Google Professional Cloud Architect Practice Test 83
Question 820
The CDC job performance regression environment at Medical supply chain faces this issue: a nightly database export finishes late despite unchanged source volume. Supporting evidence shows that telemetry shows a new transformation now triggers full-table reloads. The architect must preserve the requirement that hospital inventory reports must arrive before morning replenishment. Select TWO complementary actions: one durable target architecture and one controlled way to introduce it while preserving service. Which TWO are best?
- Monitor bytes scanned, checkpoint lag and daily delivery misses
- Compare pipeline execution plan, input cardinality and transformation deltas
- Canary revised pipeline and reconcile late-arriving updates
- Provision more dashboard servers while keeping the full reload
- Change to incremental CDC processing with checkpointed offsets
Explanation
Correct Answers
C:
Canary revised pipeline and reconcile late-arriving updates is a controlled means to introduce the approved architecture while meeting the requirement that hospital inventory reports must arrive before morning replenishment. Staged delivery, attributable change evidence and rollback guard against an otherwise avoidable service disruption. This supplies the controlled-introduction half of the two-action package.
E:
Change to incremental CDC processing with checkpointed offsets targets the underlying issue (a nightly database export finishes late despite unchanged source volume) rather than merely reporting symptoms. It also preserves the required outcome (hospital inventory reports must arrive before morning replenishment), which a target-state design must satisfy. This supplies the lasting-control half of the two-action package.
Incorrect Answers
A:
Monitoring CDC job performance regression for future drift is useful after the rollout, but it does not specify the requested target control and safe implementation for this case: a nightly database export finishes late despite unchanged source volume.
B:
This diagnoses the original issue (a nightly database export finishes late despite unchanged source volume), but the TWO requested actions are the target architecture and its implementation; discovery was already completed.
D:
Dashboard compute does not correct an inefficient ingestion path. The shortcut does not supply the requested design and safe rollout for CDC job performance regression, where telemetry shows a new transformation now triggers full-table reloads.
Question 821
Public safety communications service plans to update Secret Manager credential lifecycle. Investigators found that pipeline logs contain plaintext passwords used in production, consistent with the reported problem that a service passes a database password through the deployment YAML. The response must satisfy the requirement that credentials must rotate without emergency downtime. The team needs an operations signal rather than another design or one-time test. What should it track?
- Put the same plaintext password into a Kubernetes ConfigMap
- Rotate the database secret and verify old versions are rejected after cutover
- Track secret access anomalies and rotation-window failures
- Use Secret Manager with per-workload access and controlled secret versioning
Explanation
Correct Answer
C:
Track secret access anomalies and rotation-window failures reveals recurrence, drift or changing dependencies after launch. It provides a repeatable signal for the risk (a service passes a database password through the deployment YAML) rather than treating the first successful test as permanent assurance.
Incorrect Answers
A:
A ConfigMap alone does not secure the original secret. For this question, it is not a defensible choice for ongoing operational assurance and it does not satisfy the requirement that credentials must rotate without emergency downtime.
B:
The alternative “Rotate the database secret and verify old versions are rejected after cutover” is a one-time acceptance result rather than ongoing operational assurance. In the reported observation, pipeline logs contain plaintext passwords used in production. The option “Track secret access anomalies and rotation-window failures” instead addresses the requested ongoing operational assurance.
D:
The alternative “Use Secret Manager with per-workload access and controlled secret versioning” selects a design, but does not provide the ongoing signal required to detect drift. In the reported observation, pipeline logs contain plaintext passwords used in production. The option “Track secret access anomalies and rotation-window failures” instead addresses the requested ongoing operational assurance.
Question 822
Regional university consortium plans to update research compute funding. Investigators found that billing exports show two laboratories use most training accelerator hours, consistent with the reported problem that department costs are assigned equally although GPU workloads dominate reserved capacity. The response must satisfy the requirement that sensitive lab datasets cannot be exposed through a central chargeback dashboard. The approved change has been applied. Which validation provides the clearest evidence the original problem is fixed?
