Prepstellar

DP-600 · Mock exam · 25 questions

DP-600 mock exam: 25 free practice questions

Untimed here · timed and scored in the app

A free DP-600 practice test with 25 exam-style questions drawn from the whole course and weighted by domain like the exam guide. Answer at your pace and read why each option is right or wrong.

  • 700 or greaterPassing score
  • Maintain a data analytics solution 25–30%
  • Prepare data 45–50%
  • Implement and manage semantic models 25–30%

Exam details from the vendor’s exam guide, captured 2026-08-30.

Swipe, scroll or use ← →
  1. Q1 / 25

    What happens when City is added to both SELECT and GROUP BY in the quarterly sales example?

    1. AThe result shows a quarterly sales total for each city.
    2. BThe result ranks every city by list price within each year.
    3. CThe result returns ten arbitrary sales rows for each quarter.
    4. DThe result becomes one sales total for each product category.
    Show the answer

    Adding City as a selected grouping attribute refines the aggregation grain from quarterly totals to quarterly totals per city.

    Next → 1 / 25
  2. Q2 / 25

    What join behavior does Assume referential integrity request across a relationship?

    1. ALEFT OUTER joins instead of INNER joins
    2. BFULL OUTER joins instead of one-to-many relationships
    3. CCROSS joins instead of inactive relationship paths
    4. DINNER joins instead of LEFT OUTER joins
    Show the answer

    For Direct Lake and DirectQuery, the setting can reduce processed rows when every fact foreign key has a matching dimension key.

    Next → 2 / 25
  3. Q3 / 25

    A Sales Order column repeats an order identifier on multiple line items. A measure must return the number of business orders, not the number of lines. Which function fits?

    1. ADISTINCTCOUNT over the Sales Order column
    2. BSUM over the Sales Order column
    3. CCOUNTROWS over the Sales table
    4. DCOUNT over the Sales Order column
    Show the answer

    DISTINCTCOUNT collapses repeated order identifiers into one unique order count; COUNT and COUNTROWS retain the line-level multiplicity, while SUM adds values.

    Next → 3 / 25
  4. Q4 / 25

    What can a user with the Viewer workspace role do by default?

    1. AShare listed workspace content with underlying OneLake data access
    2. BView listed workspace content without underlying OneLake data access
    3. CCreate new workspace content with underlying SQL data access
    4. DModify listed workspace content without underlying SQL data access
    Show the answer

    Viewer is the read-oriented workspace role, but seeing listed Fabric items does not itself grant access to their underlying OneLake data.

    Next → 4 / 25
  5. Q5 / 25

    Which pair of Power Query operations is documented for handling null values?

    1. AFilter null-bearing rows or replace nulls with defaults
    2. BMerge null-bearing rows or append nulls as columns
    3. CPivot null-bearing rows or rename nulls as errors
    4. DGroup null-bearing rows or convert nulls into queries
    Show the answer

    The documented decision is between excluding rows through filtering and preserving rows by replacing null cells with approved defaults.

    Next → 5 / 25
  6. Q6 / 25

    What is the primary purpose of a calculation group?

    1. AReplace one source table with a calculated table
    2. BChoose one field from a group for a visual
    3. CApply one calculation pattern across multiple measures
    4. DDefine one numeric range for a report slicer
    Show the answer

    Calculation groups centralize reusable calculation logic in calculation items, while Fields parameters choose displayed fields and Numeric range parameters supply selectable numbers.

    Next → 6 / 25
  7. Q7 / 25

    Which clause filters groups after aggregate values have been calculated?

    1. ASELECT alias
    2. BINNER JOIN
    3. CWHERE
    4. DHAVING
    Show the answer

    HAVING applies conditions to grouped results, whereas WHERE filters individual rows before grouping begins.

    Next → 7 / 25
  8. Q8 / 25

    Which Git providers can a Fabric workspace connect to through Git integration?

    1. AFabric REST or XMLA endpoints
    2. BPower BI Desktop or local project folders
    3. CAzure DevOps or GitHub repositories
    4. DDevelopment or Test workspaces
    Show the answer

    Workspace Git integration uses Azure DevOps or GitHub as the repository provider; the other choices are interfaces, environments, or authoring locations rather than supported Git providers.

