Implementing Data Engineering Solutions Using Azure Databricks DP-750 Practice Test 68
Question 670
A production data platform experiences this condition: A daily revenue report double-counts refunds because both debit and credit rows are summed as positive numbers. Which targeted measure should be selected?
- Use a left anti-join or EXCEPT operation with aligned schemas and explicit null semantics.
- Define field-specific missing-value rules and quarantine or enrich unknown values before assigning business meaning.
- Use a compatible unpivot or stack transformation with explicit month and amount columns.
- Normalize transaction sign semantics before grouping by business date and customer.
Explanation
Correct Answer
D:The most appropriate action is: Normalize transaction sign semantics before grouping by business date and customer. It addresses the stated condition: A daily revenue report double-counts refunds because both debit and credit rows are summed as positive numbers. The technical basis is: Aggregations require correctly modeled financial direction and data types. Without this control, refund amounts treated as sales inflate net revenue even when record counts match. A direct acceptance check is: A test invoice with a linked refund produces the approved net amount.
Incorrect Answers
A:This alternative primarily addresses a different Azure Databricks problem: An analyst must find records present in the source system but absent in a target migration. The corresponding measure there would be: Use a left anti-join or EXCEPT operation with aligned schemas and explicit null semantics. The stated situation instead calls for: Normalize transaction sign semantics before grouping by business date and customer.
B:This alternative primarily addresses a different Azure Databricks problem: A customer reference dataset has missing country codes but a pipeline replaces every null with an arbitrary default country. The corresponding measure there would be: Define field-specific missing-value rules and quarantine or enrich unknown values before assigning business meaning. The stated situation instead calls for: Normalize transaction sign semantics before grouping by business date and customer.
C:This alternative primarily addresses a different Azure Databricks problem: A large orders dataset has one column per month but a dashboard expects month-value rows for filtering and grouping. The corresponding measure there would be: Use a compatible unpivot or stack transformation with explicit month and amount columns. The stated situation instead calls for: Normalize transaction sign semantics before grouping by business date and customer.
Question 671
A production data platform experiences this condition: A company needs to review table-owner changes and permissions after organizational restructuring. Which targeted measure should be selected?
- Add reviewed Unity Catalog table and column comments, owners and business definitions for consumer discovery.
- Use appropriate table history, schema change records and approved retention/audit evidence rather than relying on memory.
- Run a scheduled ownership and grant recertification process using catalog metadata and audit events.
- Develop an approved role-aware column mask and test its outputs across supported query clients.
Explanation
Correct Answer
C:The most appropriate action is: Run a scheduled ownership and grant recertification process using catalog metadata and audit events. It addresses the stated condition: A company needs to review table-owner changes and permissions after organizational restructuring. The technical basis is: Object governance is ongoing and must follow organizational role and service identity changes. Without this control, accumulated historical grants may give former project teams access to unrelated data products. A direct acceptance check is: The recertification report identifies owners, changed grants and approved remediation evidence.
Incorrect Answers
A:This alternative primarily addresses a different Azure Databricks problem: A business glossary has table names but no indication of which field is net revenue versus gross sales. The corresponding measure there would be: Add reviewed Unity Catalog table and column comments, owners and business definitions for consumer discovery. The stated situation instead calls for: Run a scheduled ownership and grant recertification process using catalog metadata and audit events.
B:This alternative primarily addresses a different Azure Databricks problem: A shared table includes an old regulated field and the business needs reproducible evidence of when it was removed. The corresponding measure there would be: Use appropriate table history, schema change records and approved retention/audit evidence rather than relying on memory. The stated situation instead calls for: Run a scheduled ownership and grant recertification process using catalog metadata and audit events.
D:This alternative primarily addresses a different Azure Databricks problem: A data product stores PII in a shared Delta table but several reports require different masking for each audience. The corresponding measure there would be: Develop an approved role-aware column mask and test its outputs across supported query clients. The stated situation instead calls for: Run a scheduled ownership and grant recertification process using catalog metadata and audit events.
Question 672
A platform architect adopts: Define a job-owned execution identity and eligible task compute separate from ad hoc analyst cluster lifecycle. For this requirement: A scheduled ETL run uses shared analyst compute and fails whenever the cluster owner restarts a notebook session. Which rationale MOST directly supports it?
- SQL warehouses are optimized for SQL analytics and isolate query serving from other compute activities.
- Spark and runtime upgrades can change supported APIs, dependency versions and execution behavior.
- Workspace compute permissions and access modes help maintain workload and security boundaries.
