cert
cert-1
cert-2

Pass Microsoft DP-700 Exam in First Attempt Guaranteed!

Get 100% Latest Exam Questions, Accurate & Verified Answers to Pass the Actual Exam!
30 Days Free Updates, Instant Download!

cert-5
cert-6
DP-700 Exam - Verified By Experts
DP-700 Premium Bundle
$14.99

DP-700 Premium Bundle

$74.99
$89.98
  • Premium File 160 Questions & Answers. Last update: Nov 10, 2026
  • Training Course 160 Video Lectures
 
$89.98
$74.99
accept 159 downloads in last 7 days
block-screenshots
Purchase Microsoft DP-700 Exam Training Products Individually
Microsoft DP-700 Premium File
Premium File 160 Q&A
$76.99$69.99
Microsoft DP-700 Video Training Course
Training Course 160 Lectures
$21.99 $19.99
Microsoft DP-700 Practice Test 29

Question 280

A distributed join runs for hours while CPU is uneven across tasks. Which metric should guide investigation?

  1. Open the Spark activity's inline monitoring details from the pipeline run.
  2. Track source and output row counts or defined quality metrics for the notebook transformation.
  3. Inspect per-task duration and shuffle/skew details in Spark stage monitoring.
  4. Use the Spark application/run ID and the matching driver and executor logs.

Explanation

Correct Answer

C:

Large task imbalances reveal partition skew or shuffle bottlenecks that overall runtime hides. Compare task duration percentiles and partition sizes.

Incorrect Answers

A:

Inline Spark monitoring connects orchestration activity with the underlying Spark application and stages.

B:

Execution success does not prove that business records were present or transformed correctly.

D:

Application IDs and logs tie an incident to its concrete execution and failure, avoiding confusion with concurrent jobs.

Question 281

An ETL notebook completes but its result is unexpectedly empty. What validation should supplement the success status?

  1. Use the Spark application/run ID and the matching driver and executor logs.
  2. Inspect per-task duration and shuffle/skew details in Spark stage monitoring.
  3. Track source and output row counts or defined quality metrics for the notebook transformation.
  4. Inspect notebook or Spark application logs and the failed cell or stage traceback.

Explanation

Correct Answer

C:

Execution success does not prove that business records were present or transformed correctly. Compare the output count to the source control total.

Incorrect Answers

A:

Application IDs and logs tie an incident to its concrete execution and failure, avoiding confusion with concurrent jobs.

B:

Large task imbalances reveal partition skew or shuffle bottlenecks that overall runtime hides.

D:

A job status is only a summary; logs reveal the exception and affected operation.

Question 282

A Spark driver was restarted and the team wants to associate its failure with a single execution. Which details matter?

  1. Track source and output row counts or defined quality metrics for the notebook transformation.
  2. Inspect per-task duration and shuffle/skew details in Spark stage monitoring.
  3. Open the Spark application details and stage metrics from the Fabric monitor experience.
  4. Use the Spark application/run ID and the matching driver and executor logs.

Explanation

Correct Answer

D:

Application IDs and logs tie an incident to its concrete execution and failure, avoiding confusion with concurrent jobs. Confirm the log timestamp and application ID align.

Incorrect Answers

A:

Execution success does not prove that business records were present or transformed correctly.

B:

Large task imbalances reveal partition skew or shuffle bottlenecks that overall runtime hides.

C:

Spark monitoring exposes tasks, stages, shuffles, and execution time rather than only notebook success.

Question 283

A Power BI semantic model must finish refresh before users see new data. What evidence confirms completion?

  1. Align orchestration so the semantic refresh follows completed transformation and load.
  2. Inspect the semantic model refresh history and its final status.
  3. Ensure model refresh is dependent on successful completion of upstream data preparation.
  4. Treat the pipeline activity starting the model refresh as confirmation that refresh finished successfully.

Explanation

Correct Answer

B:

Successful upstream ingestion alone does not prove the model has reloaded its cached or calculated state. Match the completed refresh time with the target model.

Incorrect Answers

A:

Refresh timing matters; a model can successfully capture old upstream data if run too early.

C:

Refreshing before curated tables are ready can expose inconsistent or old model data.

D:

The start event only confirms submission; the semantic model refresh can still fail or be running, so its completion status must be checked. Match the completed refresh time with the target model.

Question 284

An overnight model refresh intermittently times out. Where should operations compare past executions?

  1. Ensure model refresh is dependent on successful completion of upstream data preparation.
  2. Review semantic model refresh durations, outcomes, and detailed failure messages.
  3. Increase the refresh schedule frequency without reviewing the historical timeouts, duration trends or detailed failure messages.
  4. Verify the model's refresh or connection mode and the last successful refresh.

Explanation

Correct Answer

B:

Examine past semantic model refresh executions with their start and finish times, durations, outcomes and failure details. Compare successful runs against timeouts, then correlate the pattern with capacity, data volume, gateway and upstream dependency changes.

Incorrect Answers

A:

Refreshing before curated tables are ready can expose inconsistent or old model data. Refresh history exposes recurrent failures and timing patterns that are hidden by the latest report view. Compare failed and healthy refresh IDs.

C:

Changing the schedule before examining historical failures may intensify contention and does not explain the intermittent timeout. The investigator needs execution history and error details, not simply more attempts.

