Implementing Data Engineering Solutions Using Azure Databricks DP-750 Practice Test 9
Question 80
The team implements: Use an appropriately sized SQL warehouse with workload isolation and measured concurrency for dashboard queries. The original problem is: Analysts run short SQL dashboard queries against a shared lakehouse but their notebooks share a general-purpose Spark cluster with training experiments. Which test MOST convincingly demonstrates correctness?
- A load rehearsal meets dashboard p95 response targets while notebook resource spikes do not affect query service.
- An ordinary user can start an approved small cluster but cannot exceed policy size or runtime limits.
- The scheduled job completes within its service objective and compute billing shows no unattended idle cluster hours.
- Before-after measurements report comparable inputs, stage timings, spills and total cost.
Explanation
Correct Answer
A:The strongest acceptance proof is: A load rehearsal meets dashboard p95 response targets while notebook resource spikes do not affect query service. It specifically verifies: Use an appropriately sized SQL warehouse with workload isolation and measured concurrency for dashboard queries. The original issue was: Analysts run short SQL dashboard queries against a shared lakehouse but their notebooks share a general-purpose Spark cluster with training experiments. The underlying mechanism is: SQL warehouses are optimized for SQL analytics and isolate query serving from other compute activities. Without it, A shared notebook cluster can cause dashboards to stall when unrelated Spark notebooks consume executors.
Incorrect Answers
B:This result would validate a different measure: Create compute policies with approved node types, worker limits, tags and idle settings for the permitted users. That is relevant to the distinct problem: A department wants a self-service sandbox but must prevent users from creating huge unrestricted clusters. Here the acceptance criterion is: A load rehearsal meets dashboard p95 response targets while notebook resource spikes do not affect query service.
C:This result would validate a different measure: Move the scheduled transformation to appropriately configured jobs compute or supported serverless jobs and release resources after execution. That is relevant to the distinct problem: A nightly batch transformation uses a continuously running all-purpose cluster even though no analyst interacts with the workload between runs. Here the acceptance criterion is: A load rehearsal meets dashboard p95 response targets while notebook resource spikes do not affect query service.
D:This result would validate a different measure: Collect Spark stage profiles and test a targeted compute or Photon setting against comparable data. That is relevant to the distinct problem: A team turns on every performance setting for a workload without benchmarking whether disk spills or network shuffle dominate runtime. Here the acceptance criterion is: A load rehearsal meets dashboard p95 response targets while notebook resource spikes do not affect query service.
Question 81
An Azure Databricks architect reviews the following problem: A job's duration doubles but an administrator sees only its final success status with no phase-level metrics. What should the team implement FIRST?
- Collect job timing, task duration and relevant Spark stage metrics in a monitored baseline.
- Compare workload contention and resource reservations, then isolate or resize the critical job compute.
- Respect documented safe retention windows and coordinate VACUUM with readers, streams and time-travel needs.
- Export relevant logs and metrics through a supported Log Analytics or Azure Monitor integration with controlled retention.
Explanation
Correct Answer
A:The most appropriate action is: Collect job timing, task duration and relevant Spark stage metrics in a monitored baseline. It addresses the stated condition: A job's duration doubles but an administrator sees only its final success status with no phase-level metrics. The technical basis is: Success alone cannot show degradation against latency or cost service objectives. Without this control, unobserved performance drift can cause the pipeline to miss future delivery windows. A direct acceptance check is: A dashboard highlights task and stage regressions before the total job breaches its SLA.
Incorrect Answers
B:This alternative primarily addresses a different Azure Databricks problem: A production notebook intermittently times out only when many other jobs share the same classic cluster. The corresponding measure there would be: Compare workload contention and resource reservations, then isolate or resize the critical job compute. The stated situation instead calls for: Collect job timing, task duration and relevant Spark stage metrics in a monitored baseline.
C:This alternative primarily addresses a different Azure Databricks problem: An engineer runs VACUUM with a zero-hour retention on a shared Delta table used for historical audit queries. The corresponding measure there would be: Respect documented safe retention windows and coordinate VACUUM with readers, streams and time-travel needs. The stated situation instead calls for: Collect job timing, task duration and relevant Spark stage metrics in a monitored baseline.
D:This alternative primarily addresses a different Azure Databricks problem: An Azure Databricks pipeline has recurring errors but its logs remain only in a short-lived cluster session. The corresponding measure there would be: Export relevant logs and metrics through a supported Log Analytics or Azure Monitor integration with controlled retention. The stated situation instead calls for: Collect job timing, task duration and relevant Spark stage metrics in a monitored baseline.
