ISC CCSP Practice Test 46
Question 450
Which specific remediation is the best controlled next step? An organization implements automatic deletion after 30 days despite an active legal hold on a subset of cloud messages. Select the ONE best answer.
- Perform privacy threat assessment, minimize training attributes and test model outputs for memorization or inference leakage before release.
- Suspend normal disposal for records under a valid scoped preservation hold and audit future release of the hold.
- Plan long-term archive format, key custody, integrity checks and periodic restoration tests before decommissioning dependencies.
- Enable suitable data events with actor identity, operation, object, source and synchronized timestamp, then secure their storage.
Explanation
Correct Answer
B:Acceptance criterion: Held records remain protected and retrievable with an auditable hold, while unrelated records continue their approved lifecycle. Legal preservation duties can override ordinary retention or deletion schedules for defined evidence.
Incorrect Answers
A:Perform privacy threat assessment, minimize training attributes and test model outputs for memorization or inference leakage before release. Here the decisive dependency is: Legal preservation duties can override ordinary retention or deletion schedules for defined evidence. Use the specified intervention: Suspend normal disposal for records under a valid scoped preservation hold and audit future release of the hold.
C:Plan long-term archive format, key custody, integrity checks and periodic restoration tests before decommissioning dependencies. Here the decisive dependency is: Legal preservation duties can override ordinary retention or deletion schedules for defined evidence. Use the specified intervention: Suspend normal disposal for records under a valid scoped preservation hold and audit future release of the hold.
D:That option would enable suitable data events with actor identity, operation, object, source and synchronized timestamp, then secure their storage, but it addresses a different concern. Legal preservation duties can override ordinary retention or deletion schedules for defined evidence.
Question 451
The operator needs proof rather than a guess. Which investigation should come first? A regional disaster recovery exercise powers on database VMs but never validates inter-service authentication or dependent message queues. Select the ONE best answer.
- Trace network routes, security groups, flow logs, hypervisor port groups and service dependency inventory.
- Compare RTO, backup size, network transfer rates, dependency sequence and recent full recovery trial durations.
- Inspect architecture diagrams, available service settings, shared responsibility artifacts and the control owner matrix.
- Inspect service dependency diagrams, queues, name resolution, external identities and recovery runbook acceptance tests.
Explanation
Correct Answer
D:Infrastructure restart success is not equivalent to recovery of an application business process or its dependent services. Extend the recovery plan to include workload ordering, identity, queues, DNS and end-to-end business transaction tests.
Incorrect Answers
A:Virtual NIC boundaries do not prevent lateral movement through permitted shared networks and routed infrastructure. The focused check for the reported condition is: Inspect service dependency diagrams, queues, name resolution, external identities and recovery runbook acceptance tests.
B:A backup being available does not guarantee the workload can be restored within the business's acceptable downtime. The focused check for the reported condition is: Inspect service dependency diagrams, queues, name resolution, external identities and recovery runbook acceptance tests.
C:Cloud platform risk assessment depends on understanding where the provider's responsibilities end and tenant configuration begins. The focused check for the reported condition is: Inspect service dependency diagrams, queues, name resolution, external identities and recovery runbook acceptance tests.
Question 452
Before changing production, which focused diagnostic is most useful for identifying the true dependency? An organization implements automatic deletion after 30 days despite an active legal hold on a subset of cloud messages. Select the ONE best answer.
- Compare input and output digests, signed manifests, trusted metadata channels, and tampering test evidence.
- Trace processing jobs, data lineage records, copies, retention schedules, exports and evidence of completed deletion.
- Inspect dataset digests, source licenses, validation rules, feature distributions and anomalous sample ingestion events.
- Inspect hold case IDs, custodian and record scope, retention policies, deletion events and counsel release authorization.
Explanation
Correct Answer
D:Suspend normal disposal for records under a valid scoped preservation hold and audit future release of the hold. This observation directly isolates the suspected dependency before a risky change.
Incorrect Answers
A:A collision-resistant hash helps detect content changes, but an unauthenticated digest alone does not authenticate its claimed origin. The focused check for the reported condition is: Inspect hold case IDs, custodian and record scope, retention policies, deletion events and counsel release authorization.
