An application uses a MySQL-compatible Amazon Aurora Multi-AZ DB cluster with a cross-Region read replica for disaster recovery. The DevOps engineer needs to automate promotion of the replica to primary if the primary fails. Which solution will accomplish this?

The correct solution leverages AWS Systems Manager Parameter Store to store the Aurora endpoint, providing a centralized and dynamic way to manage the database connection string. An Amazon EventBridge event can detect database failures (e.g., via CloudWatch alarms on RDS metrics or CloudTrail events) and trigger an AWS Lambda function. This Lambda function automates the promotion of the cross-Region read replica to a standalone primary instance and updates the endpoint in Parameter Store. The application is designed to gracefully handle connection failures by reloading the endpoint from Parameter Store, ensuring high availability and automated disaster recovery. Incorrect options: Latency-based Route 53 CNAMEs are not suitable for cross-Region Aurora replica promotion as they don't inherently handle the promotion logic or the change in the primary endpoint. Aurora custom endpoints are for specific instances within a cluster, not for cross-Region failover and promotion of a replica to a new primary. Modifying CloudFormation templates dynamically for failover is overly complex and not an efficient or recommended pattern for real-time disaster recovery.