Skip to main content

AWS Pricing Models and Cost Management

In this lesson, we'll explore how AWS pricing works and learn practical strategies to manage your cloud costs effectively. Understanding AWS pricing is crucial for building cost-efficient applications and avoiding unexpected bills.

Learning Goals:

  • Understand the fundamental AWS pricing models
  • Learn to use AWS pricing calculator and cost management tools
  • Implement cost optimization strategies
  • Set up billing alerts and budgets

AWS Pricing Fundamentals​

AWS operates on a pay-as-you-go model, meaning you only pay for what you use. There are three main pricing models:

On-Demand Instances​

Pay for compute capacity by the hour or second with no long-term commitments.

Calculate EC2 On-Demand Cost
# Example: t3.micro instance in us-east-1
# $0.0104 per hour × 720 hours (30 days) = $7.49 monthly
echo "scale=2; 0.0104 * 720" | bc

Reserved Instances​

Commit to 1 or 3-year terms for significant discounts (up to 75%).

Reserved vs On-Demand Calculation
# Compare 1-year reserved vs on-demand pricing
on_demand_hourly = 0.10
reserved_hourly = 0.05
hours_per_year = 8760

savings = (on_demand_hourly - reserved_hourly) * hours_per_year
print(f"Annual savings: ${savings:.2f}")

Spot Instances​

Bid for unused EC2 capacity at up to 90% discount, but instances can be terminated with short notice.

AWS Cost Management Tools​

AWS Pricing Calculator​

tip

Always use the AWS Pricing Calculator before launching new services. It provides accurate cost estimates based on your specific configuration and region.

AWS Cost Explorer​

Visualize and analyze your AWS spending patterns with interactive charts and reports.

Cost Explorer API Query Example
{
  "TimePeriod": {
    "Start": "2024-01-01",
    "End": "2024-01-31"
  },
  "Granularity": "MONTHLY",
  "Metrics": ["BlendedCost", "UsageQuantity"],
  "GroupBy": [
    {
      "Type": "DIMENSION",
      "Key": "SERVICE"
    }
  ]
}

Cost Optimization Strategies​

Right-Sizing Resources​

Right-Sizing Analysis
def analyze_instance_utilization(cpu_utilization, memory_utilization):
    """Determine if instance needs resizing"""
    if cpu_utilization < 20 and memory_utilization < 30:
        return "DOWNGRADE - Underutilized"
    elif cpu_utilization > 80 or memory_utilization > 85:
        return "UPGRADE - Overutilized"
    else:
        return "OPTIMAL - Well sized"

# Example usage
print(analyze_instance_utilization(15, 25))  # DOWNGRADE - Underutilized
print(analyze_instance_utilization(90, 60))  # UPGRADE - Overutilized

Implementing Auto Scaling​

Create Auto Scaling Group
aws autoscaling create-auto-scaling-group \
    --auto-scaling-group-name my-asg \
    --launch-configuration-name my-launch-config \
    --min-size 2 \
    --max-size 10 \
    --desired-capacity 2 \
    --availability-zones us-east-1a us-east-1b

Setting Up Billing Alerts​

CloudWatch Billing Alarm​

Create Billing Alarm with Boto3
import boto3

def create_billing_alarm(alarm_name, threshold_amount, email_address):
    cloudwatch = boto3.client('cloudwatch')
    sns = boto3.client('sns')
    
    # Create SNS topic for notifications
    topic_arn = sns.create_topic(Name='BillingAlerts')['TopicArn']
    sns.subscribe(
        TopicArn=topic_arn,
        Protocol='email',
        Endpoint=email_address
    )
    
    # Create billing alarm
    cloudwatch.put_metric_alarm(
        AlarmName=alarm_name,
        AlarmDescription='Monthly billing alarm',
        MetricName='EstimatedCharges',
        Namespace='AWS/Billing',
        Statistic='Maximum',
        Dimensions=[
            {
                'Name': 'Currency',
                'Value': 'USD'
            }
        ],
        Period=21600,  # 6 hours
        EvaluationPeriods=1,
        Threshold=threshold_amount,
        ComparisonOperator='GreaterThanThreshold',
        AlarmActions=[topic_arn]
    )

# Usage
create_billing_alarm('MonthlyBudget', 100.0, 'admin@example.com')

AWS Budgets​

warning

Billing alarms have a delay of up to 24 hours. For real-time budget tracking, use AWS Budgets instead.

Storage Cost Optimization​

S3 Storage Classes​

S3 Storage Class Cost Comparison
def calculate_s3_costs(storage_gb, requests_per_month):
    """Compare S3 storage class costs"""
    standard_cost = (storage_gb * 0.023) + (requests_per_month * 0.0004)
    infrequent_cost = (storage_gb * 0.0125) + (requests_per_month * 0.001)
    glacier_cost = (storage_gb * 0.004) + (requests_per_month * 0.05)
    
    return {
        'STANDARD': round(standard_cost, 2),
        'STANDARD_IA': round(infrequent_cost, 2),
        'GLACIER': round(glacier_cost, 2)
    }

# Example: 100GB storage, 1000 requests
costs = calculate_s3_costs(100, 1000)
print(f"Monthly costs: {costs}")

Common Pitfalls​

  • Forgotten Resources: Leaving EC2 instances running overnight or unused EBS volumes
  • Over-Provisioning: Using larger instance types than needed "just to be safe"
  • Ignoring Data Transfer Costs: Not accounting for cross-region or internet data transfer fees
  • Manual Scaling: Not leveraging Auto Scaling for variable workloads
  • No Monitoring: Failing to set up billing alerts until receiving a large bill
  • Wrong Storage Classes: Using Standard S3 for archival data instead of Glacier

Summary​

AWS pricing follows a pay-as-you-go model with options for cost savings through Reserved and Spot Instances. Effective cost management requires using tools like Cost Explorer, setting up budgets and alerts, right-sizing resources, and choosing appropriate storage classes. Regular cost reviews and automated scaling help maintain cost efficiency as your applications evolve.

Quiz​

AWS Cost Optimization & Billing Fundamentals

What is the main advantage of Reserved Instances over On-Demand Instances?

Question 1/5