Google Professional Cloud Architect Practice Test 21
Question 200
At Financial remittance company, the Google Cloud design for card token ledger is affected because a transactional SQL database runs out of horizontal write capacity across continents. A review found that design review finds one region primary is a regulatory single point of failure. The proposed response must account for the requirement that ledger updates require strongly consistent multi-region transactions. A deployment ticket has been closed. What direct evidence would justify solution signoff?
- Stage dual-run ledger comparison and migrate one currency corridor
- Reconcile unique transfer IDs and fail a region during concurrent writes
- Create independent writable ledgers and reconcile disputed balances weekly
- Test geographic commit latency and exact data-sovereignty requirements
Explanation
Correct Answer
B:
Reconcile unique transfer IDs and fail a region during concurrent writes is a direct acceptance test of the path implicated by design review finds one region primary is a regulatory single point of failure. It checks that the original risk (a transactional SQL database runs out of horizontal write capacity across continents) has actually been controlled and tests that permitted operations still meet the requirement that ledger updates require strongly consistent multi-region transactions.
Incorrect Answers
A:
The alternative “Stage dual-run ledger comparison and migrate one currency corridor” documents change execution but not its operational acceptance result. In the reported observation, design review finds one region primary is a regulatory single point of failure. The option “Reconcile unique transfer IDs and fail a region during concurrent writes” instead addresses the requested acceptance verification.
C:
Deferred conflict reconciliation is unacceptable for money movement. For this question, it is not a defensible choice for acceptance verification and it does not satisfy the requirement that ledger updates require strongly consistent multi-region transactions.
D:
The alternative “Test geographic commit latency and exact data-sovereignty requirements” collects prechange evidence without confirming that the revised service meets its outcome. In the reported observation, design review finds one region primary is a regulatory single point of failure. The option “Reconcile unique transfer IDs and fail a region during concurrent writes” instead addresses the requested acceptance verification.
Question 201
At National rail passenger service, the Google Cloud design for disaster recovery runbook is affected because backup jobs report success but operational teams have never restored ticket data. A review found that DR rehearsal cannot identify who owns recovery steps. The proposed response must account for the requirement that tickets must be sold again within 30 minutes after a primary failure. A deployment ticket has been closed. What direct evidence would justify solution signoff?
- Increase the backup frequency and skip every restoration exercise
- Inventory recovery dependencies, owners, RTO/RPO and validated backups
- Schedule isolated recovery drills and sign off runbook changes
- Time a full restore and ticket purchase from an isolated recovery environment
Explanation
Correct Answer
D:
Time a full restore and ticket purchase from an isolated recovery environment is a direct acceptance test of the path implicated by DR rehearsal cannot identify who owns recovery steps. It checks that the original risk (backup jobs report success but operational teams have never restored ticket data) has actually been controlled and tests that permitted operations still meet the requirement that tickets must be sold again within 30 minutes after a primary failure.
Incorrect Answers
A:
Backup completion does not prove recoverability or an executable procedure. For this question, it is not a defensible choice for acceptance verification and it does not satisfy the requirement that tickets must be sold again within 30 minutes after a primary failure.
B:
The alternative “Inventory recovery dependencies, owners, RTO/RPO and validated backups” collects prechange evidence without confirming that the revised service meets its outcome. In the reported observation, DR rehearsal cannot identify who owns recovery steps. The option “Time a full restore and ticket purchase from an isolated recovery environment” instead addresses the requested acceptance verification.
C:
The alternative “Schedule isolated recovery drills and sign off runbook changes” documents change execution but not its operational acceptance result. In the reported observation, DR rehearsal cannot identify who owns recovery steps. The option “Time a full restore and ticket purchase from an isolated recovery environment” instead addresses the requested acceptance verification.
Question 202
Case study: Metropolitan bus operator: customer API throttling
Metropolitan bus operator relies on customer API throttling to deliver its operational commitments. The current architectural problem is that one external client causes an overloaded trip-planning API while most users are normal. The team has recorded this evidence: gateway logs show one tenant produces 80% of requests. Any proposed response must preserve the requirement that emergency schedule lookups must remain available to all operators.
