Eventual Consistency in a Stateful Service
summary
A presentation about designing and operating a stateful distributed storage system of over 80 TB that faces scaling challenges and cannot provide strong consistency guarantees. It explains why simple designs had to be made more complex, describes the original product requirements, and discusses practical trade-offs when implementing eventual consistency in that service.
AI-written summary · about
chapters (9)
- 0:00Introduction Metrica
- 1:56Metrics engine
- 7:35Cross-shard read challenges
- 9:01Eventual consistency
- 12:55Non-final aggregates
- 18:59Event Car optimization
- 24:13Offline event gluing
- 33:29Pros and cons
- 43:36Q&A
related talks
25mThe shell game called eventual consistency
47mA Distributed, Scalable and Highly Reliable Data Storage System
38mEventual Consistency: A Question of Time and Expectations
30mDistributed system design
53mEventual consistency in pizza production
02mWelcome to – Dasith Wijesiriwardena
27mData Consistency in Microservices Architecture
44mDistributed systems and scalability rules