Certified Data Engineer Associate Practice Test 33
Question 320
At a wholesale pharmacy group, finance analysts repeatedly execute an expensive stable aggregate over a slowly changing silver Delta table. The dataset `route_updates` is in scope, and downstream dashboards cannot accept fabricated values. Before release, which acceptance test would actually demonstrate that the root problem is controlled?
- Run only a normal, unchanged `route_updates` sample and review silver change timestamps.
- Update one silver partition, refresh the materialized view, and compare gold totals with the source
- Check that the intended configuration exists for `route_updates` and that a healthy run succeeded, without inducing the earlier failure.
- Review a platform-level availability report for `route_updates` without testing whether a conventional view recalculates the expensive query on use while a precomputed result can be refreshed deliberately.
Explanation
Correct Answer
B:Update one silver partition, refresh the materialized view, and compare gold totals with the source. This reproduces the scenario-dependent condition, verifies silver change timestamps, query cost, materialized view refresh state, and freshness SLA, and directly tests whether a conventional view recalculates the expensive query on use while a precomputed result can be refreshed deliberately is resolved. Observed results, not intent, determine approval.
Incorrect Answers
A:Run only a normal, unchanged `route_updates` sample and review silver change timestamps. A normal sample does not challenge a conventional view recalculates the expensive query on use while a precomputed result can be refreshed deliberately.
C:Check that the intended configuration exists for `route_updates` and that a healthy run succeeded, without inducing the earlier failure. Installed configuration and a healthy run are insufficient to prove the failure mode is gone. The original issue was a conventional view recalculates the expensive query on use while a precomputed result can be refreshed deliberately.
D:Review a platform-level availability report for `route_updates` without testing whether a conventional view recalculates the expensive query on use while a precomputed result can be refreshed deliberately. Aggregate availability does not establish the behavior of affected `route_updates` records. It cannot replace the direct exercise: update one silver partition, refresh the materialized view, and compare gold totals with the source.
Question 321
At a marketplace risk team, a telemetry dashboard needs a continually updated gold dataset from a streaming bronze source. The dataset `shipment_status` is in scope, and raw records must remain available for replay. What evidence should the engineer collect to establish the cause rather than a coincidental symptom?
- Compare the current configuration and the signed change approval, without source-row evidence
- Inspect normal-period success metrics for the stated dataset without reproducing the source error
- Review completed run status and aggregate cluster health without record-level correlation
- Correlate source event time, streaming-table progress across affected and unaffected `R034-CASE-065` runs.
Explanation
Correct Answer
D:Correlate source event time, streaming-table progress across affected and unaffected `R034-CASE-065` runs.. By joining source event time, streaming-table progress, processing latency, and gold query result across the affected and unaffected conditions, the investigation directly tests the recorded mechanism: manual one-time gold snapshots do not track continual source arrivals. It also sets a measurable baseline for append new telemetry events during a run and verify gold advances without duplicating previous data.
Incorrect Answers
A:Compare the current configuration and the signed change approval, without source-row evidence. A desired configuration and approval ticket are not runtime proof. In this case they omit source event time, streaming-table progress, processing latency, and gold query result, the direct means of distinguishing manual one-time gold snapshots do not track continual source arrivals.
B:Inspect normal-period success metrics for the stated dataset without reproducing the source error. A healthy-run trend can be useful, but it does not expose manual one-time gold snapshots do not track continual source arrivals. The failing-path source event time, streaming-table progress, processing latency, and gold query result has to be correlated before attributing cause.
C:Review completed run status and aggregate cluster health without record-level correlation. A successful aggregate platform state can coexist with manual one-time gold snapshots do not track continual source arrivals. This check never observes the affected `shipment_status` records or their causal path.
Question 322
At a marketplace risk team, a telemetry dashboard needs a continually updated gold dataset from a streaming bronze source. The dataset `shipment_status` is in scope, and raw records must remain available for replay. Which platform configuration or data-engineering change is most defensible?