B:Map every data lifecycle phase, including use, sharing, archiving and destruction, with accountable owners and handling controls. The focused check for the reported condition is: Inspect hold case IDs, custodian and record scope, retention policies, deletion events and counsel release authorization.
C:Artifact authenticity alone does not prevent adversarial or low-quality input data from corrupting machine-learning behavior. The focused check for the reported condition is: Inspect hold case IDs, custodian and record scope, retention policies, deletion events and counsel release authorization.
Question 453
Before changing production, which focused diagnostic is most useful for identifying the true dependency? A cloud workload is compromised through an exposed access token, but the incident playbook revokes only the human user's session. Select the ONE best answer.
- Inspect token issuance, IAM grants, API event history, secret dependencies and containment verification attempts.
- Inspect rule logic, sample correlated events, alert severity, response times and analyst workload metrics.
- Review contact matrix, contractual assistance terms, evidence request protocol and tabletop response times.
- Review incident classification, geographic data map, communication owners, known facts, breach triggers and approval trail.
Explanation
Correct Answer
A:Machine credentials can remain exploitable independently of a human web session until explicitly revoked and related trust is reviewed. Contain the affected workload credential, rotate dependent secrets, investigate access and restore under an accountable incident workflow. This check separates the suspected dependency before a risky change.
Incorrect Answers
B:SOC effectiveness depends on context and signal quality; a high alert volume alone does not demonstrate useful detection. The focused check for the reported condition is: Inspect token issuance, IAM grants, API event history, secret dependencies and containment verification attempts.
C:Cloud incident response may depend on provider-controlled telemetry and timely coordinated action across service boundaries. The focused check for the reported condition is: Inspect token issuance, IAM grants, API event history, secret dependencies and containment verification attempts.
D:Incident communication must reflect verified facts, contractual commitments and applicable notification requirements rather than speculation. The focused check for the reported condition is: Inspect token issuance, IAM grants, API event history, secret dependencies and containment verification attempts.
Question 454
The operator needs proof rather than a guess. Which investigation should come first? A business allows every employee to assign arbitrary cloud data labels, creating conflicting categories for the same regulated records. Select the ONE best answer.
- Inventory producer and consumer identities, topic permissions, intermediate retention, encryption and error handling.
- Review quasi-identifiers, linkage sources, k-anonymity style metrics where appropriate and data release tests.
- Review classification policy versions, conflicting object tags, owner approvals and effective enforcement results.
- Trace ETL dependencies, reidentification fields, classification decisions, retention rules and data-owner approvals.
Explanation
Correct Answer
C:Establish centrally governed taxonomy, owner approval and deterministic label precedence with user training and periodic review.
Incorrect Answers
A:Data in transit and temporary queues may inherit the source's security classification even when they are not persistent databases. The focused check for the reported condition is: Review classification policy versions, conflicting object tags, owner approvals and effective enforcement results.
B:Review quasi-identifiers, linkage sources, k-anonymity style metrics where appropriate and data release tests. The focused check for the reported condition is: Review classification policy versions, conflicting object tags, owner approvals and effective enforcement results.
D:Derived records may remain sensitive even when their original source table is not directly exposed. The focused check for the reported condition is: Review classification policy versions, conflicting object tags, owner approvals and effective enforcement results.
Question 455
Select TWO appropriate investigation, change, and acceptance elements for this situation. A municipal service experiences unpredictable web demand and expects cloud resource pooling to provide unlimited compute without quotas or service limits.
- Inspect ground-truth labels, false-positive rates, current features, drift alerts, and sample analyst adjudications.
- Document the AI-as-a-service responsibility boundary and implement customer data validation, prompt handling, logging, and access governance.
- Promise infinite automatic capacity solely because underlying hardware is multi-tenant and service is marketed as cloud.
- Compare regional quota limits, scale-set desired capacity, provider throttling events, and peak workload demand traces.
- Model quota, regional capacity, scaling ceilings, and workload requirements before setting a realistic elasticity objective.
Explanation
Correct Answer
D:Resource pooling and on-demand provisioning still operate within provider quotas, resource availability, and customer budget constraints. Model quota, regional capacity, scaling ceilings, and workload requirements before setting a realistic elasticity objective. This observation directly isolates the suspected dependency before a risky change.
