AWS Global Accelerator¶

Service Overview and Purpose¶

AWS Global Accelerator is a networking service that improves the performance of your users' traffic by up to 60% using Amazon Web Services' global network infrastructure. When the internet is congested, Global Accelerator's automatic routing optimizations will help keep your packet loss, jitter, and latency consistently low.

Key Characteristics¶

Global Network: Leverages AWS global infrastructure
Anycast IP: Static IP addresses for global access
Health Monitoring: Automatic health checks and failover
Traffic Optimization: Intelligent routing to optimal endpoints
DDoS Protection: Built-in DDoS protection with AWS Shield

Key Features and Capabilities¶

Core Components¶

Accelerator: The main Global Accelerator resource
Listeners: Define ports and protocols for traffic routing
Endpoint Groups: Group endpoints by geographic region
Endpoints: Application Load Balancers, Network Load Balancers, EC2 instances, or Elastic IPs

Traffic Control¶

Traffic Dials: Control percentage of traffic to endpoint groups
Weights: Distribute traffic across endpoints within a group
Client Affinity: Route requests from same client to same endpoint
Health Checks: Monitor endpoint health and route accordingly

Configuration Examples¶

Basic Global Accelerator Setup¶

# Create Global Accelerator
aws globalaccelerator create-accelerator \
  --name "Production-Global-Accelerator" \
  --ip-address-type IPV4 \
  --enabled \
  --tags Key=Environment,Value=Production Key=Application,Value=WebApp

# Create listener
aws globalaccelerator create-listener \
  --accelerator-arn arn:aws:globalaccelerator::account:accelerator/12345678-1234-1234-1234-123456789012 \
  --listener-configuration '{
    "Protocol": "TCP",
    "PortRanges": [{"FromPort": 80, "ToPort": 80}]
  }'

# Create endpoint group
aws globalaccelerator create-endpoint-group \
  --listener-arn arn:aws:globalaccelerator::account:listener/12345678-1234-1234-1234-123456789012 \
  --endpoint-group-region us-west-2 \
  --endpoint-configurations '{
    "EndpointId": "arn:aws:elasticloadbalancing:us-west-2:account:loadbalancer/app/my-alb/1234567890123456",
    "Weight": 100,
    "ClientIPPreservationEnabled": true
  }' \
  --traffic-dial-percentage 100 \
  --health-check-interval-seconds 30 \
  --health-check-protocol HTTP \
  --health-check-path "/health" \
  --threshold-count 3

Multi-Region Setup¶

# Create endpoint groups in multiple regions
# US West endpoint group
aws globalaccelerator create-endpoint-group \
  --listener-arn arn:aws:globalaccelerator::account:listener/12345678-1234-1234-1234-123456789012 \
  --endpoint-group-region us-west-2 \
  --endpoint-configurations '{
    "EndpointId": "arn:aws:elasticloadbalancing:us-west-2:account:loadbalancer/app/west-alb/1234567890123456",
    "Weight": 100
  }' \
  --traffic-dial-percentage 100

# US East endpoint group
aws globalaccelerator create-endpoint-group \
  --listener-arn arn:aws:globalaccelerator::account:listener/12345678-1234-1234-1234-123456789012 \
  --endpoint-group-region us-east-1 \
  --endpoint-configurations '{
    "EndpointId": "arn:aws:elasticloadbalancing:us-east-1:account:loadbalancer/app/east-alb/1234567890123456",
    "Weight": 100
  }' \
  --traffic-dial-percentage 100

# Europe endpoint group
aws globalaccelerator create-endpoint-group \
  --listener-arn arn:aws:globalaccelerator::account:listener/12345678-1234-1234-1234-123456789012 \
  --endpoint-group-region eu-west-1 \
  --endpoint-configurations '{
    "EndpointId": "arn:aws:elasticloadbalancing:eu-west-1:account:loadbalancer/app/europe-alb/1234567890123456",
    "Weight": 100
  }' \
  --traffic-dial-percentage 100

