Chapter 26: AWS Infrastructure

AWS Infrastructure (officially called AWS Global Infrastructure or AWS Infrastructure Regions & Availability Zones).

This is the physical & logical foundation everything else sits on — EC2, S3, Lambda, VPC, databases, you name it. If you don’t understand AWS Infrastructure, it’s like trying to build a house without knowing where the land is, how big the plot is, or where the power lines run.

Let me explain it like we’re sitting together with a big world map on the table — slow, step-by-step, lots of analogies, real Hyderabad/India examples, the current 2026 numbers, why it matters for performance/security/cost, and a simple way to “see” it in the console.

1. What is AWS Infrastructure? (The Big Picture – No Jargon First)

AWS Infrastructure = the worldwide network of physical data centers + networking that AWS owns and operates to run all its cloud services.

It’s not one big building — it’s thousands of buildings (data centers) spread across the planet, grouped in smart ways so your app can be:

• Fast (low latency for users in Hyderabad)
• Reliable (doesn’t die if one building floods or loses power)
• Compliant (data stays in India if laws require it)
• Scalable (add capacity anywhere in seconds)

Think of it like India’s electricity grid:

• Power plants = data centers
• States/regions = AWS Regions
• Local substations = Availability Zones
• Transmission lines = super-fast private fiber network
• Local transformers near your home = Edge Locations / CloudFront PoPs

You don’t see the grid — you just flip a switch and get power. Same with AWS Infrastructure — you choose a Region → AWS handles the rest.

2. The Three Main Layers of AWS Infrastructure (2026 Numbers)

3. Deep Dive: AWS Regions (The Most Important Choice You Make)

A Region = a completely separate geographic area with its own isolated power, networking, and physical security.

• Each Region has multiple AZs (minimum 3, some 4–6).
• Most services are Region-specific (your S3 bucket, EC2 instance, RDS database lives in one Region).
• Regions are chosen for:
  • Latency — closer = faster
  • Compliance — data residency (India DPDP Act, RBI rules)
  • Cost — some Regions cheaper
  • Service availability — not every service in every Region yet

India Regions in 2026 (very relevant for us):

Real example: Your Zomato-like app in Hyderabad

• Choose ap-south-2 (Hyderabad) → users in Telangana/AP get <15 ms response
• Data stays in India → complies with RBI/data residency rules
• Cheaper data transfer costs inside India vs Singapore Region

4. Deep Dive: Availability Zones (The High-Availability Secret)

An Availability Zone = one or more physically separate data centers inside a Region — far enough apart to survive floods, power outages, fires, but close enough for ultra-low-latency private fiber (metro network).

• Minimum 3 AZs per Region (most have 3–6).
• AZs are connected with redundant, high-speed fiber — you can spread your app across AZs.
• Naming: ap-south-2a, ap-south-2b, ap-south-2c (letter = AZ ID)

Analogy: AZs are different buildings in the same city (Hyderabad):

• Banjara Hills AZ → Gachibowli AZ → Hi-Tech City AZ
• Flood in one area → other buildings keep running
• Traffic between them is super-fast (private network)

Real example (multi-AZ setup):

• Your food delivery app runs:
  • 3 EC2 instances (one in each AZ) behind ALB
  • RDS database with Multi-AZ (primary in 2a, standby in 2b)
• Power failure in one data center → ALB stops routing to that AZ → RDS fails over in ~60 seconds → app stays up.

This is why AWS promises 99.99%+ availability — you design for it.

5. Edge Locations / Points of Presence (PoPs) – Speed Layer

Edge Locations = small AWS facilities (not full data centers) used mainly for:

• Amazon CloudFront (CDN) — cache images/videos close to users
• AWS Global Accelerator — faster routing
• Lambda@Edge — run code at edge

India has 20+ Edge Locations (Mumbai, Hyderabad area, Chennai, Delhi, Bangalore, Kolkata, etc.).

Example: User in Hyderabad watches Telugu short video

• Original video in S3 (ap-south-2 Hyderabad Region)
• CloudFront caches copy at Hyderabad Edge Location
• User gets video in <1 second instead of 50–100 ms round-trip to Region

6. Quick Hands-On: See the Infrastructure in Console

1. Log in to AWS Console
2. Top-right → Region dropdown → see all Regions (select ap-south-2 Hyderabad)
3. Launch EC2 → see AZ choices (a, b, c)
4. Create S3 bucket → note it’s Region-specific
5. Search “CloudFront” → create distribution → see global Edge Location map

Summary Table – AWS Infrastructure Cheat Sheet (Feb 2026)

Bottom line: AWS Infrastructure = the global, resilient, layered foundation that makes the cloud fast, reliable, and compliant. You mostly choose Region → spread across AZs → use Edge for speed.

Got it? This is the “map” you draw first before building anything on AWS.

Next?

• Why choose Hyderabad Region vs Mumbai (latency + cost deep dive)?
• How to design Multi-AZ high-availability app?
• Or latest 2026 Regions & AZ announcements?

Tell me — next class ready! 🚀🌍