What you will learn today
- What cloud computing actually means
- IaaS, PaaS, and SaaS
- Public, private, and hybrid cloud
- The six advantages (benefits) of AWS Cloud
- Regions, Availability Zones, and Edge Locations
- The wider AWS global infrastructure
- Elasticity, scalability, agility, and high availability
- AWS Well-Architected Framework introduction
- How AWS can be accessed
1. What is cloud computing?
Cloud computing is the on-demand delivery of IT resources (compute, storage, databases, networking, analytics, AI) over the internet with pay-as-you-go pricing.
Five ideas sit inside that sentence, and exam questions often test them:
| On-demand self-service | You create resources yourself in minutes, with no sales call or purchase order |
| Broad network access | Resources are reachable over the internet or private connections |
| Resource pooling | AWS serves many customers from shared, secure infrastructure (multi-tenant) |
| Rapid elasticity | Resources scale out and in quickly |
| Measured service | Usage is metered, and you pay only for what you use |
Analogy: Owning servers is like owning a car: you pay upfront, maintain it, insure it, and it sits idle most of the day. The cloud is like Uber: you pay only for the ride you take.
Traditional IT vs cloud
| Cost model | Large upfront CapEx | Pay-as-you-go OpEx |
| Capacity planning | Guess years ahead | Scale on demand |
| Time to get a server | Weeks or months | Minutes |
| Maintenance | You manage hardware, power, cooling | AWS manages the data centers |
| Global reach | Build or rent data centers | Deploy in many Regions in minutes |
| Risk of wrong guess | Idle hardware, or outages if too small | Minimal; scale up or down |
CapEx vs OpEx
| CapEx (Capital Expenditure) | Upfront spending on physical assets | Buying servers |
| OpEx (Operational Expenditure) | Ongoing spending as you consume | Paying for EC2 per hour |
2. Cloud service models: IaaS, PaaS, SaaS
| IaaS (Infrastructure as a Service) | OS, runtime, apps, data | Hardware, networking, virtualization | Amazon EC2, EBS, VPC | Renting an empty kitchen |
| PaaS (Platform as a Service) | Application code and data | OS, runtime, scaling, patching | AWS Elastic Beanstalk, Amazon RDS | Renting a kitchen with tools and ingredients |
| SaaS (Software as a Service) | Only your usage and settings | Everything | Amazon Connect, Amazon Q, Gmail, Salesforce | Ordering a finished meal at a restaurant |
| FaaS / Serverless | Only your function code | Servers, OS, scaling | AWS Lambda | Paying a taxi per ride, not owning a car |
Exam tip: More control means more responsibility. IaaS gives the most control. SaaS gives the least control and the least management work.
3. Cloud deployment models
1
2
3
4
5
6
7
8
9
10
What is your situation?
|
+------------------------+------------------------+
| | |
Everything runs Everything runs in Some in cloud,
in the cloud your own data center some on-premises
| | |
PUBLIC CLOUD PRIVATE / ON-PREMISES HYBRID CLOUD
(AWS, Azure, GCP) (your servers, or (VPN, Direct Connect,
private cloud) Outposts, Storage Gateway)
| Public cloud (cloud-native) | All resources run on a cloud provider | Speed, scale, low cost | All AWS services |
| Private cloud / on-premises | All resources in your own data center | Strict control, legacy systems | AWS Outposts can bring AWS to your site |
| Hybrid cloud | Mix of cloud and on-premises, connected | Gradual migration, data residency, low latency | Direct Connect, Site-to-Site VPN, Storage Gateway, Outposts, Systems Manager |
4. The six advantages of AWS Cloud (heavily tested)
| 1 | Trade fixed expense for variable expense | No upfront CapEx; pay only for what you use | Run a test for 3 hours and pay for 3 hours |
| 2 | Benefit from massive economies of scale | AWS aggregates usage from hundreds of thousands of customers, giving lower pay-as-you-go prices | AWS has repeatedly lowered prices |
| 3 | Stop guessing capacity | Scale up or down to match demand | No idle servers after a sale ends |
| 4 | Increase speed and agility | Provision resources in minutes; experiment cheaply | A new test environment in 10 minutes |
| 5 | Stop spending money running and maintaining data centers | Focus on business, not racks, power and cooling | Developers build features, not cable runs |
| 6 | Go global in minutes | Deploy close to customers worldwide | Launch in Mumbai and Frankfurt the same day |
Memory trick (FEGASG): Fixed to variable, Economies of scale, Guess capacity, Agility, Stop running data centers, Global in minutes.
