Microsoft DP-700 Practice Test 34

Question 330

A bad-event handling path is added to Eventstream. Which fixture validates it?

Send only valid JSON examples and conclude error handling is correct.
Compare the Eventstream name before and after deploying parser rules.
Send one valid JSON event and one deliberately malformed payload; confirm the valid record arrives and the bad record is rejected or quarantined with observable diagnostics.
Drop every message and check that zero parser errors remain.

Explanation

Correct Answer

A mixed-validity test proves the pipeline isolates malformed messages without dropping or misinterpreting all traffic.

Incorrect Answers

Without malformed input the quarantine/reject branch has never been exercised.

Name continuity provides no evidence that bad payloads are isolated and observable.

Deleting good data alongside bad data is not successful validation of selective error handling.

Question 331

A business needs telemetry less than 10 seconds old, while the source produces 50,000 messages each second. What should determine batch versus streaming architecture?

Choose an hourly batch run solely because it can transfer 50,000 messages per second when active.
Select Eventhouse just because its name sounds real-time, without benchmarking peak ingestion.
Discard queued events during congestion to make the reported freshness metric look faster.
Measure required event-to-query freshness and sustained throughput, then choose a supported Eventstream/Eventhouse or Structured Streaming design that meets both.

Explanation

Correct Answer

A nightly or hourly copy cannot meet second-level freshness even if its batch throughput is high; stream capacity and latency both matter.

Incorrect Answers

Throughput during a run doesn't meet a sub-ten-second SLA when the trigger interval is one hour.

The correct design must demonstrate the actual event-to-query latency and sustained source rate.

Silently dropping messages violates data completeness and doesn't establish a viable low-latency service.

Question 332

A team chooses hourly micro-batches for a sub-10-second alert SLA and users see stale dashboards. What fundamental mismatch explains the delay?

The batch trigger interval exceeds the freshness requirement by orders of magnitude; the pipeline lacks a continuously processing path.
The source Eventstream lacks a report title containing the word Real-Time.
The hourly job needs only a larger batch-copy concurrency value to become continuous.
The ingestion target is missing a native warehouse SQL index and so displays stale data.

Explanation

Correct Answer

Execution speed after the batch begins cannot make data visible before that batch is scheduled.

Incorrect Answers

Naming an item doesn't alter hourly scheduling or meet the ten-second event-to-query freshness SLA.

Faster execution once an hour cannot eliminate the long gap between batch triggers.

The dominant delay is scheduled triggering, not a query index causing hourly source silence.

Question 333

A continuous ingestion architecture is proposed. Which measurement distinguishes a viable streaming design from a faster batch process?

Measure the best-case event when it arrives immediately before the next scheduled batch.
Run representative peak-rate traffic and measure end-to-end event emission to query visibility, including backlog and recoverability during transient outages.
Compare a dashboard refresh timestamp without emitting unique new telemetry events.
Ignore test runs during peak traffic because normal load already meets the SLA.

Explanation

Correct Answer

The service level is not merely maximum rows per job; latency, sustained throughput and failure recovery are all part of the architecture choice.

Incorrect Answers

One favorable edge case conceals the long wait for events emitted just after a batch.

A dashboard can refresh stale data; event-to-query latency for new data is what matters.

A streaming design must sustain required throughput and freshness during representative peak volume.

Question 334

A time-series stream can replay a batch after failure. The derived aggregate should not double-count already committed event IDs. What output design should be used?

Delete durable streaming checkpoints before every restart, then append every window total again.
Use only the batch processing timestamp as an aggregate identity across retries.
Use durable streaming checkpoints and stable aggregate/window identities or idempotent sink writes so replay updates the intended bucket instead of appending a duplicate.
Assign a random new event ID and bucket key on each retry.

Explanation

Correct Answer

Checkpoint state and idempotent window keys allow a repeated input interval to converge to the same business aggregate.

Incorrect Answers

Losing offsets and appending duplicate windows causes double counting when a batch is replayed.

The same source event-time window may be replayed at a new processing time, producing duplicate logical buckets.

Unstable keys prevent idempotent merging of previously committed window aggregates.

Question 335

A replay of the same windowed telemetry batch inflates a previously correct daily total. Select THREE causes or safeguards that matter to this double-counting problem.

A dashboard was renamed after the first run.
The sink appends previously committed window results instead of matching a stable window/business key.
An unrelated lakehouse table was compacted with OPTIMIZE.
The stream restarted without its durable checkpoint, allowing previously processed source offsets to be read again.
Validate a controlled replay against unique event IDs and window keys before declaring the sink exactly-once at the business level.
A Fabric domain was assigned to the workspace.

Explanation

Correct Answers

Without an idempotent upsert boundary, identical aggregates from a retry can accumulate as duplicates.

Lost checkpoint progress increases the possibility that old events are emitted into the aggregate a second time.

A replay fixture verifies operational idempotence rather than assuming transport delivery semantics guarantee it.

Incorrect Answers

Dashboard metadata does not change stream offsets or the aggregate sink’s duplicate-write behavior.

Delta compaction does not create the repeated streaming output records in a different sink.

Domain organization cannot explain replay of an event-time aggregate.

Question 336

A nightly Data Factory Copy activity reports success but the target row count is unexpectedly low. What telemetry should the operator inspect first?