Question 82
An implementation skips: Use idempotent deployment methods or reconcile existing resource IDs before creating replacements. Engineers are observing: A REST deployment script creates duplicate jobs every time CI reruns after a timeout. What consequence is MOST likely?
- The same Git commit may run differently after a dependency's incompatible update.
- A wrong schedule can duplicate loads or miss critical reporting windows.
- Duplicate jobs may run on the same schedule and load data twice.
- A logically incorrect aggregation can pass compilation while corrupting metrics.
Explanation
Correct Answer
C:The direct residual exposure is: Duplicate jobs may run on the same schedule and load data twice. The triggering condition is: A REST deployment script creates duplicate jobs every time CI reruns after a timeout. The omitted safeguard is: Use idempotent deployment methods or reconcile existing resource IDs before creating replacements. This matters because: Infrastructure automation must tolerate retries without multiplying the intended target resources. A corrective test would show: Two identical deployment executions leave exactly one intended job with approved settings.
Incorrect Answers
A:This failure mode belongs mainly to another situation: A team cannot reproduce a prior release because package versions are resolved dynamically during deployment. Its preventative action would be: Pin package and runtime versions in the governed release artifact and record its dependency manifest. The most direct exposure in the current scenario is: Duplicate jobs may run on the same schedule and load data twice.
B:This failure mode belongs mainly to another situation: A data engineer updates a bundled job trigger but another environment's deployment silently uses the wrong schedule. Its preventative action would be: Validate per-target bundle variable resolution and compare rendered job definitions before deployment. The most direct exposure in the current scenario is: Duplicate jobs may run on the same schedule and load data twice.
D:This failure mode belongs mainly to another situation: A release modifies transformation logic but tests only whether Python files compile. Its preventative action would be: Add representative unit and integration tests asserting transformation results and schema contracts. The most direct exposure in the current scenario is: Duplicate jobs may run on the same schedule and load data twice.
Question 83
A solution applies: Maintain table and column descriptions in Unity Catalog metadata at creation and during schema changes. It addresses: Analysts create tables without column comments and the data steward cannot explain monetary units or null semantics. Which operational outcome MOST directly verifies success?
- An access review identifies the view steward and confirms a prohibited reader cannot query it.
- Catalog Explorer shows reviewed descriptions and a sample consumer can identify the column's intended units.
- Production jobs resolve fully qualified names and fail safely when pointed at a nonproduction catalog.
- Automated DDL completes in both staging and production under the same naming rules.
Explanation
Correct Answer
B:The strongest acceptance proof is: Catalog Explorer shows reviewed descriptions and a sample consumer can identify the column's intended units. It specifically verifies: Maintain table and column descriptions in Unity Catalog metadata at creation and during schema changes. The original issue was: Analysts create tables without column comments and the data steward cannot explain monetary units or null semantics. The underlying mechanism is: Discoverability requires human-readable business metadata in addition to schema types. Without it, poorly described fields lead to incorrect joins, unit conversions or disclosure decisions.
Incorrect Answers
A:This result would validate a different measure: Assign view ownership and narrowly scoped permissions under the appropriate Unity Catalog schema. That is relevant to the distinct problem: A central team creates global views that hide source objects but forgets to document ownership and access grants. Here the acceptance criterion is: Catalog Explorer shows reviewed descriptions and a sample consumer can identify the column's intended units.
C:This result would validate a different measure: Create environment-specific Unity Catalog catalogs or schemas and enforce clear ownership and naming conventions. That is relevant to the distinct problem: Production and development teams create identically named data objects in one catalog and accidentally overwrite test tables. Here the acceptance criterion is: Catalog Explorer shows reviewed descriptions and a sample consumer can identify the column's intended units.
D:This result would validate a different measure: Adopt a portable naming and quoting convention compatible with Unity Catalog SQL identifiers. That is relevant to the distinct problem: An ingestion team creates a table with a reserved or inconsistent name and automation fails in one environment. Here the acceptance criterion is: Catalog Explorer shows reviewed descriptions and a sample consumer can identify the column's intended units.
Question 84
An operator introduces: Define a supported typed schema for nested structures and expected optional fields before writing to curated Delta. The situation is: A stream produces nested JSON events with optional fields but an analyst hardcodes every field as STRING. Which validation should be required before signoff?
- Sample nested records parse into the designated types and invalid inputs follow the documented rejection path.
- Two successive batches load only new records while preserved history and counts reconcile.
