Cloud Learning Path for Beginners: A Structured Sequence That Works

Principles behind this path

A few things informed the sequence below:

• Foundations before platforms. Cloud platforms build on Linux and networking concepts. Learning AWS before you understand either means encountering explanations that assume knowledge you do not have.
• One platform deeply, not three broadly. Pick AWS, GCP, or Azure. Stick with it. The concepts transfer; shallow knowledge of all three transfers to nothing.
• Build at every stage. Each stage includes a practical output — something you built — not just something you studied. Your portfolio grows alongside your knowledge.
• Certification as a milestone, not a destination. Certifications mark checkpoints of organised knowledge. They are not the goal; a job is the goal.

Stage 0: Before you start — set yourself up correctly

Two practical things to do before any studying:

Set up a cloud free tier account

Create an account with your chosen cloud provider: AWS Free Tier, GCP Free Tier, or Azure Free Account. Set a billing alert immediately — configure it to notify you if spend exceeds £5 or $5. Free tier resources are genuinely free, but misconfigured resources can generate charges. A billing alert is your safety net.

Install a Linux environment

Download VirtualBox (free) and install Ubuntu 22.04 LTS as a virtual machine on your computer. Alternatively, use Windows Subsystem for Linux if you are on Windows. You need a real Linux environment to practise on, not a web-based terminal simulator.

Stage 1: Linux fundamentals

Duration: 4–6 weeks | Hours per week: 10–15

What to learn

• Filesystem navigation: ls, cd, pwd, find, absolute vs relative paths
• File operations: cat, less, cp, mv, rm, mkdir
• File permissions: chmod, chown, understanding octal notation
• Text editing: nano to start, then vim basics (open, insert, save, quit)
• Processes: ps aux, top, kill, systemctl
• Package management: apt-get install, keeping software up to date
• SSH: generating keys, connecting to remote machines, understanding key-based authentication
• Bash scripting: variables, loops, conditionals, functions, reading command output

Practical output

Write a Bash script that monitors disk usage and logs a warning if any partition exceeds 80% capacity. Deploy it on your local Ubuntu VM with a cron job. This exercises real skills: scripting, process management, cron, file operations.

Resources

• The Linux command line book by William Shotts (free online at linuxcommand.org)
• Official Ubuntu documentation for specific commands
• Your VM — do everything in the terminal

Stage 2: Networking fundamentals

Duration: 3–4 weeks | Can overlap with Stage 1 weeks 3–6

What to learn

• IP addresses: IPv4 notation, what each octet means, private vs public ranges
• CIDR notation: /24, /16, /8 — how to calculate the number of addresses in a range
• Subnets: what a subnet is, why you divide a network, routing between subnets
• DNS: what it is, how resolution works, A records, CNAME records, TTL
• HTTP vs HTTPS: what a request looks like, status codes (200, 301, 404, 500), TLS certificates
• Firewalls: inbound vs outbound rules, stateful vs stateless
• Load balancers: what they do, Layer 4 vs Layer 7, health checks
• NAT: what it is, why it exists, how cloud VPCs use NAT gateways

Practical output

Draw a network diagram of a simple web application: a public load balancer, web servers in a private subnet, a database in a separate private subnet, a NAT gateway for outbound internet access. This is a common interview question. Being able to draw and explain this architecture demonstrates you understand cloud networking fundamentals.

Stage 3: Cloud platform fundamentals

Duration: 6–8 weeks

What to learn

Focused on your chosen platform. Core areas in all three:

• Compute: virtual machines, managed container services, serverless functions
• Storage: object storage, block storage, file storage
• Databases: managed relational databases, managed NoSQL options
• Networking: VPCs, subnets, security groups/firewall rules, load balancers
• Identity and Access Management: users, roles, policies, service accounts
• Billing and cost management: free tier limits, billing alerts, cost explorer

Practical output

Deploy a simple three-tier web application manually through the console: a load balancer, a virtual machine running a web server, and a managed database. Connect them, verify the application works, and then — critically — delete everything cleanly and document exactly what you built. This exercise forces you to understand how the layers connect.

