Guides

DevOps Roadmap 2026: What to Learn in Order

By Ravindra Bagale, DevOps trainer · LinkedIn · Published: 28 October 2026 · Last updated: 28 October 2026

Why a DevOps roadmap in 2026? Because job posts mix twenty tools, and freshers freeze. This guide gives one clear order — Linux, Git, Docker, CI/CD, one cloud, Kubernetes, Terraform, monitoring — with weeks and exit checks used in real hiring at companies like Infosys, TCS and product teams.

Language: English · मराठी · हिंदी

Why do you need a roadmap?

Recruiters at Infosys, TCS, Amazon and Netflix-style product teams do not hire a logo collection. They hire someone who can keep a service alive. Without an order, students jump to Kubernetes on day three and cannot explain systemctl when the app is down.

Why: Order stops panic learning. What: A sequence with weeks and an exit check for each stage. How: Follow the stages below. Do not skip the exit check.

What the 2026 path looks like

Eight stages. Each stage has a clear exit check.

1. Linux and shell
2. Git and code review habits
3. Docker images and Compose
4. One CI/CD tool (GitHub Actions or Jenkins)
5. One cloud (AWS is enough for most Indian job posts)
6. Kubernetes basics
7. Terraform (infrastructure as code)
8. Monitoring and a tiny alert

Concept 1: Linux first (weeks 1-2)

Why: Every server you touch still speaks Linux. What: Users, SSH keys, systemctl, disk and logs. How: Practice on one Amazon Linux 2023 or Ubuntu lab VM you own.

Example 1: See who you are and the kernel

Type this:

whoami
uname -r

You will see:

ec2-user
6.1.119-129.201.amzn2023.x86_64

What it means: You are logged in as a normal user on Amazon Linux 2023.

Example 2: Check disk before an outage

Type this:

df -h /

You will see:

Filesystem      Size  Used Avail Use% Mounted on
/dev/nvme0n1p1   8.0G  3.1G  4.9G  39% /

What it means: Root disk is 39 percent full. Above 90 percent needs cleanup.

Example 3: Is nginx running?

Type this:

systemctl is-active nginx

You will see:

active

What it means: The service is up. If failed, open journalctl next.

Concept 2: Git every day (weeks 2-3)

Why: Mystery zip files break Infosys-style client releases. What: status, add, commit, log, branch, pull request habit. How: One tiny repo on your laptop. Commit every practice change.

Example 4: Identity once per machine

Type this:

git config --global user.name "Priya Sharma"
git config --global user.email "priya@example.com"
git config --global --get user.name

You will see:

Priya Sharma

What it means: Git stores your name on every commit.

Example 5: First commit on a lab folder

Type this:

mkdir -p ~/lab-roadmap && cd ~/lab-roadmap
printf 'hello\n' > README.md
git init -b main
git add README.md
git commit -m "docs: start roadmap lab"

You will see:

[main (root-commit) a1b2c3d] docs: start roadmap lab
 1 file changed, 1 insertion(+)

What it means: You now have history. The hash will differ on your machine.

Example 6: Read history

Type this:

git log --oneline -3

You will see:

a1b2c3d docs: start roadmap lab

What it means: One line per commit. Interviewers often ask you to read a log aloud.

Concept 3: Docker packs the same app (weeks 4-5)

Why: "Works on my laptop" is the fight between developers and operations. What: Dockerfile, image, container, Compose for two services. How: Containerize a static site, then a small API.

Example 7: Confirm Docker Engine

Type this:

docker version --format '{{.Server.Version}}'

You will see:

27.3.1

What it means: Server version means the engine is running. 24+ is fine.

Example 8: Run a tiny web container

Type this:

docker run --rm -d --name roadmap-web -p 8080:80 nginx:1.27-alpine
curl -s -o /dev/null -w '%{http_code}\n' http://127.0.0.1:8080/

You will see:

7f3a9c2e1b0d
200

What it means: Nginx answers HTTP 200 on port 8080.