Advanced Configuration¶

# Create accelerator with multiple listeners
# HTTP listener
aws globalaccelerator create-listener \
  --accelerator-arn arn:aws:globalaccelerator::account:accelerator/12345678-1234-1234-1234-123456789012 \
  --listener-configuration '{
    "Protocol": "TCP",
    "PortRanges": [{"FromPort": 80, "ToPort": 80}],
    "ClientAffinity": "SOURCE_IP"
  }'

# HTTPS listener
aws globalaccelerator create-listener \
  --accelerator-arn arn:aws:globalaccelerator::account:accelerator/12345678-1234-1234-1234-123456789012 \
  --listener-configuration '{
    "Protocol": "TCP",
    "PortRanges": [{"FromPort": 443, "ToPort": 443}],
    "ClientAffinity": "SOURCE_IP"
  }'

# UDP listener for gaming
aws globalaccelerator create-listener \
  --accelerator-arn arn:aws:globalaccelerator::account:accelerator/12345678-1234-1234-1234-123456789012 \
  --listener-configuration '{
    "Protocol": "UDP",
    "PortRanges": [{"FromPort": 3000, "ToPort": 3010}],
    "ClientAffinity": "SOURCE_IP"
  }'

Use Cases and Scenarios¶

Primary Use Cases¶

Global Applications: Improve performance for worldwide users
Gaming: Reduce latency for real-time gaming applications
IoT: Optimize connectivity for IoT devices
Media Streaming: Improve streaming quality and reduce buffering
VoIP: Enhance voice and video communication quality

Architecture Patterns¶

Global Web Application¶

# Multi-region web application with Global Accelerator
# Frontend: Global Accelerator -> ALB -> EC2/Fargate
# Backend: Regional services in each region

# Create accelerator for web traffic
aws globalaccelerator create-accelerator \
  --name "Global-Web-App" \
  --ip-address-type IPV4 \
  --enabled

# Add ALBs from multiple regions as endpoints
# Traffic automatically routed to closest healthy region

Blue-Green Deployment¶

# Use traffic dials for blue-green deployments
# Green environment (new version)
aws globalaccelerator update-endpoint-group \
  --endpoint-group-arn arn:aws:globalaccelerator::account:endpoint-group/12345678-1234-1234-1234-123456789012 \
  --traffic-dial-percentage 10  # Start with 10% traffic

# Gradually increase traffic to green environment
aws globalaccelerator update-endpoint-group \
  --endpoint-group-arn arn:aws:globalaccelerator::account:endpoint-group/12345678-1234-1234-1234-123456789012 \
  --traffic-dial-percentage 50  # 50% traffic

# Full cutover to green
aws globalaccelerator update-endpoint-group \
  --endpoint-group-arn arn:aws:globalaccelerator::account:endpoint-group/12345678-1234-1234-1234-123456789012 \
  --traffic-dial-percentage 100  # 100% traffic

Disaster Recovery¶

# Primary region with 100% traffic
aws globalaccelerator create-endpoint-group \
  --listener-arn arn:aws:globalaccelerator::account:listener/12345678-1234-1234-1234-123456789012 \
  --endpoint-group-region us-west-2 \
  --traffic-dial-percentage 100

# DR region with 0% traffic (standby)
aws globalaccelerator create-endpoint-group \
  --listener-arn arn:aws:globalaccelerator::account:listener/12345678-1234-1234-1234-123456789012 \
  --endpoint-group-region us-east-1 \
  --traffic-dial-percentage 0

# In case of disaster, shift traffic to DR region
aws globalaccelerator update-endpoint-group \
  --endpoint-group-arn arn:aws:globalaccelerator::account:endpoint-group/dr-12345678 \
  --traffic-dial-percentage 100

