Cost Management Guide

⭐⭐ Intermediate - Strategies and tools for optimizing AWS costs while maintaining performance and security.

💡 For deployment, see Deployment Guide. 💡 For monitoring, see Monitoring Guide. 💡 For architecture, see Architecture Guide.

Overview

This guide provides detailed cost analysis, optimization strategies, and management tools for the AWS Static Website Infrastructure.

Cost Philosophy

%%{init: {'theme':'default', 'themeVariables': {'fontSize':'16px'}}}%%
graph LR
    accTitle: Cost Management Philosophy and Three Pillars
    accDescr: Three-pillar cost management framework balancing visibility, optimization, and control for sustainable cloud operations. The Visibility pillar implements comprehensive cost monitoring and measurement establishing baseline understanding of spending patterns across services, environments, and resource types, enabling data-driven decision making through AWS Cost Explorer, CloudWatch metrics, and budget reports tracking expenditures in real-time. The Optimization pillar focuses on spending efficiency through right-sizing resources to match actual workload requirements, leveraging AWS free tier maximizing complimentary resources, implementing environment-specific configurations with development using cost-optimized settings and production using performance-optimized resources, and applying cost-saving strategies like S3 lifecycle policies and CloudFront caching. The Control pillar prevents cost overruns through automated budget alerts triggering notifications before limits are reached, service quotas limiting maximum resource provisioning, environment-specific spending caps with development at lower thresholds, and approval workflows for infrastructure changes impacting costs. This framework ensures you know what you spend through visibility, spend efficiently through optimization, and prevent overruns through control. The three pillars work synergistically with visibility informing optimization decisions and control mechanisms preventing budget violations, creating sustainable cost management for long-term operational success.

    A["💰 Cost Management"] --> B["📊 Visibility"]
    A --> C["🎯 Optimization"]
    A --> D["🛡️ Control"]

    B --> E["Know what you spend"]
    C --> F["Spend efficiently"]
    D --> G["Prevent overruns"]

    style A fill:#fff9c4
    style B fill:#c8e6c9
    style C fill:#e1f5fe
    style D fill:#ffccbc

    linkStyle 0 stroke:#333333,stroke-width:2px
    linkStyle 1 stroke:#333333,stroke-width:2px
    linkStyle 2 stroke:#333333,stroke-width:2px
    linkStyle 3 stroke:#333333,stroke-width:2px
    linkStyle 4 stroke:#333333,stroke-width:2px
    linkStyle 5 stroke:#333333,stroke-width:2px

Key Principles:

Visibility First - You can't optimize what you don't measure
Right-Size Resources - Pay for what you need
Leverage Free Tier - Maximize AWS free resources
Automate Controls - Set budgets and alerts
Review Regularly - Monthly cost analysis

Cost Breakdown
Environment Strategies
Service-Specific Optimization
Budget Management
Cost Analysis Tools
Optimization Checklist
Cost Forecasting

Cost Breakdown

Actual Monthly Costs by Environment

Environment	Actual Cost	Target	Services
Development	$1-5/month	$50	S3 only (cost optimized)
Staging	$15-25/month	$75	CloudFront + S3 + WAF
Production	$25-50/month	$200	Full stack with monitoring

Detailed Service Costs

Development Environment

Service Breakdown (Monthly):
├── S3 Storage (50GB)              $1.15
├── S3 GET Requests (1M)           $0.40
├── S3 PUT Requests (10K)          $0.05
├── Data Transfer Out (10GB)       $0.90
├── CloudWatch Logs (5GB)          $0.25
├── CloudWatch Metrics (Basic)     $0.00 (Free Tier)
└── Total                          ~$2.75/month

Cost Optimization:
✅ CloudFront disabled
✅ WAF disabled
✅ Basic monitoring only
✅ Single region

Staging Environment

Service Breakdown (Monthly):
├── S3 Storage (100GB)              $2.30
├── S3 Requests (5M)                $2.00
├── CloudFront Requests (10M)       $8.50
├── CloudFront Data Transfer (50GB) $4.25
├── WAF (1 WebACL + rules)          $5.00
├── CloudWatch (Enhanced)           $3.00
├── Route 53 (1 hosted zone)        $0.50
├── Data Transfer Out               $1.50
└── Total                           ~$27.05/month

Cost Optimization:
✅ CloudFront for CDN
✅ Basic WAF rules
✅ Enhanced monitoring
✅ Cross-region replication

Production Environment

Service Breakdown (Monthly):
├── S3 Storage (200GB)              $4.60
├── S3 Requests (20M)               $8.00
├── CloudFront Requests (50M)       $21.25
├── CloudFront Data Transfer (200GB) $17.00
├── WAF (Enhanced rules)            $10.00
├── CloudWatch (Full monitoring)    $8.00
├── Route 53 (Custom domain)        $1.00
├── SNS (Notifications)             $0.50
├── KMS (Encryption)                $1.00
├── Data Transfer Out               $3.00
└── Total                           ~$74.35/month

