Resiliency and Availability Design Patterns for the Cloud
summary
Describes concrete patterns, tools and numbers: use cells and thin router to limit blast radius (DynamoDB handles up to 20 million req/s), apply shuffle sharding so overlap of two reduces a 25% hit to ~3.6%, prefer exponential backoff with jitter over synchronous retries, implement circuit breakers or load‑shedding, push work to queues and read‑replicas, and run hands‑on chaos via a docker‑compose master/slave kill demo with code on GitHub.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:32Distributed systems
- 2:16What is resiliency
- 3:30How to build resilient system
- 4:42Bulk heading in isolation
- 8:08Shuffle sonic
- 14:14Default timeouts
- 15:43Default timeout example
- 27:01Chaos engineering
related talks
53mResiliency and Availability Design Patterns for the Cloud
51mPatterns for building resilient software systems
35mDesign patterns for application resiliency
30mResilience strategies to level up your workloads
40mThe Art of Embracing Failure at Scale
45mA resilient application in a partially degraded world
44mArchitectural patterns of resilient distributed systems
36mCloudy With a Chance Of Chaos: Verifying the Resiliency Of Cloud-Native Applications