At Metropolitan bus operator, the customer API throttling workload has a documented risk: one external client causes an overloaded trip-planning API while most users are normal. The review found that gateway logs show one tenant produces 80% of requests. The architecture must also ensure that emergency schedule lookups must remain available to all operators. The team must deliver the approved control safely. Which deployment sequence is most appropriate?
- Track quota denials, tenant fairness and false throttling
- Introduce preview quotas and monitor protected emergency routes
- Design per-tenant API quotas and fairness policies using Apigee
- Apply a single global rate cap that blocks every tenant equally
Explanation
Correct Answer
B:
Introduce preview quotas and monitor protected emergency routes is a controlled means to introduce the approved architecture while meeting the requirement that emergency schedule lookups must remain available to all operators. Staged delivery, attributable change evidence and rollback guard against an otherwise avoidable service disruption.
Incorrect Answers
A:
The alternative “Track quota denials, tenant fairness and false throttling” addresses later monitoring rather than the actual controlled deployment. In the reported observation, gateway logs show one tenant produces 80% of requests. The option “Introduce preview quotas and monitor protected emergency routes” instead addresses the requested controlled implementation.
C:
The alternative “Design per-tenant API quotas and fairness policies using Apigee” describes the chosen design, but not how it is introduced safely. In the reported observation, gateway logs show one tenant produces 80% of requests. The option “Introduce preview quotas and monitor protected emergency routes” instead addresses the requested controlled implementation.
D:
A global cap penalizes unaffected users rather than noisy clients. For this question, it is not a defensible choice for controlled implementation and it does not satisfy the requirement that emergency schedule lookups must remain available to all operators.
Question 203
Public grant agency plans to update executive decision analysis. Investigators found that business case omits users reliant on assistive technologies, consistent with the reported problem that project review scores optional chatbot convenience higher than accessibility and regulatory filing. The response must satisfy the requirement that benefit delivery must meet accessible filing requirements. The approved change has been applied. Which validation provides the clearest evidence the original problem is fixed?
- Translate user needs into measurable outcomes, compliance and risk
- Validate assisted-technology filing success against user acceptance criteria
- Sequence the accessibility remediation with user testing and business ownership
- Approve the chatbot solely because it uses the newest language model
Explanation
Correct Answer
B:
Validate assisted-technology filing success against user acceptance criteria is a direct acceptance test of the path implicated by business case omits users reliant on assistive technologies. It checks that the original risk (project review scores optional chatbot convenience higher than accessibility and regulatory filing) has actually been controlled and tests that permitted operations still meet the requirement that benefit delivery must meet accessible filing requirements.
Incorrect Answers
A:
The alternative “Translate user needs into measurable outcomes, compliance and risk” collects prechange evidence without confirming that the revised service meets its outcome. In the reported observation, business case omits users reliant on assistive technologies. The option “Validate assisted-technology filing success against user acceptance criteria” instead addresses the requested acceptance verification.
C:
The alternative “Sequence the accessibility remediation with user testing and business ownership” documents change execution but not its operational acceptance result. In the reported observation, business case omits users reliant on assistive technologies. The option “Validate assisted-technology filing success against user acceptance criteria” instead addresses the requested acceptance verification.
D:
Novelty does not meet the required filing accessibility outcome. For this question, it is not a defensible choice for acceptance verification and it does not satisfy the requirement that benefit delivery must meet accessible filing requirements.
Question 204
Case study: Financial records archive: legacy batch identity
Financial records archive relies on legacy batch identity to deliver its operational commitments. The current architectural problem is that migration plan provisions new service accounts without mapping on-prem scheduled-job permissions. The team has recorded this evidence: audit reveals a shared service password embedded in payroll scripts. Any proposed response must preserve the requirement that regulatory evidence must retain actor attribution after migration.
Financial records archive plans to update legacy batch identity. Investigators found that audit reveals a shared service password embedded in payroll scripts, consistent with the reported problem that migration plan provisions new service accounts without mapping on-prem scheduled-job permissions. The response must satisfy the requirement that regulatory evidence must retain actor attribution after migration. The team needs an operations signal rather than another design or one-time test. What should it track?
- Verify each job runs as intended identity and cannot read unrelated data
- Monitor privileged impersonation, job failures and revoked-account reuse
- Copy shared plaintext service passwords into cloud VM metadata
- Use separate workload identities and short-lived credentials with compatible batch roles
Explanation
Correct Answer
B:
Monitor privileged impersonation, job failures and revoked-account reuse reveals recurrence, drift or changing dependencies after launch. It provides a repeatable signal for the risk (migration plan provisions new service accounts without mapping on-prem scheduled-job permissions) rather than treating the first successful test as permanent assurance.