Cost Optimization:
✅ CloudFront with custom domain
✅ Enhanced WAF protection
✅ Full monitoring and alerts
✅ Cross-region backup
✅ Automated cost controls

Environment Strategies

Development: Cost Minimization

Philosophy: Bare minimum for functionality

Optimizations:

Disable CloudFront - Direct S3 access ($0 vs $8-10)
No WAF - Dev traffic is internal ($0 vs $5)
Single Region - No replication ($0 vs $2)
Basic Monitoring - Free tier CloudWatch only
S3 Intelligent-Tiering - Auto-move to cheaper storage

Configuration:

# terraform/environments/dev/main.tf
variable "enable_cloudfront" {
  default = false  # Save ~$10/month
}

variable "enable_waf" {
  default = false  # Save ~$5/month
}

variable "enable_cross_region_replication" {
  default = false  # Save ~$2/month
}

variable "cloudwatch_retention_days" {
  default = 7  # Shorter retention, lower costs
}

Expected Savings: ~$15-20/month vs staging

Staging: Balanced Approach

Philosophy: Production-like with some optimizations

Optimizations:

Standard CloudFront - Enable CDN, standard cache
Basic WAF Rules - Core protection only
Cross-Region Replication - For testing DR
14-day Log Retention - Balance compliance and cost
Moderate Traffic Patterns - Test with realistic load

Configuration:

# terraform/environments/staging/main.tf
variable "enable_cloudfront" {
  default = true  # Production parity
}

variable "enable_waf" {
  default = true  # Basic rules
}

variable "waf_rule_complexity" {
  default = "basic"  # vs "enhanced"
}

variable "cloudwatch_retention_days" {
  default = 14  # Balance retention and cost
}

Expected Cost: $15-25/month

Production: Performance & Reliability

Philosophy: Invest in reliability, optimize within constraints

Optimizations:

CloudFront with Custom Domain - Performance priority
Enhanced WAF Rules - Comprehensive protection
Full Monitoring - Detailed metrics and alerts
30-day Log Retention - Compliance requirements
Reserved Capacity - Where available (future)

Configuration:

# terraform/environments/prod/main.tf
variable "enable_cloudfront" {
  default = true
}

variable "enable_waf" {
  default = true
}

variable "waf_rule_complexity" {
  default = "enhanced"
}

variable "cloudwatch_retention_days" {
  default = 30
}

variable "enable_cross_region_replication" {
  default = true
}

Expected Cost: $25-50/month (scales with traffic)

Service-Specific Optimization

S3 Cost Optimization

Storage Classes:

# Check storage class distribution
aws s3api list-objects-v2 \
  --bucket BUCKET_NAME \
  --query 'Contents[].[Key,StorageClass,Size]' \
  --output table

# Enable Intelligent-Tiering
aws s3api put-bucket-intelligent-tiering-configuration \
  --bucket BUCKET_NAME \
  --id intelligent-tiering \
  --intelligent-tiering-configuration file://tiering-config.json

Lifecycle Policies:

{
  "Rules": [
    {
      "Id": "Transition old logs to Glacier",
      "Filter": {
        "Prefix": "logs/"
      },
      "Status": "Enabled",
      "Transitions": [
        {
          "Days": 90,
          "StorageClass": "GLACIER"
        }
      ],
      "Expiration": {
        "Days": 365
      }
    },
    {
      "Id": "Delete incomplete multipart uploads",
      "Status": "Enabled",
      "AbortIncompleteMultipartUpload": {
        "DaysAfterInitiation": 7
      }
    }
  ]
}

Cost Savings:

Intelligent-Tiering: 30-40% on infrequently accessed objects
Lifecycle policies: 70% on old logs (Standard → Glacier)
Multipart cleanup: Avoid storage charges for failed uploads

CloudFront Cost Optimization

Cache Configuration:

# Increase cache TTL to reduce origin requests
cache_behavior {
  default_ttl = 86400  # 24 hours (from 3600)
  max_ttl     = 604800 # 7 days
  min_ttl     = 0

  # Compress objects to reduce data transfer
  compress = true

  # Cache based on query strings selectively
  query_string_cache_keys = ["version", "lang"]
}

Cost Savings:

Longer TTL: 40-60% fewer origin requests
Compression: 50-70% reduction in data transfer
Selective query string caching: More cache hits

Price Class Optimization:

# Use fewer edge locations for dev/staging
price_class = "PriceClass_100"  # US, EU, Israel only
# vs
price_class = "PriceClass_All"  # All locations

Savings: ~20% on CloudFront costs

WAF Cost Optimization

Rule Optimization:

