Auto scaling groups associated with a load balancer should use ELB health checks

How to fix it

• Interrupts service: Auto Scaling then replaces any instance the load balancer reports unhealthy, so confirm the target group's health check passes before turning this on.

In the console

EC2 → Auto Scaling Groups → the group → Details → Health checks

Turn on Elastic Load Balancing health checks

Selected

Health check grace period

300 seconds, or as long as an instance takes to pass the load balancer's health check

With the AWS CLI

• Overwrites existing settings: The --health-check-type value replaces the group's whole list of types, so write EBS,ELB if Amazon EBS health checks are on as well.

aws autoscaling update-auto-scaling-group --auto-scaling-group-name <group-name> --health-check-type ELB --health-check-grace-period 300 --region <region>

With Terraform

resource "aws_autoscaling_group" "this" {
  name                      = var.asg_name
  min_size                  = var.min_size
  max_size                  = var.max_size
  health_check_type         = "ELB"
  health_check_grace_period = 300

  launch_template {
    id      = aws_launch_template.this.id
    version = "$Default"
  }
}