Cloud Concepts
24% of exam • Fundamental AWS principles and design
The 6 Pillars of AWS Well-Architected Framework
Design principles for building secure, high-performing, resilient, and efficient infrastructure
1. Operational Excellence
Run and monitor systems to deliver business value and continually improve processes
Key practices: IaC, frequent small changes, automate operations
2. Security
Protect information, systems, and assets while delivering business value
Key practices: IAM, encryption, least privilege, detective controls
3. Reliability
Ensure workload performs its intended function correctly and consistently when expected
Key practices: Multi-AZ, auto-recovery, change management, testing
4. Performance Efficiency
Use computing resources efficiently to meet requirements and maintain efficiency as demand changes
Key practices: Right-sizing, serverless, monitoring, experimentation
5. Cost Optimization
Avoid unnecessary costs and maximize return on investment
Key practices: Pay for what you use, Reserved Instances, analyze spend
6. Sustainability
Minimize environmental impact of running cloud workloads
Key practices: Maximize utilization, use managed services, reduce waste
Exam Tip:
Memorize all 6 pillars - questions may ask about Well-Architected Framework principles
Scaling Types:
Vertical Scaling (Scale Up/Down)
One bigger box
Change the size of a resource - resize to a bigger or smaller instance type
Visual:
1 vCPU
1 GB RAM
4 vCPU
16 GB RAM
✓ Simple to implement
✗ Limited by hardware (can't scale infinitely)
✗ Requires downtime (must stop/start instance)
Example: Upgrade RDS database from db.t3.small to db.t3.large for more CPU
Horizontal Scaling (Scale Out/In)
More boxes (distributed)
Add or remove resources - increase the number of instances
Visual:
Instance
✓ Highly available (distribute across AZs)
✓ Nearly unlimited scaling
✓ No downtime
Example: Auto Scaling adds EC2 instances behind a load balancer during Black Friday traffic
Exam Tip - Remember the Visual:
• Vertical = 📦 → 📦 (one small box becomes one BIG box)
• Horizontal = 📦 → 📦📦📦📦 (one box becomes MANY boxes)
Elasticity vs Scalability:
Elasticity
Automatically and dynamically match capacity to demand in REAL-TIME
Resources scale up during high demand and scale down during low demand
Keywords: "automatic," "dynamic," "real-time," "on-demand"
Example: Auto Scaling automatically adds EC2 instances at 9 AM when users log in, removes them at 6 PM when traffic drops
Scalability
Ability to accommodate growth OVER TIME (long-term capacity)
System is designed to handle increased load as your business grows
Keywords: "handle growth," "long-term," "accommodate"
Example: Architecture designed to support 1,000 users today can grow to support 1 million users next year
High Availability (HA)
System remains operational with minimal downtime
- • Uses multiple Availability Zones
- • Load balancers distribute traffic
- • Example: Application running in 3 AZs with ELB
Fault Tolerance
System continues operating even when components fail (zero downtime)
- • Built-in redundancy at all levels
- • Automatic failover
- • Example: S3 automatically replicates across AZs
Disaster Recovery (DR)
Strategy to recover from catastrophic failure
RTO
Recovery Time Objective
How long until system is back?
RPO
Recovery Point Objective
How much data loss is acceptable?
- Trade capital expense for variable expense: Pay only for what you use
- Benefit from massive economies of scale: Lower pay-as-you-go prices
- Stop guessing capacity: Scale up/down as needed
- Increase speed and agility: Resources available in minutes
- Stop spending money on data centers: Focus on customers, not infrastructure
- Go global in minutes: Deploy to multiple regions instantly
Regions:
Geographic area with multiple isolated Availability Zones
Availability Zones (AZ):
One or more discrete data centers with redundant power, networking, and connectivity
Edge Locations:
Data centers for caching content closer to users (CloudFront CDN)
- Vertical vs Horizontal: Vertical = bigger box, Horizontal = more boxes
- Elasticity: Look for keywords like "automatic," "dynamic," "real-time demand"
- HA vs Fault Tolerance: HA = minimal downtime, Fault Tolerance = zero downtime
- RTO vs RPO: RTO = time to recover, RPO = data loss tolerance
- AZ best practice: Always deploy across at least 2 AZs for high availability