The Actor Model and the Queue or “Batch is the New Black”
summary
Combines the Actor model with queue-based pull not push processing and shows it runs at national scale: re-architects an enrollment pipeline that must handle 1.1 billion enrollees and ~1 million enrollments/day after a database approach stalls around 500k/day. Implements tiny actors, hierarchical supervisors, supply clerk actors for serialized shared resources, idempotent messages with retry-on-timeout, redundant queues, TLS/AES cryptographic auth, and state carried in messages; result is elastic, failover-safe processing that meets the Indian Aadhaar throughput needs.
AI-written summary · about
chapters (23)
- 0:00Intro
- 9:47Concurrency
- 11:02Synchronization
- 11:26Scale
- 11:46Resource Sharing
- 12:10Actor Model
- 17:38Micro Services
- 19:42Office Worker Actor
- 20:43Supply Clerk Actor
- 22:47Supervisor Actor
- 24:24Supervisor Server
- 25:54Who Watches the Watchers
- 26:42Queues
- 30:06Human Analogy
- 30:59What happens if an actor dies
- 33:51Registration
- 35:09Push or Pull
- 36:29Dont Push
- 37:49Dont Pull Fast
- 38:20Multiple Queues
- 38:43Load Balancing
- 39:42State and External Input
- 43:32Keep Actor Small
related talks
45mSo you want to be an actor? Lessons from implementing the actor model
48mReliable messaging in the world of actors
47mUsing the Actor Model in Financial Transaction Systems
51mBuilding concurrent systems using Akka.NET
48mUse actors, Luke
48mActor, an Elegant Model for Concurrent and Distributed Computation
50mNext Generation Actors With Akka
01h 20mAkka Streams for Large Scale Data Processing