A SysOps team wants to automate recovery of an EC2 instance when the underlying host fails. The recovered instance must retain the same private IP and the same Elastic IP as the original instance. The team also must get an email when recovery begins. Which solution satisfies these requirements?

The correct solution uses a CloudWatch alarm with the StatusCheckFailedSystem metric because this metric indicates issues with the underlying host, such as loss of network connectivity, power, or hardware problems. The EC2 recover action automatically migrates the instance to a healthy host while retaining its private IP address, Elastic IP address, and all instance metadata. Publishing to an SNS topic sends the required email notification. The StatusCheckFailedInstance metric monitors issues within the instance's operating system or applications, not the underlying host, so it wouldn't trigger recovery for host failures. Auto Scaling groups are designed for scaling and high availability across multiple instances or AZs, not for recovering a single instance while preserving its private IP and Elastic IP. While Auto Scaling can launch new instances, it assigns new private IPs by default and doesn't directly re-associate existing Elastic IPs in this scenario.