AWS Regions, Availability Zones, and Edge Locations: The Global Infrastructure

An AWS Region is a physical geographic area of the world that contains multiple isolated data center clusters called Availability Zones, and edge locations are smaller sites in many more cities that cache content close to users. Regions are fully independent of one another, Availability Zones within a Region are isolated but connected by fast private links, and this layering is what lets you build applications that survive failures. CLF-C02 tests the definitions, the relationships between the layers, and the reasons you would choose one Region over another.

The three layers of AWS global infrastructure

Regions are the top layer. Each Region (such as US East in Northern Virginia or Asia Pacific in Sydney) is a separate geographic area with its own independent power, cooling, and network infrastructure. Regions do not share resources, and data you store in a Region stays in that Region unless you explicitly move or replicate it. That data-residency guarantee is a compliance cornerstone: AWS will not migrate your data out of the Region you selected.

Availability Zones (AZs) are the middle layer. Every Region contains multiple AZs, and current AWS design standards call for at least three per Region. An AZ is one or more discrete data centers with redundant power, networking, and connectivity, physically separated from the other AZs in the Region by a meaningful distance (kilometers apart, but close enough for single-digit-millisecond latency). The AZs in a Region are linked with high-bandwidth, low-latency, encrypted private fiber. The separation means a flood, fire, or power failure in one AZ should not affect another; the fast links mean applications can replicate data between AZs synchronously.

Edge locations are the bottom layer and by far the most numerous. These are small sites in hundreds of cities worldwide, far more places than there are Regions. They do not run your EC2 instances. Instead they host services that benefit from being close to end users: Amazon CloudFront (the content delivery network) caches copies of your content there, Amazon Route 53 answers DNS queries there, and AWS Global Accelerator uses them to pull user traffic onto the AWS private backbone as early as possible.

Layer	Count and scope	What runs there
Region	Dozens worldwide, geographically separate	Most AWS services and your primary workloads
Availability Zone	Multiple per Region (typically 3+)	Your instances, databases, subnets
Edge location	Hundreds of cities	CloudFront caches, Route 53, Global Accelerator

Why the layering exists: fault isolation

The hierarchy is a fault-isolation design. A single data center can fail. An AZ groups data centers but can still suffer a zone-wide event. A Region-wide failure is rare but possible. Each layer up gives you a bigger blast-radius boundary to design around:

• Deploying across multiple Availability Zones protects against data center and zone failures. This is the standard definition of a highly available architecture on AWS, and services like Elastic Load Balancing and Amazon RDS Multi-AZ deployments are built around it. Subnets are the mechanism that places resources into specific AZs, which is covered in Amazon VPC basics.
• Deploying across multiple Regions protects against regional disasters and also serves global users with lower latency. Multi-Region architectures are the domain of disaster recovery strategies and global applications.

A phrase worth knowing exactly: AWS describes fault isolation boundaries in these terms, and exam answers about “designing for failure” almost always resolve to “run the workload in multiple Availability Zones” unless the question explicitly raises disaster recovery or global reach, in which case multiple Regions is the answer.

Choosing a Region: the four factors

CLF-C02 expects you to name the considerations for selecting a Region. There are four canonical ones:

1. Compliance and data governance. If law or policy requires data to remain in a specific country, you must pick a Region in that country. This factor overrides all others when it applies.
2. Proximity to users (latency). Closer Regions mean faster round trips for your customers.
3. Service availability. New AWS services and features do not launch in every Region simultaneously. The service you need must actually be offered in the Region you pick.
4. Pricing. Costs differ between Regions because AWS’s own operating costs differ. The same instance type can cost more in one Region than another.

Also know the specialized infrastructure options at the recognition level. AWS Local Zones place compute and storage in additional metro areas close to large population centers, for single-digit-millisecond latency needs. AWS Wavelength embeds AWS compute inside telecommunications providers’ 5G networks for ultra-low-latency mobile applications. AWS Outposts delivers AWS-managed hardware into your own data center, extending AWS infrastructure and APIs on premises for workloads that must stay local.

Finally, note which services are global rather than regional. AWS Identity and Access Management (IAM), Amazon Route 53, and Amazon CloudFront operate globally; most other services, such as Amazon EC2 and Amazon S3, are regional, and you choose a Region every time you use them.

How the CLF-C02 exam tests this

• A definitional question asks what an Availability Zone is. The correct answer is “one or more discrete data centers with redundant power and networking within a Region.” Distractors define it as a Region, a data center in a different country, or an edge location.
• A scenario says an application must remain available even if an entire data center fails, and asks for the design. The answer deploys the application across multiple Availability Zones, usually behind a load balancer.
• A scenario lists requirements like “customer data must not leave Germany” and asks what determines Region choice. The answer is compliance/data residency, and the exam wants you to rank it above latency or cost.
• A scenario describes serving static content (images, video) to users worldwide with low latency and asks which infrastructure makes that possible. The answer involves edge locations via CloudFront, not adding more Regions.
• A scenario needs AWS services running on premises or single-digit-millisecond latency in a specific metro area. On premises points to Outposts; metro latency points to Local Zones; 5G mobile points to Wavelength.

Quick reference

• Region: an isolated geographic area containing multiple AZs; your data stays in the Region you choose.
• Availability Zone: one or more discrete data centers with independent power and networking; AZs in a Region are linked by low-latency private fiber.
• Edge location: a content and DNS point of presence in hundreds of cities; used by CloudFront, Route 53, and Global Accelerator.
• High availability = multiple AZs. Disaster recovery and global reach = multiple Regions.
• Region choice factors: compliance, latency to users, service availability, price.
• Local Zones (metro latency), Wavelength (5G), Outposts (your data center) extend AWS infrastructure beyond Regions.
• IAM, Route 53, and CloudFront are global services; EC2 and S3 are regional.