Building Self-Healing Cache Infrastructure
summary
Claims a deployed platform serving millions that cuts latency and DB queries by combining service‑mesh routing, event‑driven streams, and directory‑based coherency with ML for predictive cache warming and adaptive TTL management. Implementation uses declarative configs and GitHub workflows for automated pipelines, developer annotations for instant caching, and self‑filling recovery that promotes replicas, reroutes traffic, and rebuilds state from logs. Observability surfaces P50/P95/P99 latency, hit/miss ratios and anomaly alerts so the system actually adapts instead of just promising examples.
AI-written summary · about
chapters (14)
- 0:00Introduction and Speaker Background
- 0:44The Importance of Caching in Large-Scale Systems
- 1:10Challenges in Traditional Caching
- 2:45Principles of Self-Healing Systems
- 4:33Designing a Resilient Caching Architecture
- 6:22Machine Learning in Caching
- 7:24Implementation Patterns for Effective Caching
- 8:18Observability and Monitoring
- 9:11Impact and Future of Self-Healing Caching
- 9:59Advanced ML Techniques in Caching
- 10:56Edge Caching Challenges and Solutions
- 11:49Best Practices and Cultural Transformation
- 13:13Lessons Learned and Future Trends
- 14:40Conclusion and Key Takeaways
related talks
18mScaling APIs to Billions: Extreme-Scale Reliability
46mEdge Performance with In memory NoSQL
47mCaching beyond simple put and gets
25mArchitectural Caching Patterns for Kubernetes
34mAdvanced Caching Patterns used by 2000 microservices
43mCaching half a billion user transactions
32mBeyond Optimization: Resilient Cloud Microservices at Scale
38mLazy Defenses: Using Scaled TTLs to Keep Your Cache Correct