Practice Exams:

Microsoft DP-900 Dumps

EXAM: Microsoft Azure Data Fundamentals
Verified By Experts
Premium Bundle
$40.00
$59.97 $99.97

DP-900 Premium Bundle


  • Premium File 332 Questions & Answers
  • Training Course 32 Lectures
  • Study Guide 672 Pages

Microsoft DP-900 Practice Test 84

Question 830

At Silverfield Healthcare approved clinical reports need documented data provenance. The team proposes this approach: Use versioned curated datasets and lineage metadata so analysts can trace transformations and reproduce reports. Which technical principle explains the choice?

  1. Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs.
  2. A measure is evaluated under the report's current filter context while a calculated column is stored per row after refresh.
  3. Windowed streaming computations aggregate events over bounded time intervals as new data arrives.
  4. Version and lineage records connect an analytical result to its source and processing logic at a point in time.

Answer & Explanations

Correct answer: D

A.

Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs. That explains a separate situation: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. The current case is: At Silverfield Healthcare approved clinical reports need documented data provenance. The actual mechanism is: Version and lineage records connect an analytical result to its source and processing logic at a point in time.

B.

A measure is evaluated under the report's current filter context while a calculated column is stored per row after refresh. That explains a separate situation: At Silverlake Services managers compare revenue by selected region. The current case is: At Silverfield Healthcare approved clinical reports need documented data provenance. The actual mechanism is: Version and lineage records connect an analytical result to its source and processing logic at a point in time.

C.

Windowed streaming computations aggregate events over bounded time intervals as new data arrives. That explains a separate situation: At Meadowview Commerce payment anomalies are monitored in rolling five-minute windows. The current case is: At Silverfield Healthcare approved clinical reports need documented data provenance. The actual mechanism is: Version and lineage records connect an analytical result to its source and processing logic at a point in time.

D.

Version and lineage records connect an analytical result to its source and processing logic at a point in time. This principle supports the intervention: Use versioned curated datasets and lineage metadata so analysts can trace transformations and reproduce reports. The relevant scenario is: At Silverfield Healthcare approved clinical reports need documented data provenance. Verify the result by: Recompute a fictional quarter's metric using the documented dataset version and compare totals.

Question 831

At Silverfield Healthcare approved clinical reports need documented data provenance. After implementing this approach—Use versioned curated datasets and lineage metadata so analysts can trace transformations and reproduce reports—which result would most directly confirm success?

  1. Send one valid and one invalid synthetic record and confirm only the valid record reaches the curated table.
  2. Recompute a fictional quarter's metric using the documented dataset version and compare totals.
  3. Delay a fictional event by two minutes and verify the configured lateness rule places it in the correct time window.
  4. Update a synthetic source record and verify the report reflects it after a configured refresh.

Answer & Explanations

Correct answer: B

A.

Send one valid and one invalid synthetic record and confirm only the valid record reaches the curated table. This tests a separate situation: At Meadowfield Retail some supplier feeds contain negative quantities. The observed case is: At Silverfield Healthcare approved clinical reports need documented data provenance. Use this acceptance check instead: Recompute a fictional quarter's metric using the documented dataset version and compare totals.

B.

Recompute a fictional quarter's metric using the documented dataset version and compare totals. This tests the targeted action: Use versioned curated datasets and lineage metadata so analysts can trace transformations and reproduce reports. The reason is: Version and lineage records connect an analytical result to its source and processing logic at a point in time. The reported case is: At Silverfield Healthcare approved clinical reports need documented data provenance.

C.

Delay a fictional event by two minutes and verify the configured lateness rule places it in the correct time window. This tests a separate situation: At Meadowport Transit buses upload buffered GPS events after reconnecting. The observed case is: At Silverfield Healthcare approved clinical reports need documented data provenance. Use this acceptance check instead: Recompute a fictional quarter's metric using the documented dataset version and compare totals.

D.

Update a synthetic source record and verify the report reflects it after a configured refresh. This tests a separate situation: At Pinecrest Finance a morning report refresh is sufficient for yesterday's closing totals. The observed case is: At Silverfield Healthcare approved clinical reports need documented data provenance. Use this acceptance check instead: Recompute a fictional quarter's metric using the documented dataset version and compare totals.

Question 832

At Silverfield Healthcare approved clinical reports need documented data provenance. Which shortcut most directly threatens the intended data outcome?

