AWS Cloud Practitioner Study Notes · Part 4
AWS Global Infrastructure: Regions, Zones, and Edge Locations
AWS Cloud Practitioner study notes on Regions, Availability Zones, CloudFront edge locations, Local Zones, Wavelength Zones, and Direct Connect.
When an AWS exam question mentions high availability, low latency, disaster recovery, or global users, the answer often depends on choosing the right part of AWS Global Infrastructure. A Region, Availability Zone, CloudFront edge location, Local Zone, and Wavelength Zone solve different placement and connectivity problems.
This is Part 4 of the AWS Cloud Practitioner Study Notes series. The previous part explained service models and deployment strategies; this article focuses on where AWS infrastructure is located and what each location is designed to do.
The infrastructure hierarchy
The most useful mental model is:
| Component | Main purpose | Exam clue |
|---|---|---|
| Region | Geographically separate AWS area | Data residency or cross-geography disaster recovery |
| Availability Zone | Isolated location inside a Region | High availability within one Region |
| Edge location | Delivers content or edge services close to users | CloudFront and low-latency delivery |
| Regional edge cache | Larger CloudFront cache between edge locations and origins | Improve cache efficiency |
| Local Zone | Extends selected AWS resources closer to a metropolitan area | Low latency for nearby users |
| Wavelength Zone | Places AWS infrastructure at a telecommunications provider’s 5G edge | Ultra-low-latency mobile applications |
| Direct Connect location | Physical location for a dedicated connection into AWS | Consistent connectivity from on premises |
These components are complementary. A workload might run on EC2 across several Availability Zones in the Singapore Region, use CloudFront edge locations for static content, and connect an on-premises network through a Direct Connect location.
AWS Regions: separate geographic areas
An AWS Region is a separate geographic area containing AWS infrastructure. Examples include Asia Pacific (Singapore), Asia Pacific (Tokyo), Europe (Frankfurt), and US East (N. Virginia).
Regions are designed to be isolated from one another. If an application runs in the Singapore Region, its EC2 instances, databases, and storage do not automatically appear in Tokyo. Replication, backup, and failover across Regions must be designed and configured using services and processes appropriate to the workload.
Choose a Region based on more than physical distance. Important considerations include:
- Latency to the majority of users
- Availability of the AWS services and features required
- Data residency and regulatory requirements
- Disaster-recovery objectives
- Pricing and operational cost
- Network connectivity and dependencies on other systems
For users in Malaysia, the Singapore Region may provide a useful starting point because it is geographically close. That does not mean it is automatically the right Region for every workload; the required services, compliance constraints, resilience plan, and measured latency still matter.
Exam clues for Regions include:
- Different countries or geographic areas
- Keeping data in a specific country or jurisdiction
- Deploying a disaster-recovery environment far from the primary workload
- Selecting infrastructure close to the majority of customers
The key memory aid is: a Region is a geographic area, and Regions are isolated from one another.
Availability Zones: high availability within a Region
An Availability Zone (AZ) is one or more discrete data centres within a Region. AZs are separate locations with redundant power, networking, and connectivity. They are connected to other AZs in the same Region with high-bandwidth, low-latency, highly redundant networking.
The physical separation reduces the chance that a failure affecting one AZ will affect every AZ in the Region. That is why deploying a workload across multiple AZs is a common high-availability pattern.
For example, an application might run EC2 instances in two AZs behind a load balancer, while an RDS deployment uses a Multi-AZ configuration. If one AZ becomes unavailable, the architecture can continue serving traffic through resources in another AZ, provided the application and data layer have been designed for that failure mode.
An AZ is not the same as a Region:
- One Region contains multiple AZs.
- AZs provide isolation within a Region.
- Regions provide geographic separation between larger infrastructure areas.
Most AWS Regions have at least three AZs, but the exact number and service availability vary. Check the current AWS documentation rather than memorising one fixed number for every Region.
Exam clues for Availability Zones include:
- Surviving a single data-centre or AZ failure
- Deploying redundant EC2 instances in separate locations
- RDS Multi-AZ
- High availability within one Region
The key memory aid is: multiple AZs improve availability; multiple Regions improve geographic separation.
Edge locations: bring content closer to users
An edge location is part of AWS’s global edge network and is used by services such as Amazon CloudFront to serve content closer to viewers. When a user requests an object through CloudFront, DNS directs the request to an appropriate point of presence, commonly called an edge location.
If the object is cached there, CloudFront can return it without contacting the origin. If it is not cached, CloudFront retrieves it from the origin, returns it to the viewer, and may cache it for later requests according to the distribution’s rules.
Edge locations are useful for:
- Static website assets
- Images, videos, and downloads
- Cached API responses where appropriate
- Lower-latency content delivery for geographically distributed users
Edge locations are not Regions and are not a replacement for the application’s primary compute and database placement. They are part of the delivery layer. Route 53 and other AWS edge services also use globally distributed infrastructure, but the exact behaviour depends on the service.
Exam clues include:
- Delivering static content quickly to users worldwide
- Caching content close to users
- CloudFront
- Reducing the distance between users and a content-delivery point
The key memory aid is: CloudFront edge locations cache and deliver content close to viewers.
