SAG Expert Interview #12: The Art of Resilient Software Architecture
summary
Resilience becomes a strategic necessity: build timeouts, retries and fallbacks, keep organizational slack for surprises, and aim for antifragility so failures make systems better. Measure feature usage—expect 20% of features used and up to 90% waste if you don’t measure—so prioritize effectiveness over blind efficiency. Balance availability with ecology as some cloud setups show 600% CO2 increase when GenAI scales. Favor simplification and core design rules: high cohesion low coupling information hiding.
AI-written summary · about
chapters (9)
- 0:00Start
- 0:18Introduction
- 0:44Why is resilience an important aspect in software architecture?
- 2:19How can resilience be integrated with the rapidly changing demands of complex systems?
- 4:05What is the difference between effectiveness and efficiency in this context?
- 6:09What are the implications of this in terms of sustainability?
- 8:02What is your view on AI in the context of sustainability?
- 10:14What are your recommendations for a more effective and sustainable use of GenAI in the future?
- 12:39Where can we start to make things more robust, resilient and sustainable?
related talks
22mResilience Is Not Enough
14mAntifragility or taming the hydra
48mUnbreakable by Design: The Secret Tech of Antifragile Software
47mThe 7 Quests of Resilient Software Design
43mResilience, Scalability, Performance – On the Architecture of Agile Product Teams
44mThe Art of Architecting for Scale
36mDesigning for failure - architecting fault-tolerant system
01h 03mEvolutionary Software Architectures