What is Cloud-Computing?
Cloud computing is on-demand access to computing resources (servers, storage, databases, networking, software) over the internet with pay-as-you-go pricing. Provider comparison: Cloud-Provider Comparison. Provider sheets: Amazon Web Services, Google Cloud, Hetzner Cloud, DigitalOcean, Cloudflare, Oracle Cloud, Backblaze B2, Civo.
Service Models #
| MODEL | YOU MANAGE | PROVIDER MANAGES | EXAMPLES |
|---|---|---|---|
| IaaS (Infrastructure) | OS, runtime, app, data | Hardware, virtualisation, network | EC2, Azure VMs, Compute Engine, Civo instances |
| PaaS (Platform) | App and data | Everything below | Heroku, App Engine, Azure App Service, Fly.io |
| SaaS (Software) | Your usage and data | Everything | Microsoft 365, Gmail, Salesforce |
| FaaS (Functions) | Function code | Servers, scaling | AWS Lambda, Cloud Functions, Azure Functions |
| CaaS (Containers) | Containers and config | Orchestration, nodes | EKS, GKE, AKS, Civo Kubernetes |
Deployment Models #
| MODEL | DESCRIPTION |
|---|---|
| Public | Shared infrastructure run by a provider |
| Private | Dedicated to one organisation, on-premises or hosted |
| Hybrid | Public and private connected together |
| Multi-cloud | Using more than one public provider |
Shared Responsibility #
The provider secures the cloud (hardware, facilities, core services). You secure what you put in it: identities, data, configuration, patching of anything you run.
Key Concepts #
| TERM | MEANING |
|---|---|
| Region | Geographic area containing several data centres |
| Availability Zone | Isolated data centre (or group) within a region |
| Elasticity | Scale up and down automatically with demand |
| Pay-as-you-go | Billed per use, usually per second, hour or GB |
| Reserved / Savings plan | Discount for committing 1 to 3 years |
| Spot / Preemptible | Cheap spare capacity that can be reclaimed |
| Egress | Data leaving the provider, usually the surprise cost |
| IaC | Infrastructure as Code, see Terraform |
| Serverless | No servers to manage, billed per invocation |
| Object storage | Flat, API-accessed storage (S3 style) |
| Block storage | Disk volumes attached to VMs |
| CDN | Edge caching close to users |
| VPC | Your private virtual network inside a provider |
Pros and Cons #
| PROS | CONS |
|---|---|
| No upfront hardware spend | Costs can grow unnoticed |
| Fast provisioning and global reach | Vendor lock-in |
| Built-in resilience and managed services | Egress and support fees |
| Elastic scaling | Shared-responsibility mistakes (open buckets, weak IAM) |
Good Habits #
- Tag every resource for cost tracking.
- Set budget alerts on day one.
- Use least-privilege IAM and MFA on the root account.
- Define infrastructure in code (Terraform) and keep it in Git.
- Plan for a second AZ for anything important.
- Delete what you no longer use.