Question

A company runs a web application on EC2 instances in an Auto Scaling group behind an Application Load Balancer. Recently, an instance's web server process crashed, returning HTTP 503 errors to users, but the Auto Scaling group did not replace the instance for a long time. The EC2 instance itself remained running and passed all EC2 system and instance status checks. What should the CloudOps engineer configure so that Auto Scaling replaces instances that fail the application-level check?

Accepted Answer

Change the Auto Scaling group's health check type to ELB and ensure the ALB target group health check targets the application's health endpoint.

Answer

Enable EC2 detailed monitoring so that status checks are evaluated every minute instead of every five minutes.

Answer

Increase the Auto Scaling group's health check grace period so instances have more time to recover before replacement.

Answer

Attach a target-tracking scaling policy based on the ALB request count per target metric.