Question

A SysOps administrator manages an Application Load Balancer (ALB) that routes traffic to an Auto Scaling group of EC2 instances running a web application on port 8080. Users report intermittent HTTP 502 errors. The instances all pass EC2 status checks and show as 'InService' in the Auto Scaling group, but the ALB target group reports several targets as 'unhealthy'. The application process on the affected instances has crashed, though the operating system remains running. What should the administrator do to ensure unhealthy instances are automatically replaced?

Accepted Answer

Change the Auto Scaling group health check type to 'ELB' so it uses the target group health check results to replace failing instances

Answer

Enable EC2 detailed monitoring so the Auto Scaling group detects the application crash faster and replaces the instances

Answer

Configure the ALB listener to use TCP health checks instead of HTTP so the crashed application still passes health checks

Answer

Increase the target group healthy threshold count so the ALB waits longer before marking instances unhealthy