How our systems gain resilience, and why failures are useful
summary
The talk explains how modern high-load systems are made resilient by embracing and managing failures rather than avoiding them. It discusses conflicts that arise when engineers touch multiple parts of a system, why errors are inevitable, and practical approaches for detecting, isolating and recovering from faults to increase overall system robustness.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 4:41Incidents as learning events
- 7:09Resilience through recovery
- 10:01Postmortems for learning
- 13:15Alternatives to root cause
- 17:52Investing in incidents and ROI
- 22:29Starting incident analysis
- 26:15Problem aggregation and types
- 31:15Training and analysis skills
- 36:07People in the system
- 39:38Conclusion and call to action
- 40:00Q&A
related talks
43mResilience Engineering Mythbusting
32mWhat is resilience engineering?
30mYou've Lost That Process Feeling: Some Lessons from Resilience Engineering
01h 05mKeynote - Without Resilience, Nothing Else Matters
51mWithout Resilience, Nothing Else Matters
39mThe Human Nature of Failure and Resiliency
16mEngineering Failure-Resilient Systems
05mWhat is resilience engineering