Incorrect Answers
A:
The alternative “Verify each job runs as intended identity and cannot read unrelated data” is a one-time acceptance result rather than ongoing operational assurance. In the reported observation, audit reveals a shared service password embedded in payroll scripts. The option “Monitor privileged impersonation, job failures and revoked-account reuse” instead addresses the requested ongoing operational assurance.
C:
Replicating shared secrets sustains the original identity risk. For this question, it is not a defensible choice for ongoing operational assurance and it does not satisfy the requirement that regulatory evidence must retain actor attribution after migration.
D:
The alternative “Use separate workload identities and short-lived credentials with compatible batch roles” selects a design, but does not provide the ongoing signal required to detect drift. In the reported observation, audit reveals a shared service password embedded in payroll scripts. The option “Monitor privileged impersonation, job failures and revoked-account reuse” instead addresses the requested ongoing operational assurance.
Question 205
The database schema rollout environment at National blood supply service faces this issue: application instances migrate separately from the database schema and old pods fail on renamed columns. Supporting evidence shows that staging test shows older binary cannot read new tables. The architect must preserve the requirement that donation inventory cannot pause through rolling deployment. Before approving a solution, which analysis most directly clarifies the underlying dependency or control gap?
- Drop the old column before the old application instances are drained
- Analyze schema backward compatibility and the release dependency graph
- Define expand-migrate-contract schema evolution with compatible views
- Apply additive schema first, migrate services, then remove old columns after validation
Explanation
Correct Answer
B:
Analyze schema backward compatibility and the release dependency graph is the appropriate diagnostic because the observation—staging test shows older binary cannot read new tables—does not by itself establish every dependency, authorization rule or failure boundary. The findings explain the reported risk (application instances migrate separately from the database schema and old pods fail on renamed columns) before the design is approved.
Incorrect Answers
A:
A destructive schema change strands older active binaries. For this question, it is not a defensible choice for root-cause evidence and it does not satisfy the requirement that donation inventory cannot pause through rolling deployment.
C:
The alternative “Define expand-migrate-contract schema evolution with compatible views” is target architecture selection; it does not establish why the observed path failed. In the reported observation, staging test shows older binary cannot read new tables. The option “Analyze schema backward compatibility and the release dependency graph” instead addresses the requested root-cause evidence.
D:
The alternative “Apply additive schema first, migrate services, then remove old columns after validation” is rollout work that may modify the wrong dependency before root-cause evidence is collected. In the reported observation, staging test shows older binary cannot read new tables. The option “Analyze schema backward compatibility and the release dependency graph” instead addresses the requested root-cause evidence.
Question 206
Case study: Regional sports venue: deploy-time traffic splitting
Regional sports venue relies on deploy-time traffic splitting to deliver its operational commitments. The current architectural problem is that teams replace a full production service revision in one step before stadium events. The team has recorded this evidence: last release caused 15 minutes of checkout errors during a full swap. Any proposed response must preserve the requirement that ticket scanning must stay available throughout peak admission.
Consider Regional sports venue and its deploy-time traffic splitting workload. The issue is that teams replace a full production service revision in one step before stadium events; the current observation is that last release caused 15 minutes of checkout errors during a full swap. The team must ensure that ticket scanning must stay available throughout peak admission. The change committee needs to understand the current system rather than install a fix yet. What should it do?
- Introduce weighted traffic shifting and automatic rollback thresholds
- Inspect release mechanics, rollback needs and request session behavior
- Design progressive canary or blue-green release with measurable safety gates
- Use a full instantaneous replace for every release and observe failures afterward
Explanation
Correct Answer
B:
Inspect release mechanics, rollback needs and request session behavior is the appropriate diagnostic because the observation—last release caused 15 minutes of checkout errors during a full swap—does not by itself establish every dependency, authorization rule or failure boundary. The findings explain the reported risk (teams replace a full production service revision in one step before stadium events) before the design is approved.