Certification milestone

Take your introductory certification exam: AWS Cloud Practitioner, GCP Cloud Digital Leader, or Azure AZ-900. These broad exams confirm you understand the landscape. They are not technically deep, but they are worth having.

Stage 4: Infrastructure as code with Terraform

Duration: 4–6 weeks

What to learn

• HCL syntax: resources, data sources, variables, outputs, locals
• The Terraform workflow: init, validate, plan, apply, destroy
• State: what the state file is, why it matters, remote backends (S3 or GCS)
• Modules: how to organise code into reusable modules, input/output variables
• Provider documentation: how to find resource arguments in the provider docs
• Workspaces and environments: how to manage dev/staging/prod

Practical output

Re-create the three-tier application from Stage 3 entirely in Terraform. No console clicks — everything defined in code. The Terraform code should live in a Git repository with a meaningful commit history. Add a README explaining the architecture.

Stage 5: Containers and Kubernetes

Duration: 5–7 weeks

Docker (2–3 weeks)

• What a container is and how it differs from a VM
• Writing Dockerfiles: base image, copy, run, expose, entrypoint
• Building and tagging images
• Running containers: ports, environment variables, volumes
• Pushing to a registry: Docker Hub or a cloud container registry
• Docker Compose for local multi-service development

Kubernetes (3–4 weeks)

• Core objects: pods, deployments, services, namespaces
• Writing YAML manifests
• kubectl commands: apply, get, describe, logs, exec
• ConfigMaps and Secrets
• Health checks: liveness and readiness probes
• Resource requests and limits
• Using a managed Kubernetes service (GKE, EKS, or AKS)

Practical output

Containerise an application (write the Dockerfile), push it to a cloud container registry, and deploy it to a managed Kubernetes cluster using Kubernetes manifests. The manifests should be in a Git repo. Add horizontal pod autoscaling.

Stage 6: CI/CD automation

Duration: 3–4 weeks

What to learn

• What CI/CD is and why it exists
• GitHub Actions: workflow syntax, triggers, jobs, steps, runners
• Building a Docker image in a pipeline and pushing to a registry
• Deploying infrastructure changes via Terraform in a pipeline
• Managing secrets in pipelines — GitHub encrypted secrets, never hardcoded credentials
• Environment separation: dev deploys on every push; production deploys on manual approval

Practical output

Connect your Terraform project and Kubernetes deployment from stages 4 and 5 into a CI/CD pipeline. On a push to the main branch: run terraform plan, require manual approval, run terraform apply, and deploy the updated container to Kubernetes. This is a realistic representation of how infrastructure changes work in production teams.

Stage 7: Associate certification and job applications

Duration: 8–12 weeks

Certification

Work toward your associate-level certification: AWS Solutions Architect – Associate, GCP Associate Cloud Engineer, or Azure AZ-104. Use official practice exams to measure readiness. Book the exam when consistently scoring 80%+ on practice tests.

Portfolio review

Review your GitHub repositories. Each project should have a clear README, clean commit history, and sensible file organisation. Annotate your architecture choices — why did you structure your Terraform modules this way? Why did you choose that Kubernetes health check configuration? Documented reasoning demonstrates engineering judgment.

Start applying

Target: junior cloud engineer, cloud support engineer, junior DevOps engineer, infrastructure engineer. Apply before the certification exam results arrive — you can update the application once you have the result. Track all applications in a spreadsheet with notes on what was asked at each stage.

See entry-level cloud jobs explained for a full breakdown of what roles to target and what each one involves.

What comes after your first role

The path above is a path to your first job, not the end of learning. Once you are in a role:

• Deepen knowledge of the cloud platform your employer uses — you will learn more from real production systems than from any course
• Learn from your more experienced colleagues — observe how they approach incidents and architecture decisions
• Consider specialist certifications after 12–18 months: Kubernetes CKA, Terraform Associate, AWS DevOps Professional
• Build a track record of delivered work that you can articulate clearly in future interviews

For the full roadmap beyond your first role, see the career roadmaps section (coming soon).