100% Latest & Updated Google Professional Cloud Architect Practice Test Questions, Exam Dumps & Verified Answers!
30 Days Free Updates, Instant Download!
Question 260
A reservation service sees intermittent latency spikes after a cache upgrade; stakeholders blame the database without evidence. Which operational action best maintains prevention of recurring latency regressions after launch?
- Add a p95 trace-span regression gate for critical booking operations
- Keep logging the same symptoms without adding a regression test
- Increase database size every time p95 rises
- Ignore correlated frontend/cache anomalies in postmortems
Explanation
Correct Answer
A:A targeted check will catch the specific failure mode in later releases
Incorrect Answers
B:Does not prevent repeat failures. A targeted check will catch the specific failure mode in later releases.
C:May not fix cache cause. A targeted check will catch the specific failure mode in later releases.
D:Misses the true contributing tier. A targeted check will catch the specific failure mode in later releases.
Question 261
A startup provisions cloud projects through tickets; developers receive inconsistent security and monitoring configurations. Which architecture most directly meets the need for self-service standardized resource catalogs?
- Create projects from informal wiki instructions
- Copy a previous project with its old privileged identities intact
- Provide approved infrastructure modules and a governed service catalog
- Delegate all project creation to a single developer memory
Explanation
Correct Answer
C:Reusable blueprints reduce divergence between project environments
Incorrect Answers
A:Human steps remain inconsistent. Reusable blueprints reduce divergence between project environments.
B:May replicate unsafe grants. Reusable blueprints reduce divergence between project environments.
D:Not auditable or scalable. Reusable blueprints reduce divergence between project environments.
Question 262
A startup provisions cloud projects through tickets; developers receive inconsistent security and monitoring configurations. Which implementation choice most directly delivers restricted project bootstrap?
- Enforce versioned templates that create budgets, IAM roles and audit sinks together
- Provision networking first and leave required logging/budgets for optional later work
- Use mutable unversioned scripts without review
- Give every requester organization Admin privileges to bootstrap
Explanation
Correct Answer
A:Governance defaults should be installed automatically rather than remembered by each requester
Incorrect Answers
B:Templates can ship incomplete. Governance defaults should be installed automatically rather than remembered by each requester.
C:Cannot reproduce past defaults. Governance defaults should be installed automatically rather than remembered by each requester.
D:Overgrants access. Governance defaults should be installed automatically rather than remembered by each requester.
Question 263
A startup provisions cloud projects through tickets; developers receive inconsistent security and monitoring configurations. Which validation would best establish proof of template compliance?
- Check only project existence after catalog invocation
- Inspect catalog UI labels but not created resources
- Run one template lint check without provisioning test
- Provision a test project and inspect IAM, billing labels, logs and policy constraints
Explanation
Correct Answer
D:A catalog item being available does not prove it creates compliant resources
Incorrect Answers
A:Does not prove IAM/logging/budget defaults. A catalog item being available does not prove it creates compliant resources.
B:Presentation is not enforcement. A catalog item being available does not prove it creates compliant resources.
C:Misses cloud API failures. A catalog item being available does not prove it creates compliant resources.
Question 264
A startup provisions cloud projects through tickets; developers receive inconsistent security and monitoring configurations. Which operational action best maintains controlled template evolution after launch?
- Version modules and notify consuming teams about breaking changes
- Deploy template revisions to all consumers without versioning
- Treat manual post-provision corrections as permanent standard procedure
- Ignore template usage failures reported by new teams
Explanation
Correct Answer
A:Catalog updates can otherwise introduce configuration drift across projects
Incorrect Answers
B:Can create unexpected drift. Catalog updates can otherwise introduce configuration drift across projects.
C:Hides catalog design defects. Catalog updates can otherwise introduce configuration drift across projects.
D:Degrades self-service quality. Catalog updates can otherwise introduce configuration drift across projects.
Question 265
A financial institution runs failover exercises that repeatedly miss the agreed recovery time for a multi-tier application. Which architecture most directly meets the need for RTO-focused disaster recovery redesign?
- Plan recovery by database backup restoration alone
- Map every dependency in the transaction path and size recovery capacity for the target
- Add more VMs in the failed primary region only
- Define RTO by VM boot time instead of completed transactions
Explanation
Correct Answer
B:Recovering a database alone does not restore business functionality if other tiers remain unavailable
Incorrect Answers
A:Other tiers remain unavailable. Recovering a database alone does not restore business functionality if other tiers remain unavailable.
