Google Professional Cloud Architect Practice Test 74
Question 730
At International digital learning network, the Cloud Trace bottleneck workload has a documented risk: users see slow lesson page load but service-level metrics show each component as healthy. The review found that trace spans reveal sequential requests to a distant analytics service. The architecture must also ensure that lesson pages must remain responsive if nonessential analytics pauses. The team must deliver the approved control safely. Which deployment sequence is most appropriate?
- Increase every VM size without checking span-level critical path
- Introduce asynchronous recording and gradually shift user cohorts
- Move optional analytics off the synchronous critical path with trace instrumentation
- Watch trace critical-path latency and queued analytics backlog
Explanation
Correct Answer
B:
Introduce asynchronous recording and gradually shift user cohorts is a controlled means to introduce the approved architecture while meeting the requirement that lesson pages must remain responsive if nonessential analytics pauses. Staged delivery, attributable change evidence and rollback guard against an otherwise avoidable service disruption.
Incorrect Answers
A:
Generic scaling need not solve avoidable synchronous waiting. For this question, it is not a defensible choice for controlled implementation and it does not satisfy the requirement that lesson pages must remain responsive if nonessential analytics pauses.
C:
The alternative “Move optional analytics off the synchronous critical path with trace instrumentation” describes the chosen design, but not how it is introduced safely. In the reported observation, trace spans reveal sequential requests to a distant analytics service. The option “Introduce asynchronous recording and gradually shift user cohorts” instead addresses the requested controlled implementation.
D:
The alternative “Watch trace critical-path latency and queued analytics backlog” addresses later monitoring rather than the actual controlled deployment. In the reported observation, trace spans reveal sequential requests to a distant analytics service. The option “Introduce asynchronous recording and gradually shift user cohorts” instead addresses the requested controlled implementation.
Question 731
Case study: Regional manufacturing consortium: third-party API versioning
Regional manufacturing consortium relies on third-party API versioning to deliver its operational commitments. The current architectural problem is that new supplier API versions intermittently break inventory synchronization. The team has recorded this evidence: request traces show field types changed without contract testing. Any proposed response must preserve the requirement that warehouses cannot lose purchase order events.
Consider Regional manufacturing consortium and its third-party API versioning workload. The issue is that new supplier API versions intermittently break inventory synchronization; the current observation is that request traces show field types changed without contract testing. The team must ensure that warehouses cannot lose purchase order events. The target architecture is approved; execution without disruption is the remaining decision. Which rollout action is best?
- Hardcode the latest response field indexes without version checks
- Define backward-compatible contracts and contract tests per partner
- Monitor schema rejection and supplier API change notifications
- Roll out dual-version adapters with message replay and fallback
Explanation
Correct Answer
D:
Roll out dual-version adapters with message replay and fallback is a controlled means to introduce the approved architecture while meeting the requirement that warehouses cannot lose purchase order events. Staged delivery, attributable change evidence and rollback guard against an otherwise avoidable service disruption.
Incorrect Answers
A:
Unversioned mapping breaks on supplier schema evolution. For this question, it is not a defensible choice for controlled implementation and it does not satisfy the requirement that warehouses cannot lose purchase order events.
B:
The alternative “Define backward-compatible contracts and contract tests per partner” describes the chosen design, but not how it is introduced safely. In the reported observation, request traces show field types changed without contract testing. The option “Roll out dual-version adapters with message replay and fallback” instead addresses the requested controlled implementation.
C:
The alternative “Monitor schema rejection and supplier API change notifications” addresses later monitoring rather than the actual controlled deployment. In the reported observation, request traces show field types changed without contract testing. The option “Roll out dual-version adapters with message replay and fallback” instead addresses the requested controlled implementation.
Question 732
The Compute Engine API autoscaling environment at National delivery marketplace faces this issue: a regional instance group has only one VM with no health-based replacement. Supporting evidence shows that zone-outage test stops all public shipment status queries. The architect must preserve the requirement that shipment tracking needs zone-resilient read service. What test must pass before the business accepts the remediated architecture?
