Cloud concepts and global infrastructure

Foundational notes for AWS Certified Cloud Practitioner (CLF-C02). Read them in order the first time through, then use them as a revision sweep before you book.

15 study points. Everything here is exam-oriented: each point is a fact or a distinction that CLF-C02 items are built on. Test yourself against the practice exam once you can explain a section without re-reading it.

Cloud Computing / Networks

An Availability Zone (AZ) is a distinct, physically isolated data center location within an AWS Region. Each AZ has its own power supply, networking, and connectivity, so if one AZ experiences a failure, the others keep running. Think of a Region as a city and each AZ as an independent building in that city. By deploying your application across multiple AZs, you ensure that a single hardware failure or natural disaster won’t take your entire system offline.

When we say a system is “highly available,” we mean it is designed to remain operational and accessible virtually all of the time. AWS achieves this by letting you spread your resources across multiple Availability Zones and Regions. For example, if you run a website and one server goes down, a highly available setup automatically routes traffic to a healthy server. Your users never even notice the interruption.

The AWS Well-Architected Framework defines principles of sound design for cloud workloads. Two of the most important principles are scaling horizontally (adding more small machines rather than one giant machine) and stopping the practice of guessing your capacity requirements. In a traditional data center you might buy ten servers hoping they’d be enough; in the cloud, you start with what you need and let AWS scale for you automatically.

AWS has three levels of infrastructure, and the count at each level follows a pattern: there are more Edge Locations than Availability Zones, and more Availability Zones than Regions. Regions are large geographic areas like US-East or EU-West. Each Region contains multiple AZs, and scattered around the globe are hundreds of Edge Locations that cache content close to end users. Understanding this hierarchy helps you decide where to deploy resources for the best performance and resilience.

Cloud computing has three defining characteristics you should remember: pay-as-you-go pricing, on-demand delivery of resources, and access to services over the internet. Unlike traditional IT where you purchase hardware upfront, the cloud lets you pay only for what you consume, spin up resources in minutes, and manage everything through a web browser or API. This model eliminates the need for large capital expenditures and long procurement cycles.

A fault-tolerant system is one that continues to operate correctly even when some of its components fail. This goes a step beyond high availability by ensuring there is zero interruption to the user experience. For example, AWS S3 stores your files across multiple devices in multiple facilities, so even if a hard drive fails, your data is still safe and accessible. When designing on AWS, you build fault tolerance by using redundant components and automated failover.

Amazon CloudFront is AWS’s Content Delivery Network (CDN) service. It caches copies of your content at Edge Locations around the world so that users receive data from the server nearest to them, reducing latency dramatically. CloudFront also includes built-in DDoS protection through AWS Shield Standard at no extra cost. For example, if your website serves images to users in Tokyo and Paris, CloudFront ensures each user loads those images from a nearby Edge Location rather than from your origin server in Virginia.

AWS offers three cloud deployment models: Public Cloud, Private Cloud, and Hybrid Cloud. A Public Cloud means all resources run on AWS infrastructure shared among many customers. A Private Cloud (sometimes called on-premises cloud) keeps everything within your own data center using tools like VMware or AWS Outposts. A Hybrid Cloud combines both, letting you keep sensitive workloads on-premises while using AWS for everything else. Many enterprises start with a hybrid approach during their migration journey.

A NAT Gateway allows resources in a private subnet to access the internet for tasks like downloading software updates, while preventing the internet from initiating connections back to those resources. Think of it like a one-way door: your private servers can reach out, but no one on the internet can reach in. This is essential for security because databases and application servers should never be directly exposed to the public internet.

An Internet Gateway is the component that connects your Virtual Private Cloud (VPC) to the public internet. Without it, nothing in your VPC can communicate with the outside world. You attach one Internet Gateway per VPC, and it acts as the front door of your network. Resources in public subnets use the Internet Gateway to serve traffic, while resources in private subnets use a NAT Gateway that routes through it.

AWS offers several load balancers, each suited to different needs. The Classic Load Balancer performs simple round-robin distribution for HTTP and HTTPS traffic and is the legacy option. The Application Load Balancer (ALB) operates at Layer 7 and supports advanced routing rules like path-based or host-based routing. The Network Load Balancer handles millions of requests per second at ultra-low latency. For most new applications, the ALB is the recommended choice.

Auto Scaling is made up of three components working together: Launch Configurations (or Launch Templates) define what type of instance to create, Auto Scaling Groups define where and how many instances to run, and Auto Scaling Policies define when to add or remove instances based on metrics like CPU usage. For example, if your web application needs to handle traffic spikes during Black Friday, Auto Scaling automatically adds more servers as demand rises and removes them when traffic subsides, saving you money.

Elasticity is the ability of your infrastructure to automatically expand and contract based on demand. When traffic surges, AWS can divert requests to newly launched instances or route users to the instances with the least load. When demand drops, those extra instances are terminated so you stop paying for them. This is one of the most powerful advantages of cloud computing over traditional infrastructure, where you would need to manually provision and decommission servers.

Amazon SQS (Simple Queue Service) is a fully managed message queuing service that helps you decouple the components of your application. Instead of one service directly calling another, the first service places a message on a queue and the second service picks it up when it’s ready. This means if one part of your system is temporarily overwhelmed, messages simply wait in the queue rather than causing cascading failures. SQS is reliable, highly scalable, and a cornerstone of building resilient architectures.

Amazon Lightsail is AWS’s simplified Platform-as-a-Service offering, designed for developers who want to launch a small website, blog, or application without learning the full complexity of AWS. It bundles compute, storage, and networking into a single monthly price, much like traditional web hosting. Elastic Beanstalk is another managed service that lets you deploy applications written in .NET, Java, Python, and other languages while still giving you control over the underlying instance type and database type.

Where to go next

Back to the AWS Cloud Practitioner overview.
Look up any service you could not name in the AWS services glossary.
Sit the 80-item practice exam once two or three note pages are solid.