- Build a governed streaming table with an incremental processing path and measure freshness across updates
- Regenerate the entire history with a manual overwrite each minute; use that approach for `shipment_status`.
- Create a static CTAS table and never schedule refresh; use that approach for `shipment_status`.
- Use a view over untyped raw JSON and call it a curated incremental table; use that approach for `shipment_status`.
Explanation
Correct Answer
A:Build a governed streaming table with an incremental processing path and measure freshness across updates. This directly addresses manual one-time gold snapshots do not track continual source arrivals. The design should be accepted only after append new telemetry events during a run and verify gold advances without duplicating previous data; the related operational boundary is to maintain continuous freshness for the necessary dataset while bounding infrastructure and checkpoint costs.
Incorrect Answers
B:Regenerate the entire history with a manual overwrite each minute; use that approach for `shipment_status`. Full recomputation introduces unnecessary cost and risky overlapping writes. For `shipment_status`, this cannot remove the causal limitation: manual one-time gold snapshots do not track continual source arrivals.
C:Create a static CTAS table and never schedule refresh; use that approach for `shipment_status`. A static snapshot becomes stale as soon as new telemetry arrives. For `shipment_status`, this cannot remove the causal limitation: manual one-time gold snapshots do not track continual source arrivals.
D:Use a view over untyped raw JSON and call it a curated incremental table; use that approach for `shipment_status`. A plain view provides neither an incremental processing state nor typed quality enforcement. For `shipment_status`, this cannot remove the causal limitation: manual one-time gold snapshots do not track continual source arrivals.
Question 323
At a marketplace risk team, a telemetry dashboard needs a continually updated gold dataset from a streaming bronze source. The dataset `shipment_status` is in scope, and raw records must remain available for replay. A teammate proposes: "Regenerate the entire history with a manual overwrite each minute." Which technical limitation is the decisive reason not to approve that specific proposal?
- A plain view provides neither an incremental processing state nor typed quality enforcement. This would apply to another option for `shipment_status`.
- The dashboard cached the prior result; no DataFrame operation or underlying data requires correction.
- Full recomputation introduces unnecessary cost and risky overlapping writes. The proposal does not repair `shipment_status`.
- A static snapshot becomes stale as soon as new telemetry arrives. This would apply to an alternative change for `shipment_status`.
Explanation
Correct Answer
C:Full recomputation introduces unnecessary cost and risky overlapping writes. The proposal does not repair `shipment_status`. This identifies the known limitation of the exact proposed action. Full recomputation introduces unnecessary cost and risky overlapping writes. The underlying problem remains manual one-time gold snapshots do not track continual source arrivals; Build a governed streaming table with an incremental processing path and measure freshness across updates instead targets that failure.
Incorrect Answers
A:A plain view provides neither an incremental processing state nor typed quality enforcement. This would apply to another option for `shipment_status`. This limitation belongs to the different alternative "Use a view over untyped raw JSON and call it a curated incremental table". The stem calls for the specific failure of "Regenerate the entire history with a manual overwrite each minute".
B:The dashboard cached the prior result; no DataFrame operation or underlying data requires correction. The case provides a concrete failing data or control path; a mere alternate theory does not address manual one-time gold snapshots do not track continual source arrivals.
D:A static snapshot becomes stale as soon as new telemetry arrives. This would apply to an alternative change for `shipment_status`. This failure applies to the separate proposal "Create a static CTAS table and never schedule refresh".
Question 324
At a marketplace risk team, a telemetry dashboard needs a continually updated gold dataset from a streaming bronze source. The dataset `shipment_status` is in scope, and raw records must remain available for replay. What risk-and-cost compromise should the owner accept instead of the proposed shortcuts?
