100% Latest & Updated Google Professional Cloud Architect Practice Test Questions, Exam Dumps & Verified Answers!
30 Days Free Updates, Instant Download!
Question 10
A subscription firm replaces synchronous calls to a fragile partner invoicing endpoint with independently operated services. Which implementation choice most directly delivers recoverable message processing?
- Allow downstream handlers to retry with no deduplication identifiers
- Discard a message on the first partner timeout
- Disable all dead-letter handling to simplify operations
- Give consumers idempotency controls and an explicit dead-letter handling route
Explanation
Correct Answer
D:Retries and dead-letter policy avoid silent lost invoices or duplicate settlement effects
Incorrect Answers
A:Duplicates can produce repeated settlements. Retries and dead-letter policy avoid silent lost invoices or duplicate settlement effects.
B:Transient failures become lost invoices. Retries and dead-letter policy avoid silent lost invoices or duplicate settlement effects.
C:Poison messages become untraceable. Retries and dead-letter policy avoid silent lost invoices or duplicate settlement effects.
Question 11
A subscription firm replaces synchronous calls to a fragile partner invoicing endpoint with independently operated services. Which validation would best establish end-to-end delivery evidence?
- Count events published to Pub/Sub without checking subscriber outcomes
- Check partner TLS validity without simulating outages
- Run a single successful invoice in a test environment
- Test partner outage and replay while reconciling published events to processed invoices
Explanation
Correct Answer
D:Reconciliation after deliberate failure validates business completion beyond queue acknowledgement
Incorrect Answers
A:Publishing is not invoice completion. Reconciliation after deliberate failure validates business completion beyond queue acknowledgement.
B:Certificate validity does not establish retry recovery. Reconciliation after deliberate failure validates business completion beyond queue acknowledgement.
C:Does not reveal duplicate or out-of-order effects. Reconciliation after deliberate failure validates business completion beyond queue acknowledgement.
Question 12
A subscription firm replaces synchronous calls to a fragile partner invoicing endpoint with independently operated services. Which operational action best maintains partner recovery operations after launch?
- Clear failed subscriptions as soon as partner service returns
- Increase web server timeouts indefinitely
- Monitor subscription lag and replay dead-letter messages after correcting the failure
- Monitor only frontend HTTP 200 responses
Explanation
Correct Answer
C:Lag and dead-letter processing reveal delayed business transactions and enable controlled recovery
Incorrect Answers
A:Deleting backlog discards accepted work. Lag and dead-letter processing reveal delayed business transactions and enable controlled recovery.
B:Does not recover independently processed events. Lag and dead-letter processing reveal delayed business transactions and enable controlled recovery.
D:Does not show downstream invoice settlement. Lag and dead-letter processing reveal delayed business transactions and enable controlled recovery.
Question 13
An engineering company must retain a commercial database during a transitional cloud migration and has restrictive license terms. Which architecture most directly meets the need for valid migration economics and license rights?
- Model license entitlements and support costs before selecting Compute Engine or managed alternatives
- Choose the least expensive VM type based only on CPU unit price
- Move the database to a different engine without verifying application support
- Use a spot instance for the licensed database by default
Explanation
Correct Answer
A:Licensing obligations and total cost can outweigh an otherwise attractive compute price
Incorrect Answers
B:Omits licensing and support obligations. Licensing obligations and total cost can outweigh an otherwise attractive compute price.
C:Forced replacement can break payroll functionality. Licensing obligations and total cost can outweigh an otherwise attractive compute price.
D:Unplanned interruption is unsuitable for the stateful service. Licensing obligations and total cost can outweigh an otherwise attractive compute price.
Question 14
An engineering company must retain a commercial database during a transitional cloud migration and has restrictive license terms. Which implementation choice most directly delivers controlled deployment of licensed software?
- Auto-scale commercial database replicas without recording entitlement rules
- Record approved instance shapes and license assignment in deployment policy
- Store licensing approvals only in a single engineer email thread
- Use maximum VM shape for all environments regardless of entitlement
Explanation
Correct Answer
B:Documented mappings limit accidentally running unlicensed instances or unsupported configurations
Incorrect Answers
A:Scale can create unlicensed instances. Documented mappings limit accidentally running unlicensed instances or unsupported configurations.
C:Lacks durable, auditable assignment. Documented mappings limit accidentally running unlicensed instances or unsupported configurations.
D:Capacity is neither cost nor license optimized. Documented mappings limit accidentally running unlicensed instances or unsupported configurations.
Question 15
An engineering company must retain a commercial database during a transitional cloud migration and has restrictive license terms. Which validation would best establish proof of a compliant cost estimate?