- Quantify per-lab machine consumption, shared overhead and privacy restrictions
- Introduce project labeling and budgets before moving accelerator jobs
- Publish raw dataset labels and individual researcher usage to all departments
- Reconcile usage-based chargeback with billing exports without revealing dataset metadata
Explanation
Correct Answer
D:
Reconcile usage-based chargeback with billing exports without revealing dataset metadata is a direct acceptance test of the path implicated by billing exports show two laboratories use most training accelerator hours. It checks that the original risk (department costs are assigned equally although GPU workloads dominate reserved capacity) has actually been controlled and tests that permitted operations still meet the requirement that sensitive lab datasets cannot be exposed through a central chargeback dashboard.
Incorrect Answers
A:
The alternative “Quantify per-lab machine consumption, shared overhead and privacy restrictions” collects prechange evidence without confirming that the revised service meets its outcome. In the reported observation, billing exports show two laboratories use most training accelerator hours. The option “Reconcile usage-based chargeback with billing exports without revealing dataset metadata” instead addresses the requested acceptance verification.
B:
The alternative “Introduce project labeling and budgets before moving accelerator jobs” documents change execution but not its operational acceptance result. In the reported observation, billing exports show two laboratories use most training accelerator hours. The option “Reconcile usage-based chargeback with billing exports without revealing dataset metadata” instead addresses the requested acceptance verification.
C:
Overbroad chargeback visibility can reveal sensitive research details. For this question, it is not a defensible choice for acceptance verification and it does not satisfy the requirement that sensitive lab datasets cannot be exposed through a central chargeback dashboard.
Question 823
At Public safety communications service, the Secret Manager credential lifecycle workload has a documented risk: a service passes a database password through the deployment YAML. The review found that pipeline logs contain plaintext passwords used in production. The architecture must also ensure that credentials must rotate without emergency downtime. The team must deliver the approved control safely. Which deployment sequence is most appropriate?
- Use Secret Manager with per-workload access and controlled secret versioning
- Put the same plaintext password into a Kubernetes ConfigMap
- Track secret access anomalies and rotation-window failures
- Deploy consumers that read approved versions before retiring the old secret
Explanation
Correct Answer
D:
Deploy consumers that read approved versions before retiring the old secret is a controlled means to introduce the approved architecture while meeting the requirement that credentials must rotate without emergency downtime. Staged delivery, attributable change evidence and rollback guard against an otherwise avoidable service disruption.
Incorrect Answers
A:
The alternative “Use Secret Manager with per-workload access and controlled secret versioning” describes the chosen design, but not how it is introduced safely. In the reported observation, pipeline logs contain plaintext passwords used in production. The option “Deploy consumers that read approved versions before retiring the old secret” instead addresses the requested controlled implementation.
B:
A ConfigMap alone does not secure the original secret. For this question, it is not a defensible choice for controlled implementation and it does not satisfy the requirement that credentials must rotate without emergency downtime.
C:
The alternative “Track secret access anomalies and rotation-window failures” addresses later monitoring rather than the actual controlled deployment. In the reported observation, pipeline logs contain plaintext passwords used in production. The option “Deploy consumers that read approved versions before retiring the old secret” instead addresses the requested controlled implementation.
Question 824
At International cargo platform, the Google Cloud design for operational excellence ownership is affected because three independent teams deploy a critical shipment API without shared incident response rules. A review found that postmortem shows repeat outages with no accountable follow-up owner. The proposed response must account for the requirement that shipment tracking must meet a documented four-nine availability SLO. A deployment ticket has been closed. What direct evidence would justify solution signoff?
- Launch an on-call rotation and practice incident handoff before expansion
- Run an incident simulation and verify responders follow escalation and restoration steps
- Inventory incident ownership, runbooks, service objectives and bottleneck history
- Declare the system reliable because developers resolved one emergency manually
Explanation
Correct Answer
B:
Run an incident simulation and verify responders follow escalation and restoration steps is a direct acceptance test of the path implicated by postmortem shows repeat outages with no accountable follow-up owner. It checks that the original risk (three independent teams deploy a critical shipment API without shared incident response rules) has actually been controlled and tests that permitted operations still meet the requirement that shipment tracking must meet a documented four-nine availability SLO.