# Basic rules (dev/staging)
waf_rules = [
  "AWSManagedRulesCommonRuleSet",      # $1/month
  "AWSManagedRulesKnownBadInputsRuleSet" # $1/month
]

# Enhanced rules (production)
waf_rules = [
  "AWSManagedRulesCommonRuleSet",
  "AWSManagedRulesKnownBadInputsRuleSet",
  "AWSManagedRulesSQLiRuleSet",        # $1/month
  "AWSManagedRulesLinuxRuleSet",       # $1/month
  "AWSManagedRulesAnonymousIpList"      # $1/month
]

Cost Breakdown:

Web ACL: $5/month
Each rule group: $1/month
Request sampling: Free

Optimization: Use only necessary rule groups per environment

CloudWatch Cost Optimization

Log Retention:

# Reduce retention for non-prod
aws logs put-retention-policy \
  --log-group-name /aws/s3/static-site-dev \
  --retention-in-days 7

# Keep longer for production
aws logs put-retention-policy \
  --log-group-name /aws/s3/static-site-prod \
  --retention-in-days 30

Metric Filters:

# Use metric filters instead of custom metrics
aws logs put-metric-filter \
  --log-group-name /aws/s3/static-site-prod \
  --filter-name ErrorRate \
  --filter-pattern '[time, request_id, operation, key, request_uri, http_status>=400, ...]' \
  --metric-transformations \
    metricName=ErrorCount,metricNamespace=StaticSite,metricValue=1

Cost Savings:

Shorter retention: 50-70% savings on log storage
Metric filters: Free (vs $0.30 per custom metric)

Budget Management

Setting Up Budgets

Create environment-specific budgets:

# Development budget
aws budgets create-budget \
  --account-id ACCOUNT_ID \
  --budget file://dev-budget.json \
  --notifications-with-subscribers file://budget-notifications.json

dev-budget.json:

{
  "BudgetName": "static-site-dev-budget",
  "BudgetLimit": {
    "Amount": "50",
    "Unit": "USD"
  },
  "TimeUnit": "MONTHLY",
  "BudgetType": "COST",
  "CostFilters": {
    "TagKeyValue": ["Environment$dev"]
  }
}

budget-notifications.json:

{
  "Notification": {
    "NotificationType": "ACTUAL",
    "ComparisonOperator": "GREATER_THAN",
    "Threshold": 80,
    "ThresholdType": "PERCENTAGE"
  },
  "Subscribers": [
    {
      "SubscriptionType": "EMAIL",
      "Address": "alerts@example.com"
    }
  ]
}

Budget Alert Levels

Threshold	Action	Notification
80%	Review costs	Email warning
90%	Investigate	Email + Slack
100%	Immediate action	Email + SMS + Slack
110%	Emergency	All channels + on-call
120%	Auto-scale-down	Automated response

Automated Cost Controls

Lambda function for budget enforcement:

# auto-budget-control.py
import boto3

def lambda_handler(event, context):
    budget_threshold = float(event['detail']['threshold'])

    if budget_threshold >= 110:
        # Disable non-essential resources
        cf_client = boto3.client('cloudfront')

        # Disable dev CloudFront (if accidentally enabled)
        cf_client.update_distribution(
            Id=DEV_DISTRIBUTION_ID,
            DistributionConfig={'Enabled': False}
        )

        # Send critical alert
        sns = boto3.client('sns')
        sns.publish(
            TopicArn=ALERT_TOPIC,
            Subject='CRITICAL: Budget exceeded',
            Message=f'Budget at {budget_threshold}%. Auto-disabled non-essential resources.'
        )

Cost Analysis Tools

AWS Cost Explorer

Monthly cost analysis:

# Get last month's costs by service
aws ce get-cost-and-usage \
  --time-period Start=$(date -u -d 'last month' +%Y-%m-01),End=$(date -u +%Y-%m-01) \
  --granularity MONTHLY \
  --metrics BlendedCost \
  --group-by Type=DIMENSION,Key=SERVICE \
  | jq '.ResultsByTime[].Groups[] | {Service: .Keys[0], Cost: .Metrics.BlendedCost.Amount}'

# Compare month-over-month
aws ce get-cost-and-usage \
  --time-period Start=$(date -u -d '2 months ago' +%Y-%m-01),End=$(date -u +%Y-%m-01) \
  --granularity MONTHLY \
  --metrics BlendedCost \
  --group-by Type=DIMENSION,Key=SERVICE

Cost Anomaly Detection

Enable anomaly detection:

aws ce create-anomaly-monitor \
  --anomaly-monitor file://anomaly-monitor.json

aws ce create-anomaly-subscription \
  --anomaly-subscription file://anomaly-subscription.json

anomaly-monitor.json:

