Data Transfer Costs — Egress Fees, Multi-Region Costs, NAT Gateway, Direct Connect, CloudFront

# NAT Gateway pricing (per hour + per GB processed)
# $0.045/hour + $0.045/GB processed
# A NAT Gateway processing 10TB/month = $32.40 + $450 = $482.40

# Alternatives for cost savings:
# 1. Use NAT instances (EC2-based, cheaper but less reliable)
# 2. Use VPC endpoints for AWS services
# 3. Use Gateway VPC endpoints (S3, DynamoDB) — free
# 4. Use PrivateLink for specific services

def estimate_multi_region_cost(data_size_gb, regions):
    """Estimate monthly data transfer cost across regions"""

    # Inter-region transfer rates (us-east-1 ↔ eu-west-1)
    INTER_REGION_RATE = 0.02  # $/GB

    cost = data_size_gb * INTER_REGION_RATE * len(regions) * 2  # bidirectional
    return cost

# Example: 10TB data processed across 3 regions
monthly_cost = estimate_multi_region_cost(10000, ['us-east-1', 'eu-west-1', 'ap-southeast-1'])
print(f"Monthly inter-region transfer cost: ${monthly_cost:,.2f}")
# Monthly inter-region transfer cost: $600.00

# CloudFront pricing
# US/Europe: $0.085/GB first 10TB
# Asia: $0.14/GB
# South America: $0.25/GB

# Enable CloudFront for:
# 1. Static assets (images, CSS, JS)
# 2. Origin shield to reduce origin calls
# 3. Price class selection (US+Europe only vs all)

# CloudFront Origin Shield
# Adds a caching layer before origin, reducing origin load by 60-80%
aws cloudfront update-distribution \
  --id DISTRIBUTION_ID \
  --origin-shield-enabled \
  --origin-shield-origin-shield-region us-east-1

# Azure egress pricing (from US regions)
# To internet: $0.087/GB first 10TB
# Between regions: $0.01-0.05/GB
# Within region (VNet peering): $0.01/GB both directions

# Use Azure CDN for public content delivery
az cdn profile create \
  --name my-cdn \
  --resource-group my-rg \
  --sku Standard_Microsoft

# Use Azure Front Door for global routing with caching
az afd profile create \
  --profile-name my-afd \
  --resource-group my-rg \
  --sku Premium_AzureFrontDoor

# GCP egress pricing
# To internet: $0.12/GB first 10TB (premium tier)
# To internet: $0.08/GB (standard tier, lower SLA)
# Between regions: $0.01-0.08/GB

# Use Cloud CDN to reduce egress
gcp compute backend-services create my-backend \
  --global \
  --enable-cdn \
  --cdn-policy '{"cacheMode":"CACHE_ALL_STATIC","defaultTtl":86400}'

# Use Cloud NAT with static IPs for predictable egress
gcp compute routers nats create cloud-nat \
  --router=my-router \
  --region=us-central1 \
  --nat-external-ip-pool=auto \
  --nat-all-subnet-ip-ranges

# Calculate NAT Gateway vs VPC Endpoint cost
nat_cost_per_hour = 0.045
nat_cost_per_gb = 0.045
gw_endpoint_cost_per_hour = 0.01
gw_endpoint_cost_per_gb = 0.01  # Often $0 for S3/DynamoDB

data_processed_gb = 10000  # 10 TB

nat_total = (nat_cost_per_hour * 730) + (nat_cost_per_gb * data_processed_gb)
gw_total = (gw_endpoint_cost_per_hour * 730) + (gw_endpoint_cost_per_gb * data_processed_gb)

print(f"NAT Gateway cost: ${nat_total:,.2f}/month")
print(f"VPC Endpoint cost: ${gw_total:,.2f}/month")
print(f"Savings: ${nat_total - gw_total:,.2f}/month ({(1-gw_total/nat_total)*100:.0f}%)")

# Output:
# NAT Gateway cost: $482.85/month
# VPC Endpoint cost: $107.30/month
# Savings: $375.55/month (78%)

#!/bin/bash
# transfer_cost_report.sh — Analyze data transfer costs

echo "=== Data Transfer Cost Analysis ==="
echo "Date: $(date)"
echo ""

# 1. Analyze NAT Gateway costs
echo "--- NAT Gateways ---"
for nat in $(aws ec2 describe-nat-gateways --query 'NatGateways[].NatGatewayId' --output text); do
    # Get processed bytes
    bytes=$(aws cloudwatch get-metric-statistics \
      --namespace AWS/NATGateway \
      --metric-name BytesOutToDestination \
      --dimensions Name=NatGatewayId,Value=$nat \
      --start-time $(date -d '30 days ago' +%Y-%m-%dT%H:%M:%SZ) \
      --end-time $(date +%Y-%m-%dT%H:%M:%SZ) \
      --period 2592000 \
      --statistics Sum \
      --output json | jq -r '.Datapoints[0].Sum // 0')

    gb=$(echo "scale=2; $bytes / 1073741824" | bc)
    nat_cost=$(echo "scale=2; ($nat_cost_per_hour * 730) + (0.045 * $gb)" | bc 2>/dev/null || echo "N/A")

    echo "NAT Gateway: $nat"
    echo "  Data processed: ${gb} GB/month"
    echo "  Estimated cost: \$${nat_cost:-N/A}/month"
done

# 2. Analyze VPC Peering costs
echo ""
echo "--- VPC Peerings ---"
for peering in $(aws ec2 describe-vpc-peering-connections \
    --query 'VpcPeeringConnections[].VpcPeeringConnectionId' --output text); do
    status=$(aws ec2 describe-vpc-peering-connections \
      --vpc-peering-connection-ids $peering \
      --query 'VpcPeeringConnections[0].Status.Code' --output text)
    echo "Peering: $peering (Status: $status)"
done

# 3. Analyze CloudFront data transfer
echo ""
echo "--- CloudFront Distributions ---"
for dist in $(aws cloudfront list-distributions \
    --query 'DistributionList.Items[].Id' --output text); do
    aws cloudfront get-distribution --id $dist \
      --query 'Distribution.DomainName' --output text
    # Get total bytes transferred
    bytes=$(aws cloudwatch get-metric-statistics \
      --namespace AWS/CloudFront \
      --metric-name BytesDownloaded \
      --dimensions Name=DistributionId,Value=$dist,Name=Region,Value=Global \
      --start-time $(date -d '30 days ago' +%Y-%m-%dT%H:%M:%SZ) \
      --end-time $(date +%Y-%m-%dT%H:%M:%SZ) \
      --period 2592000 \
      --statistics Sum \
      --output json | jq -r '.Datapoints[0].Sum // 0' 2>/dev/null)

    cdn_gb=$(echo "scale=2; $bytes / 1073741824" | bc)
    echo "  Data delivered: ${cdn_gb} GB/month"
done

echo ""
echo "=== Recommendations ==="
echo "1. Check if S3 VPC Gateway endpoints can replace NAT"
echo "2. Verify CloudFront Origin Shield is enabled"
echo "3. Confirm inter-AZ traffic uses private IPs"