    Next → 8 / 25
  9. Q9 / 25

    A lakehouse must combine a table owned by another Fabric team with local tables. The source must remain in place and stay synchronized. What should you create?

    1. AA mirrored copy authenticated by an external cloud connection
    2. BAn internal OneLake shortcut to the source table
    3. CA report connected directly to the source semantic model
    4. DUpload a copy into the target
    Show the answer

    A shortcut gives the target lakehouse local-like access for combination while the table remains at its source; internal access still requires source read permission.

    Next → 9 / 25
  10. Q10 / 25

    How do the first and subsequent service refreshes differ under an incremental policy?

    1. AThe first removes retained history; later runs disable partitioning
    2. BThe first loads configured history; later runs update the recent window
    3. CThe first updates Desktop samples; later runs ignore service partitions
    4. DThe first loads only recent data; later runs reload all history
    Show the answer

    The initial service operation establishes the historical partitions, while later operations refresh the configured recent period and enforce retention.

    Next → 10 / 25
  11. Halfway, at your pace

    In the app the mock exam is timed and scored like the real thing.

  12. Q11 / 25

    What makes one stored procedure reusable for different monthly refresh periods?

    1. AA complete table clone created for each execution
    2. BA separate view definition created for each month
    3. CYear and month parameters supplied at execution
    4. DA fixed period embedded in every procedure statement
    Show the answer

    Parameters vary the target period without changing the saved procedure code; the procedure can then apply the same ordered processing to the supplied values.

    Next → 11 / 25
  13. Q12 / 25

    Which tool is designed for interactive DAX query testing against a published semantic model?

    1. ASemPy list_columns()
    2. BDAX Studio
    3. CTabular Editor BPA
    4. DALM Toolkit
    Show the answer

    DAX Studio supplies interactive query feedback; schema comparison, structural-rule analysis, and column inventory address different management tasks.

    Next → 12 / 25
  14. Q13 / 25

    Which destination pairing is supported when a Dataflow Gen2 query must produce a Fabric table?

    1. ADelta table in a lakehouse or schema-aware table in a warehouse
    2. BStored procedure in a lakehouse or M function in a warehouse
    3. CRead-only view in a lakehouse or temporary result in a warehouse
    4. DApplied step in a lakehouse or query preview in a warehouse
    Show the answer

    Dataflow destinations can load a query to a lakehouse Delta table or a warehouse table with schema support; the alternatives are authoring or code objects.

    Next → 13 / 25
  15. Q14 / 25

    When should the OneLake tables path be preferred over the SQL analytics endpoint?

    1. AWhen cross-database queries must combine several sources
    2. BWhen row-level security is defined in the SQL layer
    3. CWhen straightforward data resides in one Fabric data store
    4. DWhen SQL views are required as semantic-model sources
    Show the answer

    Direct OneLake tables are the simpler path for data in one place; the SQL analytics endpoint supplies the three capabilities named by the distractors.

    Next → 14 / 25
  16. Q15 / 25

    What does the Certified endorsement indicate?

    1. AA write user marked an item ready for one team's reuse
    2. BA designated user marked core data as the single source of truth
    3. CAn authorized reviewer verified organization-level quality standards
    4. DA Purview administrator assigned a sensitivity classification
    Show the answer

    Certification is the governed quality signal for broader use; Promoted, sensitivity labels, and Master data carry different meanings.

    Next → 15 / 25
  17. Q16 / 25

    Rows with a null AccountID cannot be used downstream and should not reach the destination. Which remediation meets the requirement?

    1. AFilter the query to remove rows with a null AccountID
    2. BExport the query before inspecting the AccountID field
    3. CMark the query as preview-only and keep every row
    4. DReplace null AccountID values with an arbitrary default
    Show the answer

    The requirement explicitly excludes unusable rows, so a null criterion should filter those records from the result.

    Next → 16 / 25
  18. Q17 / 25

    A model uses Delta tables in OneLake. Reports need current updates without scheduled refresh, while retaining fast interactive queries. Which storage mode should you choose first?

    1. ADirect Lake, because it reads the Delta tables without an Import refresh
    2. BImport, because its cached copy changes whenever the Delta tables change
    3. CDirectQuery, because it always provides the fastest in-memory query performance
    4. DComposite, because every Fabric-native model requires mixed table storage modes
    Show the answer

    For a Fabric-native source, the starting choice combines direct OneLake access with update visibility that does not depend on refreshing an imported copy.