- Check VM startup and assume the license is valid
- Compare disk usage only before and after migration
- Test database query response time but omit the contract review
- Reconcile utilization, license allocation, and projected operating expenditure
Explanation
Correct Answer
D:A cost model is credible only when actual capacity and entitlement constraints are represented
Incorrect Answers
A:Boot does not establish contractual entitlement. A cost model is credible only when actual capacity and entitlement constraints are represented.
B:Does not assess license costs. A cost model is credible only when actual capacity and entitlement constraints are represented.
C:Performance alone cannot validate licensing. A cost model is credible only when actual capacity and entitlement constraints are represented.
Question 16
An engineering company must retain a commercial database during a transitional cloud migration and has restrictive license terms. Which operational action best maintains ongoing compliance with purchased rights after launch?
- Review license consumption only during the initial migration
- Review license usage after autoscaling or hardware changes
- Remove licensing records after cutover to simplify operations
- Use current cloud billing cost as the complete database TCO
Explanation
Correct Answer
B:Infrastructure changes can alter license consumption and contract exposure
Incorrect Answers
A:Autoscaling may alter future use. Infrastructure changes can alter license consumption and contract exposure.
C:Destroys evidence of rights and usage. Infrastructure changes can alter license consumption and contract exposure.
D:Excludes support and contract licensing exposure. Infrastructure changes can alter license consumption and contract exposure.
Question 17
A payment processor demands recovery of its customer authorization API within 20 minutes after a full regional disruption. Which architecture most directly meets the need for business-aligned recovery topology?
- Add more serving VMs inside the primary region only
- Rely on nightly untested snapshots with no alternate-region service capacity
- Increase request timeout during regional outages
- Design an alternate-region recovery path with a tested recovery-time objective
Explanation
Correct Answer
D:A regional dependency requires a recovery architecture beyond additional zones in the same region
Incorrect Answers
A:Additional zones cannot recover full-region loss. A regional dependency requires a recovery architecture beyond additional zones in the same region.
B:Backup files alone may not meet 20-minute service restoration. A regional dependency requires a recovery architecture beyond additional zones in the same region.
C:Longer timeouts do not restore a lost region. A regional dependency requires a recovery architecture beyond additional zones in the same region.
Question 18
A payment processor demands recovery of its customer authorization API within 20 minutes after a full regional disruption. Which implementation choice most directly delivers disaster-recovery readiness?
- Copy the database to another region but leave all application and identity dependencies primary-only
- Document a recovery runbook without deploying or testing resources
- Automate infrastructure restoration and replicate necessary state to the recovery location
- Permit the disaster recovery region to write concurrently without conflict controls
Explanation
Correct Answer
C:Recoverable data and deployable capacity are both required to serve transactions after region loss
Incorrect Answers
A:Recovery path remains incomplete. Recoverable data and deployable capacity are both required to serve transactions after region loss.
B:Manual plans may not meet a short RTO. Recoverable data and deployable capacity are both required to serve transactions after region loss.
D:Can create diverging authorization state. Recoverable data and deployable capacity are both required to serve transactions after region loss.
Question 19
A payment processor demands recovery of its customer authorization API within 20 minutes after a full regional disruption. Which validation would best establish evidence that the recovery objective is met?
- Measure how quickly an alternate-region VM boots
- Conduct a timed region-failure exercise measuring usable transaction recovery
- Verify the backup checksum but never restore application traffic
- Time failover from alert generation to first cloud console login
Explanation
Correct Answer
B:Successful VM startup alone does not demonstrate restoration of customer authorization
Incorrect Answers
A:Does not prove customer authorization succeeds. Successful VM startup alone does not demonstrate restoration of customer authorization.
C:Checks a file, not the recovery path. Successful VM startup alone does not demonstrate restoration of customer authorization.
D:Does not measure business function restoration. Successful VM startup alone does not demonstrate restoration of customer authorization.
Question 20
A payment processor demands recovery of its customer authorization API within 20 minutes after a full regional disruption. Which operational action best maintains keeping failover viable as releases change after launch?
- Keep the recovery architecture frozen after major application releases
- Review recovery incidents only after customer complaints
- Exercise recovery regularly and update runbooks with actual recovery timings
- Track backup spend instead of recovery durations
Explanation
Correct Answer
C:Recurring tests detect drift between documented recovery plans and deployed dependencies
Incorrect Answers
A:New dependencies may break old plans. Recurring tests detect drift between documented recovery plans and deployed dependencies.
B:Misses proactively detected drift. Recurring tests detect drift between documented recovery plans and deployed dependencies.
D:Cost alone does not establish RTO readiness. Recurring tests detect drift between documented recovery plans and deployed dependencies.
Google Cloud Certified - Professional Cloud Architect (Professional Cloud Architect) Practice Tests