Implementing Data Engineering Solutions Using Azure Databricks DP-750 Practice Test 75
Question 740
A data product owner adopts: Configure a tested Unity Catalog row filter or suitable ABAC policy based on the authorized identity. The reason is: A user must see only the East region rows of a common sales table while managers see all territories. What observed result BEST proves effectiveness?
- An East-region identity cannot read West-region test rows while the approved manager has the intended view.
- The pipeline authenticates using the designated service principal and operates after the employee account is disabled.
- Role-based sample queries show masked values for support and exact values only for approved security operators.
- A privilege inspection shows the narrow parent usage grants and the contractor can query only the intended table.
Explanation
Correct Answer
A:The strongest acceptance proof is: An East-region identity cannot read West-region test rows while the approved manager has the intended view. It specifically verifies: Configure a tested Unity Catalog row filter or suitable ABAC policy based on the authorized identity. The original issue was: A user must see only the East region rows of a common sales table while managers see all territories. The underlying mechanism is: Row-level security evaluates access rules at query time rather than relying on users to remember a WHERE clause. Without it, A user could retrieve other regions by querying the table without an application filter.
Incorrect Answers
B:This result would validate a different measure: Use a dedicated service principal with scoped Unity Catalog grants and governed credential lifecycle. That is relevant to the distinct problem: An ETL service principal is granted permissions through a personal employee account that will leave the company. Here the acceptance criterion is: An East-region identity cannot read West-region test rows while the approved manager has the intended view.
C:This result would validate a different measure: Apply a supported column mask or ABAC masking rule to the sensitive field and test different roles. That is relevant to the distinct problem: A customer service dashboard must show the last four digits of account identifiers but never full account values. Here the acceptance criterion is: An East-region identity cannot read West-region test rows while the approved manager has the intended view.
D:This result would validate a different measure: Check required parent catalog and schema usage privileges alongside the table SELECT grant. That is relevant to the distinct problem: A contractor receives SELECT on a table yet gets an access-denied error when querying with a three-part name. Here the acceptance criterion is: An East-region identity cannot read West-region test rows while the approved manager has the intended view.
Question 741
During a lakehouse implementation review, engineers observe: An existing table must be rebuilt from corrected historical files while its readers require a consistent atomic replacement. What is the BEST next action?
- Choose a controlled CREATE OR REPLACE TABLE or compatible atomic table-replacement strategy and validate schema/permissions.
- Deduplicate by stable event identity and business time using documented state and watermark rules.
- Implement idempotent foreachBatch or compatible transactional sink logic keyed by epoch or business identity.
- Use a supported change-data-capture feed and apply insert, update and delete semantics to the target.
Explanation
Correct Answer
A:The most appropriate action is: Choose a controlled CREATE OR REPLACE TABLE or compatible atomic table-replacement strategy and validate schema/permissions. It addresses the stated condition: An existing table must be rebuilt from corrected historical files while its readers require a consistent atomic replacement. The technical basis is: Atomic table replacement can avoid exposing a partially loaded intermediate table. Without this control, dropping then recreating tables may interrupt readers and lose associated governance settings. A direct acceptance check is: Readers observe a complete validated target before or after cutover, not partial data.
Incorrect Answers
B:This alternative primarily addresses a different Azure Databricks problem: An ingestion pipeline reads cloud events with at-least-once delivery and assumes every event ID appears only once. The corresponding measure there would be: Deduplicate by stable event identity and business time using documented state and watermark rules. The stated situation instead calls for: Choose a controlled CREATE OR REPLACE TABLE or compatible atomic table-replacement strategy and validate schema/permissions.
C:This alternative primarily addresses a different Azure Databricks problem: A stream writes into an external sink without stable batch identities and causes duplicate side effects on retry. The corresponding measure there would be: Implement idempotent foreachBatch or compatible transactional sink logic keyed by epoch or business identity. The stated situation instead calls for: Choose a controlled CREATE OR REPLACE TABLE or compatible atomic table-replacement strategy and validate schema/permissions.
D:This alternative primarily addresses a different Azure Databricks problem: A SaaS database updates and deletes customer records but a pipeline appends every nightly full extract to a Delta table. The corresponding measure there would be: Use a supported change-data-capture feed and apply insert, update and delete semantics to the target. The stated situation instead calls for: Choose a controlled CREATE OR REPLACE TABLE or compatible atomic table-replacement strategy and validate schema/permissions.
