Scaling policies (target tracking, step, simple) in AWS - Mini Project: Build & Apply

Scaling policies (target tracking, step, simple)

📋 What You'll Learn

💡 Why This Matters

🌍 Real World

Auto Scaling Groups help keep web applications responsive and cost-efficient by automatically adjusting the number of servers based on demand.

💼 Career

Understanding scaling policies is essential for cloud engineers and DevOps professionals to maintain application performance and optimize resource usage.

1

Create the Auto Scaling Group and Launch Configuration

Create a launch configuration called MyAppLaunchConfig with image ID ami-12345678 and instance type t3.micro. Then create an Auto Scaling Group called MyAppASG using this launch configuration with minimum size 1, maximum size 5, and desired capacity 2.

AWS

# Your code here to create launch configuration and Auto Scaling Group

Need a hint?

Use aws_launch_configuration and aws_autoscaling_group resources. Set the exact names and values as specified.

2

Add Target Tracking Scaling Policy

Add a target tracking scaling policy named CPU50Percent to MyAppASG that keeps the average CPU utilization at 50%. Use the predefined metric ASGAverageCPUUtilization.

AWS

resource "aws_launch_configuration" "MyAppLaunchConfig" {
  name          = "MyAppLaunchConfig"
  image_id      = "ami-12345678"
  instance_type = "t3.micro"
}

resource "aws_autoscaling_group" "MyAppASG" {
  name                      = "MyAppASG"
  launch_configuration      = aws_launch_configuration.MyAppLaunchConfig.name
  min_size                  = 1
  max_size                  = 5
  desired_capacity          = 2
  vpc_zone_identifier       = ["subnet-abc12345"]
}

# Add target tracking scaling policy here

Need a hint?

Use aws_autoscaling_policy with policy_type = "TargetTrackingScaling" and set target_value to 50.

3

Add Step Scaling Policy to Increase Capacity

Add a step scaling policy named ScaleUpHighCPU to MyAppASG that increases capacity by 2 when CPU usage exceeds 70%. Use a CloudWatch alarm named HighCPUAlarm that triggers when average CPU is above 70%.

AWS

resource "aws_launch_configuration" "MyAppLaunchConfig" {
  name          = "MyAppLaunchConfig"
  image_id      = "ami-12345678"
  instance_type = "t3.micro"
}

resource "aws_autoscaling_group" "MyAppASG" {
  name                      = "MyAppASG"
  launch_configuration      = aws_launch_configuration.MyAppLaunchConfig.name
  min_size                  = 1
  max_size                  = 5
  desired_capacity          = 2
  vpc_zone_identifier       = ["subnet-abc12345"]
}

resource "aws_autoscaling_policy" "CPU50Percent" {
  name                   = "CPU50Percent"
  autoscaling_group_name = aws_autoscaling_group.MyAppASG.name
  policy_type            = "TargetTrackingScaling"

  target_tracking_configuration {
    predefined_metric_specification {
      predefined_metric_type = "ASGAverageCPUUtilization"
    }
    target_value = 50.0
  }
}

# Add step scaling policy and CloudWatch alarm here

Need a hint?

Create a CloudWatch alarm for CPU > 70%. Then create a step scaling policy linked to this alarm that increases capacity by 2.

4

Add Simple Scaling Policy to Decrease Capacity

Add a simple scaling policy named ScaleDownLowCPU to MyAppASG that decreases capacity by 1 when CPU usage falls below 30%. Use a CloudWatch alarm named LowCPUAlarm that triggers when average CPU is below 30%.

AWS

resource "aws_launch_configuration" "MyAppLaunchConfig" {
  name          = "MyAppLaunchConfig"
  image_id      = "ami-12345678"
  instance_type = "t3.micro"
}

resource "aws_autoscaling_group" "MyAppASG" {
  name                      = "MyAppASG"
  launch_configuration      = aws_launch_configuration.MyAppLaunchConfig.name
  min_size                  = 1
  max_size                  = 5
  desired_capacity          = 2
  vpc_zone_identifier       = ["subnet-abc12345"]
}

resource "aws_autoscaling_policy" "CPU50Percent" {
  name                   = "CPU50Percent"
  autoscaling_group_name = aws_autoscaling_group.MyAppASG.name
  policy_type            = "TargetTrackingScaling"

  target_tracking_configuration {
    predefined_metric_specification {
      predefined_metric_type = "ASGAverageCPUUtilization"
    }
    target_value = 50.0
  }
}

resource "aws_cloudwatch_metric_alarm" "HighCPUAlarm" {
  alarm_name          = "HighCPUAlarm"
  comparison_operator = "GreaterThanThreshold"
  evaluation_periods  = 1
  metric_name         = "CPUUtilization"
  namespace           = "AWS/EC2"
  period              = 300
  statistic           = "Average"
  threshold           = 70
  alarm_description   = "Alarm when CPU exceeds 70%"
  dimensions = {
    AutoScalingGroupName = aws_autoscaling_group.MyAppASG.name
  }
}

resource "aws_autoscaling_policy" "ScaleUpHighCPU" {
  name                   = "ScaleUpHighCPU"
  autoscaling_group_name = aws_autoscaling_group.MyAppASG.name
  policy_type            = "StepScaling"
  adjustment_type        = "ChangeInCapacity"
  cooldown               = 300
  metric_aggregation_type = "Average"

  step_adjustment {
    metric_interval_lower_bound = 0
    scaling_adjustment          = 2
  }

  alarm_names = [aws_cloudwatch_metric_alarm.HighCPUAlarm.alarm_name]
}

# Add simple scaling policy and CloudWatch alarm here

Need a hint?

Create a CloudWatch alarm for CPU < 30%. Then create a simple scaling policy linked to this alarm that decreases capacity by 1.