How does your application behaves when everything goes wrong?
summary
Practical resilience patterns and production numbers: request caching with short TTLs, client request timeouts and exponential retries, classic circuit breaker (example threshold = 1, open ≈30s), plus a new “battery backup” pattern that stores long‑TTL fallback data in Redis/MySQL/files/ETS and is populated by successful calls or async chargers. Deployed at OLX (platform moves >50 items/min, >10 phones/min) and used for credit‑limit checks, recommendation fallbacks and offline video auth; metrics and dashboards prove the approach reduces outages and keeps degraded features usable.
AI-written summary · about
chapters (13)
- 0:00Introduction and speaker
- 2:23Murphy's Law in software
- 4:37Microservices reality check
- 6:49Resilience defined
- 8:16Request cache pattern
- 10:43Request timeout importance
- 11:48Request retry strategies
- 12:51Circuit breaker explained
- 16:52Business continuity strategies
- 19:44Battery backup concept
- 26:02Lessons learned from resilience
- 28:11User experience focus
- 30:57Q&A and closing
related talks
35mDesign patterns for application resiliency
21mThe Multi Layered Cake of Resilience
34mDistributed Systems - What Can Go Wrong and How to Get It Right
38mResilience patterns in practice
15mBoosting System Resilience: Exponential Backoff
34mBuilding High Resiliency into Digital Solutions
53mResiliency and Availability Design Patterns for the Cloud
37mResilience Design Patterns