  1. Overwrite transformations without a change record while claiming prior KPI totals are reproducible.
  2. Depend on a worker to upload an uncontrolled CSV file once every morning.
  3. Paste a fixed grand total into every transaction row and expect slicers to recalculate it.
  4. Add an empty Fabric workspace and assume external live events start arriving without connectors.

Answer & Explanations

Correct answer: A

A.

Overwrite transformations without a change record while claiming prior KPI totals are reproducible. The scenario is: At Silverfield Healthcare approved clinical reports need documented data provenance. This undermines the mechanism: Version and lineage records connect an analytical result to its source and processing logic at a point in time. Instead: Use versioned curated datasets and lineage metadata so analysts can trace transformations and reproduce reports.

B.

Depend on a worker to upload an uncontrolled CSV file once every morning. This concerns another issue: At Westport Manufacturing factory transactions feed an overnight analytics store. In the stated case—At Silverfield Healthcare approved clinical reports need documented data provenance—reject: Overwrite transformations without a change record while claiming prior KPI totals are reproducible. The safer response is: Use versioned curated datasets and lineage metadata so analysts can trace transformations and reproduce reports.

C.

Paste a fixed grand total into every transaction row and expect slicers to recalculate it. This concerns another issue: At Silverlake Services managers compare revenue by selected region. In the stated case—At Silverfield Healthcare approved clinical reports need documented data provenance—reject: Overwrite transformations without a change record while claiming prior KPI totals are reproducible. The safer response is: Use versioned curated datasets and lineage metadata so analysts can trace transformations and reproduce reports.

D.

Add an empty Fabric workspace and assume external live events start arriving without connectors. This concerns another issue: At Cedarcrest Media live audience actions feed a Fabric real-time dashboard. In the stated case—At Silverfield Healthcare approved clinical reports need documented data provenance—reject: Overwrite transformations without a change record while claiming prior KPI totals are reproducible. The safer response is: Use versioned curated datasets and lineage metadata so analysts can trace transformations and reproduce reports.

Question 833

At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. Which design or Azure data approach best fits this requirement?

  1. Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.
  2. Choose event-time windows when late-arriving sensor measurements must be attributed to their occurrence time rather than receipt time.
  3. Use partitioned analytical files or tables for large date-filtered scans when data volume and query patterns justify it.
  4. Evaluate DirectQuery when Power BI must query a supported external source rather than importing all its data.

Answer & Explanations

Correct answer: A

A.

Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost. This works because: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs. The scenario is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. Verify the outcome by: Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.

B.

Choose event-time windows when late-arriving sensor measurements must be attributed to their occurrence time rather than receipt time. This addresses a different situation: At Meadowport Transit buses upload buffered GPS events after reconnecting. The present case is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. The relevant action here is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

C.

Use partitioned analytical files or tables for large date-filtered scans when data volume and query patterns justify it. This addresses a different situation: At Oakfield Utilities analysts frequently report one month of meter usage. The present case is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. The relevant action here is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

D.

Evaluate DirectQuery when Power BI must query a supported external source rather than importing all its data. This addresses a different situation: At Cedarfield Operations a governed warehouse remains the authority for interactive dashboard queries. The present case is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. The relevant action here is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

Question 834

At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. The team proposes this approach: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost. Which technical principle explains the choice?

  1. Extract-load-transform separates landing from later compute in a target platform capable of processing raw data.
  2. Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs.
  3. Position along a time axis makes change and seasonality easier to compare across ordered periods.
  4. Event time reflects when the source event happened; processing time reflects when the system saw it.

Answer & Explanations

Correct answer: B

A.

Extract-load-transform separates landing from later compute in a target platform capable of processing raw data. That explains a separate situation: At Meadowview Retail a cloud warehouse performs its own SQL cleaning after ingestion. The current case is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. The actual mechanism is: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs.

B.

Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs. This principle supports the intervention: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost. The relevant scenario is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. Verify the result by: Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.

C.

Position along a time axis makes change and seasonality easier to compare across ordered periods. That explains a separate situation: At Meadowport Media analysts inspect daily subscription signups. The current case is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. The actual mechanism is: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs.

D.

Event time reflects when the source event happened; processing time reflects when the system saw it. That explains a separate situation: At Meadowport Transit buses upload buffered GPS events after reconnecting. The current case is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. The actual mechanism is: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs.

Question 835

At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. After implementing this approach—Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost—which result would most directly confirm success?

  1. Build a fictional multi-page sales report and pin a selected KPI to a dashboard without losing the source report.
  2. Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.
  3. Send synthetic purchases across a five-minute interval and verify the window aggregate resets as designed.
  4. Update two fictional source records and confirm the second pipeline run imports only eligible changes.

