Google Professional Cloud Architect Practice Test 25
Question 240
At Agricultural crop analysis firm, the agentic data preparation workload has a documented risk: different dataset formats feed model training but lineage disappears when one parser fails. The review found that pipeline logs show successful jobs produce artifacts from incompatible schema versions. The architecture must also ensure that model decisions must be explainable against source training records. Select TWO complementary actions: one durable target architecture and one controlled way to introduce it while preserving service. Which TWO are best?
- Monitor schema rejection, data quality and drift by source
- Orchestrate data preparation and training with versioned pipeline components
- Inspect schema contracts, transformation versions and artifact references
- Stage parser updates through validation and checkpointed pipeline runs
- Let every pipeline step read and overwrite the same unversioned training table
Explanation
Correct Answers
B:
Orchestrate data preparation and training with versioned pipeline components targets the underlying issue (different dataset formats feed model training but lineage disappears when one parser fails) rather than merely reporting symptoms. It also preserves the required outcome (model decisions must be explainable against source training records), which a target-state design must satisfy. This supplies the lasting-control half of the two-action package.
D:
Stage parser updates through validation and checkpointed pipeline runs is a controlled means to introduce the approved architecture while meeting the requirement that model decisions must be explainable against source training records. 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:
Monitoring agentic data preparation for future drift is useful after the rollout, but it does not specify the requested target control and safe implementation for this case: different dataset formats feed model training but lineage disappears when one parser fails.
C:
This diagnoses the original issue (different dataset formats feed model training but lineage disappears when one parser fails), but the TWO requested actions are the target architecture and its implementation; discovery was already completed.
E:
Mutable shared state destroys lineage and reproducibility. The shortcut does not supply the requested design and safe rollout for agentic data preparation, where pipeline logs show successful jobs produce artifacts from incompatible schema versions.
Question 241
Case study: Medical appointment SaaS: serverless private database access
Medical appointment SaaS relies on serverless private database access to deliver its operational commitments. The current architectural problem is that Cloud Run service has public database IP access though all systems reside in one VPC. The team has recorded this evidence: connection logs show traffic to a globally routable database endpoint. Any proposed response must preserve the requirement that patient database must only accept internal workload connections.
Medical appointment SaaS plans to update serverless private database access. Investigators found that connection logs show traffic to a globally routable database endpoint, consistent with the reported problem that Cloud Run service has public database IP access though all systems reside in one VPC. The response must satisfy the requirement that patient database must only accept internal workload connections. The design review is over and production introduction is next. Which action is the best fit?
- Monitor egress routing drift and private connection saturation
- Cut over service revision to private egress and remove public listener safely
- Permit the database to accept all public IPs because TLS is enabled
- Use Cloud Run VPC egress integration and private database access
Explanation
Correct Answer
B:
Cut over service revision to private egress and remove public listener safely is a controlled means to introduce the approved architecture while meeting the requirement that patient database must only accept internal workload connections. Staged delivery, attributable change evidence and rollback guard against an otherwise avoidable service disruption.
Incorrect Answers
A:
The alternative “Monitor egress routing drift and private connection saturation” addresses later monitoring rather than the actual controlled deployment. In the reported observation, connection logs show traffic to a globally routable database endpoint. The option “Cut over service revision to private egress and remove public listener safely” instead addresses the requested controlled implementation.
C:
TLS does not meet the private reachability requirement. For this question, it is not a defensible choice for controlled implementation and it does not satisfy the requirement that patient database must only accept internal workload connections.
D:
The alternative “Use Cloud Run VPC egress integration and private database access” describes the chosen design, but not how it is introduced safely. In the reported observation, connection logs show traffic to a globally routable database endpoint. The option “Cut over service revision to private egress and remove public listener safely” instead addresses the requested controlled implementation.
Question 242
At Manufacturing telemetry provider, the Google Cloud design for private partner services is affected because partners traverse the public internet to reach an internal metrics API. A review found that packet traces show traffic exposed through an unnecessary external load balancer. The proposed response must account for the requirement that service access must be private and restricted to approved consumer networks. The solution is operating. Which evidence should be watched for longer-term reliability and compliance?
