View Full Salesforce Certified Integration Architect Exam Dumps and Practice Test Dumps.

Q361. Which design best reduces risk when a downstream system cannot guarantee continuous availability

1. Direct database sharing
2. Long synchronous sessions
3. Immediate permanent failure
4. Store and forward messaging

Correct Answer: 4. Store and forward messaging

Explanation

Store and forward messaging accepts a transaction and preserves it until the downstream system is available for processing. Direct database sharing creates tight coupling between applications. Long synchronous sessions are vulnerable to timeouts and service outages. Immediate permanent failure unnecessarily rejects work that could succeed later. Store and forward messaging is therefore the correct answer because temporary downstream unavailability does not have to stop the originating Salesforce process. Architects should define message durability retry limits retention monitoring and exception handling so stored transactions are eventually processed or escalated when automatic recovery is no longer appropriate.

Q362. Which practice best limits the amount of confidential information copied into middleware

1. Selective field extraction
2. Full record replication
3. Unlimited payload logging
4. Administrator access for every service

Correct Answer: 1. Selective field extraction

Explanation

Selective field extraction transfers only the fields that are required for the integration purpose. Full record replication can expose unnecessary sensitive data. Unlimited payload logging creates additional copies of confidential information. Administrator access grants excessive privileges. Selective field extraction is therefore the correct answer because reducing transferred data lowers security privacy and storage risk while often improving performance. Architects should define explicit field mappings and review them when schemas change so newly introduced confidential fields are not automatically included without a documented business need and appropriate authorization.

Q363. Which approach best detects whether an external provider has changed a required response field

1. Manual browser review
2. Ignore unknown responses
3. Response schema validation
4. Disable contract testing

Correct Answer: 3. Response schema validation

Explanation

Response schema validation checks whether an external service response contains the expected fields structures and data types. Manual browser review is inconsistent and difficult to automate. Ignoring unexpected responses can allow incorrect data into business processing. Disabling contract testing reduces protection against interface changes. Response schema validation is therefore the correct answer because external providers can introduce incompatible changes even when the service remains technically reachable. Architects should combine runtime validation with automated contract testing so breaking changes are identified early and failed messages produce clear diagnostic information.

Q364. Which pattern best handles one incoming request that must be converted into several sequential business operations

1. Data replication
2. Command orchestration
3. Browser navigation
4. Static file export

Correct Answer: 2. Command orchestration

Explanation

Command orchestration coordinates several dependent business operations in a defined sequence after receiving one initiating request. Data replication copies information rather than controlling process order. Browser navigation is a user interface behavior. Static file export does not coordinate transactional work. Command orchestration is therefore the correct answer because complex integrations may require validation creation enrichment and notification steps to occur according to business rules. Architects should define failure handling state management and compensation where appropriate so partial completion does not leave participating systems in an unknown or inconsistent condition.

Q365. Which integration control best protects Salesforce from sudden bursts generated by an external partner

1. Unlimited parallel requests
2. Disable traffic monitoring
3. Shared login credentials
4. Inbound request buffering

Correct Answer: 4. Inbound request buffering

Explanation

Inbound request buffering absorbs temporary traffic spikes and allows requests to be processed at a controlled rate instead of overwhelming Salesforce or middleware. Unlimited parallel requests increase the chance of capacity problems. Disabling monitoring hides abnormal traffic. Shared credentials weaken accountability and do not manage load. Inbound request buffering is therefore the correct answer because external partners may generate unpredictable bursts even when average traffic is acceptable. Architects should define queue capacity priority rules backpressure behavior and alert thresholds so buffering protects the platform without hiding sustained overload.

Q366. Which Salesforce capability is most suitable for exposing custom logic through a REST endpoint

1. Report API
2. Apex REST
3. Static Resources
4. Dashboard subscriptions

Correct Answer: 2. Apex REST

Explanation

Apex REST allows custom Salesforce business logic to be exposed through REST resources defined in Apex. Report API capabilities focus on analytics. Static Resources store files. Dashboard subscriptions distribute dashboard information. Apex REST is therefore the correct answer when standard Salesforce APIs do not provide the required custom business operation. Architects should prefer standard APIs where possible because custom endpoints increase development testing documentation and maintenance responsibilities. Security governor limits input validation error handling and transaction behavior should all be considered when designing an Apex REST service.