Answer & Explanations

Correct answer: B

A.

Build a fictional multi-page sales report and pin a selected KPI to a dashboard without losing the source report. This tests a separate situation: At Harborland Banking executives monitor headline risk indicators and analysts need detail pages. The observed case is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. Use this acceptance check instead: Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.

B.

Compare a synthetic nightly batch with an hourly batch against the report's promised readiness. This tests the targeted action: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost. The reason is: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs. The reported case is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts.

C.

Send synthetic purchases across a five-minute interval and verify the window aggregate resets as designed. This tests a separate situation: At Meadowview Commerce payment anomalies are monitored in rolling five-minute windows. The observed case is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. Use this acceptance check instead: Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.

D.

Update two fictional source records and confirm the second pipeline run imports only eligible changes. This tests a separate situation: At Cedarbank Payments the warehouse must refresh recent settlements without reloading years of history. The observed case is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. Use this acceptance check instead: Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.

Question 836

At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. Which shortcut most directly threatens the intended data outcome?

  1. Schedule a single weekly batch while promising a dashboard accurate every ten minutes.
  2. Store a decade of evenly used time-series data in one undifferentiated daily-export file.
  3. Calculate a 'last five minutes' alert from a table refreshed once per day.
  4. Build separate disconnected charts each using a different definition of sales.

Answer & Explanations

Correct answer: A

A.

Schedule a single weekly batch while promising a dashboard accurate every ten minutes. The scenario is: At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts. This undermines the mechanism: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs. Instead: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

B.

Store a decade of evenly used time-series data in one undifferentiated daily-export file. This concerns another issue: At Oakfield Utilities analysts frequently report one month of meter usage. In the stated case—At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts—reject: Schedule a single weekly batch while promising a dashboard accurate every ten minutes. The safer response is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

C.

Calculate a 'last five minutes' alert from a table refreshed once per day. This concerns another issue: At Meadowview Commerce payment anomalies are monitored in rolling five-minute windows. In the stated case—At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts—reject: Schedule a single weekly batch while promising a dashboard accurate every ten minutes. The safer response is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

D.

Build separate disconnected charts each using a different definition of sales. This concerns another issue: At Harborbank Retail executives need sales dashboards sharing one definition of revenue. In the stated case—At Harborland Retail nightly inventory analytics satisfy finance but not emergency stock alerts—reject: Schedule a single weekly batch while promising a dashboard accurate every ten minutes. The safer response is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

Question 837

At Meadowbrook Media morning campaign reports need data from the previous day. Which design or Azure data approach best fits this requirement?

  1. Distinguish continuous streaming ingestion from scheduled batch imports by how promptly records must be processed.
  2. Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.
  3. Add validation and a quarantine destination for records failing an agreed data quality rule.
  4. Choose a line chart for a continuous time trend rather than using a pie chart for every timestamp.

Answer & Explanations

Correct answer: B

A.

Distinguish continuous streaming ingestion from scheduled batch imports by how promptly records must be processed. This addresses a different situation: At Cedarpoint Manufacturing equipment overheating requires prompt detection. The present case is: At Meadowbrook Media morning campaign reports need data from the previous day. The relevant action here is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

B.

Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost. This works because: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs. The scenario is: At Meadowbrook Media morning campaign reports need data from the previous day. Verify the outcome by: Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.

C.

Add validation and a quarantine destination for records failing an agreed data quality rule. This addresses a different situation: At Meadowfield Retail some supplier feeds contain negative quantities. The present case is: At Meadowbrook Media morning campaign reports need data from the previous day. The relevant action here is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

D.

Choose a line chart for a continuous time trend rather than using a pie chart for every timestamp. This addresses a different situation: At Meadowport Media analysts inspect daily subscription signups. The present case is: At Meadowbrook Media morning campaign reports need data from the previous day. The relevant action here is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

Question 838

At Meadowbrook Media morning campaign reports need data from the previous day. The team proposes this approach: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost. Which technical principle explains the choice?

  1. Partition pruning lets engines skip unrelated date segments instead of scanning all years of data.
  2. RLS filters model rows according to configured role rules and identity mapping rather than merely hiding visuals.
  3. Retries and event-time disorder can produce duplicate or tardy events that require stable IDs and window rules.
  4. Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs.

Answer & Explanations

Correct answer: D

A.

