Architecting Resilient Cloud Native Applications: A Practical Guide to Deployment and Runtime
summary
Shows concrete deployment and runtime choices: blue/green and canary reduce user impact but add coordination and testing overhead; feature flags and dark launches demand extra tests and cleanup. Recommends timeout + retry with exponential backoff but warns about higher latency and cascading retries. Describes bulkhead via Kubernetes NetworkPolicy to isolate failures and enable graceful degradation. Demonstrates a real control using Istio destination rule: max 1 in-flight, 1 queued and notes circuit breakers and observability are required and can create false positives and complexity.
AI-written summary · about
chapters (8)
- 0:00Introduction to resilient cloud apps
- 1:03Feature flags and dark launches
- 2:06Handling runtime failures
- 3:53Timeouts and retries explained
- 6:02Rate limiting strategies
- 10:14Bulkhead pattern for isolation
- 13:23Circuit breakers in systems
- 17:18Recap and closing remarks
related talks
50mJourneys To Cloud Native Architecture: Sun, Sea & Emergencies
39mRunning Resilient Workloads with Istio
32mBeyond Optimization: Resilient Cloud Microservices at Scale
19mKube Me This! Kubernetes Ideas and Best Practices
44mArchitectural Patterns of Resilient Distributed Systems
53mNext Generation Microservices
37mCloud Anatomy 101: Breaking Down Cloud-Native
36mKubernetes, Decisions