Mark Burgess interviewed at Velocity 2011
summary
Argues that the hard problems are scale, complexity, and Knowledge Management and that model‑oriented, idempotent configuration (convergence) yields predictable systems. Claims cloud adds reusability but not new fundamentals, predicts a single programmatic API expressing intent, and keeps humans for architectural decisions. Demonstrates CFengine as a real tool: CFengine agent on a Nokia phone (battery is the main limit), self‑healing/desired‑state enforcement across tens of thousands of machines and embedded devices, with thermostat‑style regulation.
AI-written summary · about
chapters (6)
- 0:00What are the major issues infrastructure designers are currently facing
- 2:02Does cloud infrastructure introduce new issues
- 3:47Does the public cloud make sense for large businesses
- 5:30Applying theoretical analysis to infrastructure
- 7:18Automated infrastructure management
- 9:41Consumer grade management
related talks
01h 19mThe Business of Configuration Management
33mSystems Automation: Past, Present and Future
35mHow to Cloud Ops Like a Boss
02mO'Reilly Velocity Conference 10 Year Anniversary
23mKeynote: Enterprise DevOps: Balancing Innovation Velocity with Business Requirements
07mJoshua Bixby interviewed at Velocity 2010
07mDevOpsCon 2016: Interview Matthew Skelton
14mWelcome