- Use Private Service Connect for scoped private service consumption
- Confirm no public path exists and authorized endpoints still work
- Create a broad public IP allowlist for all partner subnets
- Watch service attachment approvals and unexpected consumer connections
Explanation
Correct Answer
D:
Watch service attachment approvals and unexpected consumer connections reveals recurrence, drift or changing dependencies after launch. It provides a repeatable signal for the risk (partners traverse the public internet to reach an internal metrics API) rather than treating the first successful test as permanent assurance.
Incorrect Answers
A:
The alternative “Use Private Service Connect for scoped private service consumption” selects a design, but does not provide the ongoing signal required to detect drift. In the reported observation, packet traces show traffic exposed through an unnecessary external load balancer. The option “Watch service attachment approvals and unexpected consumer connections” instead addresses the requested ongoing operational assurance.
B:
The alternative “Confirm no public path exists and authorized endpoints still work” is a one-time acceptance result rather than ongoing operational assurance. In the reported observation, packet traces show traffic exposed through an unnecessary external load balancer. The option “Watch service attachment approvals and unexpected consumer connections” instead addresses the requested ongoing operational assurance.
C:
A public allowlist does not provide an intrinsically private service endpoint. For this question, it is not a defensible choice for ongoing operational assurance and it does not satisfy the requirement that service access must be private and restricted to approved consumer networks.
Question 243
Multinational consulting alliance plans to update shared IAM inheritance. Investigators found that IAM policy audit shows project roles inherited from the organization root, consistent with the reported problem that a merger added interns to a group that has editor permission on financial-production resources. The response must satisfy the requirement that production access must be limited to verified operational duties. The approved change has been applied. Which validation provides the clearest evidence the original problem is fixed?
- Identify inherited member groups and smallest needed job privileges
- Create a new group but keep the old wildcard organization binding
- Stage grant changes with access simulation and validated rollback
- Verify new intern identities cannot edit production billing resources
Explanation
Correct Answer
D:
Verify new intern identities cannot edit production billing resources is a direct acceptance test of the path implicated by IAM policy audit shows project roles inherited from the organization root. It checks that the original risk (a merger added interns to a group that has editor permission on financial-production resources) has actually been controlled and tests that permitted operations still meet the requirement that production access must be limited to verified operational duties.
Incorrect Answers
A:
The alternative “Identify inherited member groups and smallest needed job privileges” collects prechange evidence without confirming that the revised service meets its outcome. In the reported observation, IAM policy audit shows project roles inherited from the organization root. The option “Verify new intern identities cannot edit production billing resources” instead addresses the requested acceptance verification.
B:
A new group does not negate an existing broad inherited permission. For this question, it is not a defensible choice for acceptance verification and it does not satisfy the requirement that production access must be limited to verified operational duties.
C:
The alternative “Stage grant changes with access simulation and validated rollback” documents change execution but not its operational acceptance result. In the reported observation, IAM policy audit shows project roles inherited from the organization root. The option “Verify new intern identities cannot edit production billing resources” instead addresses the requested acceptance verification.
Question 244
At National land registry, the Google Cloud design for admin console access is affected because staff directly expose internal admin dashboards through public VM IPs. A review found that network scan discovers admin endpoints reachable without identity-aware gateway. The proposed response must account for the requirement that authorized clerks must access remotely without public administrative ingress. The solution is operating. Which evidence should be watched for longer-term reliability and compliance?
- Add a public allow-all firewall rule with obscured URL paths
- Verify unauthorized principal denied and authorized clerk session logged
- Put supported admin applications behind Identity-Aware Proxy with appropriate IAM
- Monitor IAP access failures and unexpected direct ingress rules
Explanation
Correct Answer
D:
Monitor IAP access failures and unexpected direct ingress rules reveals recurrence, drift or changing dependencies after launch. It provides a repeatable signal for the risk (staff directly expose internal admin dashboards through public VM IPs) rather than treating the first successful test as permanent assurance.
