csv-flash-gemini-2.5pro Flashcards
(50 cards)
What is the central argument of Accelerate regarding software delivery?
High performance in software delivery is not traded off against stability; rather, speed and stability enable each other and drive organizational performance. (Chapter 2, Pages 53, 58-59)
What are the four key metrics identified to measure software delivery performance?
- Lead Time for Changes, 2. Deployment Frequency, 3. Time to Restore Service (MTTR), 4. Change Failure Rate. (Chapter 2, Pages 47-51)
According to the research, is there a trade-off between throughput (speed) and stability in software delivery?
No. High performers consistently demonstrate both higher throughput and better stability compared to low performers. They move in tandem. (Chapter 2, Pages 53, 254)
What organizational outcomes does high software delivery performance predict?
Higher profitability, productivity, market share, and achievement of noncommercial goals (e.g., efficiency, effectiveness, customer satisfaction). (Chapter 2, Pages 58-59, 253)
What is the Westrum model of organizational culture, and what are its three types?
A model describing how information flows within an organization, indicative of its culture. Types: Pathological (power-oriented), Bureaucratic (rule-oriented), and Generative (performance-oriented). (Chapter 3, Pages 64-66)
Which type of Westrum organizational culture is predictive of high software delivery and organizational performance?
Generative (performance-oriented) culture, characterized by high cooperation, shared risks, bridging encouraged, inquiry from failures, and novelty implemented. (Chapter 3, Pages 71, 250)
Can organizational culture be changed? If so, how?
Yes. Culture can be influenced and improved by changing practices. Implementing technical (like Continuous Delivery) and Lean management practices drives cultural improvements towards a generative model. (Chapter 3, Page 74)
What is Continuous Delivery (CD)?
A set of capabilities enabling changes of all kinds (features, config changes, bug fixes, experiments) to get into production or users’ hands safely, quickly, and sustainably. (Chapter 4, Page 77)
What are the five key principles of Continuous Delivery?
- Build quality in. 2. Work in small batches. 3. Computers perform repetitive tasks, people solve problems. 4. Relentlessly pursue continuous improvement. 5. Everyone is responsible. (Chapter 4, Pages 77-78)
What are the foundational technical practices required for Continuous Delivery?
- Comprehensive configuration management. 2. Continuous Integration (CI). 3. Continuous Testing. (Chapter 4, Pages 79-80)
What is the impact of Continuous Delivery practices on teams and performance?
CD practices drive higher software delivery performance, better organizational culture, lower burnout, and less deployment pain. (Chapter 4, Pages 81-85, 246)
What is Trunk-Based Development, and how does it relate to performance?
A practice involving fewer than three active branches, short branch lifetimes (<1 day), and no code freezes/stabilization periods. It’s a predictor of high delivery performance. (Chapter 4, Pages 91-92, 245)
Why is version control for all production artifacts (including configuration) important?
It’s a key component of CD. The research found that version controlling system and application configuration was even more highly correlated with performance than version controlling application code. (Chapter 4, Page 89, 245)
What characterizes effective test automation suites according to the research?
They are reliable (pass means releasable, fail means real defect), primarily created/maintained by developers, and run regularly (fast feedback on commit, comprehensive feedback daily). (Chapter 4, Pages 89-90, 246)
What architectural characteristic is the biggest driver of Continuous Delivery and IT performance?
A loosely coupled, well-encapsulated architecture, enabling teams to independently test and deploy their components/services. (Chapter 5, Pages 97, 258)
What is Conway’s Law and the ‘Inverse Conway Maneuver’?
Conway’s Law: Org structure constrains system design. Inverse Conway Maneuver: Evolve your team/org structure to achieve the desired architecture (enabling loosely coupled teams/systems). (Chapter 5, Page 98)
How does architecture affect an organization’s ability to scale its development teams?
Loosely coupled architecture allows organizations to add more developers and increase productivity linearly or better (measured by deploys/day/developer), unlike low performers where productivity decreases. (Chapter 5, Pages 100-101)
Should teams be allowed to choose their own tools?
Yes. Allowing teams to choose tools contributes positively to continuous delivery and performance. Focus should be on outcomes and enabling engineers, not mandating specific tools unless they provide clear value and ease-of-use. (Chapter 5, Pages 102-104, 247)
What does ‘Shifting Left on Security’ mean?
Integrating information security into the entire software delivery lifecycle (from design through testing and deployment), rather than treating it as a separate, downstream phase. (Chapter 6, Page 106, 246)
How does integrating information security impact performance?
It improves continuous delivery capabilities and software delivery performance, and high performers spend significantly less time remediating security issues. (Chapter 6, Pages 106, 108)
What are the key components of Lean Management practices for software delivery identified in the research?
- Limiting Work In Progress (WIP). 2. Visual Management (displaying key metrics/work status). 3. Feedback from Production (using monitoring data for decisions). (Chapter 7, Page 112, 249)
Do WIP limits alone predict performance?
No. WIP limits are effective only when combined with visual management and feedback loops from production monitoring to drive process improvement. (Chapter 7, Page 113)
What did the research find about formal change approval processes (e.g., CABs)?
External approval bodies (like CABs) negatively correlate with delivery performance (slow down delivery without improving stability). Peer review-based approaches yield higher performance. (Chapter 7, Pages 115, 249)
How can Segregation of Duties be achieved without a CAB?
Through peer review (recorded in VCS/deployment tools) and fully automated deployment pipelines that provide an audit trail. (Chapter 7, Page 116)