Elasticity, scalability, agility, high availability, and fault tolerance
| Scalability | Ability to handle a growing workload by adding resources | "Handle growth" | Move to a bigger instance (scale up) or add more instances (scale out) |
| Elasticity | Resources automatically expand and shrink with demand | "Automatically grows and shrinks", "pay only for what you use" | EC2 Auto Scaling adds servers at peak and removes them at night |
| Agility | Speed to provision, experiment, and deliver | "Quickly", "minutes" | Spin up a test environment in minutes |
| High availability (HA) | System stays accessible even when a component fails, usually through redundancy across AZs | "Minimal downtime", "multiple AZs" | App in two AZs behind a load balancer |
| Fault tolerance | System keeps operating with no interruption during failure (stronger than HA) | "No interruption at all" | Multi-AZ database with automatic failover |
| Durability | Data will not be lost | "11 nines" | Amazon S3 |
| Resilience / reliability | Recover quickly from failures | "Recover" | Backups plus automation |
Analogy: Elasticity is a rubber band that stretches and shrinks by itself. Scalability is the ability to build a bigger room. High availability is carrying a spare tire. Fault tolerance is a car that keeps driving even with a flat tire.
5. AWS Global Infrastructure
1
2
3
4
5
6
7
8
9
10
11
12
AWS GLOBAL INFRASTRUCTURE
REGION (e.g., Mumbai ap-south-1)
+-----------------------------------------------------------+
| AZ-a (ap-south-1a) | AZ-b (ap-south-1b) | AZ-c (...1c) |
| +---------------+ | +---------------+ | +-----------+ |
| | Data center(s)| | | Data center(s)| | | Data ctr | |
| +---------------+ | +---------------+ | +-----------+ |
+--------------------- low-latency private fiber ------------+
|
EDGE LOCATIONS (hundreds, in many cities)
CloudFront and Route 53 cache and answer close to users
| Region | A physical geographic area with multiple, isolated AZs (for example Mumbai, Tokyo, N. Virginia) | Data stays in the Region unless you move it | "Data residency", "compliance", "latency" |
| Availability Zone (AZ) | One or more discrete data centers with independent power, cooling and networking, connected with high-bandwidth, low-latency links | Protects against a single data center failure | "High availability" means multiple AZs |
| Edge Location | Site used by Amazon CloudFront, Route 53, and Global Accelerator to serve content near users | Lower latency for end users | "Cache content close to users" |
| Regional Edge Cache | Larger cache between origin and edge locations | Keeps content cached longer | CloudFront detail |
| Local Zones | Extension of a Region that places compute, storage and databases near large cities | Single-digit millisecond latency for specific cities | "Latency-sensitive apps near a metro area" |
| Wavelength Zones | AWS services inside telecom 5G networks | Ultra-low latency for mobile and 5G | "5G" |
| AWS Outposts | AWS-managed racks installed in your data center | Hybrid, local processing | "AWS infrastructure on-premises" |
| AWS Direct Connect locations | Facilities where you connect a dedicated line | Private connectivity | See Day 5 |
Facts to remember: AWS publishes the current number of Regions, AZs and edge locations on its infrastructure page and keeps adding more, so don't memorize counts. Remember the relationships: a Region contains several AZs (generally three or more), and edge locations are separate from Regions.
Analogy: A Region is a city. AZs are separate neighborhoods with their own power grids. Edge locations are convenience stores that stock popular items close to your home.
How to choose a Region (very common question)
1
2
3
4
5
6
7
8
9
Which Region should I use?
|
1. Compliance / data residency -- laws require data in a country? --> choose that Region
|
2. Latency to users -- closest Region wins
|
3. Service availability -- not every service is in every Region
|
4. Pricing -- prices differ by Region
| 1 | Compliance and data governance | Do regulations restrict where data can live? |
| 2 | Latency | Where are my users? |
| 3 | Service availability | Is the new service offered in that Region? |
| 4 | Cost | Is it cheaper in another allowed Region? |
Design for availability and resilience
| Survive a data center failure | Deploy across multiple AZs in one Region |
| Survive a whole Region failure (disaster recovery) | Deploy or replicate to a second Region |
| Serve users worldwide with low latency | Use CloudFront (edge locations) and multiple Regions |
| Run AWS services on your premises | Outposts |
Exam trap: AZs are inside a Region. Edge locations are not inside a Region. Multi-AZ gives high availability, multi-Region gives disaster recovery and global reach.