Q367. Which design best prevents two workers from processing the same queued task simultaneously

1. Exclusive message claim
2. Remove task identifiers
3. Shared browser session
4. Disable queue tracking

Correct Answer: 1. Exclusive message claim

Explanation

Exclusive message claim ensures that one worker owns a queued task while it is being processed so another worker does not perform the same work concurrently. Removing identifiers makes tracking more difficult. Shared browser sessions are unrelated to backend queue processing. Disabling queue tracking reduces reliability. Exclusive message claim is therefore the correct answer because parallel workers are useful for scaling but must coordinate task ownership. Architects should also define claim timeout and recovery behavior so work becomes available again if the original worker fails before completing the transaction.

Q368. Which practice best identifies performance degradation caused by one external dependency

1. Monitor only total transaction count
2. Remove dependency timing
3. Per service latency monitoring
4. Disable distributed traces

Correct Answer: 3. Per service latency monitoring

Explanation

Per service latency monitoring measures response time for each external dependency separately. Total transaction counts do not show which service is slow. Removing timing information and disabling tracing reduce diagnostic visibility. Per service latency monitoring is therefore the correct answer because one dependency can degrade while the overall integration still appears partially functional. Architects should review averages and higher percentile latency together with error rates so intermittent slowness is not hidden. Correlation with end to end transaction timing can then show how much each dependency contributes to the user or business experience.

Q369. Which strategy best prevents integration logic from depending on Salesforce record IDs that differ between environments

1. Stable external identifier
2. Hard coded production ID
3. Record position
4. Display label only

Correct Answer: 1. Stable external identifier

Explanation

A stable external identifier provides a consistent business key that can be used across development testing production and connected applications. Hard coded Salesforce record IDs can differ between environments. Record position is not a reliable identifier. Display labels may change or may not be unique. Stable external identifier is therefore the correct answer because integrations should match records using durable keys rather than environment specific Salesforce IDs. Architects should ensure the chosen identifier is unique governed and preserved during migrations or data consolidation.

Q370. Which mechanism best protects a target service when Salesforce sends requests faster than it can process them

1. Increase synchronous calls
2. Disable queueing
3. Remove rate controls
4. Controlled dispatch rate

Correct Answer: 4. Controlled dispatch rate

Explanation

A controlled dispatch rate limits how quickly queued work is delivered to a downstream service according to its sustainable capacity. Increasing synchronous calls can worsen overload. Disabling queueing removes a useful buffer. Removing rate controls exposes the target to traffic spikes. Controlled dispatch rate is therefore the correct answer because the producer and consumer may have different processing capacities. Architects should monitor throughput backlog and message age and adjust the dispatch policy when downstream capacity changes while ensuring important transactions are not delayed beyond acceptable business service levels.

Q371. Which test best verifies that retry logic does not create duplicate business records

1. User interface testing
2. Retry behavior testing
3. Report formatting test
4. Permission review

Correct Answer: 2. Retry behavior testing

Explanation

Retry behavior testing intentionally recreates timeouts failures and repeated delivery to verify that retries do not create duplicate business effects. User interface testing report formatting and permission reviews do not validate retry safety. Retry behavior testing is therefore the correct answer because integrations often face uncertain outcomes where a request may have completed before a timeout was observed. Architects should test stable request identifiers idempotency logic and response reuse so repeated attempts produce one logical business result rather than multiple orders payments or records.

Q372. Which architecture helps provide a single customer profile assembled from several source systems

1. Browser cache
2. Shared administrator login
3. Customer data aggregation layer
4. Report subscription

Correct Answer: 3. Customer data aggregation layer

Explanation

A customer data aggregation layer combines relevant customer information from multiple authoritative systems into a unified representation for consuming applications. Browser caches do not provide governed enterprise integration. Shared administrator logins are a security concern. Report subscriptions distribute analytics rather than assembling reusable customer data. Customer data aggregation layer is therefore the correct answer because many organizations store customer attributes across Salesforce billing support and other platforms. Architects should define source ownership matching rules freshness requirements and privacy controls so the assembled profile remains accurate and appropriately governed.

Q373. Which approach best supports moving an integration endpoint to a new host with minimal consumer change

1. DNS based endpoint alias
2. Hard coded host in every client
3. Manual source edits
4. Shared database access

Correct Answer: 1. DNS based endpoint alias

Explanation

A DNS based endpoint alias gives consumers a stable service name while the underlying host can change without requiring every client to modify application code. Hard coded hosts create direct infrastructure dependency. Manual source edits increase deployment effort. Shared database access does not solve endpoint portability. DNS based endpoint alias is therefore the correct answer because infrastructure changes can be isolated from consuming applications. Architects should still plan certificate configuration caching behavior and controlled cutover so the alias points to the correct healthy service during migration.

