DevOps
Origin
The traditional separation between development and operations teams created structural conflicts: developers wanted to deliver changes quickly, while operations teams wanted stability. In 2009, Patrick Debois organized the first DevOpsDays conference in Ghent (Belgium) to explore how to resolve this tension. The term went viral on Twitter under the hashtag #devops and became the name of a movement. The book The Phoenix Project (Kim, Behr, and Spafford, 2013) popularized the principles of DevOps through an accessible narrative.
The four pillars of CALMS
The CALMS model describes the pillars of the DevOps movement:
- Culture: the foundation of DevOps is a culture of collaboration and shared responsibility between development and operations. Without cultural change, tools do not deliver the expected results.
- Automation: automating the repetitive processes in the delivery pipeline—testing, integration, deployment, infrastructure—to reduce human error and speed up cycles.
- Lean: applying the principles of lean thinking to eliminate waste from the delivery flow and keep work in progress (WIP) low.
- Measurement: measuring the throughput of the delivery system—deployment frequency, time to recover from failures, change failure rate—to identify bottlenecks and improvements.
- Sharing: sharing knowledge, practices, and responsibilities among development, operations, and other functions.
Key technical practices
- Continuous integration (CI): integrating code changes into the main branch several times a day, with a suite of automated tests that validates each integration.
- Continuous delivery (CD): keeping the code in a deployable state at all times, so that a deployment to production is a business decision rather than a stressful technical event.
- Infrastructure as code (IaC): managing and provisioning infrastructure through versioned code definitions, treating servers as code rather than as manually managed assets.
- Monitoring and observability: instrumenting systems to detect problems in production before they affect users.
DevOps and Agility
DevOps and Scrum are complementary, not alternatives. Scrum organizes development work into iterations with frequent deliveries; DevOps provides the technical infrastructure and the culture that make it possible for those deliveries to reach production reliably and repeatedly.
A team that practices Scrum without DevOps can deliver increments at the end of every sprint, but if the deployment process takes days or weeks and requires manual coordination, the benefit of short sprints is partially canceled out. DevOps solves the "last mile" of agile delivery.
DevOps and AI
The adoption of generative AI has created an extension of the DevOps paradigm for AI models called MLOps (Machine Learning Operations): CI/CD, versioning, monitoring, and governance practices applied to the lifecycle of AI models, not just to code. Continuous integration of AI-generated code and automated review of the output are components of a modern DevOps pipeline.
Common mistake
Adopting DevOps tools without the cultural change. DevOps is not a tool stack: it is a way of working. Teams that implement CI/CD pipelines but keep the organizational separation between development and operations—with different managers, incentives, and responsibilities—tend to bear the costs of automation without the benefits of collaboration. Tools automate; culture transforms.
References
- Kim, Gene; Behr, Kevin; Spafford, George. (2013). The Phoenix Project: A Novel about IT, DevOps, and Helping Your Business Win. IT Revolution Press.
- Kim, Gene; Humble, Jez; Debois, Patrick; Willis, John. (2016). The DevOps Handbook. IT Revolution Press.
See also
Want to advance in agility? You can search for course and exam calls or go at your own pace by becoming a member of the Agile Club. This membership includes exclusive resources, e-learning classrooms, and access to Skill Arena: a space to practice and measure your agile skills at your own pace.