Cloud · Cloud Fundamentals
Cloud Computing
On demand delivery of compute, storage, networking and platform services over a network, billed by consumption and managed through APIs.
FoundationalUpdated 2026-09-01
Overview
Cloud computing replaces capital purchase and manual provisioning with programmable, elastic services. The defining characteristics, on demand self service, broad network access, resource pooling, rapid elasticity and measured service, come from NIST SP 800-145 and remain a useful test of whether something is genuinely cloud.
The important consequence for IT teams is not cost or elasticity but the control plane: infrastructure becomes an API. That makes automation natural and makes identity, permissions and audit logging the primary security controls.
How it works
- 01Regions are geographic deployments; availability zones are physically separate facilities within a region used for high availability.
- 02Resources are grouped for lifecycle and permission scope, subscriptions and resource groups in Azure, accounts and organisational units in AWS, projects and folders in Google Cloud.
- 03Everything is created and configured through APIs, which is why infrastructure as code and policy as code are the normal operating model.
- 04Costs accrue per resource per hour or per request, so unused resources are a direct, silent expense.
Why it matters
- Provisioning time drops from weeks to minutes, changing how projects are planned and how mistakes propagate.
- Security shifts from perimeter and hardware toward identity, configuration and control plane monitoring.
- Financial governance becomes an engineering responsibility because engineers can now spend money with an API call.
Reference table
| Model | You manage | Provider manages | Typical examples |
|---|---|---|---|
| IaaS | OS, runtime, apps, data, identity config | Hypervisor, hardware, facilities, network fabric | Virtual machines, managed disks, virtual networks |
| PaaS | Application code, data, access policy | OS, runtime, patching, scaling | App services, managed databases, functions |
| SaaS | Data, users, configuration | Everything else | Microsoft 365, Salesforce, ServiceNow |
Security considerations
- Understand the shared responsibility model per service model, the provider never manages your permissions or your data classification.
- Protect the control plane with strong administrative identity, just in time privilege and centralised audit logs.
- Assume misconfiguration is the dominant risk: public storage, permissive network rules and over broad roles cause most cloud incidents.
Common misconfigurations
- Long lived static access keys used by automation and shared between environments.
- Production and non production resources in the same account or subscription with the same permissions.
- Landing zone skipped, so naming, tagging, policy and network design are retrofitted later at high cost.