6. AWS Well-Architected Framework (introduction)
The Well-Architected Framework is AWS's collection of best practices for designing and operating good cloud workloads. It has six pillars.
| Operational Excellence | Run and monitor systems, improve continuously | Automate and learn from failures | Infrastructure as code (CloudFormation), runbooks |
| Security | Protect data, systems and assets | Least privilege, encryption, traceability | IAM, MFA, KMS, CloudTrail |
| Reliability | Recover from failure and meet demand | Test recovery, scale horizontally | Multi-AZ, backups, Auto Scaling |
| Performance Efficiency | Use resources efficiently as demand changes | Right resource for the job; go serverless | Right-size instances, use caching |
| Cost Optimization | Avoid unnecessary spending | Pay only for what you need; measure | Savings Plans, Spot, delete unused resources |
| Sustainability | Minimize environmental impact | Maximize utilization, reduce waste | Efficient instances, right-sizing |
Memory trick: Over Secure Reliable Performance Cost Sustain: O, S, R, P, C, S.
General design principles (AWS's own list)
| Stop guessing capacity | Scale on demand |
| Test systems at production scale | Create and delete test environments cheaply |
| Automate to make experimentation easier | Infrastructure as code |
| Allow evolutionary architectures | Change and improve continuously |
| Drive architectures using data | Use metrics to decide |
| Improve through game days | Practice failure scenarios |
| Design for failure | Assume anything can break; use loosely coupled components |
| Treat servers as disposable (cattle, not pets) | Replace instead of repair |
The AWS Well-Architected Tool is a free tool in the console that reviews your workload against the pillars and gives improvement guidance.
Analogy: Building a house. Strong locks are Security, a backup generator is Reliability, an energy-efficient design is Sustainability, and a monthly budget is Cost Optimization.
7. How do you interact with AWS?
| AWS Management Console | Web-based graphical interface | Learning, one-off tasks |
| AWS CLI (Command Line Interface) | Commands from a terminal | Scripting and automation |
| AWS SDKs | Libraries for languages such as Python, Java, JavaScript | Using AWS from code |
| AWS CloudShell | Browser-based shell with CLI preinstalled | Quick commands without setup |
| AWS CloudFormation / CDK | Infrastructure as code | Repeatable environments |
| Mobile Console App | Monitor resources from your phone | On-the-go checks |
Programmatic access uses an access key ID and secret access key (or better, temporary credentials from a role).
8. Exam traps for Day 1
| "Elasticity" and "scalability" are the same | Scalability = handle growth. Elasticity = automatically grow and shrink |
| A Region equals an AZ | A Region contains multiple AZs |
| Edge locations are inside Regions | They are separate, global points of presence |
| Hybrid means "cloud only" | Hybrid = cloud plus on-premises |
| Cloud always means lower cost | It means variable cost and the chance to optimize |
| PaaS means you manage the OS | In PaaS, the provider manages the OS |
| High availability needs multi-Region | Multi-AZ is enough for HA within a Region |
9. Key takeaways
- Cloud = on-demand IT resources, pay-as-you-go.
- IaaS (most control) → PaaS → SaaS (least control).
- Know all six advantages by heart.
- Multi-AZ = high availability. Multi-Region = disaster recovery.
- Region choice order: compliance, latency, service availability, price.
- Memorize the six Well-Architected pillars.
10. Day 1 practice questions
| 1 | A company wants to avoid large upfront hardware purchases. Which cloud advantage applies? | Trade fixed expense for variable expense |
| 2 | Which concept lets resources automatically grow and shrink with demand? | Elasticity |
| 3 | Which cloud model gives the customer the least management responsibility? | SaaS |
| 4 | Which AWS component is used to cache content close to end users? | Edge locations (CloudFront) |
| 5 | A company must keep customer data inside India. What is the first factor in choosing a Region? | Compliance and data residency |
| 6 | How do you design an application to survive a single data center failure? | Deploy across multiple Availability Zones |
| 7 | Which service lets you run AWS infrastructure in your own data center? | AWS Outposts |
| 8 | Which Well-Architected pillar focuses on recovering from disruptions? | Reliability |
| 9 | Which Well-Architected pillar includes right-sizing to reduce wasted spending? | Cost Optimization |
| 10 | Which free tool reviews your workload against AWS best practices? | AWS Well-Architected Tool |
Up next on Day 2: AWS Security, IAM and the Shared Responsibility Model, the highest-weighted domain on the exam.
Loading article