Incorrect Answers
A:
The alternative “Introduce weighted traffic shifting and automatic rollback thresholds” is rollout work that may modify the wrong dependency before root-cause evidence is collected. In the reported observation, last release caused 15 minutes of checkout errors during a full swap. The option “Inspect release mechanics, rollback needs and request session behavior” instead addresses the requested root-cause evidence.
C:
The alternative “Design progressive canary or blue-green release with measurable safety gates” is target architecture selection; it does not establish why the observed path failed. In the reported observation, last release caused 15 minutes of checkout errors during a full swap. The option “Inspect release mechanics, rollback needs and request session behavior” instead addresses the requested root-cause evidence.
D:
A big-bang change concentrates customer impact. For this question, it is not a defensible choice for root-cause evidence and it does not satisfy the requirement that ticket scanning must stay available throughout peak admission.
Question 207
The reservation mainframe bridge environment at International hotel operator faces this issue: reservation jobs on a mainframe call undocumented nightly batch services. Supporting evidence shows that discovery captures eleven downstream dependencies absent from design diagrams. The architect must preserve the requirement that bookings may not be lost during peak holiday weeks. Go-live and initial tests passed. Which ongoing measure is most useful to catch control drift or regression?
- Design strangler APIs around stable reservation capabilities with coexistence
- Cut all mainframe links on the first deployment without inventory reconciliation
- Compare room inventory and customer booking consistency across both systems
- Track batch reconciliation, shadow traffic failures and residual dependencies
Explanation
Correct Answer
D:
Track batch reconciliation, shadow traffic failures and residual dependencies reveals recurrence, drift or changing dependencies after launch. It provides a repeatable signal for the risk (reservation jobs on a mainframe call undocumented nightly batch services) rather than treating the first successful test as permanent assurance.
Incorrect Answers
A:
The alternative “Design strangler APIs around stable reservation capabilities with coexistence” selects a design, but does not provide the ongoing signal required to detect drift. In the reported observation, discovery captures eleven downstream dependencies absent from design diagrams. The option “Track batch reconciliation, shadow traffic failures and residual dependencies” instead addresses the requested ongoing operational assurance.
B:
A big-bang cutover risks invisible dependencies and lost bookings. For this question, it is not a defensible choice for ongoing operational assurance and it does not satisfy the requirement that bookings may not be lost during peak holiday weeks.
C:
The alternative “Compare room inventory and customer booking consistency across both systems” is a one-time acceptance result rather than ongoing operational assurance. In the reported observation, discovery captures eleven downstream dependencies absent from design diagrams. The option “Track batch reconciliation, shadow traffic failures and residual dependencies” instead addresses the requested ongoing operational assurance.
Question 208
Case study: Coastal emergency response unit: on-prem data warehouse
Coastal emergency response unit relies on on-prem data warehouse to deliver its operational commitments. The current architectural problem is that warehouse queries depend on nightly CSV copies with missing change history. The team has recorded this evidence: profile shows source records revised after files are exported. Any proposed response must preserve the requirement that emergency operations need reproducible data not stale extracts.
Consider Coastal emergency response unit and its on-prem data warehouse workload. The issue is that warehouse queries depend on nightly CSV copies with missing change history; the current observation is that profile shows source records revised after files are exported. The team must ensure that emergency operations need reproducible data not stale extracts. Which acceptance test should demonstrate both protection against the failure and continued authorized service?
- Copy the same static CSV once and label it a real-time migration
- Run source and target in parallel across a full monthly correction cycle
- Assess source CDC availability, schema drift and acceptable analytical lag
- Recompute totals with late corrections and verify audit history
Explanation
Correct Answer
D:
Recompute totals with late corrections and verify audit history is a direct acceptance test of the path implicated by profile shows source records revised after files are exported. It checks that the original risk (warehouse queries depend on nightly CSV copies with missing change history) has actually been controlled and tests that permitted operations still meet the requirement that emergency operations need reproducible data not stale extracts.
Incorrect Answers
A:
A one-time extract cannot propagate later corrections. For this question, it is not a defensible choice for acceptance verification and it does not satisfy the requirement that emergency operations need reproducible data not stale extracts.
B:
The alternative “Run source and target in parallel across a full monthly correction cycle” documents change execution but not its operational acceptance result. In the reported observation, profile shows source records revised after files are exported. The option “Recompute totals with late corrections and verify audit history” instead addresses the requested acceptance verification.
