Effective Eventual Consistency with Actor Models + Amazon Web Services

on YouTube ↗43m

" summary

Claims a practical way to sidestep CAP by using CRDTs + Event Sourcing and immutability: every clickstream event is hashed into a unique id, duplicated into S3 + Elasticsearch, and merged via set union of hash codes so order never matters. Actor systems (Akka.NET, C#) pull through SQS, use self-throttling actors to avoid OOM, and run one-per-second jobs to scale. Daily reconciliation finds missing hashes and replays S3 to recover after the AWS outage; the system works and paid back in production.

AI-written summary · about

" chapters (4)

1. 0:00Start
2. 10:27Strong Eventual Consistency
3. 28:53Use self-throttling actors
4. 41:06References Akka.NET Memory Throttling Demo

" related talks

• 36mEasy Eventual Consistency with Actor Models + Amazon Web ServicesPhilip Laureano · NDC Oslo 2017
• 46mBeyond Eventual ConsistencyJamie Allen · Code Mesh 2017
• 54mHow Events Are Reshaping Modern SystemsJonas Bonér · Devoxx Belgium 2017
• 51mImmutable data stores and CORS for safety, flexibility and profitSidney Shek · Scala Days Berlin 2016
• 46mBuilding high available systems with Akka Distributed DataJeroen Gordijn · Scala Days Copenhagen 2017
• 55mExperience report on using Akka in production: pitfalls and rewardsFrançois Laroche · ScalaIO 2019
• 39mManaging Consistency, State and Identity in Distributed MicroservicesDuncan K. DeVore · Scala Days Copenhagen 2017
• 52mEvent Sourcing and CQRSLutz Huehnken · Scala Days Copenhagen 2017