- Create a static CTAS table and never schedule refresh
- Regenerate the entire history with a manual overwrite each minute
- Maintain continuous freshness for the necessary dataset while bounding infrastructure and checkpoint costs
- Use a view over untyped raw JSON and call it a curated incremental table
Explanation
Correct Answer
C:Maintain continuous freshness for the necessary dataset while bounding infrastructure and checkpoint costs. This explicitly bounds the compromise while still addressing manual one-time gold snapshots do not track continual source arrivals. Acceptance relies on append new telemetry events during a run and verify gold advances without duplicating previous data rather than simply declaring the change complete.
Incorrect Answers
A:Create a static CTAS table and never schedule refresh. A static snapshot becomes stale as soon as new telemetry arrives. Monitoring after deployment does not make this approach a safe bounded tradeoff for `shipment_status`.
B:Regenerate the entire history with a manual overwrite each minute. Full recomputation introduces unnecessary cost and risky overlapping writes. Monitoring after deployment does not make this approach a safe bounded tradeoff for `shipment_status`.
D:Use a view over untyped raw JSON and call it a curated incremental table. A plain view provides neither an incremental processing state nor typed quality enforcement. Monitoring after deployment does not make this approach a safe bounded tradeoff for `shipment_status`.
Question 325
At a marketplace risk team, a telemetry dashboard needs a continually updated gold dataset from a streaming bronze source. The dataset `shipment_status` is in scope, and raw records must remain available for replay. Which direct test could falsify the proposed fix instead of only confirming deployment?
- Review a platform-level availability report for `shipment_status` without testing whether manual one-time gold snapshots do not track continual source arrivals.
- Append new telemetry events during a run and verify gold advances without duplicating previous data
- Run only a normal, unchanged `shipment_status` sample and review source event time.
- Check that the intended configuration exists for `shipment_status` and that a healthy run succeeded, without inducing the earlier failure.
Explanation
Correct Answer
B:Append new telemetry events during a run and verify gold advances without duplicating previous data. This reproduces the scenario-dependent condition, verifies source event time, streaming-table progress, processing latency, and gold query result, and directly tests whether manual one-time gold snapshots do not track continual source arrivals is resolved. Observed results, not intent, determine approval.
Incorrect Answers
A:Review a platform-level availability report for `shipment_status` without testing whether manual one-time gold snapshots do not track continual source arrivals. Aggregate availability does not establish the behavior of affected `shipment_status` records. It cannot replace the direct exercise: append new telemetry events during a run and verify gold advances without duplicating previous data.
C:Run only a normal, unchanged `shipment_status` sample and review source event time. A normal sample does not challenge manual one-time gold snapshots do not track continual source arrivals.
D:Check that the intended configuration exists for `shipment_status` and that a healthy run succeeded, without inducing the earlier failure. Installed configuration and a healthy run are insufficient to prove the failure mode is gone. The original issue was manual one-time gold snapshots do not track continual source arrivals.
Question 326
At a census statistics agency, a silver customer dimension occasionally contains null customer IDs that should never reach billing gold tables. The dataset `payment_decisions` is in scope, and downstream dashboards cannot accept fabricated values. Which correlated observation is the strongest basis for the change decision?
- Review aggregate cluster availability and job completion for `payment_decisions` without correlating the observed error path.
- Compare the saved configuration with the signed change ticket for `payment_decisions` while omitting source and output records.
- Examine only a normal successful run of `payment_decisions` and trend null key rate without recreating the fault.
- Correlate null key rate, rejected-record sample across affected and unaffected `R034-CASE-066` runs.
Explanation
Correct Answer
D:Correlate null key rate, rejected-record sample across affected and unaffected `R034-CASE-066` runs.. By joining null key rate, rejected-record sample, silver-to-gold counts, and quality rule events across the affected and unaffected conditions, the investigation directly tests the recorded mechanism: a null business key breaks reliable joins and can create untraceable billing records. It also sets a measurable baseline for inject a null customer ID and verify the exception record is visible but absent from trusted gold.
Incorrect Answers
A:Review aggregate cluster availability and job completion for `payment_decisions` without correlating the observed error path. A successful aggregate platform state can coexist with a null business key breaks reliable joins and can create untraceable billing records. This check never observes the affected `payment_decisions` records or their causal path.
