DOF flashcards
DOF · Build recall with a focused review session.
30 cards. Read the term and the definition here. The deck at the top of this page is optional.
- DevOps
- A collaborative approach to building and delivering software that unites development and operations teams around shared goals, automation, and rapid feedback so organizations can release reliable changes more frequently.
- CALMS framework
- A framework for assessing DevOps adoption across five areas: Culture, Automation, Lean, Measurement, and Sharing. It emphasizes that tools alone are insufficient and that people and process change matter equally.
- The Three Ways
- A set of DevOps principles describing flow from development to operations, fast feedback loops back to development, and a culture of continual experimentation and learning.
- Value stream
- The sequence of activities, people, and systems required to turn a customer request into delivered value. Mapping it reveals delays, handoffs, and waste that slow software delivery.
- Psychological safety
- A team climate in which people feel safe to speak up, admit mistakes, ask questions, and challenge ideas without fear of punishment or humiliation. It supports learning and honest problem solving.
- Westrum organizational culture
- A model classifying organizations by how they handle information: pathological (power-oriented), bureaucratic (rule-oriented), or generative (performance-oriented). Generative cultures share information well and are associated with better safety and innovation.
- Cross-functional team
- A group whose members collectively possess the skills needed to deliver a product or service end to end, reducing handoffs and dependencies on other teams.
- Continuous integration
- A practice in which developers frequently merge small code changes into a shared mainline, with automated builds and tests run on each merge to detect integration problems early.
- Continuous delivery
- A practice that keeps software in a releasable state at all times by automating build, test, and release processes, so deployments can be performed quickly, safely, and on demand.
- Continuous deployment
- Automatically deploying eligible changes that pass the configured delivery checks to production, without a separate release-approval step. User exposure may still be managed with feature controls or rollout policies.
- Deployment pipeline
- An automated sequence of stages that builds, tests, and prepares software for release. It provides fast feedback on quality and makes the delivery process repeatable and visible.
- Trunk-based development
- A version-control practice where developers integrate small changes frequently into a single shared branch, avoiding long-lived feature branches that cause merge conflicts and delay feedback.
- Kanban
- A visual work-management method that limits work in progress and makes flow visible on a board, helping teams identify bottlenecks and deliver work steadily.
- WIP limit
- A cap on how many items may be in progress at a given stage of a workflow. It reduces multitasking and congestion, encouraging the team to finish work before starting more.
- Value stream mapping
- A technique for diagramming the steps, wait times, and handoffs in a value stream to expose waste and identify improvements that shorten delivery time.
- Shift left
- Moving quality, security, and testing activities earlier in the software lifecycle so defects are found when they are cheaper and faster to fix.
- Infrastructure as Code
- Managing servers, networks, and other infrastructure through machine-readable configuration files that are versioned and applied automatically, rather than through manual setup.
- Configuration management
- The practice of defining and maintaining consistent settings for systems and applications, often through automated tools, so environments remain reproducible and drift is controlled.
- Container
- A lightweight, isolated runtime package that bundles an application with its dependencies so it runs consistently across environments. Containers share the host operating system kernel.
- Container orchestration
- Automated scheduling, deployment, scaling, and management of containers across a cluster of hosts, typically performed by a platform such as Kubernetes.
- Feature flag
- A configuration switch that enables or disables functionality at runtime without redeploying code, allowing controlled releases, experiments, and quick rollback.
- Blue-green deployment
- A release technique using two identical environments, blue and green, where traffic is switched to the new version after validation, enabling fast rollback if problems appear.
- Canary release
- A deployment strategy that exposes a small portion of users to a new version first, monitors behavior, and gradually expands traffic if the release performs well.
- Version control
- A system that records changes to files over time, allowing teams to collaborate, review history, and restore previous versions. It is foundational for code and infrastructure automation.
- Test automation
- Using software tools to execute tests, compare results, and report outcomes without manual intervention, making repeated validation fast and consistent within delivery pipelines.
- DORA metrics
- A versioned set of software-delivery performance measures. Current guidance uses deployment frequency, change lead time, failed-deployment recovery time, change fail rate and deployment rework rate; older materials often describe four keys.
- Lead time for changes
- The time from a code change being committed to it running in production. It measures how quickly work moves through the delivery process.
- Deployment frequency
- The frequency of production deployments measured for a defined service and window. Compare trends with context and other outcomes; a higher count alone does not prove customer value or reliability.
- Change failure rate
- The percentage of production changes that cause degraded service and require remediation such as a hotfix or rollback. It reflects release quality and stability.
- Mean time to restore
- An average restoration-duration measure whose population and timestamps must be specified. The historical term is broader than current DORA’s failed-deployment recovery measure, so do not silently treat every incident population as equivalent.