- Roll out an immutable instance template with rolling replacement gates
- Inspect VM startup requirements, load balancing health and zone placement
- Stop a zone and verify healthy instances continue serving requests
- Rely on a single large zonal VM with manual restart instructions
Explanation
Correct Answer
C:
Stop a zone and verify healthy instances continue serving requests is a direct acceptance test of the path implicated by zone-outage test stops all public shipment status queries. It checks that the original risk (a regional instance group has only one VM with no health-based replacement) has actually been controlled and tests that permitted operations still meet the requirement that shipment tracking needs zone-resilient read service.
Incorrect Answers
A:
The alternative “Roll out an immutable instance template with rolling replacement gates” documents change execution but not its operational acceptance result. In the reported observation, zone-outage test stops all public shipment status queries. The option “Stop a zone and verify healthy instances continue serving requests” instead addresses the requested acceptance verification.
B:
The alternative “Inspect VM startup requirements, load balancing health and zone placement” collects prechange evidence without confirming that the revised service meets its outcome. In the reported observation, zone-outage test stops all public shipment status queries. The option “Stop a zone and verify healthy instances continue serving requests” instead addresses the requested acceptance verification.
D:
One oversized VM remains a zonal single point of failure. For this question, it is not a defensible choice for acceptance verification and it does not satisfy the requirement that shipment tracking needs zone-resilient read service.
Question 733
The fare validation backend environment at Urban public transport authority faces this issue: rush-hour taps overwhelm the settlement database even though fraud checks are light. Supporting evidence shows that histogram shows 8x peak events and lock waits on identical card records. The architect must preserve the requirement that a transit pass must not be charged twice for one tap. Before approving a solution, which analysis most directly clarifies the underlying dependency or control gap?
- Scale only the front-end web server while keeping the hot database lock
- Introduce tap ingestion in station cohorts with controlled retry semantics
- Measure tap interarrival, hot-key contention and duplicate IDs
- Buffer taps durably and enforce idempotent settlement with per-pass consistency
Explanation
Correct Answer
C:
Measure tap interarrival, hot-key contention and duplicate IDs is the appropriate diagnostic because the observation—histogram shows 8x peak events and lock waits on identical card records—does not by itself establish every dependency, authorization rule or failure boundary. The findings explain the reported risk (rush-hour taps overwhelm the settlement database even though fraud checks are light) before the design is approved.
Incorrect Answers
A:
A larger front end does not remove the transactional bottleneck. For this question, it is not a defensible choice for root-cause evidence and it does not satisfy the requirement that a transit pass must not be charged twice for one tap.
B:
The alternative “Introduce tap ingestion in station cohorts with controlled retry semantics” is rollout work that may modify the wrong dependency before root-cause evidence is collected. In the reported observation, histogram shows 8x peak events and lock waits on identical card records. The option “Measure tap interarrival, hot-key contention and duplicate IDs” instead addresses the requested root-cause evidence.
D:
The alternative “Buffer taps durably and enforce idempotent settlement with per-pass consistency” is target architecture selection; it does not establish why the observed path failed. In the reported observation, histogram shows 8x peak events and lock waits on identical card records. The option “Measure tap interarrival, hot-key contention and duplicate IDs” instead addresses the requested root-cause evidence.
Question 734
Consider Airport luggage services and its service catalog onboarding workload. The issue is that new teams ask for databases by email and receive unrestricted project ownership; the current observation is that tickets show inconsistent machine sizing and missing application tags. The team must ensure that every project needs approved cost and access boundaries. Select TWO complementary actions: one durable target architecture and one controlled way to introduce it while preserving service. Which TWO are best?
- Continue granting project Owner for all simple database requests
- Review common database requests, service tiers and authorized provisioning roles
- Monitor template drift and manual overrides
- Launch approved Terraform-backed catalog items for pilot teams
- Define self-service catalog templates with IAM and budget constraints
Explanation
Correct Answers
D:
Launch approved Terraform-backed catalog items for pilot teams is a controlled means to introduce the approved architecture while meeting the requirement that every project needs approved cost and access boundaries. 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:
Define self-service catalog templates with IAM and budget constraints targets the underlying issue (new teams ask for databases by email and receive unrestricted project ownership) rather than merely reporting symptoms. It also preserves the required outcome (every project needs approved cost and access boundaries), which a target-state design must satisfy. This supplies the lasting-control half of the two-action package.
