Leveraging the AWS Global Infrastructure

AWS Global Infrastructure

Why Build a Global Application?

1. Reduced Latency

   • Latency = Time for a network packet to reach a server.
   • Deploying in multiple AWS Regions or Edge Locations reduces lag by placing servers closer to users.
   • Example: Users in India get faster response if an Asia region hosts the app instead of only US.

2. Disaster Recovery (DR)

   • Do not rely on a single Region.
   • In case of disasters (earthquake, storm, power failure, politics), failover to another Region ensures high availability.

3. Security & Attack Resistance

   • Spreading applications across multiple Regions makes it harder for attackers to take everything down at once.

AWS Global Infrastructure Components

1. Regions

   • Geographic areas where AWS has data centers.
   • Each Region contains multiple Availability Zones (AZs).

2. Availability Zones (AZs)

   • Multiple isolated data centers within a Region.
   • Designed to be physically separate but connected with high-speed private networks.
   • Provide resilience: if one AZ fails, others remain operational.
   • Example: Northern Virginia has 6 AZs, Paris has 3 AZs.

3. Edge Locations (Points of Presence)

   • Locations closer to end-users for content delivery.
   • Cannot host apps directly but support Amazon CloudFront (CDN) and caching.
   • Improve speed for users even outside AWS Regions.

4. AWS Global Network

   • Private AWS backbone connecting Regions, AZs, and Edge Locations.
   • Includes undersea fiber cables for stable and high-speed interconnectivity.

AWS Global Services (to learn in this section)

1. Amazon Route 53 (Global DNS)

   • Routes users to the closest deployment with the lowest latency.
   • Supports disaster recovery and global load balancing.

2. Amazon CloudFront (CDN)

   • Distributes and caches content at Edge Locations.
   • Reduces latency and improves user experience.

3. Amazon S3 Transfer Acceleration

   • Speeds up global uploads/downloads to S3 buckets using AWS Edge infrastructure.

4. AWS Global Accelerator

   • Improves global application availability and performance.
   • Uses AWS global network to route traffic optimally.

✅ Exam Tip:

• Regions = geographic locations.
• AZs = multiple isolated data centers per Region.
• Edge Locations = global points for caching/content delivery.
• Key global services: Route 53, CloudFront, S3 Transfer Acceleration, Global Accelerator.

Amazon Route 53

What is Route 53?

• Managed DNS (Domain Name System) service by AWS.
• Works like a phone book for the internet: maps domain names (e.g., myapp.com) to IP addresses or AWS resources.

DNS Record Types in Route 53

• A Record → Maps domain name to IPv4 address.
• AAAA Record → Maps domain name to IPv6 address.
• CNAME Record → Maps domain name to another domain name.
• Alias Record (AWS-specific) → Maps domain name to AWS resources such as:
  • ELB (Elastic Load Balancer)
  • CloudFront Distribution
  • S3 Website bucket
  • RDS instance

Exam Tip: Alias records are AWS-only and preferred over CNAME when pointing to AWS resources.

How DNS Resolution Works (simplified flow)

1. User enters myapp.mydomain.com in browser.
2. Browser queries Route 53.
3. Route 53 returns the mapped IP (from an A/AAAA record).
4. Browser connects to server at that IP.

Route 53 Routing Policies (high-level exam focus)

1. Simple Routing Policy
   • Default policy.
   • Maps domain to a single resource (one IP or instance).
   • No health checks.
2. Weighted Routing Policy
   • Distributes traffic across multiple resources.
   • Example: 70% → Server A, 20% → Server B, 10% → Server C.
   • Supports health checks (only sends traffic to healthy resources).
   • Used for load balancing, A/B testing.

1. Latency Routing Policy
   • Routes traffic to the resource with the lowest latency (closest AWS Region).
   • Example: User in Asia → Asia Region, User in US → US Region.
   • Improves performance globally.
2. Failover Routing Policy
   • Primary and secondary setup.
   • Route 53 checks health of primary.
   • If primary fails, traffic is sent to secondary.
   • Supports Disaster Recovery scenarios.

Key Exam Takeaways

• Route 53 = Managed DNS by AWS.
• Alias records = special AWS integration.
• Simple = no health checks.
• Weighted = traffic split.
• Latency = best performance.
• Failover = disaster recovery.

Amazon CloudFront – Exam Notes

What is CloudFront?

• AWS’s Content Delivery Network (CDN).
• Improves read performance by caching content at Edge Locations.
• Anytime you see CDN in the exam, think CloudFront.

Key Benefits

1. Low Latency
   • Content cached at ~216+ Points of Presence (Edge Locations).
   • Users get data from nearest Edge instead of origin → better performance.
2. Global Distribution
   • Content available worldwide without setting up multiple servers.
3. DDoS Protection
   • Spreading content globally protects against large-scale attacks.
   • Integrated with AWS Shield and WAF (Web Application Firewall).

How CloudFront Works

• User request goes to nearest Edge Location.
• If content is cached → served immediately.
• If not cached → CloudFront fetches from Origin, caches it, then serves.

S3 as an Orgin

Origins Supported by CloudFront

1. Amazon S3
   • For static files.
   • Secured via OAC (Origin Access Control) and bucket policy.
2. VPC Origins
   • Example: Application Load Balancer, Network Load Balancer, or EC2 in private subnet.
   • CloudFront can connect privately.
3. Custom Origins
   • Any HTTP backend (inside or outside AWS).
   • Example: S3 static website hosting, on-premise HTTP server.

CloudFront vs S3 Cross-Region Replication

Exam Tips

• CloudFront = CDN, caching at Edges.
• OAC used to secure S3 origin.
• Works with S3, ALB/NLB, EC2, or any HTTP server.
• CloudFront ≠ S3 replication, they solve different problems.

S3 Transfer Acceleration

Problem

• S3 Buckets are regional (linked to one Region).
• Uploading files from distant locations (e.g., USA → S3 bucket in Australia) can be slow.

What is S3 Transfer Acceleration?

• A feature that speeds up uploads and downloads to S3 buckets across large distances.
• Uses Amazon CloudFront Edge Locations as entry points.
• Data travels through AWS’s internal optimized network instead of the public internet.

How it Works

1. User uploads file to nearest Edge Location.
2. File is routed through AWS private backbone network.
3. Delivered to the destination S3 bucket in another Region.
4. Works for both uploads and downloads.

Key Points

• Best when bucket is far from users.
• Performance improvement depends on distance and internet quality.
• Typical gains range from 10% to much higher, depending on Region and location.
• AWS provides a test tool to compare standard vs accelerated transfer speeds.
• Use when you have a global application that needs to centralize uploads/downloads to one bucket.

Example

• Uploading from USA → S3 in Australia:
  • Without acceleration → slow, unreliable.
  • With acceleration → user uploads to US Edge → transferred internally to Australia.

Exam Tips

• Transfer Acceleration ≠ Cross-Region Replication.
  • Transfer Acceleration = speed up data movement into/out of S3.
  • CRR = copy data between buckets in different Regions.
• Often tested with scenario like: “Users worldwide upload to a single bucket in Region X, how to speed it up?” → Answer: Enable S3 Transfer Acceleration.

AWS Global Accelerator

Purpose

• Improves global application availability and performance.
• Uses the AWS global private network instead of the unpredictable public internet.
• Optimizes request routing, giving up to 60% faster performance.

How It Works

1. Users connect to the nearest AWS Edge Location.
2. From there, traffic is routed over the AWS private backbone network to the target region/application.
3. This reduces latency and improves reliability compared to normal internet routing.

Benefits

• Faster performance (shorter internet hops, then private AWS backbone).
• High availability with automatic rerouting.
• DDoS protection integrated with AWS Shield.
• Two static Anycast IPs (fixed entry points for your app, instead of changing regional endpoints).
• Global failover: if one region is down, traffic automatically routes to another.

Example Scenario

• App is hosted in India.
• Users in US, Europe, and Australia access it.
• Without Global Accelerator: traffic crosses many internet hops (high latency, possible failures).
• With Global Accelerator:
  • User connects to the nearest edge location (US, Europe, Australia).
  • Traffic flows via AWS private network to India.
  • Result: lower latency, more reliable connections.

Key Difference: Global Accelerator vs CloudFront

Real Performance Test (from transcript)

• Northern Virginia: 23% faster.
• Oregon: 31% faster.
• Singapore: 34% faster.
• Sydney: 53% faster.
• Ireland/Frankfurt (close regions): same performance (no real gain, since internet already short).

Exam Tips

• Use Global Accelerator when:
  • You need static IPs for global clients.
  • You want low latency and high availability for global users.
  • You need regional failover for disaster recovery.
• Use CloudFront when:
  • You want to cache content at the edge.
  • Optimizing for static/dynamic content delivery like media files and websites.

Key takeaway: Global Accelerator = faster routing & failover via AWS global network.
CloudFront = caching content at the edge (CDN).

AWS Outposts

What is Hybrid Cloud?

• Hybrid Cloud = Using both on-premises infrastructure and cloud infrastructure together.
• This means managing two environments:
  • Cloud (AWS console, CLI, APIs)
  • On-premises (local servers, physical infrastructure)
• Leads to different skillsets, different APIs, and added complexity.

AWS Outposts Overview

• AWS Outposts is a fully managed service that brings AWS infrastructure, services, APIs, and tools on-premises.
• In simple words, AWS installs server racks (Outposts racks) inside your data center.
• These racks run the same AWS hardware and software as used in AWS data centers.
• You manage them through the same AWS Console, CLI, or APIs—no difference.

How It Works

1. AWS delivers and installs Outposts racks at your site.
2. The racks are owned and managed by AWS, but physically located in your data center.
3. You can then launch AWS services locally on those racks.
4. You still manage everything using AWS tools (Console, CLI, SDKs, etc.).

Services Supported on Outposts

You can run the following AWS services locally:

• Amazon EC2 (compute instances)
• Amazon EBS (block storage)
• Amazon S3 (object storage)
• Amazon RDS (databases)
• Amazon EKS and ECS (containers)
• Amazon EMR (big data processing)

Benefits

Responsibility

• AWS manages hardware and software updates, monitoring, and scaling.
• You (the customer) are responsible for physical security of the racks since they are located in your facility.

Use Cases

• Applications requiring low-latency access to on-premises systems.
• Data residency requirements (legal or compliance).
• Industries like finance, healthcare, or manufacturing that need local processing.
• Gradual cloud migration (step-by-step move to AWS).

Exam Tip

• Outposts = AWS Cloud inside your data center.
• Fully managed, same APIs and services as AWS Cloud.
• Used when you need low latency, local data processing, or data residency.
• You are responsible for physical security of Outposts racks.

Summary Table

Key takeaway: AWS Outposts brings AWS services, APIs, and infrastructure directly into your on-premises data center, providing the same experience as the AWS Cloud but with local control and low latency.

AWS Wavelength

What is AWS Wavelength?

• AWS Wavelength Zones are AWS infrastructure deployments embedded inside telecommunication providers’ data centers at the edge of 5G networks.
• Purpose: To bring AWS services closer to end-users and provide ultra-low latency for 5G applications.

How it Works

• You can deploy AWS resources (EC2, EBS, VPC, etc.) directly into Wavelength Zones.
• These zones are connected to 5G networks via carrier gateways.
• End-users on 5G access applications with extremely low latency because the compute resources are physically near the users.
• The traffic often stays within the telecom provider’s network (CSP) and does not reach the AWS region unless needed.

Connectivity

• Wavelength Zones are connected to their parent AWS Region.
  • Example: EC2 in Wavelength Zone can access RDS or DynamoDB in the parent region.
• No extra charges or agreements are required to use Wavelength.

Benefits

• Ultra-low latency for 5G apps.
• Edge computing capabilities with AWS tools.
• Local data processing without sending traffic to AWS cloud.
• Seamless integration with AWS regions when needed.

Use Cases

• Smart cities
• Machine Learning-assisted diagnostics
• Connected vehicles
• Interactive live video streaming
• Augmented Reality (AR) / Virtual Reality (VR)
• Real-time gaming

Exam Tip

Whenever the question mentions “5G” or “ultra-low latency edge computing”, the correct answer is AWS Wavelength.