Incorrect Answers
A:
Obscure paths do not restrict authenticated administrative access. For this question, it is not a defensible choice for ongoing operational assurance and it does not satisfy the requirement that authorized clerks must access remotely without public administrative ingress.
B:
The alternative “Verify unauthorized principal denied and authorized clerk session logged” is a one-time acceptance result rather than ongoing operational assurance. In the reported observation, network scan discovers admin endpoints reachable without identity-aware gateway. The option “Monitor IAP access failures and unexpected direct ingress rules” instead addresses the requested ongoing operational assurance.
C:
The alternative “Put supported admin applications behind Identity-Aware Proxy with appropriate IAM” selects a design, but does not provide the ongoing signal required to detect drift. In the reported observation, network scan discovers admin endpoints reachable without identity-aware gateway. The option “Monitor IAP access failures and unexpected direct ingress rules” instead addresses the requested ongoing operational assurance.
Question 245
Case study: Telehealth teleconsultation: regulated regional failover
Telehealth teleconsultation relies on regulated regional failover to deliver its operational commitments. The current architectural problem is that the recovery design sends patient consultations to a region forbidden by local processing rules. The team has recorded this evidence: failover plan lists a default backup service on another continent. Any proposed response must preserve the requirement that clinical service must be recoverable inside approved national boundaries.
At Telehealth teleconsultation, the regulated regional failover workload has a documented risk: the recovery design sends patient consultations to a region forbidden by local processing rules. The review found that failover plan lists a default backup service on another continent. The architecture must also ensure that clinical service must be recoverable inside approved national boundaries. Which post-change test most reliably distinguishes a functioning control from a merely successful installation?
- Disconnect the primary and verify sessions never cross prohibited geography
- Stage a local region failover drill and update runbooks
- Review legal location rules and failover data-plane traffic
- Assume a foreign backup region is always allowed during emergencies
Explanation
Correct Answer
A:
Disconnect the primary and verify sessions never cross prohibited geography is a direct acceptance test of the path implicated by failover plan lists a default backup service on another continent. It checks that the original risk (the recovery design sends patient consultations to a region forbidden by local processing rules) has actually been controlled and tests that permitted operations still meet the requirement that clinical service must be recoverable inside approved national boundaries.
Incorrect Answers
B:
The alternative “Stage a local region failover drill and update runbooks” documents change execution but not its operational acceptance result. In the reported observation, failover plan lists a default backup service on another continent. The option “Disconnect the primary and verify sessions never cross prohibited geography” instead addresses the requested acceptance verification.
C:
The alternative “Review legal location rules and failover data-plane traffic” collects prechange evidence without confirming that the revised service meets its outcome. In the reported observation, failover plan lists a default backup service on another continent. The option “Disconnect the primary and verify sessions never cross prohibited geography” instead addresses the requested acceptance verification.
D:
Emergency operations do not automatically waive data residency laws. For this question, it is not a defensible choice for acceptance verification and it does not satisfy the requirement that clinical service must be recoverable inside approved national boundaries.
Question 246
At Financial consumer lender, the Google Cloud design for PCI data encryption separation is affected because application developers can read a cardholder encryption key and production data. A review found that role review shows one service account can decrypt and administer Cloud KMS. The proposed response must account for the requirement that PCI program requires separation of key administration and data processing. Which controlled implementation should be used to preserve operational continuity during the change?
- Set a shared spreadsheet password on exported plaintext cards
- Apply separate IAM roles for key administration and cryptographic use, with CMEK where supported
- Rotate permissions and test service continuity during KMS grant changes
- Monitor key-use logs, privileged changes and quarterly access reviews
Explanation
Correct Answer
C:
Rotate permissions and test service continuity during KMS grant changes is a controlled means to introduce the approved architecture while meeting the requirement that PCI program requires separation of key administration and data processing. Staged delivery, attributable change evidence and rollback guard against an otherwise avoidable service disruption.
Incorrect Answers
A:
A shared password does not meet secure key custody requirements. For this question, it is not a defensible choice for controlled implementation and it does not satisfy the requirement that PCI program requires separation of key administration and data processing.