Incorrect Answers
A:
The alternative “Launch an on-call rotation and practice incident handoff before expansion” documents change execution but not its operational acceptance result. In the reported observation, postmortem shows repeat outages with no accountable follow-up owner. The option “Run an incident simulation and verify responders follow escalation and restoration steps” instead addresses the requested acceptance verification.
C:
The alternative “Inventory incident ownership, runbooks, service objectives and bottleneck history” collects prechange evidence without confirming that the revised service meets its outcome. In the reported observation, postmortem shows repeat outages with no accountable follow-up owner. The option “Run an incident simulation and verify responders follow escalation and restoration steps” instead addresses the requested acceptance verification.
D:
One heroic recovery is not an operating model. For this question, it is not a defensible choice for acceptance verification and it does not satisfy the requirement that shipment tracking must meet a documented four-nine availability SLO.
Question 825
The portfolio modernization sequencing environment at Retail delivery chain faces this issue: five app teams ask to migrate together though one shared customer database will be refactored. Supporting evidence shows that dependency map shows migration plans share an incompatible schema cutover. The architect must preserve the requirement that online ordering must remain open through staged rollout. What test must pass before the business accepts the remediated architecture?
- Check transaction stability and business KPIs after each wave
- Schedule all components on one weekend without an integration dry run
- Identify business value, dependency order and rollback requirements
- Move less coupled read services first, then transactional dependencies
Explanation
Correct Answer
A:
Check transaction stability and business KPIs after each wave is a direct acceptance test of the path implicated by dependency map shows migration plans share an incompatible schema cutover. It checks that the original risk (five app teams ask to migrate together though one shared customer database will be refactored) has actually been controlled and tests that permitted operations still meet the requirement that online ordering must remain open through staged rollout.
Incorrect Answers
B:
Big-bang sequencing ignores the known shared dependency conflict. For this question, it is not a defensible choice for acceptance verification and it does not satisfy the requirement that online ordering must remain open through staged rollout.
C:
The alternative “Identify business value, dependency order and rollback requirements” collects prechange evidence without confirming that the revised service meets its outcome. In the reported observation, dependency map shows migration plans share an incompatible schema cutover. The option “Check transaction stability and business KPIs after each wave” instead addresses the requested acceptance verification.
D:
The alternative “Move less coupled read services first, then transactional dependencies” documents change execution but not its operational acceptance result. In the reported observation, dependency map shows migration plans share an incompatible schema cutover. The option “Check transaction stability and business KPIs after each wave” instead addresses the requested acceptance verification.
Question 826
At National electricity retailer, the cloud vendor contract renewal workload has a documented risk: finance only compares monthly cloud spend and ignores data-transfer commitments. The review found that network bill shows cross-region replication costs rise each quarter. The architecture must also ensure that customer billing must meet a recovery objective and fixed budget. The architecture review is paused until the team can prove the real failure path. Which investigation should happen first?
- Implement cost guardrails after recovery testing and owner approval
- Break down storage, interregion transfer, commitments and recovery requirements
- Choose a right-sized retention and replication strategy linked to RPO
- Disable all cross-region recovery copies solely to eliminate egress fees
Explanation
Correct Answer
B:
Break down storage, interregion transfer, commitments and recovery requirements is the appropriate diagnostic because the observation—network bill shows cross-region replication costs rise each quarter—does not by itself establish every dependency, authorization rule or failure boundary. The findings explain the reported risk (finance only compares monthly cloud spend and ignores data-transfer commitments) before the design is approved.
Incorrect Answers
A:
The alternative “Implement cost guardrails after recovery testing and owner approval” is rollout work that may modify the wrong dependency before root-cause evidence is collected. In the reported observation, network bill shows cross-region replication costs rise each quarter. The option “Break down storage, interregion transfer, commitments and recovery requirements” instead addresses the requested root-cause evidence.