Example 9: List running containers

Type this:

docker ps --format 'table {{.Names}}\t{{.Status}}\t{{.Ports}}'

You will see:

NAMES         STATUS         PORTS
roadmap-web   Up 20 seconds  0.0.0.0:8080->80/tcp

What it means: Name, status and published ports — ops checks these first.

Concept 4: One CI pipeline (weeks 6-7)

Why: Humans forget to run tests before Friday night. What: On every push, a runner checks out code and runs the same commands. How: Start with GitHub Actions or Jenkins — pick one.

Example 10: Local stand-in for a CI check

Type this:

cd ~/lab-roadmap
printf 'ok\n' > health.txt
test -f health.txt && echo CI_CHECK_PASS

You will see:

CI_CHECK_PASS

What it means: A pipeline is the same commands on a clean machine.

Example 11: Fail the check on purpose

Type this:

test -f missing.txt && echo CI_CHECK_PASS || echo CI_CHECK_FAIL

You will see:

CI_CHECK_FAIL

What it means: Red builds teach you to read logs.

Example 12: Record what CI should run

Type this:

printf '%s\n' '#!/bin/sh' 'set -eu' 'test -f README.md' 'echo pipeline_ok' > ~/lab-roadmap/ci-check.sh
chmod 700 ~/lab-roadmap/ci-check.sh
sh ~/lab-roadmap/ci-check.sh

You will see:

pipeline_ok

What it means: Mode 700 keeps the script private. Never use world-writable modes.

Concept 5: One cloud — AWS is enough (weeks 8-9)

Why: Most Indian MNC job descriptions list AWS. What: IAM user (not root), one EC2, one S3 bucket, billing alarm. How: Free Tier lab account. Turn on a billing alarm on day one.

Example 13: Who am I in AWS CLI?

Type this:

aws sts get-caller-identity --query Account --output text

You will see:

123456789012

What it means: A 12-digit account id means your CLI credentials work. Never paste real access keys into chats.

Example 14: List regions

Type this:

aws ec2 describe-regions --query 'Regions[0:3].RegionName' --output text

You will see:

ap-south-1  eu-west-1   us-east-1

What it means: ap-south-1 (Mumbai) is the usual first region for India-based labs.

Example 15: head-bucket safely

Type this:

aws s3api head-bucket --bucket demo-name-not-yours-example 2>&1 | head -1

You will see:

An error occurred (404) when calling the HeadBucket operation: Not Found

What it means: Create buckets only in accounts you own.

Concept 6-8: Kubernetes, Terraform, monitoring (weeks 10-16)

Why: Job posts stack these words — but only after Linux and Docker. What: K8s Deployment+Service locally; Terraform one resource; one alert. How: One small app through all three. Do not build a marketplace on week ten.

Example 16: kubectl context

Type this:

kubectl config current-context 2>/dev/null || echo no-cluster-yet

You will see:

kind-kind

What it means: A context name means kubectl knows a cluster.

Example 17: Terraform version

Type this:

terraform version | head -1

You will see:

Terraform v1.9.8

What it means: v1.x is fine. Pin versions in real teams.

Example 18: Disk alert mind-set

Type this:

python3 -c "used=91; print('ALERT_DISK' if used>=90 else 'ok')"

You will see:

ALERT_DISK

What it means: Monitoring is a threshold plus a message.

Week plan at a glance

Common mistakes

1. Skipping Linux because containers hide the OS — they do not hide outages.
2. Learning five CI tools — pick one.
3. Fake resume lines — build the lab, then write the line.
4. Ignoring cost — always set a billing alarm on AWS.

Suppose we are…

Infosys client project

The client asks who can restart nginx and read logs. Roadmap weeks 1-2 answer that.

Netflix-style releases

Many small updates need Git + CI + the same image everywhere — weeks 2-7.

Flipkart / Amazon peak traffic

Kubernetes and monitoring matter — but only after you can debug one VM.

Check yourself (quiz)

Roadmap = order + exit checks. Linux and Git before Kubernetes. One CI tool. One cloud. Then K8s, Terraform and one alert. Did you understand? Let us try it once more.