DevOps Fundamentals: Culture, Practices and Tools

what is devops

Wondering what is DevOps and whether your team needs it? DevOps is a set of practices and cultural habits that unite development and operations, so one team shares ownership of building, releasing and running software.

If you want the automation side first, read our explainer on what is ci/cd, the pipeline practice at the heart of DevOps.

Item Value
Stages in the DevOps lifecycle Eight
Elements in the CALMS framework Five (Culture, Automation, Lean, Measurement, Sharing)
Change failure rate, elite performers (DORA 2024) 5%
Change failure rate, low performers (DORA 2024) 40%
Organizations running Kubernetes in production (CNCF 2025) 80%
Organizations using CI/CD for most or all applications (CNCF 2025) 60%

The figures come from Pulumi’s DevOps explainer and the CALMS framework page. Survey results and adoption numbers change from year to year, so check the source before you rely on them.

What DevOps means in practice

DevOps is a set of practices and cultural principles that unify software development and IT operations into one continuous cycle, as Pulumi’s explainer describes it. Instead of passing a release from one team to another, the people who write software share ownership of planning, building, testing, releasing and running it.

It is easy to mistake DevOps for a tool purchase. One Opslyft guide says DevOps is a culture and practice set, not a job title or a tool, built around shared ownership of software from commit to production.

The need for it comes from a timing mismatch. The Pulumi explainer notes that development moved in weeks while operations moved in months, which made releases a recurring source of friction.

In day-to-day work, the shift looks like this:

  • Developers and operations staff share ownership of the entire lifecycle.
  • Slow manual steps between teams are automated.
  • Infrastructure is written as code and changes through the same pull-request workflow as the application.
  • Resilience, observability and compliance are built into the pipeline rather than chased down after an incident.

How DevOps evolved

DevOps grew out of overlapping shifts in how software is built, deployed and run. Knowing the timeline explains why the practices look the way they do.

  • 2001: The Agile Manifesto reorganized how teams built software, though Agile mostly stopped at the developer’s commit.
  • 2007 to 2009: Belgian engineer Patrick Debois started organizing what became DevOpsDays, and the first event was held in Ghent in 2009.
  • 2009: John Allspaw and Paul Hammond gave the talk “10+ Deploys Per Day: Dev and Ops Cooperation at Flickr”, showing that frequent deploys and stability could co-exist.
  • 2010s: Docker was open-sourced in 2013, and Kubernetes in 2014 with v1.0 shipping in 2015, making packaging and orchestration uniform.
  • 2020s: Platform engineering and DevSecOps emerged as responses to growing complexity.

According to Pulumi, platform engineering packages vetted infrastructure, CI/CD and policy into golden-path services that product teams can consume. DevSecOps moves security into every stage of the same pipeline.

The DevOps lifecycle in eight stages

The DevOps lifecycle is a continuous loop that turns an idea into running production software and feeds real-world feedback into the next iteration. Pulumi lists eight stages in order:

  1. Plan: decide what to build next and break work into small batches.
  2. Code: write application and infrastructure code with version control and code review.
  3. Build: package immutable artifacts on every commit.
  4. Test: run unit, integration, security, performance and infrastructure tests automatically in CI.
  5. Release: promote a tested artifact toward production through environment gates.
  6. Deploy: push the artifact and its infrastructure changes to the target environment, with rollback plans.
  7. Operate: run the system with on-call rotations, runbooks and defined SLOs.
  8. Monitor: collect metrics, logs, traces and user feedback, and turn them into the next planning cycle.

The release and deploy stages are where teams decide whether a human approves the last step. That choice is the heart of continuous delivery vs continuous deployment, which has its own guide.

what is devops Twisted loop
A translucent blue loop stands vertically against a purple and blue background.

Core principles: the CALMS framework

CALMS is a heuristic to evaluate and guide DevOps adoption. Each letter comes with a question your team can ask itself:

  • Culture: do we trust and collaborate across roles? Look for blameless postmortems and psychological safety.
  • Automation: are repetitive paths codified? Look for infrastructure as code, CI pipelines and policy as code.
  • Lean: are we reducing waste in flow? Look for small batch sizes and value stream mapping.
  • Measurement: do we have actionable telemetry? Look for DORA metrics, SLOs and error budgets.
  • Sharing: is knowledge spread and reusable? Look for runbooks, inner source and guilds.

The CALMS page also lists common pitfalls: treating it as a checklist certification, over-rotating on tooling before culture, and tracking vanity metrics such as deployment count without context.

Practices, metrics and tools

Measurement is what keeps DevOps honest. Pensero’s guide names four metrics: deployment frequency, lead time for changes, change failure rate, and time to restore service.

DORA’s 2024 Accelerate State of DevOps Report, as summarized by Pulumi, shows how wide the gap can be. Elite performers (19% of respondents) deploy on demand with a 5% change failure rate and recover in under an hour, while low performers (25% of respondents) fail 40% of changes and take between a week and a month to recover.

Tooling matters, but it supports the practices rather than defining them. The Opslyft guide gives typical examples:

  • CI automation with Jenkins or GitHub Actions.
  • Infrastructure as code with Terraform, Ansible or AWS CloudFormation.
  • Monitoring and observability with Prometheus, Grafana or AWS CloudWatch.

Automated checks on every change belong in the test stage, and the guide on how to automate code review shows how to take that further.

Adoption of these practices is widespread. CNCF’s 2025 Cloud Native research, reported by Pulumi, found 80% of organizations running Kubernetes in production and 60% using CI/CD for most or all applications.

How do you start adopting DevOps?

Start small and treat the work as a journey. The CALMS page suggests picking one constraint per quarter rather than chasing every dimension at once.

  1. Baseline your current state with a qualitative review and a few metrics.
  2. Introduce a lightweight postmortem template.
  3. Automate one painful manual deployment path.
  4. Map your value stream and identify the largest wait state.
  5. Reassess after each improvement and document what you learned.

Sequence matters. Pensero warns that automating broken processes just makes problems happen faster, so improve the process before you automate it.

Frequently asked questions

What is DevOps in simple terms?

Show answer

It is a way of working in which developers and operations staff share responsibility for software from planning through running it in production. Automation and feedback loops replace slow handoffs between teams.

Is DevOps a job title or a tool?

Show answer

No. According to the Opslyft guide, it is a culture and practice set, and tools only support it.

How is DevOps different from Agile?

Show answer

Agile reorganized how teams build software, but it mostly stopped at the developer’s commit. DevOps extends the same thinking through release and operations, so delivery and running software are part of one cycle.

Is CI/CD the same as DevOps?

Show answer

No. CI/CD is an automation practice inside DevOps, while DevOps also covers culture, measurement and shared ownership. Treating a pipeline as the whole of DevOps is a common mistake.