Distributed Systems at Odnoklassniki
summary
Overview of how Odnoklassniki builds fault-tolerant distributed systems in Java across thousands of servers that each hold only part of data or logic, requiring heavy inter-server network interaction. Covers common mistakes and failures, testing and diagnostic techniques, and approaches to preventing incidents in large-scale distributed services.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 12:40Distributed systems problems
- 24:00Failure handling strategies
- 35:15Speculative queries and load
- 45:50Failure testing and diagnostics
- 53:00Monitoring and takeaways
related talks
40mReliable Interaction in Distributed Systems
41mIntroduction to distributed systems
25mGrowing Distributed Systems
30mDistributed system design
33mWhy Are Distributed Systems So Hard?
52mWhat We Talk About When We Talk About Distributed Systems
50mHow We Built a Distributed Queue Service at Yandex
37mScaling a Distributed Stateful System: A LinkedIn Case Study