Incorrect Answers
A:
Broad manual grants defeat governed repeatability. The shortcut does not supply the requested design and safe rollout for service catalog onboarding, where tickets show inconsistent machine sizing and missing application tags.
B:
This diagnoses the original issue (new teams ask for databases by email and receive unrestricted project ownership), but the TWO requested actions are the target architecture and its implementation; discovery was already completed.
C:
Monitoring service catalog onboarding for future drift is useful after the rollout, but it does not specify the requested target control and safe implementation for this case: new teams ask for databases by email and receive unrestricted project ownership.
Question 735
At Online wholesale exchange, the external build pipeline identity workload has a documented risk: a third-party CI runner receives a permanent Google service account key. The review found that audit shows key is copied across vendor workspaces. The architecture must also ensure that pipeline must deploy only into one dedicated release project. What follow-up metric or alert will identify emerging failures after the initial acceptance test?
- Use OIDC-based Workload Identity Federation with repository-bound conditions
- Attempt deployment from an unapproved repository and confirm denial
- Watch issuer-condition drift and unexpected short-lived token grants
- Give the vendor a long-lived project Owner key to reduce setup work
Explanation
Correct Answer
C:
Watch issuer-condition drift and unexpected short-lived token grants reveals recurrence, drift or changing dependencies after launch. It provides a repeatable signal for the risk (a third-party CI runner receives a permanent Google service account key) rather than treating the first successful test as permanent assurance.
Incorrect Answers
A:
The alternative “Use OIDC-based Workload Identity Federation with repository-bound conditions” selects a design, but does not provide the ongoing signal required to detect drift. In the reported observation, audit shows key is copied across vendor workspaces. The option “Watch issuer-condition drift and unexpected short-lived token grants” instead addresses the requested ongoing operational assurance.
B:
The alternative “Attempt deployment from an unapproved repository and confirm denial” is a one-time acceptance result rather than ongoing operational assurance. In the reported observation, audit shows key is copied across vendor workspaces. The option “Watch issuer-condition drift and unexpected short-lived token grants” instead addresses the requested ongoing operational assurance.
D:
Permanent broad keys magnify vendor compromise impact. For this question, it is not a defensible choice for ongoing operational assurance and it does not satisfy the requirement that pipeline must deploy only into one dedicated release project.
Question 736
The retrieved prompt injection environment at Clinical AI summarization service faces this issue: untrusted patient notes include instructions asking the assistant to email records externally. Supporting evidence shows that red-team canary causes the assistant to propose unauthorized tool calls. The architect must preserve the requirement that clinical outputs may be viewed, but protected records must never be exported by retrieved text. Choose the architecture that resolves the documented weakness without breaching the operational constraint.
- Treat any instruction inside retrieved notes as an administrator command
- Use explicit tool authorization with Model Armor filtering and restricted context grounding
- Monitor blocked prompt-injection attempts and tool authorization failures
- Map retrieval-to-tool trust boundaries and external tool privileges
Explanation
Correct Answer
B:
Use explicit tool authorization with Model Armor filtering and restricted context grounding targets the underlying issue (untrusted patient notes include instructions asking the assistant to email records externally) rather than merely reporting symptoms. It also preserves the required outcome (clinical outputs may be viewed, but protected records must never be exported by retrieved text), which a target-state design must satisfy.
Incorrect Answers
A:
Retrieved patient text is untrusted data, not authorization. For this question, it is not a defensible choice for architecture selection and it does not satisfy the requirement that clinical outputs may be viewed, but protected records must never be exported by retrieved text.
C:
The alternative “Monitor blocked prompt-injection attempts and tool authorization failures” measures results without defining the preventive design now required. In the reported observation, red-team canary causes the assistant to propose unauthorized tool calls. The option “Use explicit tool authorization with Model Armor filtering and restricted context grounding” instead addresses the requested architecture selection.
