roadmaps / devops-engineer

How to become a DevOps engineer

A practical, stage-by-stage path from beginner to hireable DevOps engineer — the skills, the order to learn them in, and free resources at every step. No prior ops experience or CS degree required.

updated jul 2026·8 stages·~8–14 months at a steady pace

$ how this roadmap works
› Work the stages in order — each one assumes the last. The early stages are the foundation everything else stands on.
› Every resource here is free. Mark a stage passed when you can build something with it, not just read about it.
› DevOps isn't a single tool — it's a culture of automating the path from code to production. Learn the concepts; the tools change.

DevOps engineering is one of the best-paid, most in-demand roles in tech — and one of the few you can reach without a computer-science degree, because it's judged on what you can build and automate. The catch: it's broad. You need a bit of development, a lot of operations, and the glue between them. That breadth is exactly why a roadmap helps: it stops you drowning in tools and gives you a sane order. Below, eight stages take you from "what's a terminal?" to a portfolio that gets interviews.

STAGE 01 / 08

Foundations: Linux & networking

Almost everything in DevOps runs on Linux, and you'll spend your days in a terminal. Get comfortable there first: the file system, permissions, processes, package managers, and the everyday commands (ssh, grep, curl, systemctl). Being fluent at the command line is the single biggest force-multiplier in this career.

Pair it with the basics of networking — you can't operate systems you don't understand. Learn how DNS, HTTP(S), TCP/IP, ports, and load balancers work, and what actually happens when a request travels from a browser to a server and back. When something breaks in production, this is the mental model that lets you debug it.

SKILLSthe Linux command linefiles, permissions, processesSSHDNS / HTTP / TCP-IPports & load balancing

STAGE 02 / 08

Programming, Git & scripting

DevOps is automation, and automation is code. You don't need to be a full-stack developer, but you do need to script confidently — Bash for gluing commands together and Python for anything more involved (parsing, APIs, tooling). If you can automate a boring manual task, you can do this job.

Just as important is Git. Version control is the backbone of every DevOps workflow — pipelines trigger on commits, infrastructure lives in repos, and "GitOps" makes Git the source of truth for your whole system. Learn branching, merging, pull requests, and resolving conflicts until they're second nature.

SKILLSBash scriptingPython basicsGit & branchingpull-request workflow

STAGE 03 / 08

CI/CD pipelines

This is the heart of DevOps. A CI/CD pipeline automatically builds, tests, and ships code every time it changes — turning releases from scary events into routine, boring ones. Understanding this loop (and being able to build one) is what most "DevOps engineer" job descriptions are really asking for.

Start with the concept — why continuous integration reduces risk — then build a real pipeline. GitHub Actions is the easiest on-ramp (it's right in the repo); GitLab CI and Jenkins are everywhere in industry. Make a pipeline that runs tests on every push and deploys on merge, and you've done the core job.

SKILLScontinuous integrationcontinuous deliveryGitHub ActionsGitLab CI / Jenkinspipeline as code

STAGE 04 / 08

Containers & Kubernetes

Docker changed how software is packaged and shipped — a container bundles an app with everything it needs, so "works on my machine" finally means "works everywhere." Learn to write a Dockerfile, build images, and run containers; this is table-stakes for modern DevOps.

Then comes Kubernetes — the system that runs and scales containers across a cluster. It's the most in-demand DevOps skill and has a real learning curve, so take it in steps: pods, deployments, services, then config and scaling. You don't need to master it overnight, but you must be able to deploy an app to a cluster and explain how it works.

SKILLSDocker & Dockerfilescontainer imagesKubernetes pods/deploymentsservices & scaling

STAGE 05 / 08

Infrastructure as Code

Clicking around a cloud console doesn't scale and can't be reviewed. Infrastructure as Code (IaC) fixes that: you describe your servers, networks, and services in files you commit to Git, so your whole environment is repeatable, reviewable, and disaster-recoverable. This is a defining modern DevOps skill.

Learn Terraform to provision infrastructure (create the servers and networks) and Ansible to configure what runs on them. Build something real — a small app deployed entirely from code, no console clicking — and you'll instantly stand out to employers.

SKILLSTerraformAnsibledeclarative infrastate & idempotency

STAGE 06 / 08

Cloud platforms

Almost all modern infrastructure lives in the cloud, and DevOps is where cloud skills pay off most. You don't need all three providers — go deep on one. AWS has the biggest market share and the most jobs, so it's the safe first bet; Azure and Google Cloud are strong alternatives, especially in enterprise and data/ML respectively.

Learn the core building blocks — compute, storage, networking, IAM (permissions), and managed databases — and how they connect. Aim for a foundational certification (like AWS Cloud Practitioner); it's a concrete, résumé-worthy milestone and forces you to learn the whole landscape.

SKILLSAWS / Azure / GCPcompute & storageIAM & networkingmanaged services

STAGE 07 / 08

Monitoring, observability & SRE

Shipping fast is only half the job — you also have to know when something's wrong and why. Monitoring tells you when; observability helps you understand why. The open-source stack is remarkably consistent: Prometheus collects metrics, Grafana visualizes them, and increasingly OpenTelemetry standardizes traces and logs.

Layer on the mindset of Site Reliability Engineering (SRE) — Google's discipline for running production systems. Concepts like SLOs, error budgets, and the "four golden signals" teach you what's actually worth alerting on, so you build systems that page a human only when it matters.

SKILLSPrometheus & Grafanalogging & tracingSLOs & error budgetsalerting

STAGE 08 / 08

DevSecOps, certs & the job hunt

The final layer is security — "DevSecOps" means baking it into the pipeline instead of bolting it on at the end. Learn to scan dependencies, containers, and IaC automatically, and you'll be more valuable (and more employable) than a DevOps engineer who treats security as someone else's problem.

Now make it count. Build a portfolio: a public repo with a real app, containerized, deployed to the cloud from Terraform, with a CI/CD pipeline and monitoring — that single project demonstrates the whole roadmap. Consider a certification (AWS Cloud Practitioner or the Certified Kubernetes Administrator) as a résumé signal, then prep for interviews. You don't need to be an expert in everything — you need to prove you can automate the path from code to production.

SKILLSDevSecOpsa portfolio projectAWS / CKA certsinterview prep

Keep going

Go deep on any stage with the full DevOps & cloud study lists.

DevOps foundations CI/CD Cloud Interview prep Free certifications