics endpoint for metadata and can fall back to DirectQuery in supported scenarios. This directly matches the stated requirement.

Option review:

B: FILTER can construct the row set while CALCULATE evaluates the measure under the resulting filter context. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

D: Direct Lake on SQL uses the SQL analytics endpoint for metadata and can fall back to DirectQuery in supported scenarios. This directly matches the stated requirement.

E: Calculation groups centralize reusable calculation logic and reduce measure proliferation. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

Learning point: Choose Direct Lake on the SQL analytics endpoint

Question 54

A solution architect reviewing Fabrikam’s sales analytics solution asks the customer insights team to require OneLake-native security behavior and a Direct Lake design that does not fall back to DirectQuery. Since the team wants to keep governance centralized, which recommendation is strongest? The design decision will be reviewed by both data engineering and BI owners.

Choose Direct Lake on OneLake
Use an iterator such as SUMX over the required table expression
Store the intermediate expression in a DAX VAR and reference the variable
Reduce unnecessary visuals or interactions and lower the amount of data each visual requests
Create both relationships and use an inactive relationship for the secondary date when appropriate

Correct answer: A

Why: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This directly matches the stated requirement.

Option review:

A: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This directly matches the stated requirement.

C: Variables reduce repeated expressions and make complex DAX easier to read and maintain. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

E: Role-playing date scenarios commonly use one active relationship and an inactive relationship activated explicitly in DAX when needed. This can be valid for ‘Implement relationships, such as bridge tables and many-to-many relationships’, but it does not directly satisfy the scenario requirement being tested under ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’.

Learning point: Choose Direct Lake on OneLake

Question 55

For the supply-chain lakehouse, Trey Research has documented a business requirement to refresh only recent partitions of a very large fact table while retaining older historical partitions. The data governance group must meet it in a way where the rollout must support controlled validation. What is the best choice? Existing users should keep their current access.

Choose Direct Lake on OneLake
Configure an incremental refresh policy with RangeStart and RangeEnd parameters
Create a composite model over the remote semantic model and the local data
Use Import storage mode
Configure the model for DirectLakeOnly behavior where supported

Correct answer: B

Why: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This directly matches the stated requirement.

Option review:

B: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This directly matches the stated requirement.

D: Import loads data into the VertiPaq in-memory engine and generally provides the fastest query performance when refresh latency is acceptable. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

Learning point: Configure an incremental refresh policy with RangeStart and RangeEnd parameters

Question 56

A governance review of Alpine Ski House’s IoT telemetry solution asks for evidence that the solution can reduce refresh duration for a table whose historical rows no longer change. Because the team wants predictable performance and behavior, which action should be approved? Existing users should keep their current access.

Reduce unnecessary visuals or interactions and lower the amount of data each visual requests
Optimize the model toward a clear star schema and efficient relationships
Use incremental refresh and define a shorter refresh window than the historical retention window
Use the appropriate DAX window function with an explicit ordering definition
Create a calculation group

Correct answer: C

Why: The policy can retain a long history while refreshing only the recent period that is still subject to change. This directly matches the stated requirement.

Option review:

A: Fewer, more focused visual queries reduce rendering and query workload, especially on high-cardinality pages. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

B: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

C: The policy can retain a long history while refreshing only the recent period that is still subject to change. This directly matches the stated requirement.

E: Calculation groups centralize reusable calculation logic and reduce measure proliferation. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

Learning point: Use incremental refresh and define a shorter refresh window than the historical retention window

Question 57

Before expanding the finance reporting platform, the finance analytics squad at Contoso must preserve query folding so the source, rather than the mashup engine, restricts each incremental partition. The rollout plan says that the choice should use a native Fabric capability. Which option most directly addresses the requirement? Existing users should keep their current access.

Apply the RangeStart and RangeEnd filters in a foldable transformation
Create a field parameter
Compute the expression once in a DAX variable and reuse it
Use Dual storage mode for the compatible dimension
Use a semantic-model star schema with dimensions filtering the fact table

Correct answer: A

Why: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This directly matches the stated requirement.

Option review:

A: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This directly matches the stated requirement.

