A company uses Amazon Aurora for its global ecommerce application. Developers report poor application performance when monthly reports run. CloudWatch metrics show spikes in ReadIOPS and CPUUtilization during the reports. What is the MOST cost-effective solution?

Migrating the monthly reporting to an Aurora Replica is the most cost-effective solution because Aurora Replicas share the same underlying storage volume as the primary instance. This allows them to offload read-heavy workloads, like reporting, without impacting the performance of the primary database used by the ecommerce application. Since Aurora Replicas are read-only, they prevent the reporting queries from consuming resources on the primary instance, thus resolving the ReadIOPS and CPU utilization spikes. Migrating to Amazon Redshift would be more expensive and complex, as it's a separate data warehousing service requiring data transfer and management. Migrating to a larger Aurora instance class would increase costs for the entire database, even during non-reporting periods, and might not fully isolate the reporting workload. Increasing Provisioned IOPS would also incur higher costs and primarily addresses I/O bottlenecks, but not necessarily CPU contention from complex reporting queries.