C:Cannot survive full regional loss. Recovering a database alone does not restore business functionality if other tiers remain unavailable.
D:Understates business recovery. Recovering a database alone does not restore business functionality if other tiers remain unavailable.
Question 266
A financial institution runs failover exercises that repeatedly miss the agreed recovery time for a multi-tier application. Which implementation choice most directly delivers automated service recovery sequence?
- Recover dependencies in random order during every drill
- Codify network, identity, compute and data recovery ordering in a tested runbook
- Keep infrastructure restore procedures only in one operator memory
- Wait for an untested manual DNS change before starting recovery
Explanation
Correct Answer
B:Manual unsequenced restoration adds delay and risks invalid dependencies
Incorrect Answers
A:Slows service restoration. Manual unsequenced restoration adds delay and risks invalid dependencies.
C:Not repeatable. Manual unsequenced restoration adds delay and risks invalid dependencies.
D:May breach RTO. Manual unsequenced restoration adds delay and risks invalid dependencies.
Question 267
A financial institution runs failover exercises that repeatedly miss the agreed recovery time for a multi-tier application. Which validation would best establish proof of end-to-end time-to-recover?
- Time until the first recovered VM responds to ping
- Inspect backup timestamps but do not attempt failover
- Time a complete customer transaction through a simulated regional outage
- Measure only the monitoring alert time before any user request works
Explanation
Correct Answer
C:Infrastructure health checks may turn green before business recovery is achieved
Incorrect Answers
A:Not customer transaction restoration. Infrastructure health checks may turn green before business recovery is achieved.
B:Not an end-to-end RTO test. Infrastructure health checks may turn green before business recovery is achieved.
D:Not service recovery. Infrastructure health checks may turn green before business recovery is achieved.
Question 268
A financial institution runs failover exercises that repeatedly miss the agreed recovery time for a multi-tier application. Which operational action best maintains ongoing recovery budget control after launch?
- Review exercise timing outliers and remediate the critical recovery path
- Accept every late drill without remediating critical path
- Ignore identity service recovery dependencies
- Update disaster recovery documentation without retesting revised sequence
Explanation
Correct Answer
A:An RTO goal is enforceable only when actual drills guide corrective engineering
Incorrect Answers
B:Repeats RTO misses. An RTO goal is enforceable only when actual drills guide corrective engineering.
C:Can block logins postfailover. An RTO goal is enforceable only when actual drills guide corrective engineering.
D:Does not prove improvement. An RTO goal is enforceable only when actual drills guide corrective engineering.
Question 269
A large insurer runs infrastructure tests only after Terraform has changed shared networking in production. Which architecture most directly meets the need for pre-apply infrastructure validation?
- Run Terraform apply from each engineer laptop without a plan review
- Rely only on manual post-deployment firewall inspection
- Use copy-pasted console clicks as the source of truth
- Use versioned Terraform plan review plus automated policy validation before apply
Explanation
Correct Answer
D:Early plan checks catch unexpected network or IAM changes before they become production state
Incorrect Answers
A:Uncontrolled mutation. Early plan checks catch unexpected network or IAM changes before they become production state.
B:Unsafe changes already reached production. Early plan checks catch unexpected network or IAM changes before they become production state.
C:Not reproducible or reviewable. Early plan checks catch unexpected network or IAM changes before they become production state.
Question 270
A large insurer runs infrastructure tests only after Terraform has changed shared networking in production. Which implementation choice most directly delivers approval-bound infrastructure changes?
- Allow any applicant to bypass policy validation and apply to production
- Disable state locking for faster parallel execution
- Approve arbitrary provider plugin versions during release
- Require reviewed plan artifacts and controlled apply identities with state locking
Explanation
Correct Answer
D:Allowing any user to apply an unreviewed plan defeats separation and reproducibility
Incorrect Answers
A:Not an enforced gate. Allowing any user to apply an unreviewed plan defeats separation and reproducibility.
B:Can corrupt resource state coordination. Allowing any user to apply an unreviewed plan defeats separation and reproducibility.
C:Weakens reproducibility. Allowing any user to apply an unreviewed plan defeats separation and reproducibility.
Google Cloud Certified - Professional Cloud Architect (Professional Cloud Architect) Practice Tests