B: Field parameters let consumers dynamically choose which modeled fields a visual uses. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

C: Variables can reduce repeated evaluation and also make the measure easier to optimize. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

Learning point: Apply the RangeStart and RangeEnd filters in a foldable transformation

Question 58

Wingtip Toys is redesigning its retail performance dashboard. The analytics engineering team must refresh only recent partitions of a very large fact table while retaining older historical partitions. In addition, the design should minimize duplicated data. Which action is the best fit? No unrelated workspace or model permissions should be changed.

Use DirectQuery storage mode
Use DAX Studio or equivalent server-timing diagnostics to profile the measure
Build a composite model that uses the appropriate storage mode for each table
Store the intermediate expression in a DAX VAR and reference the variable
Configure an incremental refresh policy with RangeStart and RangeEnd parameters

Correct answer: E

Why: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This directly matches the stated requirement.

Option review:

A: DirectQuery leaves data in the source and sends queries at report time, trading source dependency for fresher access without a full import. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

B: Server timings and query plans help identify whether a DAX expression is spending time in formula-engine iteration or storage-engine scans. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

D: Variables reduce repeated expressions and make complex DAX easier to read and maintain. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

E: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This directly matches the stated requirement.

Learning point: Configure an incremental refresh policy with RangeStart and RangeEnd parameters

Question 59

During a design review for Proseware’s marketing semantic model, one requirement is non-negotiable: reduce refresh duration for a table whose historical rows no longer change. Because least privilege must be preserved, what should the operations data team implement? No unrelated workspace or model permissions should be changed.

Use an iterator such as SUMX over the required table expression
Use Power BI Performance Analyzer to capture and compare visual query durations
Use incremental refresh and define a shorter refresh window than the historical retention window
Create a field parameter
Enable large semantic model storage format for the model

Correct answer: C

Why: The policy can retain a long history while refreshing only the recent period that is still subject to change. This directly matches the stated requirement.

Option review:

A: Iterator functions evaluate an expression for each row and then aggregate the results. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

B: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

C: The policy can retain a long history while refreshing only the recent period that is still subject to change. This directly matches the stated requirement.

D: Field parameters let consumers dynamically choose which modeled fields a visual uses. This can be valid for ‘Implement calculation groups, dynamic format strings, and field parameters’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

E: Large semantic model storage format is designed for supported capacities and models that need to exceed standard size limits. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

Learning point: Use incremental refresh and define a shorter refresh window than the historical retention window

Question 60

The data governance group at Blue Yonder Airlines is preparing the next release of its supply-chain lakehouse. They need to preserve query folding so the source, rather than the mashup engine, restricts each incremental partition; the rollout must support controlled validation. Which choice most directly satisfies the requirement? No unrelated workspace or model permissions should be changed.

Choose Direct Lake on the SQL analytics endpoint
Compute the expression once in a DAX variable and reuse it
Apply the RangeStart and RangeEnd filters in a foldable transformation
Use Dual storage mode for the compatible dimension
Use direct column filters or more efficient set-based DAX where semantics allow

Correct answer: C

Why: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This directly matches the stated requirement.

Option review:

A: Direct Lake on SQL is the choice when the model must depend on supported SQL analytics endpoint security semantics. This can be valid for ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

B: Variables can reduce repeated evaluation and also make the measure easier to optimize. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

C: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This directly matches the stated requirement.

E: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

Learning point: Apply the RangeStart and RangeEnd filters in a foldable transformation

Question 61

A production readiness review at Fabrikam found a gap in the IoT telemetry solution. The remediation must refresh only recent partitions of a very large fact table while retaining older historical partitions, and the team wants predictable performance and behavior. What is the most appropriate action? The team will validate the change first in a nonproduction environment.

Use direct column filters or more efficient set-based DAX where semantics allow
Use Dual storage mode for the compatible dimension
Configure an incremental refresh policy with RangeStart and RangeEnd parameters
Use CALCULATE with an appropriate FILTER table expression
Use the appropriate DAX window function with an explicit ordering definition

Correct answer: C

Why: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This directly matches the stated requirement.