B:Compare the saved configuration with the signed change ticket for `payment_decisions` while omitting source and output records. A desired configuration and approval ticket are not runtime proof. In this case they omit null key rate, rejected-record sample, silver-to-gold counts, and quality rule events, the direct means of distinguishing a null business key breaks reliable joins and can create untraceable billing records.
C:Examine only a normal successful run of `payment_decisions` and trend null key rate without recreating the fault. A healthy-run trend can be useful, but it does not expose a null business key breaks reliable joins and can create untraceable billing records. The failing-path null key rate, rejected-record sample, silver-to-gold counts, and quality rule events has to be correlated before attributing cause.
Question 327
At a census statistics agency, a silver customer dimension occasionally contains null customer IDs that should never reach billing gold tables. The dataset `payment_decisions` is in scope, and downstream dashboards cannot accept fabricated values. What implementation should the engineer select for the failing workflow?
- Replace missing customer IDs with a common literal value; use that approach for `payment_decisions`.
- Enforce a non-null key quality expectation or validation rule and route rejected records for investigation before publishing gold
- Disable all quality checks when ingestion latency increases; use that approach for `payment_decisions`.
- Allow the rows into gold and hope a later dashboard filter hides them; use that approach for `payment_decisions`.
Explanation
Correct Answer
B:Enforce a non-null key quality expectation or validation rule and route rejected records for investigation before publishing gold. This directly addresses a null business key breaks reliable joins and can create untraceable billing records. The design should be accepted only after inject a null customer ID and verify the exception record is visible but absent from trusted gold; the related operational boundary is to keep a visible quarantine route instead of silently replacing missing identifiers.
Incorrect Answers
A:Replace missing customer IDs with a common literal value; use that approach for `payment_decisions`. A shared fabricated key merges unrelated customers into one identity. For `payment_decisions`, this cannot remove the causal limitation: a null business key breaks reliable joins and can create untraceable billing records.
C:Disable all quality checks when ingestion latency increases; use that approach for `payment_decisions`. Removing checks trades temporary performance for silent data corruption. For `payment_decisions`, this cannot remove the causal limitation: a null business key breaks reliable joins and can create untraceable billing records.
D:Allow the rows into gold and hope a later dashboard filter hides them; use that approach for `payment_decisions`. A presentation filter does not protect downstream consumers or data contracts. For `payment_decisions`, this cannot remove the causal limitation: a null business key breaks reliable joins and can create untraceable billing records.
Question 328
At a census statistics agency, a silver customer dimension occasionally contains null customer IDs that should never reach billing gold tables. The dataset `payment_decisions` is in scope, and downstream dashboards cannot accept fabricated values. A teammate proposes: "Allow the rows into gold and hope a later dashboard filter hides them." Which technical limitation is the decisive reason not to approve that specific proposal?
- The dashboard cached the prior result; no DataFrame operation or underlying data requires correction.
- A shared fabricated key merges unrelated customers into one identity. This would apply to an alternative change for `payment_decisions`.
- Removing checks trades temporary performance for silent data corruption. This would apply to another option for `payment_decisions`.
- A presentation filter does not protect downstream consumers or data contracts. The proposal does not repair `payment_decisions`.
Explanation
Correct Answer
D:A presentation filter does not protect downstream consumers or data contracts. The proposal does not repair `payment_decisions`. This identifies the known limitation of the exact proposed action. A presentation filter does not protect downstream consumers or data contracts. The underlying problem remains a null business key breaks reliable joins and can create untraceable billing records; Enforce a non-null key quality expectation or validation rule and route rejected records for investigation before publishing gold instead targets that failure.
Incorrect Answers
A:The dashboard cached the prior result; no DataFrame operation or underlying data requires correction. The case provides a concrete failing data or control path; a mere alternate theory does not address a null business key breaks reliable joins and can create untraceable billing records.
