Amazon AWS Certified Solutions Architect - Professional SAP-C02 Bundle

Enable DynamoDB point-in-time recovery and restore to a new table solves the related scenario An organization needs point-in-time rollback of a DynamoDB table after accidental writes; pITR provides continuous recovery within the available retention window.

Perform a supported encrypted snapshot copy specifying an authorized destination KMS key would address a different requirement: An encrypted database snapshot must be copied to a Region with a different customer-managed KMS key. Snapshot copies can reencrypt eligible data under the destination Region key.

This alternative, Configure cross-account snapshot or AWS Backup copy policies and destination key access, is appropriate when An audit-controlled workload needs copies of EBS backups in an independent security account. Separate-account recovery copies reduce exposure to production-account compromise.

Use Route 53 failover or Global Accelerator routing to independently healthy regional endpoints would address a different requirement: A multi-Region frontend must resume requests without hard-coding the primary Region hostname. Health-aware global ingress decouples clients from a single regional origin.

This alternative, Add circuit breaking, bounded timeouts and graceful fallback for external calls, is appropriate when A third-party SaaS dependency sometimes times out but the rest of the service must remain responsive. Dependency isolation limits cascading timeouts during upstream degradation.

Persist the payloads in S3 or another durable managed store instead of instance-local storage solves the related scenario A service writes object payloads to local instance disks that disappear on scaling events; instance-local storage cannot provide durable cross-host state for autoscaling services.

This alternative, Add circuit breaking, bounded timeouts and graceful fallback for external calls, is appropriate when A third-party SaaS dependency sometimes times out but the rest of the service must remain responsive. Dependency isolation limits cascading timeouts during upstream degradation.

Move shared session state to a multi-AZ durable or replicated session store solves the related scenario An online checkout uses in-memory session data on one host and fails when that host is replaced; external session storage prevents local host failure from invalidating user sessions.

Although Deploy zone-local NAT gateways and appropriate private subnet route tables can be valid when An application relies on a single NAT gateway in one AZ for outbound dependencies, it targets a different technical condition. AZ-local NAT removes a shared single-zone egress dependency.

For the separate situation A service writes object payloads to local instance disks that disappear on scaling events, Persist the payloads in S3 or another durable managed store instead of instance-local storage could be useful. Instance-local storage cannot provide durable cross-host state for autoscaling services.

Deploy zone-local NAT gateways and appropriate private subnet route tables solves the related scenario An application relies on a single NAT gateway in one AZ for outbound dependencies; aZ-local NAT removes a shared single-zone egress dependency.

Although Add circuit breaking, bounded timeouts and graceful fallback for external calls can be valid when A third-party SaaS dependency sometimes times out but the rest of the service must remain responsive, it targets a different technical condition. Dependency isolation limits cascading timeouts during upstream degradation.

Persist the payloads in S3 or another durable managed store instead of instance-local storage solves the related scenario A service writes object payloads to local instance disks that disappear on scaling events; instance-local storage cannot provide durable cross-host state for autoscaling services.

Although Use Route 53 failover or Global Accelerator routing to independently healthy regional endpoints can be valid when A multi-Region frontend must resume requests without hard-coding the primary Region hostname, it targets a different technical condition. Health-aware global ingress decouples clients from a single regional origin.

Move shared session state to a multi-AZ durable or replicated session store would address a different requirement: An online checkout uses in-memory session data on one host and fails when that host is replaced. External session storage prevents local host failure from invalidating user sessions.

Although Deploy zone-local NAT gateways and appropriate private subnet route tables can be valid when An application relies on a single NAT gateway in one AZ for outbound dependencies, it targets a different technical condition. AZ-local NAT removes a shared single-zone egress dependency.

Move shared session state to a multi-AZ durable or replicated session store would address a different requirement: An online checkout uses in-memory session data on one host and fails when that host is replaced. External session storage prevents local host failure from invalidating user sessions.

This alternative, Persist the payloads in S3 or another durable managed store instead of instance-local storage, is appropriate when A service writes object payloads to local instance disks that disappear on scaling events. Instance-local storage cannot provide durable cross-host state for autoscaling services.

This alternative, Use synthetic health probes and a rehearsal schedule with controlled traffic, is appropriate when A regional DR procedure must periodically validate DNS routes and endpoints without affecting real users. Proactive probes detect broken standby paths that passive backups cannot reveal.

Orchestrate the failover with Step Functions approval steps and audited transitions solves the related scenario A regional failover requires two operators to approve any data-writer promotion; human approval gates prevent automated promotion from violating recovery governance.

Although Define CloudWatch alarms on application errors and saturation signals with escalations can be valid when A cloud application must alert on elevated end-user error rate before its database exhausts connections, it targets a different technical condition. Leading health and capacity signals trigger response before failure is total.

Configure ECS service health-based replacement and load balancer health checks solves the related scenario An ECS service's health check repeatedly fails and operators currently restart tasks manually; eCS services replace unhealthy tasks to restore desired capacity.

Although Define CloudWatch alarms on application errors and saturation signals with escalations can be valid when A cloud application must alert on elevated end-user error rate before its database exhausts connections, it targets a different technical condition. Leading health and capacity signals trigger response before failure is total.

Match CloudFormation stack state-change events with EventBridge and invoke a runbook would address a different requirement: A failed CloudFormation deployment must notify operations and trigger controlled recovery steps. Event-driven remediation responds to deployment failure states without continuous polling.

Although Use synthetic health probes and a rehearsal schedule with controlled traffic can be valid when A regional DR procedure must periodically validate DNS routes and endpoints without affecting real users, it targets a different technical condition. Proactive probes detect broken standby paths that passive backups cannot reveal.

Match CloudFormation stack state-change events with EventBridge and invoke a runbook would address a different requirement: A failed CloudFormation deployment must notify operations and trigger controlled recovery steps. Event-driven remediation responds to deployment failure states without continuous polling.

This alternative, Configure ECS service health-based replacement and load balancer health checks, is appropriate when An ECS service's health check repeatedly fails and operators currently restart tasks manually. ECS services replace unhealthy tasks to restore desired capacity.

Although Match CloudFormation stack state-change events with EventBridge and invoke a runbook can be valid when A failed CloudFormation deployment must notify operations and trigger controlled recovery steps, it targets a different technical condition. Event-driven remediation responds to deployment failure states without continuous polling.

Orchestrate the failover with Step Functions approval steps and audited transitions would address a different requirement: A regional failover requires two operators to approve any data-writer promotion. Human approval gates prevent automated promotion from violating recovery governance.

This alternative, Define CloudWatch alarms on application errors and saturation signals with escalations, is appropriate when A cloud application must alert on elevated end-user error rate before its database exhausts connections. Leading health and capacity signals trigger response before failure is total.

Configure ECS service health-based replacement and load balancer health checks would address a different requirement: An ECS service's health check repeatedly fails and operators currently restart tasks manually. ECS services replace unhealthy tasks to restore desired capacity.

For the separate situation A regional failover requires two operators to approve any data-writer promotion, Orchestrate the failover with Step Functions approval steps and audited transitions could be useful. Human approval gates prevent automated promotion from violating recovery governance.

Define CloudWatch alarms on application errors and saturation signals with escalations solves the related scenario A cloud application must alert on elevated end-user error rate before its database exhausts connections; leading health and capacity signals trigger response before failure is total.