Question 742
A production data team leaves out: Grant only the required USE and SELECT privileges to the analyst group and inspect inherited grants. Its scenario is: A finance analyst must read an approved table but was granted broad MODIFY rights at the schema level. Which failure mode is MOST relevant?
- Manual per-table rules can miss newly created columns and expose sensitive values.
- Unnecessary modification rights can let a compromised analyst alter financial records.
- A user may bypass the masked presentation by reading the unmasked source table.
- A user could retrieve other regions by querying the table without an application filter.
Explanation
Correct Answer
B:The direct residual exposure is: Unnecessary modification rights can let a compromised analyst alter financial records. The triggering condition is: A finance analyst must read an approved table but was granted broad MODIFY rights at the schema level. The omitted safeguard is: Grant only the required USE and SELECT privileges to the analyst group and inspect inherited grants. This matters because: Unity Catalog privileges are scoped to securable objects and can be inherited from containers. A corrective test would show: The analyst can SELECT the authorized table but cannot UPDATE it or read an unrelated protected table.
Incorrect Answers
A:This failure mode belongs mainly to another situation: A data compliance team needs consistent redaction of columns tagged as personal data across hundreds of catalog tables. Its preventative action would be: Use centrally governed tags and compatible Unity Catalog ABAC masking policies where supported. The most direct exposure in the current scenario is: Unnecessary modification rights can let a compromised analyst alter financial records.
C:This failure mode belongs mainly to another situation: A data owner builds a masked view but analysts can still query the underlying sensitive table directly. Its preventative action would be: Revoke inappropriate base-table access and publish only the authorized view or policy-governed object. The most direct exposure in the current scenario is: Unnecessary modification rights can let a compromised analyst alter financial records.
D:This failure mode belongs mainly to another situation: A user must see only the East region rows of a common sales table while managers see all territories. Its preventative action would be: Configure a tested Unity Catalog row filter or suitable ABAC policy based on the authorized identity. The most direct exposure in the current scenario is: Unnecessary modification rights can let a compromised analyst alter financial records.
Question 743
A team omits the recommended measure: Run end-to-end contract tests for dependent readers and obtain owner approval for breaking changes. Its environment shows: A sensitive SQL view is renamed in the release but the team does not test downstream dashboards before promotion. What is the MOST direct residual risk?
- A faulty last-minute edit can create an untested release with no accountable reviewer.
- A release may finish successfully while critical reports cannot resolve their expected view.
- Duplicate jobs may run on the same schedule and load data twice.
- Unverified MERGE conditions can overwrite valid history despite passing mocked unit tests.
Explanation
Correct Answer
B:The direct residual exposure is: A release may finish successfully while critical reports cannot resolve their expected view. The triggering condition is: A sensitive SQL view is renamed in the release but the team does not test downstream dashboards before promotion. The omitted safeguard is: Run end-to-end contract tests for dependent readers and obtain owner approval for breaking changes. This matters because: Schema and object naming changes can break consumers even when new ETL succeeds. A corrective test would show: A staging dashboard loads its documented fields and a breaking-change simulation is rejected.
Incorrect Answers
A:This failure mode belongs mainly to another situation: A release process merges unreviewed notebook changes directly into the main branch minutes before a production run. Its preventative action would be: Protect the main branch with peer review, required tests and controlled promotion windows. The most direct exposure in the current scenario is: A release may finish successfully while critical reports cannot resolve their expected view.
C:This failure mode belongs mainly to another situation: A REST deployment script creates duplicate jobs every time CI reruns after a timeout. Its preventative action would be: Use idempotent deployment methods or reconcile existing resource IDs before creating replacements. The most direct exposure in the current scenario is: A release may finish successfully while critical reports cannot resolve their expected view.
D:This failure mode belongs mainly to another situation: A test harness mocks out Delta writes but the real job uses MERGE and unexpectedly changes historical rows. Its preventative action would be: Include an integration test using actual Delta table semantics and representative source changes. The most direct exposure in the current scenario is: A release may finish successfully while critical reports cannot resolve their expected view.
Question 744
A data engineering team faces this Azure Databricks requirement: An ingestion pipeline reads cloud events with at-least-once delivery and assumes every event ID appears only once. Which action is MOST appropriate?
- Use Spark-native distributed read and transformation APIs or partition-safe processing.
- Coordinate ADF orchestration with the appropriate Databricks task and Unity Catalog access model.
- Deduplicate by stable event identity and business time using documented state and watermark rules.
- Persist necessary source metadata and checkpoint lineage for troubleshooting and replay.
