AWS Certified Cloud Practitioner Training Course

Cloud computing is the on-demand delivery of IT resources over the internet, paid for as you use them. That single sentence carries three ideas the exam returns to constantly, and it is worth pulling them apart before you meet a single service name.

On-demand means you request a resource and get it in minutes, without a purchase order, a delivery, or a rack. Over the internet means the hardware belongs to someone else and lives somewhere else. Paid as you use them means the meter runs while the resource exists and stops when you delete it. If you can hold those three properties in mind, most of the value-proposition questions on this exam answer themselves.

What changes when you stop buying servers

In a traditional data center you buy capacity in advance. You forecast demand for the next three years, buy hardware to cover the peak of that forecast, and then own it whether you use it or not. Two failures follow from that model, and both are expensive.

• If you over-provision, you have paid for capacity that sits idle. The money is already spent and cannot be recovered.
• If you under-provision, you run out of capacity during exactly the event you most wanted to serve, and the fix takes weeks.

AWS calls the alternative "stop guessing capacity." You provision what you need now, add more when demand rises, and remove it when demand falls. The forecast still matters for budgeting, but a bad forecast is no longer a physical fact you are stuck with.

The six advantages AWS publishes

AWS states the value proposition of the cloud as six advantages. Learn them as a list, because exam items frequently quote one of them almost verbatim and ask you to name it.

• Trade capital expense for variable expense. Instead of a large up-front purchase of servers, you pay only for the computing resources you consume.
• Benefit from massive economies of scale. AWS aggregates the usage of hundreds of thousands of customers, buys hardware and power at that scale, and passes lower per-unit costs on.
• Stop guessing capacity. Scale up and down as demand changes, rather than committing years ahead.
• Increase speed and agility. Resources that used to take weeks to acquire take minutes, so the cost of trying an idea and being wrong drops sharply.
• Stop spending money running and maintaining data centers. Racking, stacking, powering, and cooling servers is undifferentiated work that does not distinguish your business.
• Go global in minutes. Deploy into additional geographic Regions with the same tooling you used for the first one, and serve customers there with lower latency.

Capital expense and variable expense

Two accounting terms show up in the wording of exam questions, so define them plainly. A capital expense, or capex, is money spent up front to acquire an asset you will own and depreciate — a server, a storage array, a building. A variable expense, or opex in the way the exam uses it, is money spent in proportion to consumption, recorded in the period it occurs. The cloud shifts spending from the first category toward the second. Notice the word toward: commitments such as Reserved Instances and Savings Plans deliberately reintroduce some up-front predictability in exchange for a discount, which you will meet in the billing module.

The three service models

Cloud services are conventionally grouped by how much of the stack the provider operates. The exam expects you to recognize the three names and place an AWS service into the right one.

• Infrastructure as a service (IaaS) gives you the basic building blocks — virtual machines, storage volumes, networking — and leaves the operating system and everything above it to you. Amazon EC2 is the standard example.
• Platform as a service (PaaS) removes the operating system and runtime from your workload. You supply an application or a database schema; the provider patches and operates what runs underneath. Amazon RDS and AWS Elastic Beanstalk sit here.
• Software as a service (SaaS) delivers a finished application that you configure and use rather than deploy. Amazon WorkSpaces and most AWS Marketplace applications are consumed this way.

These are labels on a spectrum, not sharp boundaries, and AWS does not require you to classify every service. What it does require is the underlying instinct: the further up the stack the provider operates, the less operational work you carry and the less low-level control you have. That trade-off is the entire subject of the shared responsibility model in the next module.

Deployment models

Alongside service models, the exam names three deployment models. A cloud deployment runs entirely on cloud infrastructure with no on-premises component. An on-premises deployment — sometimes called private cloud — runs in your own data center, possibly using virtualization and automation that resemble cloud tooling, but on hardware you own. A hybrid deployment connects the two, typically because data must stay in a specific building for latency or regulatory reasons while the rest of the application runs in AWS.

What the exam is actually testing here

This domain looks like marketing language and is not. CLF-C02 asks about elasticity, agility, and economies of scale directly, and it asks in the form of a short scenario: a company describes a problem, and you pick the cloud characteristic or benefit that addresses it. A retailer that cannot serve traffic on its busiest day of the year is a capacity-guessing problem answered by elasticity. A team waiting six weeks for hardware to test an idea is an agility problem. A finance director objecting to a large hardware purchase is the capex-to-opex shift. Train yourself to read the scenario for the underlying complaint, then name the benefit that removes it.