15 DevOps Roadmap Milestones Every Engineer Should Hit
15 DevOps milestones from Linux to platform engineering, each with a pass test and portfolio evidence, grouped for freshers, mid-level and senior.
A DevOps roadmap milestone is a checkpoint you can prove, not a topic you have read about. The 15 milestones here run from Linux and Git through containers, Kubernetes, infrastructure as code, observability, security and incident response. Each comes with a pass test and a piece of evidence, such as a repository, a pipeline or a write-up, that shows you reached it.
Plenty of roadmaps list what to learn. Our DevOps roadmaps for beginners compares those learning paths, and 12 steps to become a DevOps engineer from scratch lays out the order to study in. This page answers a different question: how do I know I am actually done with a stage and ready for the next? Use it as a self-assessment checklist, from fresher through senior.
Key takeaways
- Treat each milestone as a test you can pass or fail, with something you can show an interviewer.
- Milestones 1 to 5 are the foundation for a first DevOps or cloud job; 6 to 10 match roughly 2 to 5 years of experience; 11 to 15 are senior and lead skills.
- A portfolio of small, working projects proves more than a list of tools on a CV.
- Certifications such as RHCSA, CKA or AWS associate exams can confirm a milestone, but they do not replace the evidence.
- You do not need to finish every milestone before applying for jobs. Milestones 1 to 6 with evidence are enough for many entry-level roles.
The 15 milestones at a glance
| # | Milestone | Stage | Evidence to keep |
|---|---|---|---|
| 1 | Linux administration without notes | Foundation | Server build script and a troubleshooting log |
| 2 | Networking you can debug | Foundation | Write-up of a DNS or connectivity issue you traced |
| 3 | Shell and one programming language | Foundation | Scripts repository with README and tests |
| 4 | Git for team work | Foundation | Pull requests with reviews, a resolved conflict |
| 5 | First working CI pipeline | Foundation | Pipeline file plus a green run history |
| 6 | Containers built properly | Early career | Small, non-root, scanned images |
| 7 | One cloud platform end to end | Early career | Architecture diagram and teardown script |
| 8 | Infrastructure as code with state and modules | 2 to 5 years | Terraform repo with remote state and modules |
| 9 | Kubernetes operated, not just used | 2 to 5 years | Cluster upgrade notes, manifests or Helm chart |
| 10 | Continuous delivery with safe rollouts | 2 to 5 years | GitOps repo and a recorded rollback |
| 11 | Observability with SLOs | Senior | Dashboards, alert rules and an SLO document |
| 12 | Security built into the pipeline | Senior | Scan gates, secrets management, policy rules |
| 13 | Incident response and postmortems | Senior | A blameless postmortem you wrote |
| 14 | Cost and capacity awareness | Senior | Cost report with actions taken |
| 15 | Platform thinking and mentoring | Lead | Internal docs, templates and people you trained |
Foundation milestones (first job)
1. You can administer Linux without notes
Pass test: given a fresh VM, you create users and groups, set file permissions and ACLs, manage services with systemd, read logs with journalctl, configure a firewall, and fix a service that fails to start, without searching for each command.
Evidence: a short build script and a log of three problems you diagnosed. RHCSA is the certification that tests these skills in a hands-on exam. Useful reading: Linux permissions for DevOps engineers.
2. You can debug basic networking
Pass test: you explain what happens when a browser loads a URL (DNS, TCP, TLS, HTTP), and you can use dig, ss, curl -v, ip route and traceroute to find whether a failure is DNS, routing, firewall or the application.
Evidence: a write-up of one real issue you traced from symptom to cause. See ports and protocols every cloud engineer should know.
3. You can automate with shell and one language
Pass test: you write a Bash script with error handling (set -euo pipefail, exit codes, logging) and a Python or Go tool that calls an API, parses JSON and handles failures.
Evidence: a repository of scripts with READMEs and at least basic tests. Our Python automation examples give project ideas.
4. You use Git the way teams do
Pass test: you work on feature branches, open pull requests, respond to review, rebase or merge cleanly, resolve conflicts, and can recover a lost commit with git reflog.
Evidence: pull requests on a shared or open source repository with review comments you addressed.
5. You have built a working CI pipeline
Pass test: on every push, your pipeline installs dependencies, runs tests and linting, builds an artifact, and fails the build on errors. You can explain caching and why secrets do not belong in the pipeline file.
Evidence: the pipeline file (GitHub Actions, GitLab CI or Jenkins) and its run history, including a failed run you fixed.
Early career milestones
6. You build containers properly
Pass test: your images use multi-stage builds, pinned base image tags, a non-root user and a .dockerignore; they pass a vulnerability scan with no critical findings you cannot explain; and you can run a multi-service app with Docker Compose.
Evidence: Dockerfiles and scan output. Compare against common Dockerfile mistakes.
7. You can run an app on one cloud platform end to end
Pass test: on AWS, Azure or Google Cloud, you set up identity with least privilege, a network with public and private subnets, compute, a managed database, a load balancer with TLS, and logging. Then you delete it all cleanly.
Evidence: an architecture diagram, setup notes and a teardown script. An associate-level cloud certification can confirm this milestone.