Q374. Which pattern best prevents a permanently invalid message from blocking following messages in a queue

1. Retry it forever
2. Route poison message aside
3. Stop the entire queue
4. Delete every pending message

Correct Answer: 2. Route poison message aside

Explanation

Routing a poison message aside removes a permanently failing message from the normal processing path so later valid messages can continue. Retrying forever can repeatedly block progress. Stopping the queue allows one invalid transaction to affect unrelated work. Deleting all pending messages creates unnecessary data loss. Route poison message aside is therefore the correct answer because malformed or permanently invalid messages require separate exception handling. Architects should preserve the failed message error details and correlation information so support teams can correct and safely replay it later.

Q375. Which security practice best protects stored API secrets from being read directly by application code repositories

1. Put secrets in source files
2. Share them in documentation
3. Use a secrets manager
4. Store them in report descriptions

Correct Answer: 3. Use a secrets manager

Explanation

A secrets manager stores sensitive credentials separately from application source code and provides controlled runtime access. Putting secrets in source files can expose them through repositories and deployment packages. Sharing them in documentation creates additional uncontrolled copies. Report descriptions are not secure credential storage. Use a secrets manager is therefore the correct answer because integration credentials should be protected through access controls auditing rotation and centralized lifecycle management. Architects should also ensure applications request only the secrets they need and that retired credentials are revoked promptly.

Q376. Which design best handles a consumer that needs historical state changes rather than only the latest record state

1. Overwrite history
2. Keep only current values
3. Full daily replacement
4. Append only event history

Correct Answer: 4. Append only event history

Explanation

Append only event history preserves each significant state change so consumers can understand how a business entity evolved over time. Overwriting history or keeping only current values removes earlier transitions. Full daily replacement captures snapshots rather than every meaningful change. Append only event history is therefore the correct answer when downstream systems require an auditable sequence of business changes. Architects should define event retention ordering identifiers and privacy requirements because historical event data can grow substantially and may contain sensitive information.

Q377. Which design best prevents a temporary increase in workload from immediately requiring permanent infrastructure expansion

1. Elastic scaling
2. Fixed single worker
3. Disable queues
4. Reject all additional traffic

Correct Answer: 2. Elastic scaling

Explanation

Elastic scaling adds or removes processing capacity according to workload demand so temporary peaks can be handled without permanently running maximum resources. A fixed single worker cannot adapt to growth. Disabling queues reduces buffering options. Rejecting all additional traffic may unnecessarily interrupt business processing. Elastic scaling is therefore the correct answer because integration workloads often vary throughout the day or during scheduled imports. Architects should define scaling thresholds and ensure parallel processing does not violate ordering locking or downstream rate constraints.

Q378. Which mechanism best helps identify whether a Salesforce integration is approaching an API consumption threshold

1. Dashboard theme review
2. Record count only
3. Browser session monitoring
4. API consumption alerting

Correct Answer: 4. API consumption alerting

Explanation

API consumption alerting monitors usage and warns operations teams before available API capacity reaches an important threshold. Dashboard themes and browser sessions are unrelated. Record counts do not directly measure API request consumption. API consumption alerting is therefore the correct answer because several integrations may share Salesforce limits and one unexpected workload can affect others. Architects should establish normal usage baselines identify high consuming clients and set practical warning thresholds so corrective action can be taken before critical integrations are disrupted.

Q379. Which approach best verifies that an integration can recover after the middleware server is restarted

1. Happy path testing only
2. Report comparison
3. Restart recovery testing
4. Dashboard review

Correct Answer: 3. Restart recovery testing

Explanation

Restart recovery testing intentionally interrupts middleware and verifies that queued work processing state and transactions recover correctly after the service returns. Happy path testing does not validate interruption behavior. Report comparison and dashboard review are not substitutes for resilience testing. Restart recovery testing is therefore the correct answer because production services may restart during maintenance infrastructure failure or deployment. Architects should verify that durable messages remain available checkpoints are preserved and transactions do not become lost or duplicated when processing resumes.

Q380. Which governance practice defines how critical integration services should be restored after a major outage

1. Disaster recovery plan
2. Browser preference policy
3. Report naming standard
4. Dashboard formatting guide

Correct Answer: 1. Disaster recovery plan

Explanation

A disaster recovery plan defines how critical integration services data configurations and dependencies should be restored after a major outage. Browser preferences report naming and dashboard formatting do not address service recovery. Disaster recovery plan is therefore the correct answer because integrations may depend on middleware credentials queues network services and external endpoints that all need coordinated restoration. Architects should define recovery priorities backup requirements responsible teams testing frequency and acceptable recovery objectives so restoration procedures are practical and understood before a serious disruption occurs.