Explanation
Correct Answer
C:The most appropriate action is: Deduplicate by stable event identity and business time using documented state and watermark rules. It addresses the stated condition: An ingestion pipeline reads cloud events with at-least-once delivery and assumes every event ID appears only once. The technical basis is: Delivery guarantees alone do not remove duplicate producer or consumer events. Without this control, replayed events can double-count orders in downstream aggregates. A direct acceptance check is: A duplicate-event test yields one current business event while preserving valid distinct events.
Incorrect Answers
A:This alternative primarily addresses a different Azure Databricks problem: A Python notebook reads a massive table into pandas on the driver before processing data intended for Spark executors. The corresponding measure there would be: Use Spark-native distributed read and transformation APIs or partition-safe processing. The stated situation instead calls for: Deduplicate by stable event identity and business time using documented state and watermark rules.
B:This alternative primarily addresses a different Azure Databricks problem: A data product needs to copy a supported Azure Data Factory pipeline output into governed Databricks tables. The corresponding measure there would be: Coordinate ADF orchestration with the appropriate Databricks task and Unity Catalog access model. The stated situation instead calls for: Deduplicate by stable event identity and business time using documented state and watermark rules.
D:This alternative primarily addresses a different Azure Databricks problem: A pipeline ingests Event Hubs messages into a bronze table but never captures the partition, offset or event timestamp. The corresponding measure there would be: Persist necessary source metadata and checkpoint lineage for troubleshooting and replay. The stated situation instead calls for: Deduplicate by stable event identity and business time using documented state and watermark rules.
Question 745
The design team proposes: Keep large results distributed and replace driver collection with bounded sampling or distributed writes. The situation is: A Spark join reads large inputs but the driver reports failures after an analyst calls collect() on the result. Which technical principle justifies the decision?
- Cost optimization requires throughput per resource rather than simply maximizing available workers.
- Shuffle spills indicate data exceeded available in-memory execution resources for a stage.
- Driver memory is not an unlimited sink for results of distributed computation.
- A single hot partition can dominate a distributed stage despite idle executor capacity.
Explanation
Correct Answer
C:The decisive mechanism is: Driver memory is not an unlimited sink for results of distributed computation. That explains the selected action: Keep large results distributed and replace driver collection with bounded sampling or distributed writes. In the original scenario: A Spark join reads large inputs but the driver reports failures after an analyst calls collect() on the result. The specific unmanaged risk is: Collecting billions of rows can crash the driver despite idle executors. Validate with: The job writes results through distributed output and driver memory stays below the approved threshold.
Incorrect Answers
A:This principle concerns a separate data-engineering condition: A pipeline experiences escalating costs after a team increases node counts but execution duration barely changes. Its targeted response is: Use run-level compute utilization, cost and stage metrics to identify unused capacity. The current case is governed more directly by: Driver memory is not an unlimited sink for results of distributed computation.
B:This principle concerns a separate data-engineering condition: A large aggregation has repeated shuffle spills to disk even though source scans finish quickly. Its targeted response is: Examine shuffle metrics, memory pressure and partition sizing to target the spilling stage. The current case is governed more directly by: Driver memory is not an unlimited sink for results of distributed computation.
D:This principle concerns a separate data-engineering condition: A production Spark query is slow and its stage graph shows one task running far longer than all peers. Its targeted response is: Inspect the Spark UI DAG and skewed-key distribution before changing cluster size. The current case is governed more directly by: Driver memory is not an unlimited sink for results of distributed computation.
Question 746
An operator introduces: Use appropriate table history, schema change records and approved retention/audit evidence rather than relying on memory. The situation is: A shared table includes an old regulated field and the business needs reproducible evidence of when it was removed. Which validation should be required before signoff?
- The partner reads only the approved shared table and fails to access a nonshared table.
- The designated active service or steward owns the object and prior inappropriate rights are removed.
- A publish-time check rejects missing required descriptions, owners and classification tags.
- The change record ties the removed column to an authorized deployment and preserved evidence.
Explanation
Correct Answer
D:The strongest acceptance proof is: The change record ties the removed column to an authorized deployment and preserved evidence. It specifically verifies: Use appropriate table history, schema change records and approved retention/audit evidence rather than relying on memory. The original issue was: A shared table includes an old regulated field and the business needs reproducible evidence of when it was removed. The underlying mechanism is: Metadata history and audit trails can support investigation of structure and access changes. Without it, deleting a field without documented change history can leave unanswerable regulatory questions.