- A controlled late event appears in the correct historical period according to the approved policy.
- A MERGE transaction updates the expected records atomically and a historical version remains queryable within retention.
Explanation
Correct Answer
A:The strongest acceptance proof is: Sample nested records parse into the designated types and invalid inputs follow the documented rejection path. It specifically verifies: Define a supported typed schema for nested structures and expected optional fields before writing to curated Delta. The original issue was: A stream produces nested JSON events with optional fields but an analyst hardcodes every field as STRING. The underlying mechanism is: Correct data types preserve semantic operations and enable schema validation during ingestion. Without it, string-only columns can make numerical aggregation or timestamp ordering incorrect.
Incorrect Answers
B:This result would validate a different measure: Choose incremental append or merge based on immutable versus mutable source behavior. That is relevant to the distinct problem: A raw event lake contains append-only records but a developer chooses daily full overwrite of the curated Delta table. Here the acceptance criterion is: Sample nested records parse into the designated types and invalid inputs follow the documented rejection path.
C:This result would validate a different measure: Design event-time fields, ingestion timestamps and late-data behavior suited to replay and temporal queries. That is relevant to the distinct problem: A lakehouse receives sensor observations with event timestamps that arrive out of order. Here the acceptance criterion is: Sample nested records parse into the designated types and invalid inputs follow the documented rejection path.
D:This result would validate a different measure: Use a managed Delta table for transactional updates and controlled schema evolution. That is relevant to the distinct problem: A reporting team stores frequent updates in raw CSV files and expects atomic MERGE and historical time travel. Here the acceptance criterion is: Sample nested records parse into the designated types and invalid inputs follow the documented rejection path.
Question 85
A team omits the recommended measure: Use supported Databricks secret access with least-privilege scope access and remove secret values from logs. Its environment shows: An application uses an external Azure Key Vault secret scope but a notebook prints the secret into task logs. What is the MOST direct residual risk?
- A retired vendor workload can continue reading business data through unchanged credentials.
- A user may bypass the masked presentation by reading the unmasked source table.
- An object may be available to overprivileged identities in unintended contexts if either boundary is ignored.
- A protected secret can still leak through debug output or unrestricted notebook results.
Explanation
Correct Answer
D:The direct residual exposure is: A protected secret can still leak through debug output or unrestricted notebook results. The triggering condition is: An application uses an external Azure Key Vault secret scope but a notebook prints the secret into task logs. The omitted safeguard is: Use supported Databricks secret access with least-privilege scope access and remove secret values from logs. This matters because: Secret retrieval must be paired with permissions and safe handling to avoid disclosure after retrieval. A corrective test would show: The job authenticates successfully and audited logs contain no raw secret or token.
Incorrect Answers
A:This failure mode belongs mainly to another situation: A vendor's service principal was granted global access for a one-month integration and the project ended. Its preventative action would be: Review and revoke stale grants and any linked service credentials, preserving required internal ownership. The most direct exposure in the current scenario is: A protected secret can still leak through debug output or unrestricted notebook results.
B: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: A protected secret can still leak through debug output or unrestricted notebook results.
C:This failure mode belongs mainly to another situation: A team restricts access through a catalog-to-workspace binding but treats it as a replacement for data privileges. Its preventative action would be: Apply both the workspace binding and appropriate Unity Catalog object grants, then test each dimension. The most direct exposure in the current scenario is: A protected secret can still leak through debug output or unrestricted notebook results.
Question 86
A pipeline owner declines the action: Create a Unity Catalog view with approved permissions and verify underlying access semantics. The business situation is: An analyst wants to expose a filtered projection of a governed table without copying sensitive source rows. What specific failure can follow?
- A drop operation may remove managed table data that administrators expected to retain.
- An unqualified table reference may redirect a production job to a developer's test dataset.
- Assuming full DDL or mutation support may cause failed deployments or inadvertent source changes.
- Replicated extracts may fall outside the original source's governance and retention controls.
Explanation
Correct Answer
D:The direct residual exposure is: Replicated extracts may fall outside the original source's governance and retention controls. The triggering condition is: An analyst wants to expose a filtered projection of a governed table without copying sensitive source rows. The omitted safeguard is: Create a Unity Catalog view with approved permissions and verify underlying access semantics. This matters because: A view can define a query abstraction without requiring a second physical copy of source data. A corrective test would show: Permitted users query only expected columns and rows through the view without a separately exported file.
