DP-700 mock exam: 25 free practice questions

Q1 / 25

Which form preserves the numeric type of a parameter named `myNumber`?

1. A`@{pipeline().parameters.myNumber}`
2. B`@pipeline().parameters.myNumber`
3. C`Value: @@{pipeline().parameters.myNumber}`
4. D`Answer: @{pipeline().parameters.myNumber}`

Q2 / 25

Which window groups same-key events into an activity period that ends after a configured inactivity gap?

1. AA fixed micro-batch
2. BA session window
3. CA sliding window
4. DA tumbling window

Q3 / 25

An application must keep its existing queries and display the first character of each phone number followed by fixed padding, with no trailing digits exposed. Which implementation fits?

1. AApply a partial mask with prefix one and suffix zero
2. BApply an email mask with the `.com` suffix removed
3. CApply a random numeric mask with a one-digit range
4. DApply a default mask that returns the original first digit

Q4 / 25

A Dataflow Gen2 refresh succeeded, but destination output is lower than expected. The engineer needs both the affected entity and connector processing counts. Which investigation is most direct?

1. AOpen Historical runs and filter pipeline activities by workspace location
2. BOpen the refresh, select its table or activity, and review activity statistics
3. CRead the dataflow Status icon and export the initial history list
4. DOpen semantic model refresh details and compare Query Cache attempts

Q5 / 25

A user belongs to one role that omits a table and another role that grants the table. How does the OneLake grant model evaluate access?

1. AThe granting role still provides access to the table
2. BBoth roles are discarded because their scopes differ
3. CThe roles merge and expose only the table's first column
4. DThe role that omits the table denies access to it

Q6 / 25

Which external Delta table is within the documented query-acceleration column limit?

1. AA table containing 900 columns
2. BA table containing 1,200 columns
3. CA table containing 901 columns
4. DA table containing 2,500 columns

Q7 / 25

How long does Query Insights retain its historical query data?

1. AFifteen minutes
2. BTwenty-four hours
3. CThirty days
4. DSeven days

Q8 / 25

Which mask produces an email-shaped value that reveals the first letter and ends with `.com`?

1. AEmail mask
2. BDefault mask
3. CRandom mask
4. DPartial mask

Q9 / 25

What actions preserve full attribute history for a slowly changing dimension type 2 change?

1. ATruncate the dimension and regenerate every surrogate key.
2. BExpire the current version and insert a new current version.
3. COverwrite the current version and retain no prior value.
4. DDelete the current version and all facts that reference it.

Q10 / 25

What is the default notebook session timeout after inactivity?

1. A60 minutes
2. B20 minutes
3. C30 minutes
4. D10 minutes

Q11 / 25

Approved code has reached Dev. The organization requires automated checks in the next workspace, a manual release decision, and no production promotion before both are complete. Which sequence should the release use?

1. ARun checks against Dev, deploy Test and Prod together, then request approval.
2. BDeploy Dev to Test, run the checks, obtain approval, then deploy Test to Prod.
3. CDeploy Dev to Prod, run checks there, then copy the result back to Test.
4. DApprove the release in Dev, skip Test deployment, then deploy directly to Prod.

Q12 / 25

Which Power Query choice retains a row while substituting a default for its null value?

1. AReplace Values
2. BFilter out null rows
3. CRemove Duplicates
4. DRemove Errors

Q13 / 25

A semantic model refresh appears unusually long. The team must determine whether Power BI retried failures and whether data loaded before cache processing failed. What should they inspect?

1. AThe pipeline Gantt view and its activity Input links
2. BThe Dataflow Gen2 table list and destination byte counts
3. CThe monitoring hub owner filter and workspace location
4. DRefresh history attempts and their Data and Query Cache operations

Q14 / 25

How can a team automate a Fabric deployment pipeline from an existing release workflow?

1. ACall Fabric Items APIs without invoking the deployment pipeline.
2. BChange variable-library values and expect promotion to start automatically.
3. CCall deployment pipeline APIs from Azure Pipelines or GitHub workflows.
4. DCall Fabric Git APIs and treat repository sync as stage promotion.

Q15 / 25

A research platform must retain petabytes of Parquet, JSON, and image data. Engineers require PySpark transformations and both batch and real-time ingestion. Which store should own the data?

1. AA mirrored database providing a read-only replica
2. BA Lakehouse using Delta storage with Spark processing
3. CA SQL database using transactional relational storage
4. DA Warehouse using structured tables and broad T-SQL

Q16 / 25

A large query fails for lack of tempdb space during peak concurrency. Statistics were updated after the latest DML, but the query still has many GROUP BY and ORDER BY columns. Which response addresses both remaining constraints?

1. AIncrease user-configured tempdb capacity, then rerun the statement beside every competing query
2. BSwitch client tools, then use CTAS solely to test result-set delivery
3. CAdd more ordering columns, then pause and resume only the client application
4. DReduce grouping and ordering columns, then rerun when no other queries are active

Q17 / 25

What is the result when a OneLake row-level or column-level rule references a column that is absent from its table?

1. AThe query fails and returns no data
2. BThe rule becomes object security
3. CThe invalid condition is skipped and all data returns
4. DThe missing column is created with null values

Q18 / 25

How does a typical KQL transformation pipeline begin and continue?

1. AA materialized view followed by notebook magic commands
2. BA stored function followed by tables
3. CA table name followed by pipe-separated operators
4. DA SELECT keyword followed by a mandatory GROUP BY clause

Q19 / 25

A notebook reports excessive query complexity after many joins and aggregations. The transformations must remain, but the optimizer must handle a smaller plan at each stage. Which change is appropriate?

1. ASplit the logic and materialize intermediate results
2. BKeep one chain and refresh the authentication token
3. CKeep one chain and enable notebook AutoSave
4. DKeep one chain and lengthen the inactivity timeout

Q20 / 25

A task must move supported source data directly to a destination without transforming it. Which choice is the best fit?

1. AA notebook with custom Spark algorithms
2. BA Copy Data activity in a pipeline
3. CA control flow loop around variables
4. DA Dataflow Gen2 with merged queries

Q21 / 25

What happens to target data when its shortcut object is deleted?

1. AThe target data is converted automatically to Delta.
2. BThe target data and its storage remain unchanged.
3. CThe target data is deleted with the shortcut.
4. DThe target data is copied into the Lakehouse.

Q22 / 25

Which Query Insights view returns information about each completed SQL request?

1. Aqueryinsights.exec_requests_history
2. Bqueryinsights.sql_pool_insights
3. Cqueryinsights.frequently_run_queries
4. Dqueryinsights.exec_sessions_history

Q23 / 25

What does an empty Users field include in an audit search?

1. AOnly service accounts assigned to the current capacity
2. BOnly interactive users in the current Fabric workspace
3. CNo identities until an email address is entered
4. DAll users and service accounts in the organization

Q24 / 25

Which store natively combines structured and unstructured data with Spark processing?

1. AFabric Lakehouse
2. BSQL database
3. CData warehouse
4. DMirrored database

Q25 / 25

Which error metric indicates that incoming events do not match the expected readable format?

1. ADeserialization Errors
2. BData Conversion Errors
3. CRuntime Errors
4. DConnector Events Skipped