DevOps Foundation Roadmap

Twelve sequential steps from Git to cloud delivery the standard path before specializing in security or SRE.

DevOps Foundation Roadmap

The canonical DevOps engineer path: version control → automation → containers → cloud → CI/CD → observability. Security awareness is woven in early; full DevSecOps depth comes in the dedicated DevSecOps roadmap.

The twelve steps

Step	Topic	Time box	Outcome
1	Git & collaboration	1 week	PR workflow on a shared repo
2	Python or Go basics	2 weeks	Automation script in repo
3	Linux & shell	2 weeks	User management + cron job
4	Networking	2 weeks	Explain DNS + TLS handshake
5	Web servers & proxies	1 week	Nginx reverse proxy lab
6	Docker	2 weeks	Multi-container Compose stack
7	Kubernetes	3 weeks	Deploy + expose an app
8	Terraform / Ansible	2 weeks	Reproducible environment
9	CI/CD	2 weeks	Pipeline on every PR
10	Prometheus / Grafana	2 weeks	Dashboard + one alert
11	AWS / Azure / GCP	4+ weeks	One production-like project
12	DevOps culture	ongoing	Blameless postmortems, DORA metrics

Step details

1 Git

Master branching strategies (trunk-based vs GitFlow for your context). Understand why merge commits differ from rebase. Use signed commits if your org requires them.

2 Programming language

Pick one language and stay with it for six months. DevOps value comes from reading SDKs and writing glue not algorithm contests.

3 Linux

Comfort on Ubuntu or RHEL. Know where logs live, how to debug disk full, and how sudo and groups interact.

4 Networking

Draw your home lab: client → DNS → load balancer → pod. Know common ports (22, 80, 443, 5432, 6379).

5 Server management

Configure Nginx or Caddy with TLS. Understand forward vs reverse proxy and when to use a CDN.

6 Containers

Optimize Dockerfile layer cache. Never run as root in production images. Pin base image digests.

7 Kubernetes

kubectl fluency: debug CrashLoopBackOff, inspect events, port-forward. Learn Helm for packaging, not as a crutch for missing GitOps discipline.

8 Infrastructure as Code

Terraform for provisioning; Ansible (or cloud-init) for configuration. Never edit production by hand without a ticket.

9 CI/CD

Stages: lint → unit test → integration test → build artifact → deploy. Cache dependencies. Fail fast.

10 Observability

RED/USE methods. One SLI per service. Alerts must be actionable if it wakes someone at 3 a.m., it must be fixable.

11 Cloud

Start with free tier. Build: VPC, compute, managed DB, object storage, IAM roles (not long-lived keys).

12 Culture

Read The DevOps Handbook and Accelerate. Measure lead time, deployment frequency, MTTR, change failure rate.

What to build as proof

Artifact	Demonstrates
Public GitHub repo with Actions	CI/CD
Terraform module with README	IaC
Blog post or Research Core note	Communication
Mini K8s demo with ingress + TLS	Platform skills

Next steps

Security-heavy roles → DevSecOps Learning Roadmap
Senior staff track → Senior Platform Gap Checklist