    Next → 17 / 25
  19. Q18 / 25

    A shared semantic model column will be renamed, and several reports may consume it. What should the team do first?

    1. ARun impact analysis and review downstream consumers
    2. BUse lineage view only after reports have broken
    3. CRename the column and wait for refresh failures
    4. DPromote the model and review its commit authors
    Show the answer

    The proposed semantic-model schema change creates dependency risk, so the team should identify exposed report consumers before renaming the column.

    Next → 18 / 25
  20. Q19 / 25

    A warehouse fact load receives customer IDs from staging, but the fact table must reference warehouse-generated customer keys. Which loading pattern is appropriate?

    1. ACopy the source customer IDs into every descriptive dimension attribute
    2. BJoin staging to the customer dimension and insert the matched surrogate keys
    3. CInsert customer names into the measure columns and omit all keys
    4. DAggregate staging first and discard the customer relationship before loading
    Show the answer

    The fact load resolves each source natural key against its dimension and inserts the corresponding surrogate key with the event measurements.

    Next → 19 / 25
  21. Q20 / 25

    Which storage mode is the default for semantic models created in Microsoft Fabric?

    1. AComposite
    2. BDirectQuery
    3. CImport
    4. DDirect Lake
    Show the answer

    Fabric-native semantic models begin with the mode that reads OneLake Delta tables directly; the other modes address cached, source-query, or mixed requirements.

    Next → 20 / 25
  22. Q21 / 25

    Which workspace role can create and modify all content without adding sharing or permission-management capabilities?

    1. AMember
    2. BViewer workspace role
    3. CAdmin
    4. DContributor
    Show the answer

    Contributor supplies creation and modification while stopping below Member's sharing capability and Admin's permission management.

    Next → 21 / 25
  23. Q22 / 25

    A warehouse dashboard repeatedly reads monthly sales by region. Recomputing the summary from daily transactions is expensive, and the result should be stored. Which implementation fits?

    1. AUse a view so the daily aggregation is recalculated when queried
    2. BNormalize region attributes into additional tables before querying
    3. CUse CTAS to materialize a monthly regional aggregate table
    4. DUse a window function so each daily transaction row remains
    Show the answer

    A CTAS aggregation creates and populates the precomputed higher-grain table that recurring dashboard reads can consume.

    Next → 22 / 25
  24. Q23 / 25

    How are fact and dimension tables divided in a semantic star schema?

    1. AFacts hold refresh schedules; dimensions hold query translations
    2. BFacts hold descriptive context; dimensions hold measurable events
    3. CFacts hold storage modes; dimensions hold source connections
    4. DFacts hold measurable events; dimensions hold descriptive context
    Show the answer

    The star separates business-event measurements from the descriptive attributes that users employ to slice those measurements.

    Next → 23 / 25
  25. Q24 / 25

    How does the pipeline comparison identify content that differs between stages?

    1. AIt marks the item with a synced status
    2. BIt marks the item with a green icon
    3. CIt marks the item with an orange icon
    4. DIt marks the item with a branch name
    Show the answer

    The comparison uses orange for differences, green for identical items, and a new indicator for source-only items.

    Next → 24 / 25
  26. Q25 / 25

    A semantic model contains Import tables, DirectQuery tables, measures, and calculation groups. What does OneLake integration export?

    1. AThe Import-mode tables as Delta tables
    2. BThe complete model in Direct Lake mode
    3. CThe DirectQuery tables with their measures
    4. DThe measures and calculation-group tables
    Show the answer

    Eligibility is based on model object and storage mode: only Import tables are exported as Delta tables for downstream Fabric use.

    Next → 25 / 25
  27. That’s the whole mock exam

    Every question you miss comes back exactly when you’re about to forget it.

How to use this mock exam

Sit all 25 questions in one go: the mix covers every domain in the same proportion as the exam, so a low score points at the domain you skipped rather than at bad luck.

Read the explanation under every question, including the ones you got right — the reason an option is wrong is usually the thing being tested.

Then retake it in the app, where the mock exam is timed and scored and the questions you miss come back on a schedule.

The whole course, on your phone

Lessons you can read, audio you can listen to on the way to work, and practice that remembers what you got wrong.