Question

A company runs a stateless web application on an Amazon EC2 Auto Scaling group behind an Application Load Balancer (ALB) across three Availability Zones. Occasionally, an instance's application process hangs and returns HTTP 500 errors, but the EC2 instance itself continues running and passes the default EC2 status checks. The Auto Scaling group therefore does not replace these unhealthy instances, and users intermittently receive errors. What should the solutions architect do to ensure unhealthy instances are automatically replaced?

Accepted Answer

Configure the Auto Scaling group to use ELB health checks so it replaces instances the ALB target group marks as unhealthy.

Answer

Increase the Auto Scaling group's desired capacity so there are always enough healthy instances to serve traffic.

Answer

Enable ALB access logging and create a CloudWatch alarm that emails the operations team when 500 errors occur.

Answer

Reduce the EC2 status check grace period so the Auto Scaling group replaces instances more aggressively.