Cloud Fundamentals: Understanding the Foundation of Modern Infrastructure
📅 Published: Feb 2026
⏱️ Estimated Reading Time: 16 minutes
🏷️ Tags: Cloud Computing, IaaS, PaaS, SaaS, AWS, Azure, GCP
Introduction: What is Cloud Computing?
The Simple Explanation
Cloud computing is the delivery of computing services over the internet. Instead of buying and maintaining your own physical servers, you rent computing power, storage, and applications from a provider like Amazon Web Services (AWS), Microsoft Azure, or Google Cloud.
Think of it like electricity. Before the electrical grid, every factory had to generate its own power. They built power plants, maintained generators, and paid for capacity whether they used it or not. When the grid arrived, factories could plug in and pay only for what they used.
Cloud computing is the same transformation for technology. Instead of buying servers, you rent them. Instead of predicting capacity years in advance, you scale up and down in minutes. Instead of managing physical hardware, you focus on building applications.
The Traditional Way vs. The Cloud Way
Traditional data centers required you to:
Buy servers years before you needed them
Pay for full capacity even when idle
Hire staff to maintain hardware
Wait weeks for new servers to arrive
Manage cooling, power, and physical security
The cloud changes everything:
Provision servers in minutes, not weeks
Pay only for what you use
No hardware to maintain
Scale up during demand, scale down when quiet
Access services from anywhere
The Five Essential Characteristics
The National Institute of Standards and Technology (NIST) defines five essential characteristics of cloud computing:
1. On-demand self-service
You can provision computing resources without human interaction with the service provider. Through a web portal or API, you can spin up servers, add storage, or configure networking instantly.
2. Broad network access
Resources are available over the network and accessible through standard protocols. You can access your cloud resources from anywhere with an internet connection, using any device.
3. Resource pooling
The provider pools computing resources to serve multiple customers. Physical and virtual resources are dynamically assigned based on demand. You don't know exactly which physical server your workload is running on, and you don't need to.
4. Rapid elasticity
Resources can be scaled up or down quickly, often automatically. If your application suddenly gets millions of users, the cloud can add more servers. When traffic drops, it can remove them. To you, capacity appears unlimited.
5. Measured service
Cloud systems automatically control and optimize resource usage. You are billed only for what you use, often by the hour or even by the second. Usage can be monitored, reported, and billed with transparency.
IaaS, PaaS, SaaS: The Three Service Models
Cloud services are delivered in three primary models. They represent different levels of control and responsibility.
Infrastructure as a Service (IaaS)
IaaS provides the fundamental building blocks of computing: virtual servers, storage, and networking. You are responsible for everything above the infrastructure—the operating system, middleware, applications, and data.
Think of IaaS as renting an empty apartment. You get the walls, floors, and utilities. You bring your own furniture, paint, and decide how to arrange it.
Examples:
Amazon EC2 (virtual servers)
Amazon S3 (object storage)
Google Compute Engine
Azure Virtual Machines
You manage:
Operating systems
Applications
Data
Runtime
Middleware
Provider manages:
Virtualization
Servers
Storage
Networking
Data center facilities
Best for:
Lift-and-shift migrations
Applications requiring custom configurations
Workloads with specific compliance requirements
Teams that want maximum control
Platform as a Service (PaaS)
PaaS provides a platform for developing and deploying applications without managing the underlying infrastructure. You focus on your code; the platform handles servers, operating systems, and runtime environments.
Think of PaaS as renting a fully furnished apartment. The furniture, appliances, and utilities are included. You bring your personal belongings (your code) and arrange them how you like.
Examples:
AWS Elastic Beanstalk
Google App Engine
Azure App Service
Heroku
You manage:
Applications
Data
Provider manages:
Runtime
Middleware
Operating system
Virtualization
Servers
Storage
Networking
Best for:
Web applications and APIs
Development teams wanting to focus on code
Applications that scale automatically
Teams without dedicated infrastructure expertise
Software as a Service (SaaS)
SaaS delivers complete applications over the internet. You don't manage anything—you just use the software. The provider handles everything from infrastructure to application features.
Think of SaaS as staying in a hotel. Everything is provided. You simply use the room and amenities. You don't worry about maintenance, utilities, or furniture.
