Phase Verbs / Action Phrases
| Phrase | Meaning |
|---|---|
| externalize | move something internal outside (e.g., outsource IT) |
| launch (a service) | make a product publicly available |
| expand (into) | grow presence into a new market or geography |
| leverage | use to gain an advantage |
| deploy (into a region) | release and run an application in a specific location |
| cascade (into) | spread or propagate a failure to another system |
| deliver content | send data/media to end users |
Technical Vocabulary
| Term | Definition |
|---|---|
| AWS Region | A cluster of data centers in a specific geographic area (e.g., us-east-1, eu-west-3) |
| Availability Zone (AZ) | One or more discrete data centers within a region with redundant power, networking, and connectivity |
| Data center | Physical facility housing the servers and network equipment |
| Point of Presence (PoP) / Edge Location | AWS infrastructure endpoints distributed globally for low-latency content delivery (400+ in 90 cities, 40 countries) |
| Latency | The delay between a request and response — how “fast” the connection feels |
| Compliance | Meeting legal or regulatory rules about where data must be stored |
| Redundant | Duplicated to prevent failure — if one fails, another takes over |
| High bandwidth | Large capacity for data transfer |
| Ultra-low latency networking | Extremely fast connections between data centers |
| Global service | AWS service not tied to a specific region (e.g., IAM, Route 53, CloudFront, WAF) |
| Region-scoped service | AWS service that exists per region (e.g., EC2, Lambda, Rekognition) |
| SQS | Simple Queue Service — AWS’s first publicly launched service (2004) |
Key Concepts
AWS History
- 2002: Launched internally at Amazon
- 2004: First public service — SQS
- 2006: Added S3 and EC2
- Today: $90B revenue, 31% market share (Q1 2024), 13+ consecutive years as market leader
AWS Regions
- A region = cluster of data centers in one geographic area
- Each region has a code:
us-east-1,ap-southeast-2, etc. - Services used in one region are separate from the same service in another region
How to Choose a Region — 4 Factors
- Compliance — data sovereignty laws may require data to stay in-country (e.g., France)
- Latency — deploy close to your users to reduce lag
- Service availability — not all services exist in all regions
- Pricing — costs vary by region
Availability Zones (AZs)
- Each region has 3–6 AZs (usually 3)
- Example: Sydney (
ap-southeast-2) → AZs:2a,2b,2c - Each AZ = one or more data centers with redundant power & networking
- AZs are isolated from each other (disasters in one don’t cascade to others)
- AZs within a region are connected by high bandwidth, ultra-low latency private network
Points of Presence / Edge Locations
- 400+ PoPs in 90 cities across 40 countries
- Used to deliver content to end users with lowest possible latency
- Powers services like CloudFront (CDN)
Global vs Region-Scoped Services
| Global (no region) | Region-scoped |
|---|---|
| IAM | EC2 |
| Route 53 | Lambda |
| CloudFront | Elastic Beanstalk |
| WAF | Rekognition |
Exam Tips
- Know the difference between Regions, AZs, and Edge Locations
- Know the 4 factors for choosing a region (compliance, latency, availability, pricing)
- AZs exist to provide high availability and fault isolation