Scaling to 500k concurrent customers with orleans
summary
Kaizen Gaming adopts the actor model with Orleans grains and a runtime that auto-activates, provides location transparency and single-threaded grain execution to handle highly bursty sports traffic and a goal of 500k concurrent customers. Session management migration completes in three weeks and handled World Cup spikes. A live SQLite demo shows a simple HTTP→DB approach at ~80 req/s versus grain-mediated updates at ~232 req/s because lock contention moves from DB to single-threaded actors.
AI-written summary · about
chapters (10)
- 0:00Scaling challenges
- 1:32Company traffic patterns
- 2:38Actor model
- 4:51Actor model vs services
- 6:36Actor model implementations overview
- 7:43Why Microsoft Orleans
- 9:10Virtual actors features
- 10:55Orleans use cases
- 12:01Live demo
- 15:26Q&A
related talks
01h 04mOrleans Architecture Patterns
52mWould Your System Survive a Million Concurrent Players?
40mReal Time Performance at Massive Scale
01h 07mStairway to Cloud: Orleans Framework for building Halo-scale systems
48mUse actors, Luke
18m"Kafka's Metamorphosis"
57mConsuming Events Like There's No Tomorrow
25mGo distributed (and scale out) with Actors and Akka Clustering