Incorrect Answers
A:This failure mode belongs mainly to another situation: A team expects dropping a managed table to preserve all physical data as though it were an external table. Its preventative action would be: Select managed or external Unity Catalog table ownership intentionally and verify lifecycle semantics. The most direct exposure in the current scenario is: Replicated extracts may fall outside the original source's governance and retention controls.
B:This failure mode belongs mainly to another situation: Production and development teams create identically named data objects in one catalog and accidentally overwrite test tables. Its preventative action would be: Create environment-specific Unity Catalog catalogs or schemas and enforce clear ownership and naming conventions. The most direct exposure in the current scenario is: Replicated extracts may fall outside the original source's governance and retention controls.
C:This failure mode belongs mainly to another situation: A foreign catalog is connected to a remote database but an analyst expects standard managed Delta writes against it. Its preventative action would be: Check federation connector capabilities and distinguish read-only remote access from supported writable local tables. The most direct exposure in the current scenario is: Replicated extracts may fall outside the original source's governance and retention controls.
Question 87
A platform architect adopts: Propagate nonrecoverable exceptions to Lakeflow Jobs and attach actionable error metadata. For this requirement: A workflow uses a notebook task that catches exceptions and returns success even when a write step fails. Which rationale MOST directly supports it?
- Job status must accurately reflect failed required operations for alerting and dependency logic.
- Separating pipeline stages creates observable failure boundaries and supports targeted recovery.
- Staging plus validation separates construction of data from its declaration as trusted.
- A recoverable pipeline should produce the same logical state after an equivalent retry.
Explanation
Correct Answer
A:The decisive mechanism is: Job status must accurately reflect failed required operations for alerting and dependency logic. That explains the selected action: Propagate nonrecoverable exceptions to Lakeflow Jobs and attach actionable error metadata. In the original scenario: A workflow uses a notebook task that catches exceptions and returns success even when a write step fails. The specific unmanaged risk is: A swallowed exception can allow downstream tasks to publish stale or missing outputs. Validate with: A deliberately failed write marks the job unsuccessful and prevents dependent publish tasks.
Incorrect Answers
B:This principle concerns a separate data-engineering condition: A multi-step notebook workflow handles source extraction and final reporting in one cell sequence with no recoverable checkpoints. Its targeted response is: Split ingestion, validation and publishing into explicit orchestrated tasks with durable intermediate outputs. The current case is governed more directly by: Job status must accurately reflect failed required operations for alerting and dependency logic.
C:This principle concerns a separate data-engineering condition: A finance transformation writes its output directly into a published table before reconciling totals against the source. Its targeted response is: Write to a staging target, validate row totals and invariants, then perform the approved publication step. The current case is governed more directly by: Job status must accurately reflect failed required operations for alerting and dependency logic.
D:This principle concerns a separate data-engineering condition: A nightly pipeline rerun appends the same orders to the target and doubles totals after an operator presses Retry. Its targeted response is: Use idempotent writes or MERGE keyed by stable order identity for repeatable task execution. The current case is governed more directly by: Job status must accurately reflect failed required operations for alerting and dependency logic.
Question 88
The team implements: Use an incremental append pattern with suitable deduplication when source events cannot change. The original problem is: An append-only telemetry feed includes newly observed events but engineers use MERGE with expensive full-table scans on every insert. Which test MOST convincingly demonstrates correctness?
- Mixed-case, whitespace and non-ASCII test records resolve according to the documented mapping policy.
- A seeded repeated-ID sample yields one chosen current order version and reconciled totals.
- A timed append comparison meets the freshness and duplicate controls with lower compute consumption.
- The merged dataset preserves typed values and column names from both input feeds.
Explanation
Correct Answer
C:The strongest acceptance proof is: A timed append comparison meets the freshness and duplicate controls with lower compute consumption. It specifically verifies: Use an incremental append pattern with suitable deduplication when source events cannot change. The original issue was: An append-only telemetry feed includes newly observed events but engineers use MERGE with expensive full-table scans on every insert. The underlying mechanism is: Append is generally simpler for immutable events than key-by-key mutation logic. Without it, unnecessary MERGE operations can add substantial cost and contention without correctness benefit.
Incorrect Answers
A:This result would validate a different measure: Define and test consistent trimming, Unicode normalization and case comparison for the business key. That is relevant to the distinct problem: A customer-name cleansing script uses an arbitrary lowercase conversion without preserving Unicode and leading/trailing whitespace rules. Here the acceptance criterion is: A timed append comparison meets the freshness and duplicate controls with lower compute consumption.