B:
The alternative “Apply separate IAM roles for key administration and cryptographic use, with CMEK where supported” describes the chosen design, but not how it is introduced safely. In the reported observation, role review shows one service account can decrypt and administer Cloud KMS. The option “Rotate permissions and test service continuity during KMS grant changes” instead addresses the requested controlled implementation.
D:
The alternative “Monitor key-use logs, privileged changes and quarterly access reviews” addresses later monitoring rather than the actual controlled deployment. In the reported observation, role review shows one service account can decrypt and administer Cloud KMS. The option “Rotate permissions and test service continuity during KMS grant changes” instead addresses the requested controlled implementation.
Question 247
The content recommendation evolution environment at European media distributor faces this issue: recommendation service is tied to one proprietary model endpoint and cannot adapt to new languages. Supporting evidence shows that product roadmap adds five languages and stricter content ranking rules next year. The architect must preserve the requirement that existing subscribers must keep playback uninterrupted. Select TWO complementary actions: one durable target architecture and one controlled way to introduce it while preserving service. Which TWO are best?
- Adopt versioned model contracts and independent feature pipelines that can evolve
- Forecast model and language demand against content-governance obligations
- Introduce compatibility tests for new language models with staged cohort activation
- Track deprecation notices, bias drift and API costs
- Hardcode one unversioned provider response into the user profile schema
Explanation
Correct Answers
A:
Adopt versioned model contracts and independent feature pipelines that can evolve targets the underlying issue (recommendation service is tied to one proprietary model endpoint and cannot adapt to new languages) rather than merely reporting symptoms. It also preserves the required outcome (existing subscribers must keep playback uninterrupted), which a target-state design must satisfy. This supplies the lasting-control half of the two-action package.
C:
Introduce compatibility tests for new language models with staged cohort activation is a controlled means to introduce the approved architecture while meeting the requirement that existing subscribers must keep playback uninterrupted. 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
B:
This diagnoses the original issue (recommendation service is tied to one proprietary model endpoint and cannot adapt to new languages), but the TWO requested actions are the target architecture and its implementation; discovery was already completed.
D:
Monitoring content recommendation evolution for future drift is useful after the rollout, but it does not specify the requested target control and safe implementation for this case: recommendation service is tied to one proprietary model endpoint and cannot adapt to new languages.
E:
A brittle API contract blocks planned language expansion. The shortcut does not supply the requested design and safe rollout for content recommendation evolution, where product roadmap adds five languages and stricter content ranking rules next year.
Question 248
Case study: Regional online pharmacy: Cloud API quota-aware batching
Regional online pharmacy relies on Cloud API quota-aware batching to deliver its operational commitments. The current architectural problem is that client uploads product catalog through thousands of concurrent write requests and hits API quotas. The team has recorded this evidence: errors are 429 and retries synchronize into periodic bursts. Any proposed response must preserve the requirement that inventory updates must complete without losing products.
The Cloud API quota-aware batching environment at Regional online pharmacy faces this issue: client uploads product catalog through thousands of concurrent write requests and hits API quotas. Supporting evidence shows that errors are 429 and retries synchronize into periodic bursts. The architect must preserve the requirement that inventory updates must complete without losing products. Select TWO complementary actions: one durable target architecture and one controlled way to introduce it while preserving service. Which TWO are best?
- Ship a rate-aware batch worker with retry and dead-letter handling
- Track retry age, permanent failures and quota change requests
- Retry every failed request immediately with no backoff or limit
- Inspect quota headers, rate limits and idempotent write behavior
- Implement exponential backoff with jitter and bounded concurrency using client libraries
Explanation
Correct Answers
A:
Ship a rate-aware batch worker with retry and dead-letter handling is a controlled means to introduce the approved architecture while meeting the requirement that inventory updates must complete without losing products. 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:
Implement exponential backoff with jitter and bounded concurrency using client libraries targets the underlying issue (client uploads product catalog through thousands of concurrent write requests and hits API quotas) rather than merely reporting symptoms. It also preserves the required outcome (inventory updates must complete without losing products), which a target-state design must satisfy. This supplies the lasting-control half of the two-action package.
