ALB and ASG Health Check Recovery

In this lab you'll build a small but realistic web tier in your TeKanAid AWS lab account: an Application Load Balancer fronting an Auto Scaling Group that spans two Availability Zones. You'll then break one instance on purpose to prove that ELB health checks and ASG instance replacement work end-to-end.

You will:

• Create a VPC, two public subnets in two AZs, security groups, an ALB, and a target group with an HTTP `/healthz` health check
• Build a launch template (t3.small, Amazon Linux 2023, nginx via user-data) and an Auto Scaling Group attached to the target group
• Configure the ASG with `HealthCheckType=ELB` and `HealthCheckGracePeriod=180` so the ASG trusts the ALB's health check verdict
• Verify all targets reach `healthy` and confirm traffic flows through the ALB DNS name
• Use SSM Session Manager to terminate nginx on one instance, observe the ALB mark it unhealthy, and watch the ASG terminate and replace it automatically

The whole exercise is driven from the AWS CLI on the workstation. Every task is idempotent and produces files in `/root/alb-asg/` that the next task reads.