Option review:

A: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

B: Dual can act as Import or DirectQuery depending on the query path and can improve composite-model dimension behavior. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

C: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This directly matches the stated requirement.

D: FILTER can construct the row set while CALCULATE evaluates the measure under the resulting filter context. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

E: DAX window functions support ordered row-window calculations without manually reconstructing the entire sequence logic. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

Learning point: Configure an incremental refresh policy with RangeStart and RangeEnd parameters

Question 62

For a new phase of the finance reporting platform, Litware asks the finance analytics squad to reduce refresh duration for a table whose historical rows no longer change. The architecture decision record also states that the choice should use a native Fabric capability. Which approach should be selected? The team will validate the change first in a nonproduction environment.

Enable large semantic model storage format for the model
Use the appropriate DAX window function with an explicit ordering definition
Use incremental refresh and define a shorter refresh window than the historical retention window
Create a field parameter
Store the intermediate expression in a DAX VAR and reference the variable

Correct answer: C

Why: The policy can retain a long history while refreshing only the recent period that is still subject to change. This directly matches the stated requirement.

Option review:

A: Large semantic model storage format is designed for supported capacities and models that need to exceed standard size limits. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

C: The policy can retain a long history while refreshing only the recent period that is still subject to change. This directly matches the stated requirement.

E: Variables reduce repeated expressions and make complex DAX easier to read and maintain. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

Learning point: Use incremental refresh and define a shorter refresh window than the historical retention window

Question 63

Woodgrove Bank is standardizing how the retail performance dashboard is managed. The immediate goal is to preserve query folding so the source, rather than the mashup engine, restricts each incremental partition. Given that the design should minimize duplicated data, which option should the analytics engineering team choose? The team will validate the change first in a nonproduction environment.

Use the appropriate DAX window function with an explicit ordering definition
Check whether the query is falling back to DirectQuery and remove the condition that triggers fallback
Enable large semantic model storage format for the model
Apply the RangeStart and RangeEnd filters in a foldable transformation
Use Power BI Performance Analyzer to capture and compare visual query durations

Correct answer: D

Why: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This directly matches the stated requirement.

Option review:

A: DAX window functions support ordered row-window calculations without manually reconstructing the entire sequence logic. This can be valid for ‘Write calculations that use DAX variables and functions, such as iterators, table filtering, windowing, and information functions’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

B: Direct Lake on SQL can fall back to DirectQuery for unsupported scenarios, which can materially change query performance. This can be valid for ‘Configure Direct Lake, including default fallback and refresh behavior’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

C: Large semantic model storage format is designed for supported capacities and models that need to exceed standard size limits. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

D: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This directly matches the stated requirement.

E: Performance Analyzer provides per-visual timing information that helps isolate slow visuals and DAX queries. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

Learning point: Apply the RangeStart and RangeEnd filters in a foldable transformation

Question 64

An internal audit of Coho Winery’s marketing semantic model identifies this requirement: refresh only recent partitions of a very large fact table while retaining older historical partitions. The operations data team also notes that least privilege must be preserved. What should they do? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Configure the model for DirectLakeOnly behavior where supported
Choose Direct Lake on OneLake
Configure an incremental refresh policy with RangeStart and RangeEnd parameters
Use Import storage mode
Use DirectQuery storage mode

Correct answer: C

Why: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This directly matches the stated requirement.

Option review:

A: DirectLakeOnly prevents DirectQuery fallback, causing unsupported queries to fail instead of silently using DirectQuery. This can be valid for ‘Configure Direct Lake, including default fallback and refresh behavior’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

B: Direct Lake on OneLake uses OneLake directly and operates without DirectQuery fallback. This can be valid for ‘Choose between Direct Lake on OneLake and Direct Lake on SQL analytics endpoint’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

C: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This directly matches the stated requirement.

