Certified Data Engineer Associate Practice Test 4
Question 30
At an airport services provider, a nightly batch transform starts a costly interactive cluster and leaves it running idle after the successful run. The dataset `invoice_lines` is in scope, and the control board requires an auditable data lineage. What verification provides evidence of correct behavior under the original adverse condition?
- Schedule two equivalent batches and check task results, runtime startup, and cost until cluster termination
- Run only a normal, unchanged `invoice_lines` sample and review DBU cost per successful batch.
- Check that the intended configuration exists for `invoice_lines` and that a healthy run succeeded, without inducing the earlier failure.
- Review a platform-level availability report for `invoice_lines` without testing whether the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle.
Explanation
Correct Answer
A:Schedule two equivalent batches and check task results, runtime startup, and cost until cluster termination. This reproduces the scenario-dependent condition, verifies DBU cost per successful batch, cluster lifetime, job status, and library compatibility, and directly tests whether the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle is resolved. Observed results, not intent, determine approval.
Incorrect Answers
B:Run only a normal, unchanged `invoice_lines` sample and review DBU cost per successful batch. A normal sample does not challenge the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle.
C:Check that the intended configuration exists for `invoice_lines` 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 the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle.
D:Review a platform-level availability report for `invoice_lines` without testing whether the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle. Aggregate availability does not establish the behavior of affected `invoice_lines` records. It cannot replace the direct exercise: schedule two equivalent batches and check task results, runtime startup, and cost until cluster termination.
Question 31
At a digital publishing service, interactive SQL dashboards compete with hourly ingestion notebooks on an always-on general-purpose cluster. The dataset `baggage_scans` is in scope, and workspace users must retain least-privilege access. Which investigation best tests the stated constraint with direct data?
- Compare the current configuration and the signed change approval, without source-row evidence
- Review completed run status and aggregate cluster health without record-level correlation
- Examine only a normal successful run of `baggage_scans` and trend query p95 without recreating the fault.
- Correlate query p95, cluster utilization by workload across affected and unaffected `R034-CASE-007` runs.
Explanation
Correct Answer
D:Correlate query p95, cluster utilization by workload across affected and unaffected `R034-CASE-007` runs.. By joining query p95, cluster utilization by workload, idle spend, and job run duration across the affected and unaffected conditions, the investigation directly tests the recorded mechanism: one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low. It also sets a measurable baseline for run dashboard and ETL load concurrently, compare p95 and DBUs by workload, and observe independent scaling.
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 query p95, cluster utilization by workload, idle spend, and job run duration, the direct means of distinguishing one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low.
B:Review completed run status and aggregate cluster health without record-level correlation. A successful aggregate platform state can coexist with one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low. This check never observes the affected `baggage_scans` records or their causal path.
C:Examine only a normal successful run of `baggage_scans` and trend query p95 without recreating the fault. A healthy-run trend can be useful, but it does not expose one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low. The failing-path query p95, cluster utilization by workload, idle spend, and job run duration has to be correlated before attributing cause.
Question 32
At a digital publishing service, interactive SQL dashboards compete with hourly ingestion notebooks on an always-on general-purpose cluster. The dataset `baggage_scans` is in scope, and workspace users must retain least-privilege access. Which option targets the actual technical dependency rather than an incidental symptom?
- Run dashboards from the ingestion notebook driver; use that approach for `baggage_scans`.
- Schedule all dashboard users to query only after the hourly load; use that approach for `baggage_scans`.
- Increase the shared interactive cluster size permanently; use that approach for `baggage_scans`.
- Route BI queries to an appropriately sized SQL warehouse and scheduled ingestion to job compute with independent cost controls
Explanation
Correct Answer
D:Route BI queries to an appropriately sized SQL warehouse and scheduled ingestion to job compute with independent cost controls. This directly addresses one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low. The design should be accepted only after run dashboard and ETL load concurrently, compare p95 and DBUs by workload, and observe independent scaling; the related operational boundary is to spend separately on demand-driven SQL serving and ephemeral job execution instead of one large idle shared cluster.
Incorrect Answers
A:Run dashboards from the ingestion notebook driver; use that approach for `baggage_scans`. Driver-hosted BI queries entangle scheduled ETL failures with interactive service. For `baggage_scans`, this cannot remove the causal limitation: one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low.
B:Schedule all dashboard users to query only after the hourly load; use that approach for `baggage_scans`. Serial access violates interactive dashboard expectations without fixing compute separation. For `baggage_scans`, this cannot remove the causal limitation: one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low.