Partition pruning lets engines skip unrelated date segments instead of scanning all years of data. That explains a separate situation: At Oakfield Utilities analysts frequently report one month of meter usage. The current case is: At Meadowbrook Media morning campaign reports need data from the previous day. The actual mechanism is: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs.

B.

RLS filters model rows according to configured role rules and identity mapping rather than merely hiding visuals. That explains a separate situation: At Cedarcrest Finance regional managers should see only their branch transactions. The current case is: At Meadowbrook Media morning campaign reports need data from the previous day. The actual mechanism is: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs.

C.

Retries and event-time disorder can produce duplicate or tardy events that require stable IDs and window rules. That explains a separate situation: At Silverpoint Banking an event publisher retries after a brief timeout. The current case is: At Meadowbrook Media morning campaign reports need data from the previous day. The actual mechanism is: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs.

D.

Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs. This principle supports the intervention: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost. The relevant scenario is: At Meadowbrook Media morning campaign reports need data from the previous day. Verify the result by: Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.

Question 839

At Meadowbrook Media morning campaign reports need data from the previous day. After implementing this approach—Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost—which result would most directly confirm success?

  1. Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.
  2. Load fictional sales and finance facts and verify both reports share a consistent date dimension.
  3. Select a fictional product category and verify only related sales transactions contribute to the visual.
  4. Run two test consumer groups and confirm each sees the intended events without resetting the other's offset.

Answer & Explanations

Correct answer: A

A.

Compare a synthetic nightly batch with an hourly batch against the report's promised readiness. This tests the targeted action: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost. The reason is: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs. The reported case is: At Meadowbrook Media morning campaign reports need data from the previous day.

B.

Load fictional sales and finance facts and verify both reports share a consistent date dimension. This tests a separate situation: At Silvercrest Commerce executives compare sales performance across regions. The observed case is: At Meadowbrook Media morning campaign reports need data from the previous day. Use this acceptance check instead: Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.

C.

Select a fictional product category and verify only related sales transactions contribute to the visual. This tests a separate situation: At Harborfield Commerce sales facts reference a product dimension. The observed case is: At Meadowbrook Media morning campaign reports need data from the previous day. Use this acceptance check instead: Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.

D.

Run two test consumer groups and confirm each sees the intended events without resetting the other's offset. This tests a separate situation: At Oakcrest Mobility live tracking and reporting consume the same event source. The observed case is: At Meadowbrook Media morning campaign reports need data from the previous day. Use this acceptance check instead: Compare a synthetic nightly batch with an hourly batch against the report's promised readiness.

Question 840

At Meadowbrook Media morning campaign reports need data from the previous day. Which shortcut most directly threatens the intended data outcome?

  1. Deploy a static SQL query in a disconnected relational database and assume it automatically subscribes to the sensor stream.
  2. Schedule a single weekly batch while promising a dashboard accurate every ten minutes.
  3. Build separate disconnected charts each using a different definition of sales.
  4. Assume configuring a storage account alone creates a daily ingestion workflow.

Answer & Explanations

Correct answer: B

A.

Deploy a static SQL query in a disconnected relational database and assume it automatically subscribes to the sensor stream. This concerns another issue: At Pinebrook Facilities sensors send readings to an event hub. In the stated case—At Meadowbrook Media morning campaign reports need data from the previous day—reject: Schedule a single weekly batch while promising a dashboard accurate every ten minutes. The safer response is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

B.

Schedule a single weekly batch while promising a dashboard accurate every ten minutes. The scenario is: At Meadowbrook Media morning campaign reports need data from the previous day. This undermines the mechanism: Larger batch jobs can be efficient but delay availability, so cadence must match business SLAs. Instead: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

C.

Build separate disconnected charts each using a different definition of sales. This concerns another issue: At Harborbank Retail executives need sales dashboards sharing one definition of revenue. In the stated case—At Meadowbrook Media morning campaign reports need data from the previous day—reject: Schedule a single weekly batch while promising a dashboard accurate every ten minutes. The safer response is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

D.

Assume configuring a storage account alone creates a daily ingestion workflow. This concerns another issue: At Harborpoint Education SQL and file extracts feed a shared Azure reporting lake. In the stated case—At Meadowbrook Media morning campaign reports need data from the previous day—reject: Schedule a single weekly batch while promising a dashboard accurate every ten minutes. The safer response is: Select appropriate batch processing windows and compute scale based on data freshness requirements and job cost.

Microsoft Azure Data Fundamentals (DP-900) Practice Tests