E:Before making it, verify: Compare regional quota limits, scale-set desired capacity, provider throttling events, and peak workload demand traces. Acceptance criterion: Peak-load tests meet service objectives within documented account limits and show controlled capacity when quotas are reached.
Incorrect Answers
A:This check investigates a different symptom (A cloud threat-detection model starts marking many legitimate payroll sessions as attacks after a seasonal change in user behavior.). For this incident, the targeted check is: Compare regional quota limits, scale-set desired capacity, provider throttling events, and peak workload demand traces.
B:Here the decisive dependency is: Resource pooling and on-demand provisioning still operate within provider quotas, resource availability, and customer budget constraints. This action belongs to a different situation: A data science group deploys an AI model through a provider API but assumes provider infrastructure controls also validate customer prompts and sensitive training uploads.
C:Resource pooling and on-demand provisioning still operate within provider quotas, resource availability, and customer budget constraints. The required remediation is: Model quota, regional capacity, scaling ceilings, and workload requirements before setting a realistic elasticity objective.
Question 456
Which acceptance evidence would show that the original fault or risk was actually remediated? A healthcare application provides public read access to a storage bucket while relying on client-side file names as the privacy boundary. Select the ONE best answer.
- Approved document access recovers after controlled certificate renewal without granting broad unreviewed privileges.
- A poisoned test sample is quarantined with lineage evidence while vetted datasets continue producing reproducible outputs.
- A test sensitive prompt is redacted from analytics views and any retained record follows the approved lifecycle and access rules.
- An anonymous or unrelated account cannot retrieve a protected sample while approved clinicians retain audited access.
Explanation
Correct Answer
D:Object storage has native authorization controls; obscure object names provide neither confidentiality nor controlled sharing. It validates the controlled change: Apply private object policies, verified request authorization, access logging and data classification before storing patient artifacts.
Incorrect Answers
A:Rights enforcement may rely on certificate and key lifecycle, so validity and revocation directly affect authorized consumption. This is a different acceptance outcome; the required result here is: An anonymous or unrelated account cannot retrieve a protected sample while approved clinicians retain audited access.
B:Artifact authenticity alone does not prevent adversarial or low-quality input data from corrupting machine-learning behavior. This is a different acceptance outcome; the required result here is: An anonymous or unrelated account cannot retrieve a protected sample while approved clinicians retain audited access.
C:AI telemetry may reproduce personal information and requires lifecycle protection like other derived cloud data products. This is a different acceptance outcome; the required result here is: An anonymous or unrelated account cannot retrieve a protected sample while approved clinicians retain audited access.
Question 457
Before changing production, which focused diagnostic is most useful for identifying the true dependency? A legacy application concatenates user search text into SQL statements running against a managed cloud database. Select the ONE best answer.
- Inspect database query construction, ORM bindings, user input samples, DB grants and injected test payload behavior.
- Inspect gateway routes, quotas, JWT validation, backend authorization tests and request telemetry under load.
- Review data flow diagrams, trust boundaries, service tokens, privilege levels and threat-to-control traceability.
- Review service account grants, API key inventories, token expiration and machine identity authentication logs.
Explanation
Correct Answer
A:SQL injection follows unsafe command construction and remains possible even when the database is a managed cloud service. Use parameterized queries, least-privileged database identities and tests for unsafe input handling before release.
Incorrect Answers
B:An API gateway can enforce shared entry controls but does not replace application authorization and secure backend logic. The focused check for the reported condition is: Inspect database query construction, ORM bindings, user input samples, DB grants and injected test payload behavior.
C:This examines a different symptom (A team models only external attackers while a managed database application also faces threats from compromised internal service identities.).
D:Human MFA and workload authentication address different actors and require distinct assurance and revocation mechanisms. The focused check for the reported condition is: Inspect database query construction, ORM bindings, user input samples, DB grants and injected test payload behavior.
Question 458
A change engineer must select a justified corrective intervention. Which should be approved? A company connects a private datacenter to two public providers and assumes the design has no shared identity or transit dependencies. Select the ONE best answer.
- Request the module-specific validated security certificate and check its exact version, operating mode and relevant algorithm claims.
- Build eligible service replicas and data recovery across independent zones and test traffic failover under an agreed recovery objective.
- Classify the deployment as a community cloud when access and governance are restricted to organizations with shared concerns.
