Amazon AWS Certified Solutions Architect - Professional SAP-C02 Bundle

Choose a pilot-light recovery pattern for persistent essential services solves the related scenario An application has a four-hour RTO but recovery requires core infrastructure ready with no application fleet running; pilot light retains critical foundations while scaling application capacity during recovery.

Select a backup-and-restore recovery pattern with tested restore runbooks would address a different requirement: A noncritical nightly batch service can lose several hours of work and recover the following day. A long RTO and RPO justify inexpensive recovery from stored backups.

This alternative, Design a suitably replicated multi-Region active-active architecture with tested conflict handling, is appropriate when An essential payments service requires near-zero recovery downtime across regional disruptions. Active-active serves traffic in multiple Regions without waiting to create standby infrastructure.

Restore the RDS automated backup to the desired earlier point into a recovered instance would address a different requirement: An RDS workload needs point-in-time recovery after an operator mistakenly drops a table. PITR replays retained transaction logs to restore database state before the deletion.

This alternative, Add read replicas with regional read routing and account for replication lag, is appropriate when A database must serve read-heavy global queries while retaining one primary writable Region. Read replicas offload reads but do not inherently deliver synchronous cross-Region writes.

For the separate situation A multi-Region DynamoDB application needs writable local replicas and global synchronization, Configure DynamoDB global tables and design for eventual cross-Region replication conflict rules could be useful. Global tables replicate multi-active writes asynchronously between configured Regions.

This alternative, Deploy an RDS Multi-AZ database configuration, is appropriate when An RDS PostgreSQL database must survive an Availability Zone failure with automatic recovery and unchanged writer identity. Multi-AZ provides managed synchronous standby and automatic failover in another AZ.

For the separate situation An Aurora application needs disaster recovery in another AWS Region with low replication lag, Configure an Aurora Global Database with deliberate writer failover procedures could be useful. Aurora Global Database supports storage-based cross-Region replication and planned failover.

Restore the RDS automated backup to the desired earlier point into a recovered instance solves the related scenario An RDS workload needs point-in-time recovery after an operator mistakenly drops a table; pITR replays retained transaction logs to restore database state before the deletion.

For the separate situation An Aurora application needs disaster recovery in another AWS Region with low replication lag, Configure an Aurora Global Database with deliberate writer failover procedures could be useful. Aurora Global Database supports storage-based cross-Region replication and planned failover.

Although Restore the RDS automated backup to the desired earlier point into a recovered instance can be valid when An RDS workload needs point-in-time recovery after an operator mistakenly drops a table, it targets a different technical condition. PITR replays retained transaction logs to restore database state before the deletion.

Add read replicas with regional read routing and account for replication lag would address a different requirement: A database must serve read-heavy global queries while retaining one primary writable Region. Read replicas offload reads but do not inherently deliver synchronous cross-Region writes.

Configure DynamoDB global tables and design for eventual cross-Region replication conflict rules solves the related scenario A multi-Region DynamoDB application needs writable local replicas and global synchronization; global tables replicate multi-active writes asynchronously between configured Regions.

Although Deploy an RDS Multi-AZ database configuration can be valid when An RDS PostgreSQL database must survive an Availability Zone failure with automatic recovery and unchanged writer identity, it targets a different technical condition. Multi-AZ provides managed synchronous standby and automatic failover in another AZ.

Add read replicas with regional read routing and account for replication lag would address a different requirement: A database must serve read-heavy global queries while retaining one primary writable Region. Read replicas offload reads but do not inherently deliver synchronous cross-Region writes.

Although Deploy an RDS Multi-AZ database configuration can be valid when An RDS PostgreSQL database must survive an Availability Zone failure with automatic recovery and unchanged writer identity, it targets a different technical condition. Multi-AZ provides managed synchronous standby and automatic failover in another AZ.

Add read replicas with regional read routing and account for replication lag would address a different requirement: A database must serve read-heavy global queries while retaining one primary writable Region. Read replicas offload reads but do not inherently deliver synchronous cross-Region writes.

This alternative, Configure an Aurora Global Database with deliberate writer failover procedures, is appropriate when An Aurora application needs disaster recovery in another AWS Region with low replication lag. Aurora Global Database supports storage-based cross-Region replication and planned failover.

For the separate situation A private application must survive one AZ becoming unavailable without moving stateful sessions to a dead node, Store sessions in a resilient shared service and spread stateless application hosts across AZs could be useful. Externalizing session state permits healthy hosts to serve requests after AZ loss.

Enable the cache engine's supported multi-AZ replica and automatic failover features solves the related scenario A cache cluster loses its primary node and needs fast automatic recovery; replicas and failover promote healthy capacity when the primary fails.

Although Distribute EKS node groups and relevant workloads across multiple AZs with placement constraints can be valid when A regional Kubernetes platform must keep nodes scheduled during an AZ interruption, it targets a different technical condition. Zonal spread and replicas prevent losing every workload replica in one AZ.

Enable the cache engine's supported multi-AZ replica and automatic failover features solves the related scenario A cache cluster loses its primary node and needs fast automatic recovery; replicas and failover promote healthy capacity when the primary fails.

Although Scale consumer capacity from queue metrics using target tracking or step scaling can be valid when A service must scale gradually when sustained queue depth increases rather than only on average CPU, it targets a different technical condition. Queue-driven scaling aligns processing workers with pending demand.

Use an Auto Scaling group spanning AZs behind an Application Load Balancer would address a different requirement: A web application needs to replace failed EC2 hosts while serving traffic through two or more AZs. ASG health replacement and ALB target routing reduce single-instance failures.

Although Scale consumer capacity from queue metrics using target tracking or step scaling can be valid when A service must scale gradually when sustained queue depth increases rather than only on average CPU, it targets a different technical condition. Queue-driven scaling aligns processing workers with pending demand.

Use an Auto Scaling group spanning AZs behind an Application Load Balancer would address a different requirement: A web application needs to replace failed EC2 hosts while serving traffic through two or more AZs. ASG health replacement and ALB target routing reduce single-instance failures.

This alternative, Distribute EKS node groups and relevant workloads across multiple AZs with placement constraints, is appropriate when A regional Kubernetes platform must keep nodes scheduled during an AZ interruption. Zonal spread and replicas prevent losing every workload replica in one AZ.

Although Store sessions in a resilient shared service and spread stateless application hosts across AZs can be valid when A private application must survive one AZ becoming unavailable without moving stateful sessions to a dead node, it targets a different technical condition. Externalizing session state permits healthy hosts to serve requests after AZ loss.

This alternative, Enable the cache engine's supported multi-AZ replica and automatic failover features, is appropriate when A cache cluster loses its primary node and needs fast automatic recovery. Replicas and failover promote healthy capacity when the primary fails.

For the separate situation A service must scale gradually when sustained queue depth increases rather than only on average CPU, Scale consumer capacity from queue metrics using target tracking or step scaling could be useful. Queue-driven scaling aligns processing workers with pending demand.

Store sessions in a resilient shared service and spread stateless application hosts across AZs would address a different requirement: A private application must survive one AZ becoming unavailable without moving stateful sessions to a dead node. Externalizing session state permits healthy hosts to serve requests after AZ loss.

This alternative, Enable the cache engine's supported multi-AZ replica and automatic failover features, is appropriate when A cache cluster loses its primary node and needs fast automatic recovery. Replicas and failover promote healthy capacity when the primary fails.

Use an Auto Scaling group spanning AZs behind an Application Load Balancer solves the related scenario A web application needs to replace failed EC2 hosts while serving traffic through two or more AZs; aSG health replacement and ALB target routing reduce single-instance failures.