A company runs an Amazon Aurora MySQL DB cluster with six Aurora Replicas. They want near-real-time reporting queries to be automatically distributed across three specific Aurora Replicas that use different compute and memory specs than the other cluster nodes. Which solution meets these requirements?

Creating a custom endpoint allows you to group specific Aurora Replicas and direct read traffic to them. This is ideal for isolating reporting workloads to replicas with different compute and memory specifications. The custom endpoint will automatically distribute queries among the three designated replicas. Creating a cluster clone would involve unnecessary data duplication and management overhead. Using individual instance endpoints would require manual management of connections and load balancing, defeating the purpose of automatic distribution. The default reader endpoint distributes traffic across all Aurora Replicas, not just a specific subset, which wouldn't meet the requirement of using only three specific replicas.