Incorrect Answers
A:This result would validate a different measure: Use Delta Sharing with an authorized recipient, narrowly scoped shares and lifecycle management. That is relevant to the distinct problem: A data-sharing project emails storage keys to an external partner for reading curated Delta results. Here the acceptance criterion is: The change record ties the removed column to an authorized deployment and preserved evidence.
B:This result would validate a different measure: Transfer ownership through approved governance workflow and review grants after reassignment. That is relevant to the distinct problem: An audit asks who owns a catalog table but object ownership still points to an inactive employee. Here the acceptance criterion is: The change record ties the removed column to an authorized deployment and preserved evidence.
C:This result would validate a different measure: Implement a catalog stewardship process to review metadata and governed tags at dataset publication. That is relevant to the distinct problem: A central repository contains schemas with missing column owners and no sensitivity classifications. Here the acceptance criterion is: The change record ties the removed column to an authorized deployment and preserved evidence.
Question 747
A data product owner adopts: Use controlled task-level timeouts and isolate unpredictable external calls from bounded transform stages. The reason is: A scheduled notebook contains both transformation SQL and a long-running external API loop that occasionally hangs. What observed result BEST proves effectiveness?
- A delayed-ingestion test prevents gold refresh until bronze and silver are completed successfully.
- A reviewed graph lists every expected parent and a source change propagates to the correct targets.
- A simulated hung request times out, alerts the owner and avoids publishing incomplete data.
- A seeded mismatch prevents cutover and the prior certified table remains queryable.
Explanation
Correct Answer
C:The strongest acceptance proof is: A simulated hung request times out, alerts the owner and avoids publishing incomplete data. It specifically verifies: Use controlled task-level timeouts and isolate unpredictable external calls from bounded transform stages. The original issue was: A scheduled notebook contains both transformation SQL and a long-running external API loop that occasionally hangs. The underlying mechanism is: Task boundaries make duration limits and operational ownership explicit. Without it, A stalled API call can hold an entire pipeline indefinitely and consume compute.
Incorrect Answers
A:This result would validate a different measure: Define explicit Lakeflow Jobs task dependencies so silver and gold steps run only after their required predecessors succeed. That is relevant to the distinct problem: A sales pipeline aggregates gold metrics before the bronze ingestion task finishes and occasionally reports incomplete totals. Here the acceptance criterion is: A simulated hung request times out, alerts the owner and avoids publishing incomplete data.
B:This result would validate a different measure: Declare consistent table names and dependencies and validate the generated pipeline graph. That is relevant to the distinct problem: A declarative pipeline transforms several datasets but its source names are ambiguous and a data team cannot reconstruct dependencies. Here the acceptance criterion is: A simulated hung request times out, alerts the owner and avoids publishing incomplete data.
D:This result would validate a different measure: Write to a staging target, validate row totals and invariants, then perform the approved publication step. That is relevant to the distinct problem: A finance transformation writes its output directly into a published table before reconciling totals against the source. Here the acceptance criterion is: A simulated hung request times out, alerts the owner and avoids publishing incomplete data.
Question 748
An Azure Databricks architect reviews the following problem: A large Delta fact table is filtered primarily on high-cardinality account IDs with changing query patterns. What should the team implement FIRST?
- Model a slowly changing dimension Type 2 with effective dates, current flags and correctly managed surrogate keys.
- Enable supported liquid clustering with appropriate keys and run compatible optimization.
- Evaluate Parquet or supported open-table formats against reader compatibility and mutation requirements.
- Expose stable SQL views over a governed base table when only presentation aliases differ.
Explanation
Correct Answer
B:The most appropriate action is: Enable supported liquid clustering with appropriate keys and run compatible optimization. It addresses the stated condition: A large Delta fact table is filtered primarily on high-cardinality account IDs with changing query patterns. The technical basis is: Liquid clustering can organize files for selective filters without static partition boundaries. Without this control, rigid directory partitioning can cause imbalance and require costly redesign as filters change. A direct acceptance check is: A representative account-filter query reads fewer irrelevant files after optimized clustering.
Incorrect Answers
A:This alternative primarily addresses a different Azure Databricks problem: A retail customer table changes delivery address repeatedly but analysts must reconstruct the address known at any prior invoice date. The corresponding measure there would be: Model a slowly changing dimension Type 2 with effective dates, current flags and correctly managed surrogate keys. The stated situation instead calls for: Enable supported liquid clustering with appropriate keys and run compatible optimization.
