Cloud & Deployment
How apps go from your computer to the internet
Introduction
You have written the code, tested it locally, and your app works perfectly on your own computer. But how does it get from your machine to the internet where millions of users can access it? That is what deployment is all about.
In the early days of the web, companies had to buy, set up, and maintain their own physical servers — racks of computers in a data centre. Today, most apps run on the cloud, which means renting computing power from companies like Amazon (AWS), Microsoft (Azure), or Google (Google Cloud).
The cloud has transformed how software is built and delivered. Instead of buying hardware upfront, developers can spin up servers in minutes, scale to millions of users, and pay only for what they use. This chapter explores how deployment works and the different cloud services available.
How It Works
Deployment is the process of copying your application code to a server that is connected to the internet 24/7. In a modern CI/CD pipeline, the process is automated: you push code to GitHub, the pipeline runs tests and builds your app, then deploys it to a cloud server. DNS (Domain Name System) routes your domain name (like myapp.com) to the server's IP address so users can find it.
Apartment Analogy
On-premise is like owning a house — you buy the land, build the structure, and handle all maintenance yourself. Cloud (IaaS) is like renting an apartment — the landlord handles the building maintenance, but you still furnish and manage your unit. PaaS is like a fully furnished apartment — everything is set up for you, just move in with your code. Serverless is like a hotel room — you pay per night, someone else cleans, and you can check out anytime.
Deeper Dive
A modern deployment follows these steps, often automated through a CI/CD pipeline:
Write Code
Develop locally on your machine
Push to Git
Upload code to GitHub/GitLab
CI/CD Pipeline
Tests run, app is built
Deploy to Cloud
Copy to AWS/Azure/Google
DNS Routes
Users reach your app
Cloud Service Models
Not all cloud services are the same. They range from raw infrastructure to fully managed platforms:
VPS / IaaS
Virtual Private Server. You rent a virtual machine and have full control over the OS, software, and security. Examples: AWS EC2, DigitalOcean, Linode.
PaaS
Platform as a Service. You just deploy your code — the platform manages servers, scaling, and updates. Examples: Heroku, Vercel, Google App Engine.
Serverless
You write functions that run on-demand. No server management at all — you pay only when your code runs. Examples: AWS Lambda, Cloudflare Workers.
Key Insight
"The cloud" is just someone else's computer. But the real value of cloud computing is not the hardware — it is the services built around it: automatic scaling, managed databases, load balancers, monitoring, and global content delivery networks (CDNs) that make apps fast everywhere.
Advanced
At a deeper level, cloud & deployment involves rules and patterns that engineers use worldwide. Cloud follows standards so different brands and devices can still work together. That is why your phone, school laptop, and game console can all connect to the same network or use the same apps.
Deployment does not happen in a straight line. Systems often use backup paths, error checking, and retries so information arrives correctly. When something fails, smart Server design helps the system recover instead of shutting down completely.
Scientists and engineers keep improving these systems every year — making them faster, safer, and more energy-efficient. The ideas you learn in this chapter are the same building blocks used in real data centers, robots, apps, and websites around the world.
Vocabulary Table
| Term | Definition |
|---|---|
| Cloud | On-demand computing resources (servers, storage, databases) delivered over the internet |
| Deployment | The process of making an application available for users on a server |
| Server | A computer that provides data or services to other computers over a network |
| CI/CD | Continuous Integration / Continuous Deployment — automated pipeline for testing and deploying code |
| DNS | Domain Name System — translates domain names (google.com) into IP addresses |
| Hosting | Providing the infrastructure and services to make a website or app accessible online |
| Scalability | The ability of a system to handle increasing amounts of traffic or data |
Fun Facts
AWS (Amazon Web Services) launched in 2006 and now powers over a third of the internet. Netflix, Airbnb, and Slack all run on AWS.
The term "cloud computing" was popularised by Google CEO Eric Schmidt in 2006, but the concept dates back to the 1960s with J.C.R. Licklider's "intergalactic computer network."
Heroku, one of the first PaaS platforms, was created in 2007. It let developers deploy apps with a single "git push" command — revolutionary at the time.
AWS Lambda changed the game in 2014 by introducing serverless computing. You can run code without provisioning a single server — it scales from zero to millions of requests instantly.
Netflix deploys code thousands of times per day using a custom CI/CD tool called Spinnaker. If a deployment causes issues, it automatically rolls back in seconds.
Interactive Diagram
Launch the interactive diagram to see the deployment pipeline and cloud services.
Open Interactive DiagramThe interactive diagram for this chapter demonstrates Cloud and Deployment. It shows the deployment pipeline from code commit to production server, with cloud infrastructure.
What to explore:
- click to commit code; watch it go through build, test, and deploy stages; see the app go live on cloud servers
- cloud deployment automates the process of getting code from a developer's machine into production servers where users can access it
Knowledge Check
1. What does CI/CD stand for?
Answer: Continuous Integration / Continuous Deployment
2. Which cloud service model lets you run code without managing any servers?
Answer: Serverless
3. What does DNS do?
Answer: Translates domain names into IP addresses