- Production job reliability benefits from stable identities and workload-specific compute ownership.
Explanation
Correct Answer
D:The decisive mechanism is: Production job reliability benefits from stable identities and workload-specific compute ownership. That explains the selected action: Define a job-owned execution identity and eligible task compute separate from ad hoc analyst cluster lifecycle. In the original scenario: A scheduled ETL run uses shared analyst compute and fails whenever the cluster owner restarts a notebook session. The specific unmanaged risk is: An analyst restart or entitlement change can interrupt a production data delivery. Validate with: A forced restart of interactive compute leaves the scheduled production task unaffected.
Incorrect Answers
A:This principle concerns a separate data-engineering condition: Analysts run short SQL dashboard queries against a shared lakehouse but their notebooks share a general-purpose Spark cluster with training experiments. Its targeted response is: Use an appropriately sized SQL warehouse with workload isolation and measured concurrency for dashboard queries. The current case is governed more directly by: Production job reliability benefits from stable identities and workload-specific compute ownership.
B:This principle concerns a separate data-engineering condition: A developer selects a new Databricks Runtime version for production despite an untested JDBC driver dependency. Its targeted response is: Pin a supported Databricks Runtime version and validate application libraries against it before deployment. The current case is governed more directly by: Production job reliability benefits from stable identities and workload-specific compute ownership.
C:This principle concerns a separate data-engineering condition: A multi-team interactive cluster permits users to attach arbitrary unapproved libraries and run notebooks under broad shared privileges. Its targeted response is: Apply suitable compute access modes, policies and entitlements that restrict who can attach or modify shared compute. The current case is governed more directly by: Production job reliability benefits from stable identities and workload-specific compute ownership.
Question 673
The design team proposes: Compare incoming event rate, processed records, watermark progress and state-store metrics before scaling. The situation is: A streaming pipeline's throughput falls although event arrival increases and its input backlog grows over time. Which technical principle justifies the decision?
- Maintenance operations themselves consume compute and should correspond to data growth and read patterns.
- Full scans caused by poor layout require data organization or predicate changes rather than only more compute.
- Sustained source-to-sink throughput below arrival rate causes accumulating data freshness debt.
- Different failure layers require different diagnostic paths instead of generic cluster restarts.
Explanation
Correct Answer
C:The decisive mechanism is: Sustained source-to-sink throughput below arrival rate causes accumulating data freshness debt. That explains the selected action: Compare incoming event rate, processed records, watermark progress and state-store metrics before scaling. In the original scenario: A streaming pipeline's throughput falls although event arrival increases and its input backlog grows over time. The specific unmanaged risk is: Increasing source throughput without repairing processing bottlenecks can worsen queue age. Validate with: A peak-load rehearsal stabilizes backlog and keeps end-to-end event age within target.
Incorrect Answers
A:This principle concerns a separate data-engineering condition: A new Delta table maintenance job combines repeated OPTIMIZE and VACUUM at every microbatch even though changes are tiny. Its targeted response is: Set a measured maintenance cadence or evaluate supported predictive optimization for eligible managed tables. The current case is governed more directly by: Sustained source-to-sink throughput below arrival rate causes accumulating data freshness debt.
B:This principle concerns a separate data-engineering condition: A warehouse query remains slow even after larger compute is allocated and the profile reports scanning nearly every file. Its targeted response is: Inspect query predicates and Delta data skipping or clustering layout before increasing warehouse size again. The current case is governed more directly by: Sustained source-to-sink throughput below arrival rate causes accumulating data freshness debt.
D:This principle concerns a separate data-engineering condition: An operations team sees Dataflow-like ingestion failures but cannot identify whether the problem is a Unity Catalog grant or a Spark stage. Its targeted response is: Correlate error messages, job run IDs, catalog authorization logs and Spark execution evidence before remediation. The current case is governed more directly by: Sustained source-to-sink throughput below arrival rate causes accumulating data freshness debt.
Question 674
A team omits the recommended measure: Use incremental file ingestion with explicit late-arrival cutoffs and a measurable source-to-target watermark. Its environment shows: A Databricks job receives daily structured extracts but an upstream schedule occasionally delivers files late. What is the MOST direct residual risk?
- Uncoordinated retries can run transformations before ingestion finishes or under the wrong identity.
- Too-short watermark drops legitimate delayed events; too-long state retention may exhaust memory.
- A bespoke long-running stream for a one-time load adds complexity and idle cost.
- A fixed midnight file list can omit valid delayed uploads.