Regional edge caches: an intermediate CloudFront layer
CloudFront regional edge caches sit between origin servers and points of presence. They have larger caches than individual edge locations and can retain objects that are not popular enough to remain in a specific point of presence.
The request path can therefore involve several layers:
- A viewer requests an object through CloudFront.
- CloudFront checks the nearby point of presence.
- On a miss, CloudFront may check the regional edge cache.
- If the object is still unavailable, CloudFront requests it from the origin.
Regional edge caches are an optimisation inside CloudFront’s delivery architecture. They are less common as the direct answer in Cloud Practitioner questions than Regions, AZs, and edge locations, but the distinction is useful when a question asks about an intermediate caching layer.
Local Zones: extend a Region towards a city
An AWS Local Zone is an extension of an AWS Region placed in geographic proximity to users. It allows supported resources, such as selected EC2 and EBS configurations, to run closer to a metropolitan area while remaining connected to the parent Region.
Local Zones are useful when an application requires lower latency than the parent Region can provide and the supported services meet the workload’s needs. Possible use cases include interactive media, real-time gaming, and applications with strict response-time requirements.
Local Zones are not independent Regions. They are associated with a parent Region, must be enabled before use, and support only a subset of AWS services and features. A Local Zone should therefore be selected for a measured latency or placement requirement, not simply because it sounds more local.
Exam clues include:
- Extending AWS infrastructure closer to a large city
- Lower latency for users near that metropolitan area
- A workload that can use supported Local Zone services
The key memory aid is: Local Zones extend a parent Region closer to users in a specific metropolitan area.
Wavelength Zones: AWS infrastructure at the 5G edge
AWS Wavelength places selected AWS compute and storage services at the edge of telecommunications providers’ 5G networks. A Wavelength Zone is associated with an AWS Region and is designed for applications that need ultra-low latency to 5G devices and users.
Potential use cases include:
- Augmented and virtual reality
- Connected vehicles
- Smart factories
- Real-time gaming
- Mobile edge computing
Wavelength Zones are not available in every Region and are not a general-purpose replacement for ordinary AZs. The application must also account for the supported carrier network, device connectivity, service availability, and the rest of the workload’s dependencies.
Exam clues include:
- 5G mobile devices
- Ultra-low latency to telecom users
- Mobile edge computing
- AWS infrastructure located inside a telecommunications network
The key memory aid is: Wavelength means AWS at the 5G edge.
Direct Connect locations: dedicated connectivity into AWS
An AWS Direct Connect location is a physical location where a customer can establish a Direct Connect connection into AWS. Direct Connect links an on-premises network to AWS over a dedicated connection and bypasses the Internet Service Provider in the network path.
Direct Connect can provide a more consistent network experience than an internet-based connection. It supports different virtual interface types, including private connectivity to a VPC and public connectivity to AWS public services. The exact design, routing, resilience, and billing depend on the connection and virtual interface chosen.
Direct Connect is different from a Site-to-Site VPN:
- Site-to-Site VPN uses encrypted tunnels over the internet.
- Direct Connect uses a dedicated network connection into AWS and does not use the ISP path for that connection.
For resilient architecture, a single Direct Connect connection should not automatically be treated as a complete high-availability design. Organisations may need redundant connections, diverse locations, or a VPN backup depending on the business requirement.
Exam clues include:
- A dedicated connection from an on-premises data centre to AWS
- More consistent network performance
- Private connectivity requirements
- Avoiding the normal internet service-provider path
The key memory aid is: Direct Connect is dedicated connectivity; VPN is encrypted connectivity over the internet.
Putting the components together
Consider an e-commerce platform serving users across Asia and Europe:
- EC2 instances run across multiple AZs in the Singapore Region for application availability.
- A database uses a Multi-AZ design for resilience within that Region.
- Amazon S3 stores static assets in the Region.
- CloudFront edge locations deliver images and other cacheable content close to users.
- A Local Zone is considered only if a supported workload needs lower latency near a particular city.
- Wavelength is considered only for a mobile 5G use case with an ultra-low-latency requirement.
- Direct Connect links an on-premises system to AWS when dedicated connectivity is required.
- A second Region is added when the recovery, data-residency, or global-availability objective justifies the additional architecture and operating cost.
This design uses different infrastructure components for different failure and latency boundaries. No single component provides every form of availability or performance.
Exam-ready memory aid
Use the question’s strongest keyword:
- Different geographic area or country → Region
- Survive one data-centre failure → multiple Availability Zones
- Cache content near global users → CloudFront edge locations
- Larger cache between points of presence and origin → regional edge cache
- Lower latency near a metropolitan area → Local Zone
- 5G and ultra-low mobile latency → Wavelength Zone
- Dedicated connection from on premises → Direct Connect
Conclusion
AWS Global Infrastructure is a set of distinct placement and connectivity layers. Regions provide geographic separation, AZs provide isolated locations within a Region, edge locations support global delivery, and regional edge caches improve CloudFront caching efficiency. Local Zones extend selected resources towards cities, Wavelength Zones place infrastructure at the 5G edge, and Direct Connect locations support dedicated network connections into AWS.
For the Cloud Practitioner exam, identify the failure boundary, user location, network requirement, or latency requirement first. Then choose the smallest AWS infrastructure concept that directly answers the scenario.