D:
The alternative “Map retrieval-to-tool trust boundaries and external tool privileges” investigates the cause without specifying the durable architecture to resolve it. In the reported observation, red-team canary causes the assistant to propose unauthorized tool calls. The option “Use explicit tool authorization with Model Armor filtering and restricted context grounding” instead addresses the requested architecture selection.
Question 737
Consider International hotel chain and its central incident notification API workload. The issue is that operations migrates a notification service before updating caller authentication and message retry rules; the current observation is that test customers see duplicate room-change notifications. The team must ensure that customer booking updates must remain unique and attributable. The team has completed discovery. Which preventive or corrective architecture is most appropriate?
- Review retry, authentication and idempotency semantics across partners
- Disable all retries and discard messages when the first send fails
- Design idempotent notification delivery with scoped API identities
- Monitor retry storms and unexpected duplicate IDs
Explanation
Correct Answer
C:
Design idempotent notification delivery with scoped API identities targets the underlying issue (operations migrates a notification service before updating caller authentication and message retry rules) rather than merely reporting symptoms. It also preserves the required outcome (customer booking updates must remain unique and attributable), which a target-state design must satisfy.
Incorrect Answers
A:
The alternative “Review retry, authentication and idempotency semantics across partners” investigates the cause without specifying the durable architecture to resolve it. In the reported observation, test customers see duplicate room-change notifications. The option “Design idempotent notification delivery with scoped API identities” instead addresses the requested architecture selection.
B:
At-most-once loss does not meet delivery commitments. For this question, it is not a defensible choice for architecture selection and it does not satisfy the requirement that customer booking updates must remain unique and attributable.
D:
The alternative “Monitor retry storms and unexpected duplicate IDs” measures results without defining the preventive design now required. In the reported observation, test customers see duplicate room-change notifications. The option “Design idempotent notification delivery with scoped API identities” instead addresses the requested architecture selection.
Question 738
Case study: Streaming fitness platform: event analytics ingestion
Streaming fitness platform relies on event analytics ingestion to deliver its operational commitments. The current architectural problem is that incoming device events arrive late and in duplicates and dashboards need windows. The team has recorded this evidence: sample payloads show device clocks skew ten minutes across regions. Any proposed response must preserve the requirement that analytics should not interfere with real-time device safety controls.
At Streaming fitness platform, the Google Cloud design for event analytics ingestion is affected because incoming device events arrive late and in duplicates and dashboards need windows. A review found that sample payloads show device clocks skew ten minutes across regions. The proposed response must account for the requirement that analytics should not interfere with real-time device safety controls. Which controlled implementation should be used to preserve operational continuity during the change?
- Monitor watermark lag, discarded events and slot spending
- Use Pub/Sub with Dataflow event-time processing and BigQuery analytics sinks
- Roll out schema-versioned consumers with dead-letter handling
- Use request arrival timestamps as ground truth without clock handling
Explanation
Correct Answer
C:
Roll out schema-versioned consumers with dead-letter handling is a controlled means to introduce the approved architecture while meeting the requirement that analytics should not interfere with real-time device safety controls. Staged delivery, attributable change evidence and rollback guard against an otherwise avoidable service disruption.
Incorrect Answers
A:
The alternative “Monitor watermark lag, discarded events and slot spending” addresses later monitoring rather than the actual controlled deployment. In the reported observation, sample payloads show device clocks skew ten minutes across regions. The option “Roll out schema-versioned consumers with dead-letter handling” instead addresses the requested controlled implementation.
B:
The alternative “Use Pub/Sub with Dataflow event-time processing and BigQuery analytics sinks” describes the chosen design, but not how it is introduced safely. In the reported observation, sample payloads show device clocks skew ten minutes across regions. The option “Roll out schema-versioned consumers with dead-letter handling” instead addresses the requested controlled implementation.
D:
Arrival time alone distorts late-event analytics. For this question, it is not a defensible choice for controlled implementation and it does not satisfy the requirement that analytics should not interfere with real-time device safety controls.