Open the specific activity run details and compare source rows read, sink rows written, mappings, filters and warnings.
Treat the pipeline's Succeeded status as proof all expected source records reached the sink.
Compare only the destination's last modified timestamp without reading Copy output counters.
Increase the lakehouse SQL analytics endpoint query timeout before examining Copy row metrics.

Explanation

Correct Answer

Activity input/output metrics pinpoint where records were filtered or rejected and provide more evidence than pipeline status alone.

Incorrect Answers

A run can complete its configured extraction successfully while filtering out or skipping eligible rows.

Timestamp freshness cannot isolate whether the source returned too few rows or the sink rejected them.

A read-only query timeout does not explain a Copy job that reports fewer source rows read.

Question 337

A pipeline finished green but only half the expected rows reached the lakehouse. Copy output shows fewer rows read than source totals. Where is the discrepancy likely introduced?

The Copy activity succeeded, so every row from the source must have been ingested.
The Copy source selection/filter or input extraction stage is narrower than the expected source dataset, before target writing.
The low row count proves the lakehouse SQL endpoint lost stored Delta rows.
The pipeline does not have a custom Spark environment for the Copy activity.

Explanation

Correct Answer

Rows missing from the activity input cannot be created by sink tuning, so the read-side row count is the first diagnostic boundary.

Incorrect Answers

A success status only means the configured transfer completed; source-row filters or partitions can still omit intended records.

When Copy reports fewer rows read, investigate the extraction and query predicate before blaming the destination SQL reader.

Copy runs with its connector/integration mechanisms, not a notebook Spark environment; that is not the source-count discrepancy.

Question 338

A low-row-count Copy incident is resolved. Which validation demonstrates both telemetry and data reconcile?

Validate only that the Copy activity status is green on the replay.
Compare only output file names without checking source predicates and row counts.
Run a controlled input selection and match activity rows read/written, target key counts and rejected-row/error detail for that run.
Inspect destination total rows using a privileged query but ignore the job's execution window.

Explanation

Correct Answer

Comparing execution metrics with the physical target confirms whether records reached the intended sink, not merely that a status badge turned green.

Incorrect Answers

Green status can coexist with incomplete extraction; compare rows read, written, skipped and source totals.

File names do not identify omitted source partitions or show all eligible rows were transferred.

A total can hide missing period-specific records; validate the exact source scope and run metrics.

Question 339

An overnight Dataflow Gen2 refresh is expected to update an analytics table, but the dashboard is stale. Where should operations check for a failed refresh?

Inspect only the Power Query preview and assume its presence proves scheduled destination refresh.
Refresh the dashboard visual manually without checking the failing Dataflow run.
Change the Dataflow's display name and trust its latest published definition.
Inspect Dataflow Gen2 recent run/refresh history and monitoring hub status, then drill into failing query and destination steps.

Explanation

Correct Answer

Refresh history reports start time, status and detailed logs, which reveal failures not evident from merely opening the saved dataflow.

Incorrect Answers

Preview execution is not the same as the scheduled Dataflow run and persisted output refresh.

Refreshing a report won't repair the upstream Dataflow Gen2 error or stale target table.

A published name/definition does not identify the failed source, step, gateway or destination activity.

Question 340

A dataflow has a valid saved Power Query definition, but its last scheduled refresh failed and downstream rows stayed old. What explains the discrepancy?

A saved dataflow definition is not evidence that the most recent refresh and destination write completed successfully.
The Dataflow's saved query definition automatically refreshes all published destinations.
The dashboard must be in a Fabric domain with the same name as the Dataflow.
The dataflow's preview used a small sample, which deletes all production destination rows.

Explanation

Correct Answer

Refresh execution has its own status and per-activity diagnostics; design-time validity alone does not update the output.

Incorrect Answers

Persisting Power Query steps doesn't mean a scheduled refresh succeeded, so source rows may stay unchanged.

Domain organization does not execute Power Query refreshes or guarantee updated table contents.

A small preview does not ordinarily delete destination tables; inspect the failed refresh error chain.

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

Microsoft DP-700 Practice Test 32 - 10 Questions (Last Reviewed 03-Oct-2026)
Microsoft DP-700 Practice Test 33 - 10 Questions (Last Reviewed 02-Oct-2026)
Microsoft DP-700 Practice Test 34 - 10 Questions (Last Reviewed 01-Oct-2026)
Microsoft DP-700 Practice Test 35 - 10 Questions (Last Reviewed 30-Sep-2026)
Microsoft DP-700 Practice Test 36 - 10 Questions (Last Reviewed 29-Sep-2026)

Pass Your Microsoft DP-700 Exam Easy!

Question 330

Explanation

Correct Answer

Incorrect Answers

Question 331

Explanation

Correct Answer

Incorrect Answers

Question 332

Explanation

Correct Answer

Incorrect Answers

Question 333

Explanation

Correct Answer

Incorrect Answers

Question 334

Explanation

Correct Answer

Incorrect Answers

Question 335

Explanation

Correct Answers

Incorrect Answers

Question 336

Explanation

Correct Answer

Incorrect Answers

Question 337

Explanation

Correct Answer

Incorrect Answers

Question 338

Explanation

Correct Answer

Incorrect Answers

Question 339

Explanation

Correct Answer

Incorrect Answers

Question 340

Explanation

Correct Answer

Incorrect Answers

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