C:
The alternative “Assess source CDC availability, schema drift and acceptable analytical lag” collects prechange evidence without confirming that the revised service meets its outcome. In the reported observation, profile shows source records revised after files are exported. The option “Recompute totals with late corrections and verify audit history” instead addresses the requested acceptance verification.
Question 209
At Climate modeling consortium, the GPU inference profiling workload has a documented risk: AI inference costs spike although request counts stay flat. The review found that profile shows a new preprocessing step copies tensors repeatedly between CPU and GPU. The architecture must also ensure that cost optimization must not lower model output quality. What follow-up metric or alert will identify emerging failures after the initial acceptance test?
- Reduce redundant tensor transfer and right-size accelerators based on measured bottleneck
- Compare output parity, GPU utilization and p99 inference latency
- Monitor transfer bandwidth, idle accelerator time and model drift
- Buy more GPUs without measuring the memory-transfer bottleneck
Explanation
Correct Answer
C:
Monitor transfer bandwidth, idle accelerator time and model drift reveals recurrence, drift or changing dependencies after launch. It provides a repeatable signal for the risk (AI inference costs spike although request counts stay flat) rather than treating the first successful test as permanent assurance.
Incorrect Answers
A:
The alternative “Reduce redundant tensor transfer and right-size accelerators based on measured bottleneck” selects a design, but does not provide the ongoing signal required to detect drift. In the reported observation, profile shows a new preprocessing step copies tensors repeatedly between CPU and GPU. The option “Monitor transfer bandwidth, idle accelerator time and model drift” instead addresses the requested ongoing operational assurance.
B:
The alternative “Compare output parity, GPU utilization and p99 inference latency” is a one-time acceptance result rather than ongoing operational assurance. In the reported observation, profile shows a new preprocessing step copies tensors repeatedly between CPU and GPU. The option “Monitor transfer bandwidth, idle accelerator time and model drift” instead addresses the requested ongoing operational assurance.
D:
More compute may increase spend while copies still dominate. For this question, it is not a defensible choice for ongoing operational assurance and it does not satisfy the requirement that cost optimization must not lower model output quality.
Question 210
Case study: Global research cloud: project-to-project data movement
Global research cloud relies on project-to-project data movement to deliver its operational commitments. The current architectural problem is that a sensitive BigQuery dataset is read by a broad reporting service account. The team has recorded this evidence: query logs show access from an unrelated project outside the research boundary. Any proposed response must preserve the requirement that only approved analytical workloads may join identifiable records.
Global research cloud plans to update project-to-project data movement. Investigators found that query logs show access from an unrelated project outside the research boundary, consistent with the reported problem that a sensitive BigQuery dataset is read by a broad reporting service account. The response must satisfy the requirement that only approved analytical workloads may join identifiable records. The security and reliability teams dispute the root cause. Which evidence-gathering action is justified now?
- Combine dataset IAM with supported VPC Service Controls restrictions
- Inspect principals, service perimeter scope and destination datasets
- Assume a private IP alone prevents data exfiltration through API calls
- Apply access changes in dry-run and validate legitimate research exports
Explanation
Correct Answer
B:
Inspect principals, service perimeter scope and destination datasets is the appropriate diagnostic because the observation—query logs show access from an unrelated project outside the research boundary—does not by itself establish every dependency, authorization rule or failure boundary. The findings explain the reported risk (a sensitive BigQuery dataset is read by a broad reporting service account) before the design is approved.
Incorrect Answers
A:
The alternative “Combine dataset IAM with supported VPC Service Controls restrictions” is target architecture selection; it does not establish why the observed path failed. In the reported observation, query logs show access from an unrelated project outside the research boundary. The option “Inspect principals, service perimeter scope and destination datasets” instead addresses the requested root-cause evidence.
C:
Network privacy alone cannot enforce Google API data boundaries. For this question, it is not a defensible choice for root-cause evidence and it does not satisfy the requirement that only approved analytical workloads may join identifiable records.
D:
The alternative “Apply access changes in dry-run and validate legitimate research exports” is rollout work that may modify the wrong dependency before root-cause evidence is collected. In the reported observation, query logs show access from an unrelated project outside the research boundary. The option “Inspect principals, service perimeter scope and destination datasets” instead addresses the requested root-cause evidence.