C:Increase the shared interactive cluster size permanently; use that approach for `baggage_scans`. More always-on capacity raises idle cost and leaves the workloads competing. For `baggage_scans`, this cannot remove the causal limitation: one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low.
Question 33
At a digital publishing service, interactive SQL dashboards compete with hourly ingestion notebooks on an always-on general-purpose cluster. The dataset `baggage_scans` is in scope, and workspace users must retain least-privilege access. A teammate proposes: "Increase the shared interactive cluster size permanently." Which technical limitation is the decisive reason not to approve that specific proposal?
- The anomaly is entirely a compute sizing problem; the file format and catalog design are not involved.
- Serial access violates interactive dashboard expectations without fixing compute separation. This would apply to another option for `baggage_scans`.
- Driver-hosted BI queries entangle scheduled ETL failures with interactive service. This would apply to an alternative change for `baggage_scans`.
- More always-on capacity raises idle cost and leaves the workloads competing. The proposal does not repair `baggage_scans`.
Explanation
Correct Answer
D:More always-on capacity raises idle cost and leaves the workloads competing. The proposal does not repair `baggage_scans`. This identifies the known limitation of the exact proposed action. More always-on capacity raises idle cost and leaves the workloads competing. The underlying problem remains one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low; Route BI queries to an appropriately sized SQL warehouse and scheduled ingestion to job compute with independent cost controls instead targets that failure.
Incorrect Answers
A:The anomaly is entirely a compute sizing problem; the file format and catalog design are not involved. The case provides a concrete failing data or control path; a mere alternate theory does not address one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low.
B:Serial access violates interactive dashboard expectations without fixing compute separation. This would apply to another option for `baggage_scans`. This limitation belongs to the different alternative "Schedule all dashboard users to query only after the hourly load". The stem calls for the specific failure of "Increase the shared interactive cluster size permanently".
C:Driver-hosted BI queries entangle scheduled ETL failures with interactive service. This would apply to an alternative change for `baggage_scans`. This failure applies to the separate proposal "Run dashboards from the ingestion notebook driver".
Question 34
At a digital publishing service, interactive SQL dashboards compete with hourly ingestion notebooks on an always-on general-purpose cluster. The dataset `baggage_scans` is in scope, and workspace users must retain least-privilege access. What operational boundary should be documented for approval?
- Spend separately on demand-driven SQL serving and ephemeral job execution instead of one large idle shared cluster
- Increase the shared interactive cluster size permanently
- Schedule all dashboard users to query only after the hourly load
- Run dashboards from the ingestion notebook driver; monitor results only after launch of `baggage_scans`.
Explanation
Correct Answer
A:Spend separately on demand-driven SQL serving and ephemeral job execution instead of one large idle shared cluster. This explicitly bounds the compromise while still addressing one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low. Acceptance relies on run dashboard and ETL load concurrently, compare p95 and DBUs by workload, and observe independent scaling rather than simply declaring the change complete.
Incorrect Answers
B:Increase the shared interactive cluster size permanently. More always-on capacity raises idle cost and leaves the workloads competing. Monitoring after deployment does not make this approach a safe bounded tradeoff for `baggage_scans`.
C:Schedule all dashboard users to query only after the hourly load. Serial access violates interactive dashboard expectations without fixing compute separation. Monitoring after deployment does not make this approach a safe bounded tradeoff for `baggage_scans`.
D:Run dashboards from the ingestion notebook driver; monitor results only after launch of `baggage_scans`. Driver-hosted BI queries entangle scheduled ETL failures with interactive service. Monitoring after deployment does not make this approach a safe bounded tradeoff for `baggage_scans`.
Question 35
At a digital publishing service, interactive SQL dashboards compete with hourly ingestion notebooks on an always-on general-purpose cluster. The dataset `baggage_scans` is in scope, and workspace users must retain least-privilege access. Which validation should be required in the release record?
- Check that the intended configuration exists for `baggage_scans` and that a healthy run succeeded, without inducing the earlier failure.
- Run only a normal, unchanged `baggage_scans` sample and review query p95.
- Run dashboard and ETL load concurrently, compare p95 and DBUs by workload, and observe independent scaling
- Review a platform-level availability report for `baggage_scans` without testing whether one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low.
