Microsoft DP-700 Practice Test 15

Question 140

A downstream notebook starts even though the preceding Copy activity failed. The pipeline uses a completion dependency. What explains the execution?

An Upon completion dependency continues after both success and failure, rather than requiring success from the upstream activity.
The Copy activity has no retry interval, so the transform is allowed to start on success only.
The transform is bound to a success-only condition but the monitoring hub mislabels failure.
The upstream Copy and downstream notebook have identical activity names.

Explanation

Correct Answer

Completion describes finalization, not successful extraction. The downstream step needs the success condition to enforce this business rule.

Incorrect Answers

The configured Upon completion branch fires on success and failure. Retry timing cannot turn that dependency into an Upon success gate.

The scenario expressly identifies an Upon completion condition, which includes both terminal outcomes. There is no evidence of a monitoring labeling defect.

Name similarity is unrelated to dependency outcome semantics. The completion link, not activity naming, permitted the downstream run.

Question 141

A pipeline now includes a success-only dependency from Copy to transform. What failure test demonstrates its behavior?

Verify only that the downstream transformation runs after one successful Copy.
Cause Copy to fail in a controlled test and confirm the downstream transformation is skipped rather than executing on partial data.
Read the pipeline diagram but never execute a failing upstream activity.
Check the transformed table row count after a successful run.

Explanation

Correct Answer

Negative-path execution is essential evidence that the dependency enforces an upstream success condition.

Incorrect Answers

Running on success is compatible with both Upon success and Upon completion. A controlled failure is needed to distinguish them.

An apparently correct arrow does not prove the deployed pipeline correctly skips dependent work on Copy failure.

A count on a success path does not demonstrate that the transform is protected from processing incomplete data after an upstream failure.

Question 142

A Fabric pipeline should run a cleanup routine only when the source extraction fails, while a separate audit write should occur after any outcome. Which dependency selection is appropriate?

Link cleanup using Upon success and link audit using Upon failure.
Run cleanup on Upon completion and audit only on success.
Attach cleanup to the failure path and the audit step to the completion path, keeping each dependency outcome explicit.
Set both branches to Upon skipped and rely on a later retry to execute them.

Explanation

Correct Answer

Failure and completion dependencies have different trigger conditions and must be connected to their intended downstream work.

Incorrect Answers

This reverses the intended result: cleanup should not run after a healthy extraction and audit must also run after successful extraction.

Completion would execute exception cleanup on success, while the success-only audit would omit failed extraction attempts.

Skipped is not the desired failure/completion outcome, so this design would not reliably execute cleanup after failure or audit after every termination.

Question 143

An exception-cleanup notebook runs after every successful Copy, although it should run only on failure. Its dependency is Upon completion. Why?

The cleanup notebook is configured to run on the success branch but the pipeline redraws it under completion.
The completion branch is restricted to failures when its name contains Cleanup.
The audit write being unconditional causes the failure cleanup to inherit audit semantics.
Upon completion executes when the preceding activity finishes with either result; the cleanup requires the failure-specific dependency.

Explanation

Correct Answer

Completion is not synonymous with error handling, so the current edge also fires on successful Copy runs.

Incorrect Answers

The actual dependency is Upon completion; that condition legitimately executes after either outcome and explains the unwanted successful-run cleanup.

Activity labels cannot change the semantics of the dependency condition. The configured outcome type governs execution.

Each activity can have its own dependency condition. The cleanup runs because it is explicitly completion-linked, independently of the audit path.

Question 144

The ETL team changes cleanup to Upon failure and audit to Upon completion. Select THREE observations that validate the conditional graph.

Run only a success case and confirm both audit and cleanup executed.
Run only a failure case and assume the audit always executes on success.
On a successful extraction, the failure-only cleanup is skipped.
On a controlled failed extraction, the cleanup routine starts.
Inspect the activity names to infer the branch outcome without examining run states.
The completion-linked audit runs once the extraction has completed, regardless of whether extraction succeeded or failed.

Explanation

Correct Answers

This confirms that the cleanup branch no longer runs after a nonfailing prerequisite.

The negative-path test confirms that failure is a real trigger for the exception handler.

The audit dependency is deliberately outcome-independent and should record both terminal states.

Incorrect Answers

Executing cleanup on success is the defect. A valid acceptance check requires cleanup skipped for success and started for failure.

The two paths must be tested separately; a single failed extraction cannot demonstrate the audit-on-success requirement.

Names such as Cleanup or Audit do not establish actual trigger conditions. The execution graph’s success/failure/completion transitions must be observed.

Question 145

A pipeline receives a manifest with ten independent partition dates and must run Copy for each entry. Which activity pattern is suitable?

Use a ForEach activity whose items expression iterates the manifest array, invoking the Copy step with each current partition.
Pass the manifest as a string to one Copy activity without an iteration step.
Use a single Until loop without checking any manifest index or termination condition.
Set ForEach batch count to ten but leave its items expression empty.

Explanation

Correct Answer

ForEach consumes a collection and evaluates the iteration item, allowing one per-partition run without ten manually duplicated activities.

Incorrect Answers

A Copy source configured once cannot automatically bind each of ten independent partition dates. The manifest needs a supported ForEach items array.

An unbounded or condition-free loop does not guarantee one iteration per supplied manifest date and can repeat the wrong partition.

Batch count is a concurrency limit, not a supplier of data to iterate. The date array must be bound to the loop items.

Question 146

Only one partition is copied although the manifest contains ten. The pipeline ForEach items expression is bound to a constant array containing one value. Why?