C:
The alternative “Choose a right-sized retention and replication strategy linked to RPO” is target architecture selection; it does not establish why the observed path failed. In the reported observation, network bill shows cross-region replication costs rise each quarter. The option “Break down storage, interregion transfer, commitments and recovery requirements” instead addresses the requested root-cause evidence.
D:
Deleting recovery ignores the stated continuity obligation. For this question, it is not a defensible choice for root-cause evidence and it does not satisfy the requirement that customer billing must meet a recovery objective and fixed budget.
Question 827
Climate modeling consortium plans to update GPU inference profiling. Investigators found that profile shows a new preprocessing step copies tensors repeatedly between CPU and GPU, consistent with the reported problem that AI inference costs spike although request counts stay flat. The response must satisfy the requirement that cost optimization must not lower model output quality. Select TWO complementary actions: one durable target architecture and one controlled way to introduce it while preserving service. Which TWO are best?
- Buy more GPUs without measuring the memory-transfer bottleneck
- Reduce redundant tensor transfer and right-size accelerators based on measured bottleneck
- Roll out optimized preprocessing behind a versioned serving endpoint
- Inspect profiler traces, memory transfer sizes and latency distributions
- Monitor transfer bandwidth, idle accelerator time and model drift
Explanation
Correct Answers
B:
Reduce redundant tensor transfer and right-size accelerators based on measured bottleneck targets the underlying issue (AI inference costs spike although request counts stay flat) rather than merely reporting symptoms. It also preserves the required outcome (cost optimization must not lower model output quality), which a target-state design must satisfy. This supplies the lasting-control half of the two-action package.
C:
Roll out optimized preprocessing behind a versioned serving endpoint is a controlled means to introduce the approved architecture while meeting the requirement that cost optimization must not lower model output quality. Staged delivery, attributable change evidence and rollback guard against an otherwise avoidable service disruption. This supplies the controlled-introduction half of the two-action package.
Incorrect Answers
A:
More compute may increase spend while copies still dominate. The shortcut does not supply the requested design and safe rollout for GPU inference profiling, where profile shows a new preprocessing step copies tensors repeatedly between CPU and GPU.
D:
This diagnoses the original issue (AI inference costs spike although request counts stay flat), but the TWO requested actions are the target architecture and its implementation; discovery was already completed.
E:
Monitoring GPU inference profiling for future drift is useful after the rollout, but it does not specify the requested target control and safe implementation for this case: AI inference costs spike although request counts stay flat.
Question 828
Consider Online wholesale exchange and its external build pipeline identity workload. The issue is that a third-party CI runner receives a permanent Google service account key; the current observation is that audit shows key is copied across vendor workspaces. The team must ensure that pipeline must deploy only into one dedicated release project. The target architecture is approved; execution without disruption is the remaining decision. Which rollout action is best?
- Give the vendor a long-lived project Owner key to reduce setup work
- Watch issuer-condition drift and unexpected short-lived token grants
- Introduce federated deploy identities and revoke the permanent key
- Use OIDC-based Workload Identity Federation with repository-bound conditions
Explanation
Correct Answer
C:
Introduce federated deploy identities and revoke the permanent key is a controlled means to introduce the approved architecture while meeting the requirement that pipeline must deploy only into one dedicated release project. Staged delivery, attributable change evidence and rollback guard against an otherwise avoidable service disruption.
Incorrect Answers
A:
Permanent broad keys magnify vendor compromise impact. For this question, it is not a defensible choice for controlled implementation and it does not satisfy the requirement that pipeline must deploy only into one dedicated release project.
B:
The alternative “Watch issuer-condition drift and unexpected short-lived token grants” addresses later monitoring rather than the actual controlled deployment. In the reported observation, audit shows key is copied across vendor workspaces. The option “Introduce federated deploy identities and revoke the permanent key” instead addresses the requested controlled implementation.
