Extreme Engineering-Elements of Building a High -scale Broking Application
summary
A high-scale broking platform that handles 22 million clients and onboards a million users a month processes about 10 million orders daily and peaks at 3 million RPS by sharding users into replicated cell-based shards, running frontends that do MVVM+Room offline sync with eventual-idempotent order queues, and isolating external dependencies with sized thread-pools. Tech choices are Go for low-overhead concurrency (tiny Go goroutines), Python for batch ML, SQL for transactions, Redis/Redshift for cache/analytics. Quarterly regulator DR drills validate the design.
AI-written summary · about
chapters (7)
- 0:00Introduction to high-scale broking
- 3:18Core mechanics of equity trading
- 5:26Design pattern: Eggs and basket
- 8:13Design pattern: Always on
- 11:48Design pattern: External distrust
- 13:35Language and tech choices
- 16:08Cross-platform UI approach
related talks
01h 06mPatterns for scalability and availability in (trading) systems
46mDigital scale-up architectural patterns
55mFrom the engine to the bridge - building systems that scale
01h 03mArchitecting a large social network without breaking the bank
56mMastering Cloud Scalability and System Designs Deep dive
34mBuilding High Resiliency into Digital Solutions
39mBuilding Scalable Orchestration
54mArchitecture Foundations: Styles & Patterns (2021)