Death to Latency: Building Reactive, Cloud Native Apps with Akka NET
summary
Shows how Akka.NET builds soft real‑time apps by keeping state in memory instead of the DB, using the actor model to map one actor per business entity, and distributing them with Akka.Cluster and cluster sharding so at most one actor owns an entity. Demonstrates event sourcing via Akka.Persistence, runs ~80 million messages/sec in a process with microsecond latencies, meets per‑operation deadlines like 5 milliseconds, and recommends one‑by‑one rolling deploys on Kubernetes with gRPC or websockets.
AI-written summary · about
chapters (13)
- 0:00🚀 Build Stuff Conference – Get Your Tickets!
- 2:06Introduction to Real-Time Systems & Akka.NET
- 4:44Key Differences Between Hard and Soft Real-Time Systems
- 6:00Examples of Real-Time Applications & Industries
- 10:11How Real-Time Applications Impact Business and Safety
- 14:00Key Questions to Build a Real-Time Application
- 17:36Problem with Traditional Paradigms: Databases in Real-Time
- 20:00Introduction to Stateful vs Stateless Architecture
- 25:12Using Akka.NET to Manage State and Reduce Latency
- 28:00The Actor Model for Concurrency and Scalability
- 32:34Akka.NET Cluster & Cluster Sharding for Fault-Tolerance
- 37:02Implementing Real-Time Applications in a Cloud Environment
- 40:12Challenges in Deployment and Operationalizing Real-Time Apps
related talks
58mDeath to Latency: Building Reactive, Cloud Native Apps with Akka.NET
01h 14mAkka.NET: The Future of Distributed Programming in .NET
27mFour Principles of Reactive Architectures
44mAkka post-mortem in a microservice architecture
39mAkka.net: The actor programming model
37mAkka Present and Future
40mTop Mistakes When Writing Reactive Applications
52mBuilding Reactive Apps