Mid-level milestones (2 to 5 years)
8. You manage infrastructure as code with state and modules
Pass test: your Terraform or OpenTofu code uses remote state with locking, reusable modules, separate environments, and a plan step reviewed in pull requests. You can import an existing resource and recover from a state problem.
Evidence: the repository and a pull request showing a plan output that was reviewed before apply.
9. You operate Kubernetes, not just deploy to it
Pass test: you can explain the control plane components, write Deployments, Services, Ingress and probes, set resource requests and limits, debug CrashLoopBackOff and Pending pods, back up etcd, and upgrade a cluster one minor version.
Evidence: manifests or a Helm chart and notes from an upgrade you performed. The CKA exam from the Linux Foundation tests these tasks hands-on. See Kubernetes architecture components.
10. You ship with continuous delivery and safe rollouts
Pass test: releases go out through a pipeline or GitOps tool such as Argo CD or Flux, use rolling, blue-green or canary strategies, and you have rolled back a bad release in minutes.
Evidence: a GitOps repository and a short record of a rollback, including what triggered it.
Senior milestones
11. You run observability with SLOs
Pass test: you collect metrics, logs and traces (for example Prometheus, Loki or ELK, and OpenTelemetry), define service level indicators and objectives with the product owner, and write alerts based on user impact rather than CPU percentages.
Evidence: an SLO document, dashboards and alert rules. Read the role of SREs in DevOps teams.
12. You build security into delivery
Pass test: pipelines scan code, dependencies and images; secrets live in a manager such as Vault or a cloud secrets service; clusters enforce policies through Pod Security Admission or a policy engine; and access follows least privilege.
Evidence: pipeline gates, policy files and a documented exception process. See DevSecOps tools for CI/CD.
13. You handle incidents and write postmortems
Pass test: you have taken part in on-call, followed or improved a runbook, coordinated during an incident, and written a blameless postmortem with action items that were completed.
Evidence: a postmortem (anonymised if needed) and the runbook changes that followed.
14. You understand cost and capacity
Pass test: you can read a cloud bill, tag resources for cost allocation, rightsize instances and pods, choose between on-demand, reserved and spot capacity, and plan capacity for expected growth.
Evidence: a cost report with the changes you made and their measured effect.
15. You think in platforms and grow other engineers
Pass test: you build reusable templates, modules or golden paths that other teams adopt; you write documentation people use; and you mentor juniors through the earlier milestones.
Evidence: internal docs or templates and adoption numbers from your own team, plus people you have coached.
How long do the milestones take?
It depends on prior experience and hours per week, so we do not give fixed timelines. A useful rule: you have passed a milestone when you can do the pass test on a new machine, explain each step to someone else, and fix it when one step breaks. If you need a tutorial open to complete it, you are still learning that stage.
Common mistakes on the roadmap
- Jumping to Kubernetes before Linux and networking are solid. Most Kubernetes debugging is Linux and network debugging.
- Collecting tutorials without building anything you can show.
- Learning five CI tools shallowly instead of one deeply.
- Treating certifications as the goal instead of proof of skills you already practised.
- Skipping security and cost until "later." Interviewers for mid and senior roles ask about both.
Turning milestones into a portfolio
| Portfolio project | Milestones it proves |
|---|---|
| Three-tier app with CI, containers and Compose | 3, 4, 5, 6 |
| Same app on one cloud with Terraform | 7, 8 |
| Same app on Kubernetes with Helm and Argo CD, canary rollout | 9, 10 |
| Add Prometheus, Grafana, SLOs, alerts and scans | 11, 12 |
| Simulated outage with postmortem and cost report | 13, 14 |
One project that grows across milestones is easier to explain in an interview than ten unrelated demos. When you reach the interview stage, our 200 DevOps interview questions guide maps questions to these same topics.
The DevOps course follows a similar sequence with labs for each stage, in our Kharadi, Pune classroom or live online, and the Linux administration course covers milestone 1 in depth.
Frequently asked questions
What is a DevOps roadmap milestone?
A checkpoint you can prove with a pass test and evidence, such as a working pipeline or repository, rather than a topic you have only read about.
Which milestones do I need for my first DevOps job?
Milestones 1 to 5, ideally with 6 as well: Linux, networking, scripting, Git, a CI pipeline and properly built containers, each with evidence.
Should I learn Kubernetes before Linux?
No. Kubernetes runs on Linux and most troubleshooting involves Linux processes, files and networking. Finish milestones 1 and 2 first.
Do certifications count as milestones?
They can confirm one. RHCSA fits milestone 1, cloud associate exams fit milestone 7, and CKA fits milestone 9, but keep project evidence too.
How is this different from a DevOps learning roadmap?
A learning roadmap lists topics in order. These milestones define when you are done with each stage, using tests and evidence.
What evidence should I show in interviews?
Repositories, pipeline runs, architecture diagrams, a postmortem, and short write-ups of problems you solved. Be ready to explain each decision.
Which cloud should I pick for milestone 7?
Pick the one most used by employers you are targeting. The concepts of identity, networking, compute and managed services transfer between AWS, Azure and Google Cloud.
What separates senior milestones from mid-level ones?
Senior milestones focus on reliability targets, security, incidents, cost and helping other teams, not just building and deploying.
References
What's Your Reaction?
Like
0
Dislike
0
Love
0
Funny
0
Angry
0
Sad
0
Wow
0