Question

A company runs a stateful order-processing service on EC2 instances behind an Application Load Balancer. AWS sometimes issues scheduled maintenance events (planned retirement or reboot) for these instances. Operations wants an automated response that, upon receiving an AWS Health scheduled-change notification for a specific instance, gracefully deregisters that instance from the target group, waits for in-flight orders to complete, and then triggers a replacement before the maintenance window begins. Which approach BEST meets these requirements with the least custom infrastructure?

Accepted Answer

Create an EventBridge rule matching AWS Health events with detail-type 'AWS Health Event' and the relevant event type code, targeting a Lambda function that deregisters the affected instance from the target group, polls until connection draining completes, and marks the instance unhealthy so the Auto Scaling group launches a replacement.

Answer

Enable ALB connection draining and rely on the Auto Scaling group's built-in health checks to detect and replace the instance automatically once the maintenance reboot occurs.

Answer

Configure a CloudWatch alarm on the EC2 StatusCheckFailed_System metric that invokes an SSM Automation runbook to stop and start the instance before AWS performs maintenance.

Answer

Create an EventBridge rule for EC2 instance-state-change 'stopping' events that triggers a Lambda function to move the instance out of the target group after the maintenance stop begins.