D:
The alternative “Use OIDC-based Workload Identity Federation with repository-bound conditions” describes the chosen design, but not how it is introduced safely. In the reported observation, audit shows key is copied across vendor workspaces. The option “Introduce federated deploy identities and revoke the permanent key” instead addresses the requested controlled implementation.
Question 829
Case study: International cinema platform: secure software supply chain
International cinema platform relies on secure software supply chain to deliver its operational commitments. The current architectural problem is that a deployment references mutable container tags from an unaudited third-party registry. The team has recorded this evidence: build provenance report shows tag digest changed after approval. Any proposed response must preserve the requirement that production media services must run only approved signed artifacts.
International cinema platform plans to update secure software supply chain. Investigators found that build provenance report shows tag digest changed after approval, consistent with the reported problem that a deployment references mutable container tags from an unaudited third-party registry. The response must satisfy the requirement that production media services must run only approved signed artifacts. The team needs an operations signal rather than another design or one-time test. What should it track?
- Use Artifact Registry provenance and Binary Authorization enforcement for approved images
- Attempt deployment with unsigned and digest-swapped images and verify rejection
- Watch policy bypasses, signer changes and rejected admissions
- Allow every external image when its tag says latest-stable
Explanation
Correct Answer
C:
Watch policy bypasses, signer changes and rejected admissions reveals recurrence, drift or changing dependencies after launch. It provides a repeatable signal for the risk (a deployment references mutable container tags from an unaudited third-party registry) rather than treating the first successful test as permanent assurance.
Incorrect Answers
A:
The alternative “Use Artifact Registry provenance and Binary Authorization enforcement for approved images” selects a design, but does not provide the ongoing signal required to detect drift. In the reported observation, build provenance report shows tag digest changed after approval. The option “Watch policy bypasses, signer changes and rejected admissions” instead addresses the requested ongoing operational assurance.
B:
The alternative “Attempt deployment with unsigned and digest-swapped images and verify rejection” is a one-time acceptance result rather than ongoing operational assurance. In the reported observation, build provenance report shows tag digest changed after approval. The option “Watch policy bypasses, signer changes and rejected admissions” instead addresses the requested ongoing operational assurance.
D:
A mutable tag does not establish artifact authenticity. For this question, it is not a defensible choice for ongoing operational assurance and it does not satisfy the requirement that production media services must run only approved signed artifacts.
Question 830
The regulated data classification environment at Hospital imaging research unit faces this issue: an AI training workflow ingests medical reports with accidental patient identifiers. Supporting evidence shows that dataset profiler finds patient IDs inside free-text radiology comments. The architect must preserve the requirement that only approved deidentified inputs may enter model training. Choose the architecture that resolves the documented weakness without breaching the operational constraint.
- Use Sensitive Data Protection inspection/deidentification with governance controls
- Trust a generic prompt to remove every patient ID from raw input
- Track newly detected sensitive patterns, policy versions and false negatives
- Classify identifiable fields and exposure paths with sample inspections
Explanation
Correct Answer
A:
Use Sensitive Data Protection inspection/deidentification with governance controls targets the underlying issue (an AI training workflow ingests medical reports with accidental patient identifiers) rather than merely reporting symptoms. It also preserves the required outcome (only approved deidentified inputs may enter model training), which a target-state design must satisfy.
Incorrect Answers
B:
Prompt guidance is not a deterministic privacy-control boundary. For this question, it is not a defensible choice for architecture selection and it does not satisfy the requirement that only approved deidentified inputs may enter model training.
C:
The alternative “Track newly detected sensitive patterns, policy versions and false negatives” measures results without defining the preventive design now required. In the reported observation, dataset profiler finds patient IDs inside free-text radiology comments. The option “Use Sensitive Data Protection inspection/deidentification with governance controls” instead addresses the requested architecture selection.
D:
The alternative “Classify identifiable fields and exposure paths with sample inspections” investigates the cause without specifying the durable architecture to resolve it. In the reported observation, dataset profiler finds patient IDs inside free-text radiology comments. The option “Use Sensitive Data Protection inspection/deidentification with governance controls” instead addresses the requested architecture selection.