Client IP Preservation¶

Client IP Configuration¶

# Enable client IP preservation (ALB/NLB endpoints)
aws globalaccelerator update-endpoint-group \
  --endpoint-group-arn arn:aws:globalaccelerator::account:endpoint-group/12345678-1234-1234-1234-123456789012 \
  --endpoint-configurations '{
    "EndpointId": "arn:aws:elasticloadbalancing:us-west-2:account:loadbalancer/app/my-alb/1234567890123456",
    "Weight": 100,
    "ClientIPPreservationEnabled": true
  }'

# For EC2 endpoints, client IP is always preserved
# Original client IP available in X-Forwarded-For header for ALB

Application Integration¶

# Python Flask application to handle client IP
from flask import Flask, request
import json

app = Flask(__name__)

@app.route('/api/user-info')
def get_user_info():
    # Get client IP from various sources
    client_ip = request.headers.get('X-Forwarded-For', request.remote_addr)

    # Global Accelerator adds custom headers
    ga_edge_location = request.headers.get('X-Amzn-Ga-Edge-Location')
    ga_request_id = request.headers.get('X-Amzn-Ga-Request-Id')

    return json.dumps({
        'client_ip': client_ip,
        'edge_location': ga_edge_location,
        'request_id': ga_request_id,
        'user_agent': request.headers.get('User-Agent')
    })

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=80)

Monitoring and Troubleshooting¶

CloudWatch Metrics¶

# Monitor Global Accelerator performance
aws cloudwatch get-metric-statistics \
  --namespace AWS/GlobalAccelerator \
  --metric-name NewFlowCount \
  --dimensions Name=Accelerator,Value=12345678-1234-1234-1234-123456789012 \
  --start-time 2023-01-01T00:00:00Z \
  --end-time 2023-01-01T23:59:59Z \
  --period 300 \
  --statistics Sum

# Monitor processed bytes
aws cloudwatch get-metric-statistics \
  --namespace AWS/GlobalAccelerator \
  --metric-name ProcessedBytesIn \
  --dimensions Name=Accelerator,Value=12345678-1234-1234-1234-123456789012 \
  --start-time 2023-01-01T00:00:00Z \
  --end-time 2023-01-01T23:59:59Z \
  --period 300 \
  --statistics Sum

# Monitor endpoint health
aws cloudwatch get-metric-statistics \
  --namespace AWS/GlobalAccelerator \
  --metric-name HealthyEndpointCount \
  --dimensions Name=EndpointGroup,Value=12345678-1234-1234-1234-123456789012 \
  --start-time 2023-01-01T00:00:00Z \
  --end-time 2023-01-01T23:59:59Z \
  --period 300 \
  --statistics Average

Flow Logs¶

# Enable Flow Logs for Global Accelerator
aws globalaccelerator update-accelerator \
  --accelerator-arn arn:aws:globalaccelerator::account:accelerator/12345678-1234-1234-1234-123456789012 \
  --flow-logs-enabled \
  --flow-logs-s3-bucket my-ga-flow-logs \
  --flow-logs-s3-prefix production/

# Analyze flow logs with Athena
CREATE EXTERNAL TABLE ga_flow_logs (
  version string,
  account_id string,
  accelerator_id string,
  client_ip string,
  client_port int,
  accelerator_ip string,
  accelerator_port int,
  endpoint_ip string,
  endpoint_port int,
  protocol string,
  ip_address_type string,
  packets bigint,
  bytes bigint,
  start_time bigint,
  end_time bigint,
  action string
)
PARTITIONED BY(year string, month string, day string)
STORED AS INPUTFORMAT 'org.apache.hadoop.mapred.TextInputFormat'
OUTPUTFORMAT 'org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat'
LOCATION 's3://my-ga-flow-logs/production/';

Health Check Monitoring¶

# Check endpoint health status
aws globalaccelerator describe-endpoint-group \
  --endpoint-group-arn arn:aws:globalaccelerator::account:endpoint-group/12345678-1234-1234-1234-123456789012 \
  --query 'EndpointGroup.EndpointDescriptions[*].[EndpointId,HealthState,HealthReason]'