Explanation
Correct Answer
C:Run dashboard and ETL load concurrently, compare p95 and DBUs by workload, and observe independent scaling. This reproduces the scenario-dependent condition, verifies query p95, cluster utilization by workload, idle spend, and job run duration, and directly tests whether one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low is resolved. Observed results, not intent, determine approval.
Incorrect Answers
A:Check that the intended configuration exists for `baggage_scans` 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 one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low.
B:Run only a normal, unchanged `baggage_scans` sample and review query p95. A normal sample does not challenge one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low.
D:Review a platform-level availability report for `baggage_scans` without testing whether one fixed general-purpose cluster shares contention and charges for idle capacity even when query demand is low. Aggregate availability does not establish the behavior of affected `baggage_scans` records. It cannot replace the direct exercise: run dashboard and ETL load concurrently, compare p95 and DBUs by workload, and observe independent scaling.
Question 36
At a medical device company, a nightly batch transform starts a costly interactive cluster and leaves it running idle after the successful run. The dataset `program_enrollments` is in scope, and the control board requires an auditable data lineage. Before choosing a remedy, which investigation actually distinguishes the failure mechanism from a green dashboard?
- Review completed run status and aggregate cluster health without record-level correlation
- Inspect normal-period success metrics for the stated dataset without reproducing the source error
- Compare the current configuration and the signed change approval, without source-row evidence
- Correlate DBU cost per successful batch, cluster lifetime across affected and unaffected `R034-CASE-008` runs.
Explanation
Correct Answer
D:Correlate DBU cost per successful batch, cluster lifetime across affected and unaffected `R034-CASE-008` runs.. By joining DBU cost per successful batch, cluster lifetime, job status, and library compatibility across the affected and unaffected conditions, the investigation directly tests the recorded mechanism: the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle. It also sets a measurable baseline for schedule two equivalent batches and check task results, runtime startup, and cost until cluster termination.
Incorrect Answers
A:Review completed run status and aggregate cluster health without record-level correlation. A successful aggregate platform state can coexist with the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle. This check never observes the affected `program_enrollments` records or their causal path.
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 the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle. The failing-path DBU cost per successful batch, cluster lifetime, job status, and library compatibility has to be correlated before attributing cause.
C: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 DBU cost per successful batch, cluster lifetime, job status, and library compatibility, the direct means of distinguishing the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle.
Question 37
At a medical device company, a nightly batch transform starts a costly interactive cluster and leaves it running idle after the successful run. The dataset `program_enrollments` is in scope, and the control board requires an auditable data lineage. Which implementation best resolves the stated defect without breaking the business constraint?
- Use a Lakeflow Jobs task with suitable job or serverless compute and release resources when the workload ends
- Set the all-purpose cluster to a much larger worker count; use that approach for `program_enrollments`.
- Convert the pipeline to manual notebook runs on the same cluster; use that approach for `program_enrollments`.
- Use a SQL warehouse for arbitrary Python libraries without checking execution support; use that approach for `program_enrollments`.
Explanation
Correct Answer
A:Use a Lakeflow Jobs task with suitable job or serverless compute and release resources when the workload ends. This directly addresses the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle. The design should be accepted only after schedule two equivalent batches and check task results, runtime startup, and cost until cluster termination; the related operational boundary is to accept predictable job startup overhead to avoid continuous idle compute charges.
Incorrect Answers
B:Set the all-purpose cluster to a much larger worker count; use that approach for `program_enrollments`. Larger idle instances exacerbate cost without improving shutdown behavior. For `program_enrollments`, this cannot remove the causal limitation: the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle.
C:Convert the pipeline to manual notebook runs on the same cluster; use that approach for `program_enrollments`. Manual execution loses repeatable orchestration while retaining idle spend. For `program_enrollments`, this cannot remove the causal limitation: the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle.
D:Use a SQL warehouse for arbitrary Python libraries without checking execution support; use that approach for `program_enrollments`. A SQL warehouse is not a substitute for validating the Python job execution environment. For `program_enrollments`, this cannot remove the causal limitation: the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle.
Question 38
At a medical device company, a nightly batch transform starts a costly interactive cluster and leaves it running idle after the successful run. The dataset `program_enrollments` is in scope, and the control board requires an auditable data lineage. A teammate proposes: "Use a SQL warehouse for arbitrary Python libraries without checking execution support." Which technical limitation is the decisive reason not to approve that specific proposal?
- The anomaly is entirely a compute sizing problem; the file format and catalog design are not involved.
