Patterns for Scalability and Availability in Trading Systems

Explore architectural patterns for building highly scalable and available trading systems in this conference talk. Delve into nontraditional approaches that balance high throughput, low latency, and high availability requirements. Learn how to process hundreds of thousands of business events per second, react to thousands of transactions, and distribute data to tens of thousands of concurrent users in millisecond-critical environments. Discover techniques used in large-scale trading systems, including synchronous and asynchronous transaction registrations, casual eventual consistency, and graceful failure handling. Examine the challenges of exponential business growth, reactive systems, and technical environments. Gain insights into incident management, realtime reactive systems, and important learning points for implementing these patterns effectively.

Intro
About Saxo Bank
Business model
Asset types
Service requirements
Business challenges
Critical business flows
Exponential business growth
Reactive systems
Technical environment
Architecture
Patterns
The humble beginnings
Synchronous transaction registrations
Casual eventual consistency
Transport latency
Scalability resilient
Race conditions
Asynchronous transaction registration
Fail theory
Commercial products
Price updates
Graceful failure
Traditional mirroring
Use cases
Culture
Incident management
Realtime reactive
Learning points
Be careful
Questions
Best days