# Monitor health check failures
aws cloudwatch put-metric-alarm \
  --alarm-name "GA-Unhealthy-Endpoints" \
  --alarm-description "Global Accelerator unhealthy endpoints" \
  --metric-name HealthyEndpointCount \
  --namespace AWS/GlobalAccelerator \
  --statistic Average \
  --period 300 \
  --threshold 1 \
  --comparison-operator LessThanThreshold \
  --dimensions Name=EndpointGroup,Value=12345678-1234-1234-1234-123456789012

Security Considerations¶

DDoS Protection¶

AWS Shield Standard: Automatic protection included
AWS Shield Advanced: Enhanced protection (optional)
Rate Limiting: Implement at application level
WAF Integration: Use with ALB endpoints

Access Control¶

{
  "Version": "2012-10-17",
  "Statement": [{
    "Effect": "Allow",
    "Action": [
      "globalaccelerator:Describe*",
      "globalaccelerator:List*"
    ],
    "Resource": "*"
  }, {
    "Effect": "Allow",
    "Action": [
      "globalaccelerator:UpdateAccelerator",
      "globalaccelerator:UpdateEndpointGroup"
    ],
    "Resource": "arn:aws:globalaccelerator::account:accelerator/12345678-*"
  }]
}

Network Security¶

# Security group rules for ALB endpoints
aws ec2 authorize-security-group-ingress \
  --group-id sg-alb123 \
  --protocol tcp \
  --port 80 \
  --source-group sg-ga-prefix-list

# Use Global Accelerator IP ranges for security groups
# Download AWS IP ranges and filter for Global Accelerator
curl -s https://ip-ranges.amazonaws.com/ip-ranges.json | \
  jq -r '.prefixes[] | select(.service=="GLOBALACCELERATOR") | .ip_prefix'

Cost Optimization¶

Pricing Components¶

Fixed Fee: $0.025 per hour per accelerator
Data Transfer Premium: $0.015 per GB for first 10TB/month
Dominion-Specific: Additional costs for some regions

Cost Optimization Strategies¶

# Monitor data transfer costs
aws cloudwatch get-metric-statistics \
  --namespace AWS/GlobalAccelerator \
  --metric-name ProcessedBytesOut \
  --dimensions Name=Accelerator,Value=12345678-1234-1234-1234-123456789012 \
  --start-time 2023-01-01T00:00:00Z \
  --end-time 2023-01-31T23:59:59Z \
  --period 86400 \
  --statistics Sum

# Optimize by:
# 1. Using CloudFront for static content
# 2. Implementing efficient caching
# 3. Compressing data at application level
# 4. Regional content optimization

Exam-Specific Tips¶

Key Exam Topics¶

Global Performance: How Global Accelerator improves performance
Anycast IPs: Static IP addresses for global access
Traffic Control: Traffic dials and endpoint weights
Health Checks: Automatic failover and health monitoring
Client IP Preservation: When and how it works

Common Exam Scenarios¶

Global Application: Improve performance for worldwide users
Gaming Application: Reduce latency for real-time applications
Disaster Recovery: Use traffic dials for failover
Blue-Green Deployment: Gradual traffic shifting
Multi-Region Setup: Distribute traffic across regions

Exam Tips¶

Know the difference between Global Accelerator and CloudFront
Understand anycast IPs and their benefits
Remember traffic dial functionality for deployments
Know client IP preservation capabilities and limitations
Understand health check behavior and failover

Global Accelerator vs CloudFront¶

Feature	Global Accelerator	CloudFront
Use Case	Dynamic content, TCP/UDP	Static content, HTTP/HTTPS
Protocol	TCP, UDP	HTTP/HTTPS
Caching	No caching	Edge caching
IP Addresses	Static anycast IPs	Dynamic edge IPs
Client IP	Preserved	Via headers