Explanation
Correct Answer
D:The direct residual exposure is: A fixed midnight file list can omit valid delayed uploads. The triggering condition is: A Databricks job receives daily structured extracts but an upstream schedule occasionally delivers files late. The omitted safeguard is: Use incremental file ingestion with explicit late-arrival cutoffs and a measurable source-to-target watermark. This matters because: Data pipeline completeness requires tracking arrivals rather than assuming an exact wall-clock boundary. A corrective test would show: A deliberately delayed file is ingested exactly once within the allowed arrival window.
Incorrect Answers
A:This failure mode belongs mainly to another situation: A data product needs to copy a supported Azure Data Factory pipeline output into governed Databricks tables. Its preventative action would be: Coordinate ADF orchestration with the appropriate Databricks task and Unity Catalog access model. The most direct exposure in the current scenario is: A fixed midnight file list can omit valid delayed uploads.
B:This failure mode belongs mainly to another situation: A stream must group clicks into five-minute event-time windows while some mobile devices transmit data several minutes late. Its preventative action would be: Configure event-time windowing and an appropriate watermark based on observed lateness. The most direct exposure in the current scenario is: A fixed midnight file list can omit valid delayed uploads.
C:This failure mode belongs mainly to another situation: An integration performs a one-off initial import of an existing CSV folder into a Delta table. Its preventative action would be: Use a suitable COPY INTO or supported batch ingestion statement with validated format and path options. The most direct exposure in the current scenario is: A fixed midnight file list can omit valid delayed uploads.
Question 675
A production data team leaves out: Review explicit dependency and run-if settings and validate skip behavior through a staged failure. Its scenario is: A job owner cannot explain why a task skipped execution after a predecessor failed. Which failure mode is MOST relevant?
- A clock-only trigger can start before the dependency has produced the expected dataset.
- Mixed processing dates across tasks can create unreconciled reports and repeated records.
- Misunderstood skip conditions can leave expected cleanup or alerts unexecuted.
- Replaying an entire DAG without need can duplicate side effects and increase cost.
Explanation
Correct Answer
C:The direct residual exposure is: Misunderstood skip conditions can leave expected cleanup or alerts unexecuted. The triggering condition is: A job owner cannot explain why a task skipped execution after a predecessor failed. The omitted safeguard is: Review explicit dependency and run-if settings and validate skip behavior through a staged failure. This matters because: Job task state transitions follow dependency and conditional execution rules. A corrective test would show: A staged predecessor failure shows the approved downstream skip and cleanup path in run history.
Incorrect Answers
A:This failure mode belongs mainly to another situation: A pipeline must run when an upstream Databricks table is updated but analysts have hardcoded a daily time estimate. Its preventative action would be: Evaluate supported table update triggers or a clear upstream job dependency based on actual availability. The most direct exposure in the current scenario is: Misunderstood skip conditions can leave expected cleanup or alerts unexecuted.
B:This failure mode belongs mainly to another situation: A workflow needs to pass a date parameter from a scheduled trigger into all its downstream notebook tasks. Its preventative action would be: Define validated job parameters and propagate them through supported task parameter bindings. The most direct exposure in the current scenario is: Misunderstood skip conditions can leave expected cleanup or alerts unexecuted.
D:This failure mode belongs mainly to another situation: A team assumes that automatic job retries are the same as recovering a failed multi-step run at its last unsuccessful task. Its preventative action would be: Choose the appropriate retry and repair-run functionality based on failed-task state and valid dependencies. The most direct exposure in the current scenario is: Misunderstood skip conditions can leave expected cleanup or alerts unexecuted.
Question 676
The following control has NOT been implemented: Design a compatible atomic publication or versioned readiness contract so consumers see consistent results. In this case: A regulated pipeline needs consistent transaction boundaries across multiple table writes but updates are published one at a time. Which exposure remains?
- A bad transformation may make a broken dataset immediately visible to consumers.
- Readers can combine incompatible versions when related tables update independently.
- Full reruns can repeat expensive historical ingestion and duplicate side effects.
- A swallowed exception can allow downstream tasks to publish stale or missing outputs.
Explanation
Correct Answer
B:The direct residual exposure is: Readers can combine incompatible versions when related tables update independently. The triggering condition is: A regulated pipeline needs consistent transaction boundaries across multiple table writes but updates are published one at a time. The omitted safeguard is: Design a compatible atomic publication or versioned readiness contract so consumers see consistent results. This matters because: Multi-table orchestration requires a defined consistency point even if each individual Delta write is atomic. A corrective test would show: A controlled consumer test sees only one approved publication version across dependent datasets.