Examples:
Google Workspace (Gmail, Docs, Drive)
Microsoft 365
Salesforce
Dropbox
Slack
You manage:
User accounts
Configuration settings
Data you create
Provider manages:
Everything else—application, runtime, middleware, operating system, virtualization, servers, storage, networking
Best for:
Standard business applications
Teams wanting to avoid any infrastructure management
Applications where customization is not critical
The Responsibility Spectrum
As you move from IaaS to PaaS to SaaS, your control decreases but your management burden also decreases.
| Responsibility | On-Premises | IaaS | PaaS | SaaS |
|---|---|---|---|---|
| Applications | You | You | You | Provider |
| Data | You | You | You | You (mostly) |
| Runtime | You | You | Provider | Provider |
| Middleware | You | You | Provider | Provider |
| OS | You | You | Provider | Provider |
| Virtualization | You | Provider | Provider | Provider |
| Servers | You | Provider | Provider | Provider |
| Storage | You | Provider | Provider | Provider |
| Networking | You | Provider | Provider | Provider |
No model is universally better. The right choice depends on your team's expertise, control requirements, and business needs.
Public vs Private vs Hybrid Cloud
Public Cloud
Public cloud is the most common model. Services are delivered over the public internet and shared across multiple customers. Each customer's data is logically isolated, but the physical infrastructure is shared.
Examples: AWS, Microsoft Azure, Google Cloud Platform
Advantages:
No capital expenditure (pay as you go)
Nearly unlimited scale
No hardware to manage
Global footprint available instantly
Disadvantages:
Less control over physical infrastructure
Potential compliance challenges for regulated industries
Shared infrastructure (though logically isolated)
Costs can become unpredictable at scale
Best for:
Startups and growing companies
Applications with variable demand
Development and testing environments
Companies without data center expertise
Private Cloud
Private cloud is cloud computing dedicated to a single organization. It can be hosted on-premises or by a third-party provider, but the infrastructure is not shared.
Examples: VMware Cloud, OpenStack, AWS Outposts, Azure Stack
Advantages:
Complete control over infrastructure
Predictable costs at scale
Easier compliance with regulations
No noisy neighbors (shared resources)
Disadvantages:
High capital expenditure
You manage hardware lifecycle
Limited elasticity (you can't exceed physical capacity)
Requires specialized staff
Best for:
Regulated industries (finance, healthcare, government)
Organizations with predictable, stable workloads
Companies with existing data centers
Workloads requiring complete isolation
Hybrid Cloud
Hybrid cloud combines public and private cloud, allowing data and applications to move between them. This model gives you the best of both worlds: the security of private cloud for sensitive workloads and the scale of public cloud for variable demand.
Examples: AWS with on-premises, Azure Arc, Google Anthos
Advantages:
Flexibility to place workloads where they make sense
Burst to public cloud during peak demand
Keep sensitive data on-premises
Gradual migration path to cloud
Disadvantages:
Complex networking and security
Requires expertise in both environments
Higher operational complexity
Integration challenges
Best for:
Enterprises with existing data centers
Organizations with compliance requirements that don't fit all workloads
Companies migrating gradually to cloud
Workloads with variable demand that must keep some data on-premises
Multi-Cloud
A related concept is multi-cloud—using multiple public cloud providers simultaneously. This is different from hybrid cloud, which combines public and private.
Examples: Using AWS for compute, Google Cloud for analytics, and Azure for Microsoft workloads
Advantages:
Avoid vendor lock-in
Use best-of-breed services from each provider
Geographic redundancy across providers
Negotiation leverage with providers
Disadvantages:
Significant operational complexity
Multiple security models to manage
Higher staffing requirements
Data transfer costs between clouds
Best for:
Large enterprises with diverse workloads
Organizations with specific compliance requirements across regions
Teams with expertise across multiple clouds
Regions & Availability Zones
Why Geography Matters
Cloud providers build data centers all over the world. These data centers are grouped into regions and availability zones. Understanding this geography is essential for designing resilient, low-latency applications.
Regions
A region is a geographic area that contains multiple availability zones. Regions are completely independent of each other. A failure in one region does not affect other regions.
Examples:
US East (N. Virginia) — us-east-1
US West (Oregon) — us-west-2
EU (Ireland) — eu-west-1
Asia Pacific (Singapore) — ap-southeast-1
Choosing a region:
Latency: Choose regions close to your users
Cost: Prices vary by region
Compliance: Some data must stay in specific geographic boundaries
Service availability: New services launch in some regions before others
Disaster recovery: Use multiple regions for redundancy
Availability Zones
An availability zone is one or more data centers within a region. Zones are physically separate but connected by low-latency links.
Key characteristics:
Each zone has independent power, cooling, and networking
Zones are miles apart (preventing correlated failures)
Latency between zones is low (milliseconds)
You can run applications across zones for high availability
Example: us-east-1 region
us-east-1a (one or more data centers)
us-east-1b (one or more data centers)
us-east-1c (one or more data centers)
us-east-1d, 1e, 1f (additional zones)
Designing for Resilience
Single Availability Zone
A single server running in one availability zone. If that zone has a power outage, your application goes down. Suitable for development and testing.
Multiple Availability Zones (Within One Region)
Servers running in two or more zones. If one zone fails, traffic routes to the others. This is the standard for production workloads.