C:This alternative primarily addresses a different Azure Databricks problem: An external sharing integration needs interoperable columnar read files but not native Delta updates. The corresponding measure there would be: Evaluate Parquet or supported open-table formats against reader compatibility and mutation requirements. The stated situation instead calls for: Enable supported liquid clustering with appropriate keys and run compatible optimization.
D:This alternative primarily addresses a different Azure Databricks problem: A data product creates multiple physical copies of a curated dimension because teams need different column names. The corresponding measure there would be: Expose stable SQL views over a governed base table when only presentation aliases differ. The stated situation instead calls for: Enable supported liquid clustering with appropriate keys and run compatible optimization.
Question 749
A platform architect adopts: Preserve diagnostic events and Spark UI stage evidence before restart where possible. For this requirement: An operator restarts a failed Databricks cluster before collecting the Spark executor and driver logs. Which rationale MOST directly supports it?
- Severe key skew causes uneven task work even with many executors.
- Maintenance operations themselves consume compute and should correspond to data growth and read patterns.
- A single hot partition can dominate a distributed stage despite idle executor capacity.
- Transient cluster restarts can remove evidence necessary to distinguish dependency, data and resource failures.
Explanation
Correct Answer
D:The decisive mechanism is: Transient cluster restarts can remove evidence necessary to distinguish dependency, data and resource failures. That explains the selected action: Preserve diagnostic events and Spark UI stage evidence before restart where possible. In the original scenario: An operator restarts a failed Databricks cluster before collecting the Spark executor and driver logs. The specific unmanaged risk is: A recurring failure may never be diagnosed if its only logs are destroyed repeatedly. Validate with: A simulated failure yields archived logs, failed stage IDs and a documented root-cause path.
Incorrect Answers
A:This principle concerns a separate data-engineering condition: A DataFrame join spills heavily because a single join key occurs in most rows. Its targeted response is: Use query-profile evidence and a supported skew mitigation such as adaptive execution or key-aware repartitioning. The current case is governed more directly by: Transient cluster restarts can remove evidence necessary to distinguish dependency, data and resource failures.
B: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: Transient cluster restarts can remove evidence necessary to distinguish dependency, data and resource failures.
C:This principle concerns a separate data-engineering condition: A production Spark query is slow and its stage graph shows one task running far longer than all peers. Its targeted response is: Inspect the Spark UI DAG and skewed-key distribution before changing cluster size. The current case is governed more directly by: Transient cluster restarts can remove evidence necessary to distinguish dependency, data and resource failures.
Question 750
An implementation skips: Use an appropriate intersect or semi-join operation based on set and duplicate requirements. Engineers are observing: A team wants only customer IDs present in both systems but uses an inner join that multiplies repeated rows. What consequence is MOST likely?
- Refund amounts treated as sales inflate net revenue even when record counts match.
- Orders near midnight may be reported under the wrong business day.
- Unknown statuses may be dropped or duplicated if pivot categories are assumed without validation.
- An unrestricted join can multiply keys and exaggerate customer counts.
Explanation
Correct Answer
D:The direct residual exposure is: An unrestricted join can multiply keys and exaggerate customer counts. The triggering condition is: A team wants only customer IDs present in both systems but uses an inner join that multiplies repeated rows. The omitted safeguard is: Use an appropriate intersect or semi-join operation based on set and duplicate requirements. This matters because: Set operations and joins differ in duplicate behavior and result cardinality. A corrective test would show: A test with repeated IDs produces the expected unique or multiset intersection by policy.
Incorrect Answers
A:This failure mode belongs mainly to another situation: A daily revenue report double-counts refunds because both debit and credit rows are summed as positive numbers. Its preventative action would be: Normalize transaction sign semantics before grouping by business date and customer. The most direct exposure in the current scenario is: An unrestricted join can multiply keys and exaggerate customer counts.
B:This failure mode belongs mainly to another situation: A date field mixes UTC timestamps and local store times but a pipeline groups all values by the raw date string. Its preventative action would be: Normalize time zone semantics before extracting business-local reporting dates. The most direct exposure in the current scenario is: An unrestricted join can multiply keys and exaggerate customer counts.
C:This failure mode belongs mainly to another situation: A quality report needs one column for each product status category but stores status as separate rows. Its preventative action would be: Apply a controlled pivot with a defined set of expected categories and aggregation rule. The most direct exposure in the current scenario is: An unrestricted join can multiply keys and exaggerate customer counts.