Incorrect Answers
A:This failure mode belongs mainly to another situation: A finance transformation writes its output directly into a published table before reconciling totals against the source. Its preventative action would be: Write to a staging target, validate row totals and invariants, then perform the approved publication step. The most direct exposure in the current scenario is: Readers can combine incompatible versions when related tables update independently.
C:This failure mode belongs mainly to another situation: A bulk repair executes all pipeline tasks again even when only the final reporting notebook failed. Its preventative action would be: Separate tasks and use the supported repair-run workflow at the failed stage when dependencies remain valid. The most direct exposure in the current scenario is: Readers can combine incompatible versions when related tables update independently.
D:This failure mode belongs mainly to another situation: A workflow uses a notebook task that catches exceptions and returns success even when a write step fails. Its preventative action would be: Propagate nonrecoverable exceptions to Lakeflow Jobs and attach actionable error metadata. The most direct exposure in the current scenario is: Readers can combine incompatible versions when related tables update independently.
Question 677
For this Microsoft DP-750 case: A data compliance team needs consistent redaction of columns tagged as personal data across hundreds of catalog tables. Which design or operational response BEST fits?
- Review and revoke stale grants and any linked service credentials, preserving required internal ownership.
- Check required parent catalog and schema usage privileges alongside the table SELECT grant.
- Use centrally governed tags and compatible Unity Catalog ABAC masking policies where supported.
- Configure a tested Unity Catalog row filter or suitable ABAC policy based on the authorized identity.
Explanation
Correct Answer
C:The most appropriate action is: Use centrally governed tags and compatible Unity Catalog ABAC masking policies where supported. It addresses the stated condition: A data compliance team needs consistent redaction of columns tagged as personal data across hundreds of catalog tables. The technical basis is: Attribute-based policies can apply common decisions at scale rather than managing every table individually. Without this control, manual per-table rules can miss newly created columns and expose sensitive values. A direct acceptance check is: A new tagged test column receives the required masking behavior without a separate manual table exception.
Incorrect Answers
A:This alternative primarily addresses a different Azure Databricks problem: A vendor's service principal was granted global access for a one-month integration and the project ended. The corresponding measure there would be: Review and revoke stale grants and any linked service credentials, preserving required internal ownership. The stated situation instead calls for: Use centrally governed tags and compatible Unity Catalog ABAC masking policies where supported.
B:This alternative primarily addresses a different Azure Databricks problem: A contractor receives SELECT on a table yet gets an access-denied error when querying with a three-part name. The corresponding measure there would be: Check required parent catalog and schema usage privileges alongside the table SELECT grant. The stated situation instead calls for: Use centrally governed tags and compatible Unity Catalog ABAC masking policies where supported.
D:This alternative primarily addresses a different Azure Databricks problem: A user must see only the East region rows of a common sales table while managers see all territories. The corresponding measure there would be: Configure a tested Unity Catalog row filter or suitable ABAC policy based on the authorized identity. The stated situation instead calls for: Use centrally governed tags and compatible Unity Catalog ABAC masking policies where supported.
Question 678
An operator introduces: Compare workload contention and resource reservations, then isolate or resize the critical job compute. The situation is: A production notebook intermittently times out only when many other jobs share the same classic cluster. Which validation should be required before signoff?
- A short benign delay does not page staff while a sustained backlog triggers the documented severity.
- A simulated failure yields archived logs, failed stage IDs and a documented root-cause path.
- The isolated benchmark meets its percentile duration under simultaneous unrelated workload pressure.
- A peak-load rehearsal stabilizes backlog and keeps end-to-end event age within target.
Explanation
Correct Answer
C:The strongest acceptance proof is: The isolated benchmark meets its percentile duration under simultaneous unrelated workload pressure. It specifically verifies: Compare workload contention and resource reservations, then isolate or resize the critical job compute. The original issue was: A production notebook intermittently times out only when many other jobs share the same classic cluster. The underlying mechanism is: Shared compute competition can cause unpredictable task latency independent of input correctness. Without it, A critical SLA may depend on unrelated teams' scheduling and resource usage.
Incorrect Answers
A:This result would validate a different measure: Define meaningful latency, error and backlog thresholds with alert routing based on the business SLO. That is relevant to the distinct problem: An alert triggers on every harmless microbatch delay while real hour-long ingestion stalls go unnoticed. Here the acceptance criterion is: The isolated benchmark meets its percentile duration under simultaneous unrelated workload pressure.