Question 739
Case study: Healthcare diagnostics startup: private clinical AI roadmap
Healthcare diagnostics startup relies on private clinical AI roadmap to deliver its operational commitments. The current architectural problem is that new model versions require retraining with patient data but existing architecture exposes raw records to developers. The team has recorded this evidence: governance review finds manual exports from a protected clinical warehouse. Any proposed response must preserve the requirement that clinical data processing must remain consent-bound and auditable.
At Healthcare diagnostics startup, the private clinical AI roadmap workload has a documented risk: new model versions require retraining with patient data but existing architecture exposes raw records to developers. The review found that governance review finds manual exports from a protected clinical warehouse. The architecture must also ensure that clinical data processing must remain consent-bound and auditable. Which post-change test most reliably distinguishes a functioning control from a merely successful installation?
- Give all model developers project Owner to speed experimentation
- Pilot a new model using approved deidentified datasets and audit trails
- Verify new training jobs cannot export unapproved patient identifiers
- Map future model use cases, consent, access and lifecycle expectations
Explanation
Correct Answer
C:
Verify new training jobs cannot export unapproved patient identifiers is a direct acceptance test of the path implicated by governance review finds manual exports from a protected clinical warehouse. It checks that the original risk (new model versions require retraining with patient data but existing architecture exposes raw records to developers) has actually been controlled and tests that permitted operations still meet the requirement that clinical data processing must remain consent-bound and auditable.
Incorrect Answers
A:
Broad project ownership bypasses patient-data isolation. For this question, it is not a defensible choice for acceptance verification and it does not satisfy the requirement that clinical data processing must remain consent-bound and auditable.
B:
The alternative “Pilot a new model using approved deidentified datasets and audit trails” documents change execution but not its operational acceptance result. In the reported observation, governance review finds manual exports from a protected clinical warehouse. The option “Verify new training jobs cannot export unapproved patient identifiers” instead addresses the requested acceptance verification.
D:
The alternative “Map future model use cases, consent, access and lifecycle expectations” collects prechange evidence without confirming that the revised service meets its outcome. In the reported observation, governance review finds manual exports from a protected clinical warehouse. The option “Verify new training jobs cannot export unapproved patient identifiers” instead addresses the requested acceptance verification.
Question 740
The BigQuery scripted exports environment at Digital university archive faces this issue: a nightly bq CLI export uses wildcard datasets and leaks unrelated faculty research. Supporting evidence shows that audit log shows query scope expanded after a new dataset was added. The architect must preserve the requirement that exports must include only approved course statistics. Before approving a solution, which analysis most directly clarifies the underlying dependency or control gap?
- Give the export service account BigQuery Admin over the whole organization
- Identify exact dataset boundaries, IAM roles and query requirements
- Use parameterized explicit BigQuery dataset/table scope with read-only job identity
- Deploy export script with bounded project context and approved schedules
Explanation
Correct Answer
B:
Identify exact dataset boundaries, IAM roles and query requirements is the appropriate diagnostic because the observation—audit log shows query scope expanded after a new dataset was added—does not by itself establish every dependency, authorization rule or failure boundary. The findings explain the reported risk (a nightly bq CLI export uses wildcard datasets and leaks unrelated faculty research) before the design is approved.
Incorrect Answers
A:
Broad admin rights violate the limited read-only extraction purpose. For this question, it is not a defensible choice for root-cause evidence and it does not satisfy the requirement that exports must include only approved course statistics.
C:
The alternative “Use parameterized explicit BigQuery dataset/table scope with read-only job identity” is target architecture selection; it does not establish why the observed path failed. In the reported observation, audit log shows query scope expanded after a new dataset was added. The option “Identify exact dataset boundaries, IAM roles and query requirements” instead addresses the requested root-cause evidence.
D:
The alternative “Deploy export script with bounded project context and approved schedules” is rollout work that may modify the wrong dependency before root-cause evidence is collected. In the reported observation, audit log shows query scope expanded after a new dataset was added. The option “Identify exact dataset boundaries, IAM roles and query requirements” instead addresses the requested root-cause evidence.