B:This result would validate a different measure: Apply a deterministic deduplication rule based on stable order identity and source version before aggregation. That is relevant to the distinct problem: An incoming orders table contains repeated order IDs because a vendor retransmits the same daily extract. Here the acceptance criterion is: A timed append comparison meets the freshness and duplicate controls with lower compute consumption.
D:This result would validate a different measure: Align compatible schemas and use name-based union semantics with explicit type checks. That is relevant to the distinct problem: Two quarterly sales feeds have the same logical fields but different column ordering and a union shifts amounts into dates. Here the acceptance criterion is: A timed append comparison meets the freshness and duplicate controls with lower compute consumption.
Question 89
The team implements: Choose and document a supported format and catalog integration with explicit read/write compatibility tests. The original problem is: A cross-engine consumer expects a Delta table but source is plain Iceberg and engineers assume identical mutation semantics. Which test MOST convincingly demonstrates correctness?
- Sample invoice lines reconcile exactly to the fact table and roll up to the approved daily total.
- Storage metrics show a manageable file distribution and point lookups meet query SLAs.
- All required engines pass a controlled read and supported mutation matrix before production.
- A supported optimization command completes and query plans show the chosen layout works.
Explanation
Correct Answer
C:The strongest acceptance proof is: All required engines pass a controlled read and supported mutation matrix before production. It specifically verifies: Choose and document a supported format and catalog integration with explicit read/write compatibility tests. The original issue was: A cross-engine consumer expects a Delta table but source is plain Iceberg and engineers assume identical mutation semantics. The underlying mechanism is: Delta and Iceberg are different transactional table formats with version-dependent feature support. Without it, assuming every engine can use every feature can result in invalid writes or unreadable metadata.
Incorrect Answers
A:This result would validate a different measure: Define fact-table grain at the required transaction or line-item level before aggregation. That is relevant to the distinct problem: A facts table groups product sales at daily level although auditors must inspect individual transaction adjustments. Here the acceptance criterion is: All required engines pass a controlled read and supported mutation matrix before production.
B:This result would validate a different measure: Prefer appropriate clustering or lower-cardinality partitioning and assess file sizes for the dimension. That is relevant to the distinct problem: A data engineer chooses the distinct customer ID as a partition directory for an SCD dimension with millions of customers. Here the acceptance criterion is: All required engines pass a controlled read and supported mutation matrix before production.
D:This result would validate a different measure: Choose one compatible data layout strategy based on table format, platform support and query evidence. That is relevant to the distinct problem: A team attempts to apply both ZORDER and liquid clustering to the same Delta table. Here the acceptance criterion is: All required engines pass a controlled read and supported mutation matrix before production.
Question 90
For this Microsoft DP-750 case: A pipeline joins a huge fact table with a tiny dimension and a normal shuffle join creates excessive network traffic. Which design or operational response BEST fits?
- Evaluate a suitable broadcast join based on measured dimension size and Spark execution plans.
- Align compatible schemas and use name-based union semantics with explicit type checks.
- Use a left anti-join or EXCEPT operation with aligned schemas and explicit null semantics.
- Use an incremental append pattern with suitable deduplication when source events cannot change.
Explanation
Correct Answer
A:The most appropriate action is: Evaluate a suitable broadcast join based on measured dimension size and Spark execution plans. It addresses the stated condition: A pipeline joins a huge fact table with a tiny dimension and a normal shuffle join creates excessive network traffic. The technical basis is: Spark can reduce shuffle costs by broadcasting a sufficiently small dimension where appropriate. Without this control, blindly broadcasting a large table can exhaust executor memory while unnecessary shuffles slow joins. A direct acceptance check is: Spark UI shows reduced shuffle bytes and matching results without executor out-of-memory.
Incorrect Answers
B:This alternative primarily addresses a different Azure Databricks problem: Two quarterly sales feeds have the same logical fields but different column ordering and a union shifts amounts into dates. The corresponding measure there would be: Align compatible schemas and use name-based union semantics with explicit type checks. The stated situation instead calls for: Evaluate a suitable broadcast join based on measured dimension size and Spark execution plans.
C: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: Evaluate a suitable broadcast join based on measured dimension size and Spark execution plans.
D:This alternative primarily addresses a different Azure Databricks problem: An append-only telemetry feed includes newly observed events but engineers use MERGE with expensive full-table scans on every insert. The corresponding measure there would be: Use an incremental append pattern with suitable deduplication when source events cannot change. The stated situation instead calls for: Evaluate a suitable broadcast join based on measured dimension size and Spark execution plans.