Copy activity treats the entire JSON manifest as one source and automatically detects ten subpartitions.
The ForEach parallelism is too low to create multiple iterations at all.
ForEach is iterating the hard-coded one-element collection instead of the manifest parameter or lookup output.
The lakehouse SQL endpoint rejects array-typed parameters for the Copy sink.

Explanation

Correct Answer

The iteration count is determined by the evaluated items expression, not by the manifest existing elsewhere in the pipeline.

Incorrect Answers

A dataset reference is not an iteration over the manifest array. ForEach must use the correct items expression to bind each partition.

Concurrency controls how many existing iterations may run in parallel; it does not supply a missing manifest items binding.

The source problem is that the loop is not consuming the provided list of partition dates, rather than a sink SQL type check.

Question 147

After parameterizing ForEach over a partition manifest, which validation is useful?

Check that the manifest contains ten dates but do not run the ForEach activity.
Inspect only aggregate Copy bytes after one run, ignoring iteration inputs.
Set batch count to ten without confirming the loop’s items expression.
Run a controlled manifest with unique partitions and correlate ForEach iteration input and output records to ensure each expected partition is processed once.

Explanation

Correct Answer

Checking each iteration’s effective item and sink identity exposes dropped or duplicated partitions beyond the overall success status.

Incorrect Answers

A valid manifest alone does not prove the loop evaluates it or uses the current date inside each Copy instance.

A single aggregate can conceal repeated copying of the same partition or missing partitions. The effective item for each iteration must match the manifest.

Batch count limits concurrency; it does not verify that the loop received ten distinct partition identifiers.

Question 148

A pipeline must execute existing PySpark transformations with custom Python libraries and emit a Delta table before downstream validation. Which activity should run the code?

Use a Fabric Notebook activity wired into the pipeline dependencies and attach the required published Spark environment to that notebook.
Use Copy data alone and assume it executes the existing Python transformation cells.
Use a Dataflow Gen2 refresh without rewriting any of the required PySpark logic in Power Query.
Create a OneLake shortcut and assume its presence runs custom Spark transformations.

Explanation

Correct Answer

Notebook activities execute PySpark code and can be orchestrated alongside Copy and other pipeline activities without converting all transformations to Power Query.

Incorrect Answers

Copy is primarily a supported data movement mechanism and does not execute an arbitrary PySpark notebook with its custom libraries.

A dataflow executes Power Query transformations rather than the unchanged Python code specified in the scenario.

A shortcut makes an existing data path available but does not execute notebook code or materialize the transformed Delta output.

Question 149

A team migrates a complex PySpark transformation into a Dataflow Gen2 refresh with no equivalent implementation, and its custom Python logic never runs. Why?

Dataflow Gen2 automatically executes a notebook’s `%%pyspark` cells when given its file name.
Dataflow Gen2 executes Power Query transformations rather than arbitrary existing PySpark notebook code; a Notebook activity is needed for that code path.
The notebook’s output folder contains no Microsoft Purview endorsement.
The PySpark logic would run if the Dataflow Gen2 destination table used V-Order.

Explanation

Correct Answer

Selecting an engine incompatible with the transformation language leaves the intended Python logic outside the actual execution graph.

Incorrect Answers

Dataflow Gen2 uses Power Query transformations and does not act as a generic interpreter for existing custom PySpark code.

Governance metadata does not cause missing Spark computations in a dataflow. The unsupported execution-engine substitution is the critical mismatch.

Storage layout is separate from which transformation engine executes. V-Order cannot convert M/Power Query into Python code.

Question 150

An ingestion pipeline now runs a Notebook activity before a quality check. Which test proves that the intended Spark code executed?

Check only the pipeline activity run state without inspecting notebook execution artifacts.
Read the output Delta table without checking whether the Notebook activity ran in this invocation.
Inspect the notebook activity’s run and configured environment, then verify the expected Delta output exists before the dependent quality activity starts.
Confirm that Dataflow Gen2 has a refresh schedule even though the pipeline calls a notebook.

Explanation

Correct Answer

The notebook run evidence plus downstream data and dependency ordering tests actual execution, not only that a notebook item was linked.

Incorrect Answers

A green pipeline state can coexist with an empty or wrong Delta output; the actual notebook activity, environment and resulting table must be examined.

An old table may already exist. Correlating activity execution and current output is necessary before attributing the transformation to this run.

A dataflow schedule does not prove the custom PySpark notebook was invoked and completed before the quality dependency.

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

Microsoft DP-700 Practice Test 13 - 10 Questions (Last Reviewed 22-Oct-2026)
Microsoft DP-700 Practice Test 14 - 10 Questions (Last Reviewed 21-Oct-2026)
Microsoft DP-700 Practice Test 15 - 10 Questions (Last Reviewed 20-Oct-2026)
Microsoft DP-700 Practice Test 16 - 10 Questions (Last Reviewed 19-Oct-2026)
Microsoft DP-700 Practice Test 17 - 10 Questions (Last Reviewed 18-Oct-2026)

Pass Your Microsoft DP-700 Exam Easy!

Question 140

Explanation

Correct Answer

Incorrect Answers

Question 141

Explanation

Correct Answer

Incorrect Answers

Question 142

Explanation

Correct Answer

Incorrect Answers

Question 143

Explanation

Correct Answer

Incorrect Answers

Question 144

Explanation

Correct Answers

Incorrect Answers

Question 145

Explanation

Correct Answer

Incorrect Answers

Question 146

Explanation

Correct Answer

Incorrect Answers

Question 147

Explanation

Correct Answer

Incorrect Answers

Question 148

Explanation

Correct Answer

Incorrect Answers

Question 149

Explanation

Correct Answer

Incorrect Answers

Question 150

Explanation

Correct Answer

Incorrect Answers

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