Cloud Concepts: Value, Economics, and the AWS Global Model
Why the cloud changes the economics
Cloud Concepts questions test whether you understand why organizations move to AWS, not just what services exist. The core ideas are elasticity, agility, and the shift from capital expense to variable operating expense. On-premises capacity is purchased ahead of demand and sized for peaks, so it sits idle most of the time. AWS lets you provision capacity in minutes, scale it with demand, and stop paying when you stop using it. Related concepts you must be able to name: economies of scale (AWS aggregates demand and passes savings along), global reach (deploy close to users in minutes), and speed of experimentation (failed pilots cost little).
The Well-Architected pillars
Know the six pillars and what each emphasizes: Operational Excellence (run and improve processes), Security (protect data and systems), Reliability (recover from failure and meet demand), Performance Efficiency (use resources efficiently), Cost Optimization (avoid unnecessary spend), and Sustainability (minimize environmental impact). Questions often describe a scenario and ask which pillar a decision supports. A trap: reliability and performance sound similar, but reliability is about surviving failure, performance is about doing work efficiently.
The global infrastructure model
Regions are separate geographic areas containing multiple isolated Availability Zones. Availability Zones are one or more discrete data centers with independent power, cooling, and networking, connected by low-latency links. Edge locations sit closer to users for content delivery. A common trap is assuming an AZ is a single data center or that regions share failure domains. They do not.
Worked example
A team runs two web servers in one data center and wants to survive a facility outage. Rule of thumb: if the requirement is survive a data center failure, spread across at least two Availability Zones in one Region; if the requirement is survive a regional disaster or serve local users with low latency, use multiple Regions. Match the scope of the failure domain to the scope of the design.