1. The three ways of DevOps #
The framework of The Phoenix Project and the DevOps Handbook.
Flow, feedback, learning #
| Way | Goal | Practices |
|---|---|---|
| 1. Flow | Speed up the movement of work from development to production | small batches, CI/CD, limiting work in progress, removing queues |
| 2. Feedback | Detect and fix problems as early as possible | automated tests, observability, alerts, fast reviews |
| 3. Continuous learning | Experiment and learn from failures | blameless post-mortems, time dedicated to improvement, sharing |
DevOps is neither a tool nor a job title: it is a way of organizing work so that shipping often is both fast and safe. The tools (CI/CD, IaC) serve these three ways.
2. Seeing the value stream #
You only improve what you can see.
Value stream mapping #
Value stream mapping follows a request from idea to user and measures, at each step, the processing time and the waiting time. In most organizations, actual work represents less than 15% of the total lead time: gains come from removing waits (manual approvals, unavailable environments, handoffs between teams).
Good practice: Run the workshop with all the teams involved (dev, ops, security, business) and start from a recent real request, not from the theoretical process.
Small batches and work in progress #
- A smaller batch moves through the system faster and its defects are easier to find.
- Limit work in progress (WIP): finish before starting.
- Make work visible: a shared kanban board including operations tasks and technical debt.
3. Organizing teams #
Who builds, who runs.
You build it, you run it #
In the historical model, developers « throw over the wall » what operations must run (see the wall of confusion). DevOps brings the two together: the team that builds a service is also responsible for running it in production, on-call included. In return, it gets the autonomy and tools to deploy on its own.
Anti-pattern: Renaming the operations team « DevOps team » and giving it all the pipelines recreates a silo: DevOps is a shared responsibility, not a new intermediary team.
Team Topologies #
| Type | Role |
|---|---|
| Stream-aligned | delivers value on a business stream, end to end |
| Platform | provides self-service internal services (CI/CD, clusters, observability) |
| Enabling | temporarily helps teams acquire a skill |
| Complicated subsystem | masters a highly specialized component |
4. A learning culture #
Psychological safety and post-mortems.
Blameless post-mortem #
After an incident, the goal is not to find a culprit but to understand how the system (tools, processes, available information) made the error possible. A blame culture pushes people to hide problems, which makes them worse.
- Factual timeline of the incident (detection, actions, resolution).
- Impact: users, duration, data.
- Contributing factors, looking for several causes rather than a single « root cause ».
- What worked well.
- Corrective actions with an owner and a deadline, tracked like the rest of the work.
Did you know?: DORA research shows that a generative culture (trust, information sharing, failures seen as learning opportunities) is a predictor of delivery performance.
My progress
Sign in to track your progress and take notes. Sign in
💬 Questions about this lesson
No questions yet: be the first to ask for help or share your configuration.
Recommended Articles
End-to-end CI/CD (4/5): deployment strategies and rollback
Deploy without downtime and without risk: recreate, rolling update, blue/green, canary, feature f...
Measuring DevOps performance: DORA metrics, SPACE and continuous improvement
Measure and improve delivery performance: the DORA metrics (deployment frequency, lead time, chan...
SRE best practices: useful SLOs and respected budgets
Checklist for setting up SLOs that matter: choosing user journeys, realistic objectives, budget p...
End-to-end CI/CD (3/5): artifacts, registries and environments
Deliver continuously: immutable, versioned artifacts, image and package registries, promoting the...
End-to-end CI/CD (5/5): GitHub Actions, GitLab CI, Jenkins and Argo CD in practice
Compare and use CI/CD tools: GitHub Actions, GitLab CI, Jenkins, Argo CD and Flux; a complete pip...
Infrastructure as Code and GitOps: Terraform, Ansible and Argo CD
Manage infrastructure as code: declarative or imperative, Terraform (providers, state, plan, modu...
DevSecOps: securing the software supply chain end to end
Integrate security into the DevOps pipeline: shift-left, threat modeling, secret detection, SAST,...
Containers, Kubernetes and cloud native in a DevOps approach
The role of containers and Kubernetes in DevOps: reproducible images, multi-stage Dockerfiles, re...
Integrated Logistics Support
Integrated Logistics Support (ILS) is a systematic, integrated approach to logistics management, ...