B:A shared fabricated key merges unrelated customers into one identity. This would apply to an alternative change for `payment_decisions`. This failure applies to the separate proposal "Replace missing customer IDs with a common literal value".
C:Removing checks trades temporary performance for silent data corruption. This would apply to another option for `payment_decisions`. This limitation belongs to the different alternative "Disable all quality checks when ingestion latency increases". The stem calls for the specific failure of "Allow the rows into gold and hope a later dashboard filter hides them".
Question 329
At a census statistics agency, a silver customer dimension occasionally contains null customer IDs that should never reach billing gold tables. The dataset `payment_decisions` is in scope, and downstream dashboards cannot accept fabricated values. Which decision preserves the stated data contract while keeping the change scope proportionate?
- Keep a visible quarantine route instead of silently replacing missing identifiers
- Disable all quality checks when ingestion latency increases; monitor results only after launch of `payment_decisions`.
- Replace missing customer IDs with a common literal value; monitor results only after launch of `payment_decisions`.
- Allow the rows into gold and hope a later dashboard filter hides them; monitor results only after launch of `payment_decisions`.
Explanation
Correct Answer
A:Keep a visible quarantine route instead of silently replacing missing identifiers. This explicitly bounds the compromise while still addressing a null business key breaks reliable joins and can create untraceable billing records. Acceptance relies on inject a null customer ID and verify the exception record is visible but absent from trusted gold rather than simply declaring the change complete.
Incorrect Answers
B:Disable all quality checks when ingestion latency increases; monitor results only after launch of `payment_decisions`. Removing checks trades temporary performance for silent data corruption. Monitoring after deployment does not make this approach a safe bounded tradeoff for `payment_decisions`.
C:Replace missing customer IDs with a common literal value; monitor results only after launch of `payment_decisions`. A shared fabricated key merges unrelated customers into one identity. Monitoring after deployment does not make this approach a safe bounded tradeoff for `payment_decisions`.
D:Allow the rows into gold and hope a later dashboard filter hides them; monitor results only after launch of `payment_decisions`. A presentation filter does not protect downstream consumers or data contracts. Monitoring after deployment does not make this approach a safe bounded tradeoff for `payment_decisions`.
Question 330
At a census statistics agency, a silver customer dimension occasionally contains null customer IDs that should never reach billing gold tables. The dataset `payment_decisions` is in scope, and downstream dashboards cannot accept fabricated values. What verification provides evidence of correct behavior under the original adverse condition?
- Run only a normal, unchanged `payment_decisions` sample and review null key rate.
- Review a platform-level availability report for `payment_decisions` without testing whether a null business key breaks reliable joins and can create untraceable billing records.
- Check that the intended configuration exists for `payment_decisions` and that a healthy run succeeded, without inducing the earlier failure.
- Inject a null customer ID and verify the exception record is visible but absent from trusted gold
Explanation
Correct Answer
D:Inject a null customer ID and verify the exception record is visible but absent from trusted gold. This reproduces the scenario-dependent condition, verifies null key rate, rejected-record sample, silver-to-gold counts, and quality rule events, and directly tests whether a null business key breaks reliable joins and can create untraceable billing records is resolved. Observed results, not intent, determine approval.
Incorrect Answers
A:Run only a normal, unchanged `payment_decisions` sample and review null key rate. A normal sample does not challenge a null business key breaks reliable joins and can create untraceable billing records.
B:Review a platform-level availability report for `payment_decisions` without testing whether a null business key breaks reliable joins and can create untraceable billing records. Aggregate availability does not establish the behavior of affected `payment_decisions` records. It cannot replace the direct exercise: inject a null customer ID and verify the exception record is visible but absent from trusted gold.
C:Check that the intended configuration exists for `payment_decisions` and that a healthy run succeeded, without inducing the earlier failure. Installed configuration and a healthy run are insufficient to prove the failure mode is gone. The original issue was a null business key breaks reliable joins and can create untraceable billing records.