Learning point: Configure an incremental refresh policy with RangeStart and RangeEnd parameters

Question 65

The supply-chain lakehouse at Adventure Works is moving from proof of concept to production. Before rollout, the data governance group must reduce refresh duration for a table whose historical rows no longer change, while ensuring that the rollout must support controlled validation. Which action best meets both needs? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Compute the expression once in a DAX variable and reuse it
Use DAX Studio or equivalent server-timing diagnostics to profile the measure
Use direct column filters or more efficient set-based DAX where semantics allow
Introduce a bridge table at the customer-segment membership grain
Use incremental refresh and define a shorter refresh window than the historical retention window

Correct answer: E

Why: The policy can retain a long history while refreshing only the recent period that is still subject to change. This directly matches the stated requirement.

Option review:

C: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

E: The policy can retain a long history while refreshing only the recent period that is still subject to change. This directly matches the stated requirement.

Learning point: Use incremental refresh and define a shorter refresh window than the historical retention window

Question 66

Tailspin Toys has a change request for the IoT telemetry solution: preserve query folding so the source, rather than the mashup engine, restricts each incremental partition. The enterprise reporting group wants a solution where the team wants predictable performance and behavior. Which implementation is most suitable? The solution must work with the current Fabric architecture rather than a parallel custom platform.

Apply the RangeStart and RangeEnd filters in a foldable transformation
Use DirectQuery storage mode
Use Power BI Performance Analyzer to capture and compare visual query durations
Use Direct Lake storage mode
Choose Direct Lake on OneLake

Correct answer: A

Why: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This directly matches the stated requirement.

Option review:

A: Query folding lets the source execute the partition predicates efficiently, which is important for scalable incremental refresh. This directly matches the stated requirement.

B: DirectQuery leaves data in the source and sends queries at report time, trading source dependency for fresher access without a full import. This can be valid for ‘Choose a storage mode’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

Learning point: Apply the RangeStart and RangeEnd filters in a foldable transformation

Question 67

A solution architect reviewing Fourth Coffee’s finance reporting platform asks the finance analytics squad to refresh only recent partitions of a very large fact table while retaining older historical partitions. Since the choice should use a native Fabric capability, which recommendation is strongest? The design decision will be reviewed by both data engineering and BI owners.

Configure an incremental refresh policy with RangeStart and RangeEnd parameters
Optimize the model toward a clear star schema and efficient relationships
Use direct column filters or more efficient set-based DAX where semantics allow
Create a composite model over the remote semantic model and the local data
Enable large semantic model storage format for the model

Correct answer: A

Why: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This directly matches the stated requirement.

Option review:

A: Incremental refresh partitions the table so scheduled refresh can process only the required recent range instead of reloading all history. This directly matches the stated requirement.

B: A well-designed star schema reduces relationship complexity and helps the engine resolve filters efficiently. This can be valid for ‘Implement performance improvements in queries and report visuals’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

C: Avoiding expensive iterator patterns over large tables can reduce formula-engine work and improve query performance. This can be valid for ‘Improve DAX performance’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

D: Composite models can extend a remote semantic model with additional local model objects when supported. This can be valid for ‘Design and build composite models’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

E: Large semantic model storage format is designed for supported capacities and models that need to exceed standard size limits. This can be valid for ‘Identify use cases for and configure large semantic model storage format’, but it does not directly satisfy the scenario requirement being tested under ‘Implement incremental refresh for semantic models’.

Learning point: Configure an incremental refresh policy with RangeStart and RangeEnd parameters

Question 1

Question 2

Question 3

Question 4

Question 5

Question 6

Question 7

Question 8

Question 9

Question 10

Question 11

Question 12

Question 13

Question 14

Question 15

Question 16

Question 17

Question 18

Question 19

Question 20

Question 21

Question 22

Question 23

Question 24

Question 25

Question 26

Question 27

Question 28

Question 29

Question 30

Question 31

Question 32

Question 33

Question 34

Question 35

Question 36

Question 37

Question 38

Question 39

Question 40

Question 41

Question 42

Question 43

Question 44

Question 45

Question 46

Question 47

Question 48

Question 49

Question 50

Question 51

Question 52

Question 53

Question 54

Question 55

Question 56

Question 57

Question 58

Question 59

Question 60

Question 61

Question 62

Question 63

Question 64

Question 65

Question 66

Question 67

Related posts: