DESCRIPTION
Before launching your first server or storage bucket, it is important to understand where AWS physically exists. This is not some abstract “cloud” — behind every service are real data centers at specific points on the planet. Understanding how AWS infrastructure is organized directly impacts how reliably, quickly, and cost-effectively your application will run.
In this lesson, we will explore the AWS global infrastructure and get acquainted with the key services that form the foundation of the platform.
WHAT YOU WILL LEARN
– What AWS Regions, Availability Zones, and Edge Locations are
– How to choose the right region for your project
– What key AWS services exist and what they are used for
– What to know for the Solutions Architect Associate exam
– How the Free Tier is structured and how AWS services are priced
AWS GLOBAL INFRASTRUCTURE
AWS REGIONS
A Region is a geographic location where AWS has deployed a cluster of data centers. Think of a Region as a separate, independent cloud: it is completely isolated from other Regions, has its own services, its own network, and its own pricing.
Key characteristics of Regions:
– Each Region is fully isolated from all other Regions
– Every Region contains a minimum of 3 Availability Zones
– Data is not replicated between Regions automatically — only if you explicitly configure it
– Regions are geographically dispersed to provide maximum resilience
As of 2026:
– 33 geographic Regions worldwide
– 105+ Availability Zones
– Several additional Regions in various stages of opening
REGION EXAMPLES
us-east-1 (Northern Virginia, USA) — the first and largest Region; new services appear here first
us-east-2 (Ohio, USA)
us-west-1 (Northern California, USA)
us-west-2 (Oregon, USA)
eu-west-1 (Ireland)
eu-central-1 (Frankfurt, Germany)
eu-north-1 (Stockholm, Sweden)
ap-southeast-1 (Singapore)
ap-northeast-1 (Tokyo, Japan)
ca-central-1 (Montreal, Canada)
sa-east-1 (São Paulo, Brazil)
AVAILABILITY ZONES
An Availability Zone (AZ) is one or more physically isolated data centers within a single Region. This is where real fault tolerance begins.
Key characteristics of AZs:
– Physically separated: different buildings, different areas, sufficient distance to isolate from localized disasters
– Independent power supplies, cooling systems, and network connectivity
– Interconnected by high-speed private links with less than 2 ms latency
– Named with letter suffixes: us-east-1a, us-east-1b, us-east-1c
Why this matters in practice:
If you run two servers — one in us-east-1a and one in us-east-1b — and one zone fails (fire, power outage, network issue), the second server keeps running. A load balancer automatically redirects traffic. Users notice nothing.
This is why multi-AZ architecture is the standard for any production application.
EDGE LOCATIONS
An Edge Location is a point of presence placed closer to end users than full Regions. Edge Locations are not data centers for running servers — they are caching nodes that accelerate content delivery.
Key characteristics:
– More than 600 points of presence in 90+ cities worldwide
– Used by CloudFront (CDN) to cache content close to users
– When a user in Vilnius requests a file from your site, they receive it from the nearest Edge Location rather than from a data center in the US
The key distinction: Edge Locations do not support running EC2, RDS, or other compute services. They serve only as caching and content delivery nodes.
HOW TO CHOOSE A REGION
Choosing a Region is not just a technical decision. Here are four key factors:
1. PROXIMITY TO YOUR USERS
Choose the Region closest to your customers. This directly reduces latency and improves the user experience. If your audience is in Europe, use eu-west-1 or eu-central-1.
2. REGULATORY COMPLIANCE
Many countries and industries require data to be stored within a specific jurisdiction. GDPR in Europe, national data residency laws in various countries — all of these affect Region choice. Ignoring these requirements can lead to significant fines.
3. SERVICE AVAILABILITY
New AWS services first appear in us-east-1 and then roll out gradually to other Regions. If you need a specific service, check its availability in your target Region: https://aws.amazon.com/about-aws/global-infrastructure/regional-product-services/
4. PRICING
Prices vary between Regions. us-east-1 is traditionally one of the most cost-effective options. Asia-Pacific Regions tend to be more expensive. At high usage volumes, the pricing difference can be substantial — always compare using the AWS Pricing Calculator.
CORE AWS SERVICES
AWS offers more than 240 services. Below are the key ones, organized by category. For the Solutions Architect Associate exam, you need a solid understanding of 20-30 core services.
COMPUTE
AMAZON EC2 (Elastic Compute Cloud)
Virtual servers in the cloud — the foundation of AWS compute infrastructure. Each server is called an instance. EC2 gives you full control over the operating system, configuration, and software.
Supported operating systems: Linux, Windows, macOS
Key concepts:
– Instance Types — categories with different CPU/RAM ratios: t3.micro, m5.large, c5.xlarge, r5.2xlarge, etc.
– AMI (Amazon Machine Image) — pre-built OS images for launching instances
– Security Groups — virtual firewalls controlling inbound and outbound traffic
– Key Pairs — SSH keys for accessing Linux instances
ELASTIC LOAD BALANCING (ELB)
Load balancers distribute incoming traffic across multiple instances, automatically routing around unhealthy servers.
Three types of load balancers:
– Application Load Balancer (ALB) — Layer 7, HTTP/HTTPS, routing by URL path and headers
– Network Load Balancer (NLB) — Layer 4, TCP/UDP, extremely high performance, supports Security Groups
– Classic Load Balancer (CLB) — legacy type, not recommended for new projects
AWS ELASTIC BEANSTALK
Platform as a Service (PaaS): you upload your code, and AWS handles everything else — creating EC2 instances, load balancer, Auto Scaling Group, and monitoring.
Supported platforms: Java, .NET, PHP, Node.js, Python, Ruby, Go, Docker
Ideal for rapid application deployment without deep infrastructure management knowledge.
AWS LAMBDA
Serverless compute: code runs in response to events without managing servers. You pay only for execution time, billed to the millisecond.
Supported languages: Python, Node.js, Java, C#, Go, Ruby
Typical use cases: API request handling, automation, event processing from S3/DynamoDB/SQS
STORAGE
AMAZON S3 (Simple Storage Service)
Object storage for files of any type and size. One of the foundational AWS services — used in virtually every architecture.
Key characteristics:
– Durability: 99.999999999% (11 nines)
– Virtually unlimited capacity
– Files are stored in “buckets”
– Object versioning support
– Can be used as a static website host
Storage classes (from frequent access to long-term archiving):
– S3 Standard — frequent access, high performance
– S3 Intelligent-Tiering — automatic movement between classes based on access patterns
– S3 Standard-IA — infrequent access, lower storage cost
– S3 One Zone-IA — infrequent access, single AZ, most affordable for non-critical data
– S3 Glacier Instant Retrieval — archive with millisecond access
– S3 Glacier Flexible Retrieval — archive, 1-5 minute retrieval time
– S3 Glacier Deep Archive — long-term archive, up to 12-hour retrieval, minimum cost
AMAZON EBS (Elastic Block Store)
Block storage — a virtual hard drive attached to EC2 instances. It works like a regular disk: file system, real-time read/write operations.
Key characteristics:
– Bound to a specific Availability Zone (cannot attach to an instance in a different AZ)
– Snapshot support for backup and recovery
– Types: SSD (gp3, io2) for performance workloads; HDD (st1, sc1) for streaming workloads
AMAZON EFS (Elastic File System)
A network file system (NFS) that can be simultaneously mounted on multiple EC2 instances across different AZs. Unlike EBS — a shared resource, not a personal disk.
Ideal for: shared storage in clustered applications, web servers with shared content, containerized workloads.
DATABASES
AMAZON RDS (Relational Database Service)
Managed relational databases: AWS handles installation, patching, backups, and monitoring. You work with the database without worrying about its administration.
Supported engines:
– MySQL
– PostgreSQL
– MariaDB
– Oracle
– Microsoft SQL Server
– Amazon Aurora
Key features:
– Automated daily backups
– Multi-AZ: automatic failover to a standby database on failure
– Read Replicas: read-only replicas for scaling read-heavy workloads
AMAZON AURORA
AWS’s own database engine, compatible with MySQL and PostgreSQL. Designed from the ground up for cloud workloads.
Characteristics:
– Up to 5x faster than standard MySQL
– Up to 3x faster than standard PostgreSQL
– Automatic storage scaling up to 128 TB
– Automatic data replication across 3 AZs (6 copies)
– Available in Provisioned and Serverless modes
AMAZON DYNAMODB
Fully managed NoSQL database with millisecond latency at any scale. Data is stored as key-value pairs or JSON documents.
Characteristics:
– No servers or schema to manage
– Automatic scaling under load
– Built-in multi-AZ replication
– Essential service for the Developer Associate exam
AMAZON REDSHIFT
Cloud data warehouse for analytical workloads. Designed to process queries against petabytes of data using standard SQL.
Used for: business intelligence, reporting, ETL pipelines.
NETWORKING
AMAZON VPC (Virtual Private Cloud)
An isolated virtual network in AWS — your own network space in the cloud. Full control over IP addresses, subnets, routing, and access rules.
Key concepts:
– CIDR blocks — defining IP address ranges (e.g., 10.0.0.0/16)
– Public Subnet — subnet with direct internet access
– Private Subnet — subnet without direct internet access
– Internet Gateway — gateway for inbound and outbound internet traffic
– NAT Gateway — allows resources in Private Subnets to reach the internet without being directly reachable from it
– Security Groups — stateful firewalls at the instance level
– Network ACLs — stateless rules at the subnet level
AMAZON ROUTE 53
AWS’s DNS service and domain registrar. Manages domain name resolution and supports multiple routing strategies.
Routing policies:
– Simple — one record, one resource
– Weighted — distribute traffic by weight across multiple resources (useful for A/B testing)
– Latency-based — route users to the Region with the lowest latency
– Failover — automatic switch to a standby resource on failure
– Geolocation — route based on the user’s geographic location
Supports Health Checks to monitor endpoint availability.
AMAZON CLOUDFRONT
Global Content Delivery Network (CDN). Caches content at more than 600 points of presence worldwide, reducing delivery time to end users.
Capabilities:
– Integration with S3, EC2, ALB, and any HTTP-based server
– HTTPS support with SSL certificate management through ACM
– Built-in DDoS protection via AWS Shield
– Lambda@Edge support for processing requests at edge nodes
DEVOPS AND AUTOMATION
AWS CLOUDFORMATION
Infrastructure as Code (IaC): describe your entire infrastructure in templates (JSON or YAML) and deploy it automatically. CloudFormation creates resources in the correct order, manages dependencies, and supports automatic rollback on failure.
Benefits:
– Reproducible deployments: the same template deploys identical infrastructure in any Region
– Infrastructure versioning in source control (Git)
– Automatic rollback on error
– Resource group management through stacks
MONITORING AND MANAGEMENT
AMAZON CLOUDWATCH
Centralized monitoring for AWS resources and applications. Collects metrics, stores logs, and sends alerts.
Capabilities:
– Metrics: CPU Utilization, Network In/Out, Disk I/O for EC2; DB Connections, CPU for RDS; Request Count, Latency for ELB
– Alarms: notifications when values exceed thresholds (e.g., CPU > 80%)
– CloudWatch Logs: centralized log collection from EC2, Lambda, and containers
– Dashboards: real-time metric visualization
AWS IAM (Identity and Access Management)
Access management for AWS resources. IAM defines who can do what in your account.
Key entities:
– Users — accounts for individual people
– Groups — user groups with shared permissions
– Roles — temporary permissions for services and applications (e.g., an EC2 instance accessing S3)
– Policies — JSON documents describing permissions
Security principles:
– Least Privilege — grant only the permissions that are necessary, nothing more
– MFA — enable multi-factor authentication for all users
– Regular access key rotation
APPLICATION SERVICES
AMAZON SNS (Simple Notification Service)
A pub/sub service for sending notifications. A message is published to a “topic,” and SNS delivers it to all subscribers.
Supported channels: Email, SMS, HTTP/HTTPS, Lambda, SQS
Typical use: CloudWatch alerts, system event notifications
AMAZON SQS (Simple Queue Service)
Managed message queue for asynchronous communication between system components. Decouples producers and consumers of messages.
Two queue types:
– Standard Queue — high throughput, delivery order not guaranteed
– FIFO Queue — strict processing order, exactly-once delivery guaranteed
AMAZON SES (Simple Email Service)
Service for sending and receiving email at scale: transactional messages, marketing campaigns, notifications.
ANALYTICS
AMAZON EMR (Elastic MapReduce)
Managed cluster for big data processing using Apache Hadoop, Spark, Hive, and other frameworks. Enables analysis of petabytes of data using S3 as the data source.
SERVICES BY LEARNING PRIORITY
MUST KNOW (Solutions Architect Associate exam):
EC2 — virtual servers
S3 — object storage
VPC — virtual networks
IAM — access management
RDS — relational databases
ELB (ALB/NLB) — load balancers
CloudWatch — monitoring
Route 53 — DNS
CloudFront — CDN
Lambda — serverless compute
SHOULD KNOW WELL:
EBS and EFS — block and file storage
DynamoDB — NoSQL database
CloudFormation — Infrastructure as Code
Auto Scaling — automatic scaling
Elastic Beanstalk — PaaS
SNS and SQS — notifications and message queues
BASIC UNDERSTANDING REQUIRED:
Aurora — AWS database engine
Redshift — Data Warehouse
EMR — Big Data
SES — email service
AWS FREE TIER AND PRICING
FREE TIER
AWS provides a free tier for new accounts. Some services are free for the first 12 months; others are free indefinitely.
First 12 months:
– 750 hours of EC2 t3.micro per month
– 5 GB of S3 Standard storage
– 750 hours of RDS db.t2.micro per month
– And much more
Always Free:
– Lambda: 1 million requests per month
– DynamoDB: 25 GB of storage
– CloudWatch: basic metrics
Current conditions always available at: https://aws.amazon.com/free
ACCOUNT REGISTRATION
To register, you will need:
– Email address
– Phone number
– Credit card (Visa, MasterCard, American Express) — for identity verification only; no charge occurs at registration
PRICING AND COST CONTROL
AWS operates on a Pay-as-you-go model: you pay only for what you use. Prices vary by Region.
Cost control tools:
– AWS Pricing Calculator — estimate costs before deployment
– AWS Budgets — set budget limits with alerts as you approach them
– Cost Explorer — analyze actual spending
Cost-saving tips:
– Stop unused EC2 instances (a stopped instance incurs no compute charges)
– Use Auto Scaling: pay for exactly the capacity you need
– Reserved Instances for long-term predictable workloads (savings of up to 72% vs On-Demand)
– S3 Lifecycle Policies to automatically move older files to cheaper storage classes
IMPORTANT TO KNOW
Do not confuse Regions and Availability Zones. A Region is a geographic area (such as Ireland). An Availability Zone is an isolated data center within that Region. Many beginners mix these up — do not be one of them.
us-east-1 is special. New AWS services appear here first. If a service is unavailable in your preferred Region, check whether it exists in us-east-1.
Edge Locations are not Regions. You cannot run EC2 or RDS in an Edge Location. They exist only as caching nodes for CloudFront.
Data does not leave a Region without your consent. AWS does not automatically replicate data between Regions. This is important from a GDPR and regulatory compliance perspective.
CHECK YOURSELF
1. What is an Availability Zone, and how does it differ from a Region?
2. What is the minimum number of Availability Zones in each AWS Region?
3. What is an Edge Location and which AWS service uses them?
4. Name the three types of load balancers in AWS.
5. What is the difference between S3 and EBS?
6. What is a VPC and why is it needed?
7. What is CloudWatch used for?
8. What is the difference between RDS and DynamoDB?
9. What four factors are important when choosing an AWS Region?
10. What does Multi-AZ mean in the context of RDS?
CONCLUSIONS
AWS global infrastructure is the foundation on which everything else is built. Understanding Regions, Availability Zones, and Edge Locations directly shapes architectural decisions.
Key facts:
– 33 Regions, 105+ Availability Zones — and the infrastructure continues to expand
– Every Region is isolated and contains a minimum of 3 AZs
– Edge Locations (600+) are used to cache content via CloudFront
– Region selection is driven by user proximity, regulatory requirements, service availability, and cost
– AWS offers 240+ services; a solid grasp of 20-30 key ones is sufficient to get started
– The Free Tier allows you to learn with real services at no cost for 12 months
Next lesson: registering an AWS account and taking your first steps in the AWS Management Console.