B:This result would validate a different measure: Preserve diagnostic events and Spark UI stage evidence before restart where possible. That is relevant to the distinct problem: An operator restarts a failed Databricks cluster before collecting the Spark executor and driver logs. Here the acceptance criterion is: The isolated benchmark meets its percentile duration under simultaneous unrelated workload pressure.
D:This result would validate a different measure: Compare incoming event rate, processed records, watermark progress and state-store metrics before scaling. That is relevant to the distinct problem: A streaming pipeline's throughput falls although event arrival increases and its input backlog grows over time. Here the acceptance criterion is: The isolated benchmark meets its percentile duration under simultaneous unrelated workload pressure.
Question 679
For this Azure Databricks issue: A daily data job runs twice during an incident and both executions write to the same target. Engineers choose: Configure appropriate maximum concurrent runs and idempotent target writes to control overlap. What acceptance evidence is BEST?
- A controlled dual-trigger test respects the run limit and produces one correct final target.
- The historical batch completes with reconciled counts and next daily scheduled run remains unchanged.
- A staged predecessor failure shows the approved downstream skip and cleanup path in run history.
- A test file arrives and starts one valid job run without repeated empty polling.
Explanation
Correct Answer
A:The strongest acceptance proof is: A controlled dual-trigger test respects the run limit and produces one correct final target. It specifically verifies: Configure appropriate maximum concurrent runs and idempotent target writes to control overlap. The original issue was: A daily data job runs twice during an incident and both executions write to the same target. The underlying mechanism is: Concurrency management prevents competing executions from corrupting shared mutable state. Without it, overlapping runs may race to overwrite outputs or duplicate appended records.
Incorrect Answers
B:This result would validate a different measure: Run a controlled parameterized manual invocation for the historical interval while retaining the regular trigger. That is relevant to the distinct problem: An ingestion job requires a one-time backfill without changing the future daily production schedule. Here the acceptance criterion is: A controlled dual-trigger test respects the run limit and produces one correct final target.
C:This result would validate a different measure: Review explicit dependency and run-if settings and validate skip behavior through a staged failure. That is relevant to the distinct problem: A job owner cannot explain why a task skipped execution after a predecessor failed. Here the acceptance criterion is: A controlled dual-trigger test respects the run limit and produces one correct final target.
D:This result would validate a different measure: Use a supported file-arrival trigger where its prerequisites and source location are compatible. That is relevant to the distinct problem: A cloud file feed needs a job triggered by arriving files instead of polling every minute. Here the acceptance criterion is: A controlled dual-trigger test respects the run limit and produces one correct final target.
Question 680
A production data team leaves out: Use query-profile evidence and a supported skew mitigation such as adaptive execution or key-aware repartitioning. Its scenario is: A DataFrame join spills heavily because a single join key occurs in most rows. Which failure mode is MOST relevant?
- A small number of hot tasks can dominate the join while extra workers remain idle.
- Unnecessary replay may duplicate writes and extend incident recovery time.
- Overlapping backlogs can amplify compute consumption and create conflicting writes.
- Increasing source throughput without repairing processing bottlenecks can worsen queue age.
Explanation
Correct Answer
A:The direct residual exposure is: A small number of hot tasks can dominate the join while extra workers remain idle. The triggering condition is: A DataFrame join spills heavily because a single join key occurs in most rows. The omitted safeguard is: Use query-profile evidence and a supported skew mitigation such as adaptive execution or key-aware repartitioning. This matters because: Severe key skew causes uneven task work even with many executors. A corrective test would show: Post-mitigation stage durations are balanced and joined row counts remain correct.
Incorrect Answers
B:This failure mode belongs mainly to another situation: A job fails during a temporary Azure storage outage but operators rerun every successful upstream task. Its preventative action would be: Use Lakeflow Jobs repair at the failed task when dependency outputs remain valid. The most direct exposure in the current scenario is: A small number of hot tasks can dominate the join while extra workers remain idle.
C:This failure mode belongs mainly to another situation: A job scheduled each hour starts running for 80 minutes and overlaps the next run. Its preventative action would be: Investigate stage runtime growth and enforce appropriate concurrency policy before raising frequency. The most direct exposure in the current scenario is: A small number of hot tasks can dominate the join while extra workers remain idle.
D:This failure mode belongs mainly to another situation: A streaming pipeline's throughput falls although event arrival increases and its input backlog grows over time. Its preventative action would be: Compare incoming event rate, processed records, watermark progress and state-store metrics before scaling. The most direct exposure in the current scenario is: A small number of hot tasks can dominate the join while extra workers remain idle.