Incorrect Answers
B:
Monitoring Cloud API quota-aware batching for future drift is useful after the rollout, but it does not specify the requested target control and safe implementation for this case: client uploads product catalog through thousands of concurrent write requests and hits API quotas.
C:
Synchronized retries amplify quota exhaustion. The shortcut does not supply the requested design and safe rollout for Cloud API quota-aware batching, where errors are 429 and retries synchronize into periodic bursts.
D:
This diagnoses the original issue (client uploads product catalog through thousands of concurrent write requests and hits API quotas), but the TWO requested actions are the target architecture and its implementation; discovery was already completed.
Question 249
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.
Coastal emergency response unit plans to update on-prem data warehouse. Investigators found that profile shows source records revised after files are exported, consistent with the reported problem that warehouse queries depend on nightly CSV copies with missing change history. The response must satisfy the requirement that emergency operations need reproducible data not stale extracts. The design review is over and production introduction is next. Which action is the best fit?
- Run source and target in parallel across a full monthly correction cycle
- Track CDC lag, rejected schemas and incomplete snapshots
- Copy the same static CSV once and label it a real-time migration
- Design controlled CDC or incremental warehouse ingestion with lineage
Explanation
Correct Answer
A:
Run source and target in parallel across a full monthly correction cycle is a controlled means to introduce the approved architecture while meeting the requirement that emergency operations need reproducible data not stale extracts. Staged delivery, attributable change evidence and rollback guard against an otherwise avoidable service disruption.
Incorrect Answers
B:
The alternative “Track CDC lag, rejected schemas and incomplete snapshots” addresses later monitoring rather than the actual controlled deployment. In the reported observation, profile shows source records revised after files are exported. The option “Run source and target in parallel across a full monthly correction cycle” instead addresses the requested controlled implementation.
C:
A one-time extract cannot propagate later corrections. For this question, it is not a defensible choice for controlled implementation and it does not satisfy the requirement that emergency operations need reproducible data not stale extracts.
D:
The alternative “Design controlled CDC or incremental warehouse ingestion with lineage” describes the chosen design, but not how it is introduced safely. In the reported observation, profile shows source records revised after files are exported. The option “Run source and target in parallel across a full monthly correction cycle” instead addresses the requested controlled implementation.
Question 250
At Community farming platform, the gcloud project configuration workload has a documented risk: engineers repeatedly use their personal projects when running deployment scripts. The review found that shell history shows wrong active project and accidental unrelated resource creation. The architecture must also ensure that deployments must target designated staging projects only. The architecture review is paused until the team can prove the real failure path. Which investigation should happen first?
- Run deployment automation in isolated configuration environments
- Inspect gcloud configuration, account identity and project context
- Assume whatever project is active in Cloud Shell is always staging
- Use explicit project arguments, scoped identities and scripted preflight checks
Explanation
Correct Answer
B:
Inspect gcloud configuration, account identity and project context is the appropriate diagnostic because the observation—shell history shows wrong active project and accidental unrelated resource creation—does not by itself establish every dependency, authorization rule or failure boundary. The findings explain the reported risk (engineers repeatedly use their personal projects when running deployment scripts) before the design is approved.
Incorrect Answers
A:
The alternative “Run deployment automation in isolated configuration environments” is rollout work that may modify the wrong dependency before root-cause evidence is collected. In the reported observation, shell history shows wrong active project and accidental unrelated resource creation. The option “Inspect gcloud configuration, account identity and project context” instead addresses the requested root-cause evidence.
C:
Ambient CLI context can target the wrong project. For this question, it is not a defensible choice for root-cause evidence and it does not satisfy the requirement that deployments must target designated staging projects only.
D:
The alternative “Use explicit project arguments, scoped identities and scripted preflight checks” is target architecture selection; it does not establish why the observed path failed. In the reported observation, shell history shows wrong active project and accidental unrelated resource creation. The option “Inspect gcloud configuration, account identity and project context” instead addresses the requested root-cause evidence.