- Larger idle instances exacerbate cost without improving shutdown behavior. This would apply to an alternative change for `program_enrollments`.
- Manual execution loses repeatable orchestration while retaining idle spend. This would apply to another option for `program_enrollments`.
- A SQL warehouse is not a substitute for validating the Python job execution environment. The proposal does not repair `program_enrollments`.
Explanation
Correct Answer
D:A SQL warehouse is not a substitute for validating the Python job execution environment. The proposal does not repair `program_enrollments`. This identifies the known limitation of the exact proposed action. A SQL warehouse is not a substitute for validating the Python job execution environment. The underlying problem remains the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle; Use a Lakeflow Jobs task with suitable job or serverless compute and release resources when the workload ends instead targets that failure.
Incorrect Answers
A:The anomaly is entirely a compute sizing problem; the file format and catalog design are not involved. The case provides a concrete failing data or control path; a mere alternate theory does not address the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle.
B:Larger idle instances exacerbate cost without improving shutdown behavior. This would apply to an alternative change for `program_enrollments`. This failure applies to the separate proposal "Set the all-purpose cluster to a much larger worker count".
C:Manual execution loses repeatable orchestration while retaining idle spend. This would apply to another option for `program_enrollments`. This limitation belongs to the different alternative "Convert the pipeline to manual notebook runs on the same cluster". The stem calls for the specific failure of "Use a SQL warehouse for arbitrary Python libraries without checking execution support".
Question 39
At a medical device company, a nightly batch transform starts a costly interactive cluster and leaves it running idle after the successful run. The dataset `program_enrollments` is in scope, and the control board requires an auditable data lineage. Which bounded tradeoff would preserve the core outcome and control operational risk?
- Convert the pipeline to manual notebook runs on the same cluster
- Accept predictable job startup overhead to avoid continuous idle compute charges
- Set the all-purpose cluster to a much larger worker count
- Use a SQL warehouse for arbitrary Python libraries without checking execution support
Explanation
Correct Answer
B:Accept predictable job startup overhead to avoid continuous idle compute charges. This explicitly bounds the compromise while still addressing the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle. Acceptance relies on schedule two equivalent batches and check task results, runtime startup, and cost until cluster termination rather than simply declaring the change complete.
Incorrect Answers
A:Convert the pipeline to manual notebook runs on the same cluster. Manual execution loses repeatable orchestration while retaining idle spend. Monitoring after deployment does not make this approach a safe bounded tradeoff for `program_enrollments`.
C:Set the all-purpose cluster to a much larger worker count. Larger idle instances exacerbate cost without improving shutdown behavior. Monitoring after deployment does not make this approach a safe bounded tradeoff for `program_enrollments`.
D:Use a SQL warehouse for arbitrary Python libraries without checking execution support. A SQL warehouse is not a substitute for validating the Python job execution environment. Monitoring after deployment does not make this approach a safe bounded tradeoff for `program_enrollments`.
Question 40
At a medical device company, a nightly batch transform starts a costly interactive cluster and leaves it running idle after the successful run. The dataset `program_enrollments` is in scope, and the control board requires an auditable data lineage. Before release, which acceptance test would actually demonstrate that the root problem is controlled?
- Check that the intended configuration exists for `program_enrollments` and that a healthy run succeeded, without inducing the earlier failure.
- Schedule two equivalent batches and check task results, runtime startup, and cost until cluster termination
- Run only a normal, unchanged `program_enrollments` sample and review DBU cost per successful batch.
- Review a platform-level availability report for `program_enrollments` without testing whether the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle.
Explanation
Correct Answer
B:Schedule two equivalent batches and check task results, runtime startup, and cost until cluster termination. This reproduces the scenario-dependent condition, verifies DBU cost per successful batch, cluster lifetime, job status, and library compatibility, and directly tests whether the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle is resolved. Observed results, not intent, determine approval.
Incorrect Answers
A:Check that the intended configuration exists for `program_enrollments` 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 the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle.
C:Run only a normal, unchanged `program_enrollments` sample and review DBU cost per successful batch. A normal sample does not challenge the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle.
D:Review a platform-level availability report for `program_enrollments` without testing whether the scheduled pipeline inherits a persistent development compute lifecycle instead of an on-demand job lifecycle. Aggregate availability does not establish the behavior of affected `program_enrollments` records. It cannot replace the direct exercise: schedule two equivalent batches and check task results, runtime startup, and cost until cluster termination.