- Map hybrid and multi-cloud control planes, trust relationships, routing, and outage dependencies before declaring independent resilience.
Explanation
Correct Answer
D:Using multiple clouds can improve options but shared authentication, network, or key services can remain common failure points. Acceptance criterion: Failure testing of one provider maintains required workflows using independent authorized identity and connectivity dependencies.
Incorrect Answers
A:Request the module-specific validated security certificate and check its exact version, operating mode and relevant algorithm claims. Here the decisive dependency is: Using multiple clouds can improve options but shared authentication, network, or key services can remain common failure points. Use the specified intervention: Map hybrid and multi-cloud control planes, trust relationships, routing, and outage dependencies before declaring independent resilience.
B:That option would build eligible service replicas and data recovery across independent zones and test traffic failover under an agreed recovery objective, but it addresses a different concern. Using multiple clouds can improve options but shared authentication, network, or key services can remain common failure points.
C:Classify the deployment as a community cloud when access and governance are restricted to organizations with shared concerns. Here the decisive dependency is: Using multiple clouds can improve options but shared authentication, network, or key services can remain common failure points. Use the specified intervention: Map hybrid and multi-cloud control planes, trust relationships, routing, and outage dependencies before declaring independent resilience.
Question 459
Which acceptance evidence would show that the original fault or risk was actually remediated? A backup architect stores a regulated archive in ephemeral compute scratch disks because they appear cheaper than persistent object storage. Select the ONE best answer.
- An archive remains recoverable after source-instance deletion and a sample restore matches the retention and integrity requirements.
- The retirement record identifies the approved sanitization method and demonstrates loss of the protected data decryption capability.
- Conflicting test labels resolve under documented rules and sensitive records trigger the intended handling restrictions.
- An anonymous or unrelated account cannot retrieve a protected sample while approved clinicians retain audited access.
Explanation
Correct Answer
A:Ephemeral storage can disappear with a workload lifecycle event and cannot be assumed durable for long-lived records. It validates the controlled change: Select durable storage with documented retention, redundancy and restore support rather than instance-tied ephemeral scratch storage.
Incorrect Answers
B:Logical deletion and physical media sanitization are distinct processes whose evidence depends on the service and control boundary. This is a different acceptance outcome; the required result here is: An archive remains recoverable after source-instance deletion and a sample restore matches the retention and integrity requirements.
C:The alternative observation is: Conflicting test labels resolve under documented rules and sensitive records trigger the intended handling restrictions. The acceptance test for this case is: An archive remains recoverable after source-instance deletion and a sample restore matches the retention and integrity requirements.
D:Object storage has native authorization controls; obscure object names provide neither confidentiality nor controlled sharing. This is a different acceptance outcome; the required result here is: An archive remains recoverable after source-instance deletion and a sample restore matches the retention and integrity requirements.
Question 460
Which observed outcome provides the strongest verification for this requirement? A provider offers separate cloud regions but a business workload uses one centralized corporate DNS resolver with no recovery alternative. Select the ONE best answer.
- The reachable privileged endpoint is protected under the emergency window and retesting demonstrates reduced exploitability.
- A simulated regional loss completes an authorized end-to-end transaction and meets the signed recovery service objective.
- An adversarial test tenant fails to read another tenant's data or management state across each agreed trust boundary.
- During loss of the primary resolver, approved clients continue resolving application endpoints within the recovery objective.
Explanation
Correct Answer
D:Service availability can be limited by shared DNS or identity dependencies outside the service's main compute failure domains. It validates the controlled change: Redesign critical name-resolution dependencies with appropriate secondary resolvers and validation of failover across regions.
Incorrect Answers
A:A useful risk assessment prioritizes realistic loss exposure rather than sorting only by a vulnerability score or age. This is a different acceptance outcome; the required result here is: During loss of the primary resolver, approved clients continue resolving application endpoints within the recovery objective.
B:Infrastructure restart success is not equivalent to recovery of an application business process or its dependent services. This is a different acceptance outcome; the required result here is: During loss of the primary resolver, approved clients continue resolving application endpoints within the recovery objective.
C:Tenant isolation is multilayered; separating names or projects alone does not guarantee prevention of cross-tenant access. This is a different acceptance outcome; the required result here is: During loss of the primary resolver, approved clients continue resolving application endpoints within the recovery objective.