Multiple Regions
Infrastructure running in two or more regions. If an entire region has a major failure, your application continues running elsewhere. Necessary for mission-critical applications.
Service Types by Resilience
| Service Category | Behavior | Examples |
|---|---|---|
| Global | Runs across all regions | CloudFront, Route 53, IAM |
| Regional | Runs within a region, across zones | S3, VPC, Lambda |
| Zonal | Runs within a single zone | EC2 instances, EBS volumes |
Understanding these distinctions helps you design applications that survive failures.
Real-World Cloud Scenarios
Scenario 1: Startup Launch
A startup is launching a mobile app. They expect unpredictable traffic. They have a small team and no data center expertise.
Recommended approach:
Use public cloud (AWS, Google Cloud, or Azure)
Use PaaS or serverless to minimize infrastructure management
Deploy across multiple availability zones for resilience
Use auto-scaling to handle traffic spikes
Monitor costs carefully as they grow
Scenario 2: Financial Services Company
A bank wants to move some workloads to cloud but must comply with strict regulations. Customer data cannot leave the country.
Recommended approach:
Use private cloud for sensitive customer data
Use public cloud for development and non-sensitive workloads
Deploy in the specific region required by regulations
Use hybrid cloud to connect environments
Implement strict security controls and auditing
Scenario 3: E-commerce Platform
An online retailer experiences huge traffic spikes during holiday seasons but steady traffic the rest of the year. They have existing data centers.
Recommended approach:
Hybrid cloud: maintain baseline capacity on-premises
Burst to public cloud during peak seasons
Use auto-scaling to handle variable demand
Deploy across availability zones for availability
Use content delivery networks for static assets
Cloud Service Providers
Amazon Web Services (AWS)
AWS is the market leader with the largest number of services and the broadest global footprint.
Strengths:
Largest ecosystem of services
Most mature with longest track record
Best documentation and community support
Broadest geographic coverage
Common services:
Compute: EC2, Lambda, ECS, EKS
Storage: S3, EBS, EFS
Database: RDS, DynamoDB, Aurora
Networking: VPC, CloudFront, Route 53
Microsoft Azure
Azure is the second-largest provider, with strong enterprise integration, especially for Microsoft shops.
Strengths:
Excellent integration with Microsoft products
Strong hybrid cloud capabilities
Enterprise-friendly contracting
Good for Windows workloads
Common services:
Compute: Virtual Machines, App Service, AKS
Storage: Blob Storage, File Storage
Database: SQL Database, Cosmos DB
Networking: Virtual Network, Front Door
Google Cloud Platform (GCP)
GCP leverages Google's expertise in data, machine learning, and global networking.
Strengths:
Leadership in data analytics and AI/ML
Best-in-class networking infrastructure
Strong Kubernetes offerings
Transparent pricing
Common services:
Compute: Compute Engine, GKE, Cloud Run
Storage: Cloud Storage, Persistent Disk
Database: Cloud SQL, BigQuery, Spanner
Networking: VPC, Cloud CDN, Cloud Load Balancing
Cloud Computing Vocabulary
| Term | Meaning |
|---|---|
| Elasticity | The ability to scale resources up and down automatically |
| Scalability | The ability to handle increasing load by adding resources |
| High Availability | Designing systems to remain operational even when components fail |
| Disaster Recovery | The ability to recover from catastrophic failures |
| Consumption-based pricing | Pay only for what you use, no upfront costs |
| CAPEX | Capital expenditure (buying hardware upfront) |
| OPEX | Operational expenditure (paying for usage over time) |
| SLA | Service Level Agreement—the guaranteed availability level |
| Serverless | Running code without provisioning or managing servers |
Summary
Cloud computing transforms how we build and operate technology. The key concepts to remember:
IaaS, PaaS, SaaS represent different levels of control and responsibility
Public, private, hybrid, multi-cloud represent different deployment models
Regions and availability zones provide the physical infrastructure for resilience
Elasticity and consumption-based pricing are fundamental advantages of cloud
Whether you are a developer, operations engineer, or architect, understanding these fundamentals is essential for modern technology work.
Practice Questions
A startup needs to deploy a web application with unpredictable traffic. They have no data center staff. Which cloud model is most appropriate?
A hospital needs to store patient records that must remain in the country and cannot be shared with other organizations. Which deployment model is most appropriate?
An e-commerce company has existing data centers but needs to handle holiday traffic spikes. Which deployment model is most appropriate?
A development team wants to deploy code without managing servers. Which service model is most appropriate?
You need to design an application that remains available even if one data center fails. What cloud architecture should you use?
Learn More
Practice cloud fundamentals with hands-on exercises in our interactive labs:
https://devops.trainwithsky.com/
Comments
Post a Comment