10.2.4 · Skill in Reliability engineering
Resilience
Designing systems that degrade gracefully when parts fail.
Timeouts, retries with backoff, circuit breakers and fallbacks keep one failure from taking everything down. External services, including AI providers, will fail. A product that degrades gracefully keeps users' trust.
Resources
1- bookGo deeperAddressing cascading failures (Site Reliability Engineering, chapter 22) ↗Google SRE · 40 min
How overload spreads through a system, and the timeouts, retries and load shedding that stop it.
Tools
none yetTools for this topic are being researched.