Question 680 A business wants to switch back from a disaster recovery region after primary services are rebuilt. What should determine the timing? The first moment the primary server responds to ping Which region has the lower cloud billing rate that day A tested failback plan, synchronized data state, verified security posture and business-authorized cutover An arbitrary deadline set before checking data reconciliation Explanation Correct Answer C:Failback can cause data conflict or renewed outage if dependencies are not validated. Incorrect Answers A:Network reachability is not sufficient business recovery. B:Cost alone cannot establish safe failback readiness. D:Premature cutover can lose recent transactions.
Question 681 The incident team has restored endpoint fleets but has not verified the security monitoring agent is operational. What does this mean for recovery acceptance? Recovery remains incomplete until minimum detection and logging coverage is verified for the restored assets Monitoring can be ignored because the operating system was reinstalled Close the incident and remove escalation contacts Record the issue as a cosmetic interface preference only Explanation Correct Answer A:A service may run while being unable to detect recurrence. Incorrect Answers B:Reinstallation does not prove telemetry is active. C:No visibility creates residual risk. D:Missing detection capability is substantive operational exposure.
Question 682 An executive asks when a compromised identity store can be restored. The team has a backup but cannot establish which accounts and secrets were altered. What should precede cutover? Restore all identities without review to minimize downtime Remove all role audit events after restoration Revoke every legitimate employee permanently without an access recovery path Reconcile identities and privileged changes to a trusted baseline and rotate at-risk secrets before enabling production access Explanation Correct Answer D:Restoring raw directory data may retain malicious principals or credentials. Incorrect Answers A:Attacker-created access could be reactivated. B:Those records are needed to identify unauthorized changes. C:An indiscriminate permanent block prevents necessary business operation.
Question 683 After recovery, a formerly compromised server exhibits repeated beacon attempts to a suspicious host. What is the most defensible decision? Call the attempts harmless because the server was reimaged Suspend reconnection, reinvestigate persistence and validate the effective controls before declaring eradication Disable detection for that destination to make the recovery dashboard green Wait until the next quarterly audit to assess active beacons Explanation Correct Answer B:Continuing attacker-like behavior is evidence that recovery criteria have not been met. Incorrect Answers A:Imaging does not resolve all potential external or identity footholds. C:Suppressing evidence does not address the behavior. D:Active recurrence needs prompt assessment.
Question 684 A contact center has switched to a temporary cloud platform after an attack. Customers can call, but recordings and transaction audit trails are missing. How should management classify the state? Full recovery because inbound calls are answered No recovery because the original site is still offline Partial recovery with identified compliance and evidence gaps requiring remediation or formal risk acceptance Permanent waiver of all recordkeeping rules without legal review Explanation Correct Answer C:Operational availability does not automatically restore mandated records and controls. Incorrect Answers A:Critical recording and audit functions remain unavailable. B:An alternate service may meet some recovery functions. D:Legal duties cannot be unilaterally discarded for convenience.
Question 685 A security team rebuilt a compromised application but reused the same exposed signing private key. What is the material remaining risk? Only the application server's hostname may be wrong There is no risk if the new application binary hash differs Renaming the certificate file is sufficient because users will not recognize it Attackers may still produce trusted-looking signed artifacts; revoke and replace the affected key and update trust appropriately Explanation Correct Answer D:Rebuilding software does not invalidate a stolen signing credential. Incorrect Answers A:A signing-key compromise creates independent trust risk. B:The same private key may sign new malicious artifacts. C:Key identity and cryptographic trust are not changed by filenames.
Question 686 A post-incident review after phishing finds response was delayed because the on-call roster was outdated. What is the best corrective action? Record 'communicate better' without an owner or test Dismiss the failure because the incident was eventually contained Assign ownership for roster maintenance and test the updated escalation path with evidence of successful contact Replace the full security strategy based solely on one stale telephone number Explanation Correct Answer C:A lesson learned should produce a measurable control improvement. Incorrect Answers A:A vague note cannot demonstrate improvement. B:A delayed response remains a process gap. D:Broad unrelated change is disproportionate to the finding.
Question 687 A ransomware response exceeded its recovery objective by eight hours. Which post-incident analysis is most useful? Map actual recovery milestones and dependencies against the approved RTO to identify the critical delay Use the number of incident emails as a proxy for restoration time Declare the objective met because the team worked overtime Omit supplier-related delays to avoid difficult contract discussions Explanation Correct Answer A:A measured timeline can locate the bottleneck for remediation. Incorrect Answers B:Message volume does not show business recovery timing. C:Effort does not change the measured outcome. D:External dependencies can be material root causes.
Question 688 A postmortem blames a junior analyst for an alert that went untriaged, but records show the case queue silently discarded high-priority notifications. What should review leadership prioritize? Punish the analyst without examining technical evidence Erase the queue logs before conducting interviews An evidence-based, blameless analysis of queue routing and escalation controls with assigned fixes Classify every future alert as highest priority to avoid triage decisions Explanation Correct Answer C:Process and tooling deficiencies need correction beyond attributing fault to one individual. Incorrect Answers A:Unjustified blame does not fix alert loss. B:Removing records prevents finding the real cause. D:Indiscriminate severity creates noise without solving routing.
Question 689 A vendor outage revealed that the company's backup contact route depended on the same cloud identity tenant as its primary route. What lesson should become a control change? Establish and exercise an independent communications channel with separate authentication dependencies Add more channels on the same tenant and assume redundancy is achieved Remove the incident communications plan because it did not work Require only one trusted employee to memorize every contact Explanation Correct Answer A:A fallback sharing the same failure domain is not resilient. Incorrect Answers B:A common identity dependency can disable all channels. C:Response planning still needs improvement, not deletion. D:Personal memory is not a tested continuity control.
Question 690 A recovery team completed a tabletop exercise but never tested failover of the actual transaction service. What gap should the review report? Declare both kinds of testing equivalent because participants were experienced The exercise validated decision coordination, not technical service recovery; schedule a risk-managed functional test Cancel future exercise testing once a tabletop report is signed Use the tabletop attendee count as proof the recovery path works Explanation Correct Answer B:Tabletops cannot establish failover mechanics or transaction integrity. Incorrect Answers A:Discussion does not exercise the system. C:Functional gaps remain unvalidated. D:Attendance does not measure failover success.