AWS Local Zones

What is AWS Local Zones?

• Local Zones extend AWS Regions closer to end users to run latency-sensitive applications.
• They place compute, storage, database, and other AWS services physically near large population or industrial centers.
• Think of them as mini extensions of an AWS Region located in different cities.

How It Works

• Each Local Zone is an extension of a specific AWS Region.
• You can extend your VPC from the region into one or more Local Zones.
• You can then launch AWS resources (like EC2, RDS, ECS, EBS, etc.) directly in the Local Zone.
• Example:
  • Region: us-east-1 (Northern Virginia)
  • Local Zones: Boston, Chicago, Dallas, Miami, Houston, etc.
• These local zones are connected to their parent region through high-bandwidth private AWS network links.

Key Features

Example from Hands-On

1. Go to EC2 Console → pick a region (e.g., us-east-1).
2. Under Settings → Zones, enable a Local Zone (like Boston).
3. Create a subnet for that Local Zone (e.g., Boston Subnet).
4. Launch an EC2 instance and select the Boston Local Zone subnet.
5. This places compute physically closer to users in Boston.

Benefits

• Lower latency for users far from AWS Regions.
• Improved performance for real-time applications.
• No need to build your own edge infrastructure.
• Integrated with AWS ecosystem (VPC, IAM, CloudWatch, etc.).

Exam Tip

If the question says:

• “Run applications with ultra-low latency for users in a specific city.”
• “Need AWS services close to end-users, but not inside a 5G network.”

The answer is AWS Local Zones.

(If the question mentions 5G or telecom edge, that’s AWS Wavelength instead.)

Global Application Architectures in AWS

Single Region, Single Availability Zone (AZ)

Single Region, Multi-AZ

Multi-Region, Active–Passive

Multi-Region, Active–Active

Summary Table

AWS Example Services

Exam Tip

If the question says:

• “Users across the world should have low latency for both reads and writes.” → ✅ Active–Active
• “Disaster recovery setup, passive region acts as backup.” → ✅ Active–Passive
• “Single region fault tolerance only.” → ✅ Single Region Multi-AZ

AWS Global Applications – Summary Notes

Route 53 (Global DNS)

• Routes users to nearest deployment for lowest latency.
• Helps in disaster recovery (DR) setups.
• Can map hostnames → IPs.
• Supports routing policies like latency-based, geolocation, failover, etc.

Amazon CloudFront (CDN)

• A Content Delivery Network that delivers content with low latency and high transfer speed.
• Connected to S3 in the example (to serve static content).
• Uses AWS Edge Locations to cache and deliver content near users.
• Reduces latency and improves user experience.

S3 Transfer Acceleration

• Speeds up global uploads/downloads to S3.
• Uses Edge Locations to route data through AWS’s global network backbone.
• Ideal for users uploading data from far away regions.

AWS Global Accelerator

• Improves global application performance and availability.
• Uses AWS Edge Network to route user traffic via optimal AWS backbone paths.
• Reduces latency and avoids internet congestion.

AWS Outposts

• Extends AWS on-premises.
• AWS installs Outposts Racks in your data center.
• You can run AWS services locally while connecting seamlessly to the AWS Cloud.
• Ideal for workloads needing low latency to on-prem systems or local data processing.

AWS Wavelength

• Brings AWS services to the 5G edge (telecom provider networks).
• Enables ultra-low latency applications like IoT, AR/VR, real-time gaming.
• Integrates with 5G networks for milliseconds-level latency.

AWS Local Zones

• Extend AWS infrastructure to specific metro areas (like Boston, Dallas, Miami, Chicago).
• Bring compute, database, and storage closer to end users.
• Great for latency-sensitive apps (e.g., media, gaming, finance).
• Currently focused on the US, expanding globally.

Summary Table

Key Takeaway:

• AWS provides multiple global services to improve latency, performance, and availability for worldwide users.
• Each service addresses a different part of global infrastructure — DNS, content delivery, data transfer, networking, or edge computing.