{
  "MonitorName": "static-site-anomaly-monitor",
  "MonitorType": "DIMENSIONAL",
  "MonitorDimension": "SERVICE"
}

Cost Allocation Tags

Tag all resources:

# terraform/modules/common/tags.tf
locals {
  common_tags = {
    Project     = "static-site"
    Environment = var.environment
    ManagedBy   = "terraform"
    CostCenter  = "engineering"
    Owner       = "devops"
  }
}

Generate cost reports by tag:

aws ce get-cost-and-usage \
  --time-period Start=$(date -u -d '30 days ago' +%Y-%m-%d),End=$(date -u +%Y-%m-%d) \
  --granularity DAILY \
  --metrics BlendedCost \
  --group-by Type=TAG,Key=Environment

Optimization Checklist

Monthly Review

Quarterly Review

Analyze 3-month cost trends
Review and adjust budgets
Evaluate service pricing changes
Consider Reserved Capacity (if available)
Review architectural decisions
Benchmark against industry standards
Update cost forecasts
Document optimization wins

Annual Review

Full infrastructure cost analysis
ROI calculation for services
Strategic planning for next year
Budget allocation review
Service consolidation opportunities
Multi-year pricing commitments

Cost Forecasting

Predictive Analysis

# 30-day forecast
aws ce get-cost-forecast \
  --time-period Start=$(date -u +%Y-%m-%d),End=$(date -u -d '+30 days' +%Y-%m-%d) \
  --metric BLENDED_COST \
  --granularity DAILY

# 6-month forecast
aws ce get-cost-forecast \
  --time-period Start=$(date -u +%Y-%m-%d),End=$(date -u -d '+6 months' +%Y-%m-%d) \
  --metric BLENDED_COST \
  --granularity MONTHLY

Growth Modeling

Traffic-based cost projection:

# cost-forecast.py
def project_costs(current_monthly, growth_rate_pct, months):
    """
    Project future costs based on growth rate

    current_monthly: Current monthly cost
    growth_rate_pct: Expected monthly growth (e.g., 10 for 10%)
    months: Number of months to project
    """
    costs = [current_monthly]
    for month in range(1, months + 1):
        next_cost = costs[-1] * (1 + growth_rate_pct / 100)
        costs.append(next_cost)
    return costs

# Example: 10% monthly growth
projection = project_costs(
    current_monthly=30,
    growth_rate_pct=10,
    months=12
)

print(f"Year-end projection: ${projection[-1]:.2f}/month")
# Output: Year-end projection: $94.20/month

Best Practices

Cost-Conscious Architecture

Design for efficiency from day one
Use feature flags to control costs per environment
Leverage free tiers where possible
Right-size from the start - don't over-provision
Monitor continuously - catch issues early
Automate cleanup - remove unused resources
Tag everything - enable detailed cost tracking

Team Practices

Cost awareness training - Everyone understands impact
Budget ownership - Clear accountability
Regular reviews - Monthly cost meetings
Celebrate wins - Recognize optimization efforts
Document learnings - Share cost-saving strategies

Emergency Cost Reduction

If you exceed budget:

Immediate actions (< 1 hour):
- Disable dev CloudFront
- Reduce log retention to 1 day
- Disable non-essential CloudWatch alarms
Short-term actions (< 1 day):
- Review and optimize WAF rules
- Increase cache TTLs
- Enable S3 Intelligent-Tiering
Medium-term actions (< 1 week):
- Analyze traffic patterns
- Optimize S3 storage classes
- Review architectural decisions
- Consider multi-year commitments

Additional Resources

Monitoring Guide - Cost monitoring and alerts
Architecture Guide - Cost-optimized design
Feature Flags - Conditional resource creation
Deployment Guide - Environment-specific configs
AWS Cost Management - Official AWS tools

Last Updated: 2025-10-07 Version: 1.0.0

FilesExpand file tree

cost-management.md

Latest commit

History

cost-management.md

File metadata and controls

Cost Management Guide

Overview

Cost Philosophy

Table of Contents

Cost Breakdown

Actual Monthly Costs by Environment

Detailed Service Costs

Development Environment

Staging Environment

Production Environment

Environment Strategies

Development: Cost Minimization

Staging: Balanced Approach

Production: Performance & Reliability

Service-Specific Optimization

S3 Cost Optimization

CloudFront Cost Optimization

WAF Cost Optimization

CloudWatch Cost Optimization

Budget Management

Setting Up Budgets

Budget Alert Levels

Automated Cost Controls

Cost Analysis Tools

AWS Cost Explorer

Cost Anomaly Detection

Cost Allocation Tags

Optimization Checklist

Monthly Review

Quarterly Review

Annual Review

Cost Forecasting

Predictive Analysis

Growth Modeling

Best Practices

Cost-Conscious Architecture

Team Practices

Emergency Cost Reduction

Additional Resources