D:

Data landing and model visibility have separate freshness boundaries; rerunning ingestion may be unnecessary. Refresh history exposes recurrent failures and timing patterns that are hidden by the latest report view. Compare failed and healthy refresh IDs.

Question 285

A dashboard looks stale even though its lakehouse ingested current records. What should be checked before rerunning ingestion?

  1. Verify the model's refresh or connection mode and the last successful refresh.
  2. Ensure model refresh is dependent on successful completion of upstream data preparation.
  3. Restart the lakehouse ingestion immediately without checking the semantic model refresh or connection mode.
  4. Align orchestration so the semantic refresh follows completed transformation and load.

Explanation

Correct Answer

A:

Data landing and model visibility have separate freshness boundaries; rerunning ingestion may be unnecessary. Compare model and source timestamps.

Incorrect Answers

B:

Refreshing before curated tables are ready can expose inconsistent or old model data.

C:

Reingestion is not justified when source records already exist in the lakehouse; the semantic model may have a separate refresh or serving delay. Compare model and source timestamps.

D:

Refresh timing matters; a model can successfully capture old upstream data if run too early.

Question 286

A pipeline invokes a model refresh after loading a dimension. Which ordering condition is required?

  1. Review semantic model refresh durations, outcomes, and detailed failure messages.
  2. Use the recorded semantic model refresh status and error, not the trigger initiation result.
  3. Ensure model refresh is dependent on successful completion of upstream data preparation.
  4. Configure the supported model refresh schedule and monitor its completion history.

Explanation

Correct Answer

C:

Refreshing before curated tables are ready can expose inconsistent or old model data. Simulate a delayed dimension load in test.

Incorrect Answers

A:

Refresh history exposes recurrent failures and timing patterns that are hidden by the latest report view.

B:

A successfully sent refresh request does not establish that processing finished successfully.

D:

Imported model data needs its own refresh cadence and status checks independent of source ingestion.

Question 287

An analyst says the model was refreshed but the refresh job actually failed. What should be communicated?

  1. Ensure model refresh is dependent on successful completion of upstream data preparation.
  2. Use the recorded semantic model refresh status and error, not the trigger initiation result.
  3. Configure the supported model refresh schedule and monitor its completion history.
  4. Verify the model's refresh or connection mode and the last successful refresh.

Explanation

Correct Answer

B:

A successfully sent refresh request does not establish that processing finished successfully. Retrieve the terminal status for the requested refresh.

Incorrect Answers

A:

Refreshing before curated tables are ready can expose inconsistent or old model data.

C:

Imported model data needs its own refresh cadence and status checks independent of source ingestion.

D:

Data landing and model visibility have separate freshness boundaries; rerunning ingestion may be unnecessary.

Question 288

A semantic model uses import storage and should update nightly. Which scheduling and monitoring pair fits?

  1. Configure the supported model refresh schedule and monitor its completion history.
  2. Review semantic model refresh durations, outcomes, and detailed failure messages.
  3. Ensure model refresh is dependent on successful completion of upstream data preparation.
  4. Align orchestration so the semantic refresh follows completed transformation and load.

Explanation

Correct Answer

A:

Imported model data needs its own refresh cadence and status checks independent of source ingestion. Confirm the next scheduled refresh and output timestamp.

Incorrect Answers

B:

Refresh history exposes recurrent failures and timing patterns that are hidden by the latest report view.

C:

Refreshing before curated tables are ready can expose inconsistent or old model data.

D:

Refresh timing matters; a model can successfully capture old upstream data if run too early.

Question 289

A refreshed model still displays yesterday's numbers because the upstream transformation finished afterward. What should be corrected?

  1. Verify the model's refresh or connection mode and the last successful refresh.
  2. Align orchestration so the semantic refresh follows completed transformation and load.
  3. Configure the supported model refresh schedule and monitor its completion history.
  4. Inspect the semantic model refresh history and its final status.

Explanation

Correct Answer

B:

Refresh timing matters; a model can successfully capture old upstream data if run too early. Verify source completion precedes refresh start.

Incorrect Answers

A:

Data landing and model visibility have separate freshness boundaries; rerunning ingestion may be unnecessary.

C:

Imported model data needs its own refresh cadence and status checks independent of source ingestion.

D:

Successful upstream ingestion alone does not prove the model has reloaded its cached or calculated state.

Question 290

Operators need notification when a scheduled ingestion job fails. Which feature should they configure?

  1. Publish a validated data-quality metric and alert on the specified variance threshold.
  2. Require the threshold condition to persist for an approved observation window before triggering an alert.
  3. Set a supported Fabric monitoring failure alert or notification for the scheduled job.
  4. Detect execution-duration or SLA breach using supported monitoring telemetry and an alert rule.

Explanation

Correct Answer

C:

A failure alert moves the response from manual dashboard polling to event-driven operations. Cause a controlled failure and confirm notification delivery.

Incorrect Answers

A:

Business completeness is not implied by a pipeline's technical success, so the metric must be computed explicitly.

B:

A sustained-duration condition distinguishes persistent breaches from one-off spikes and reduces nuisance notifications.

D:

Duration-based detection can identify slow-but-not-failed runs that success/failure status alone misses.

Implementing Data Engineering Solutions Using Microsoft Fabric (DP-700) Practice Tests