Reactive microsystems

on YouTube ↗49m

" summary

Argues for a middle path between monolith and microservices by making an append-only event journal the source of truth, splitting reads and writes, and projecting eventual-consistent views; shows a real pattern where projections and state in-process shards let a hot auction (example: Justin Bieber hair) serialize bids without DB locks. Mentions Legume framework, avoids Kafka by using simple DB append rows, and cites a customer case that shrank infrastructure from ~500 nodes to 10.

AI-written summary · about

" chapters (13)

1. 0:00Introduction and setup
2. 1:27Reactive systems vs reactive programming
3. 3:19The monolith and its limitations
4. 7:03From monolith to micro leaves
5. 10:46Event sourcing fundamentals
6. 15:05Asynchronous event propagation
7. 18:47What are reactive microsystems
8. 23:11Scaling with stateful services
9. 27:48Handling high concurrency
10. 33:12Legume framework overview
11. 37:08Applying lessons to monoliths
12. 39:18Conclusion and trade-offs
13. 40:24Q&A: Fault tolerance and scaling

" related talks

• 55mRethinking Reactive ArchitecturesDavid Leitner · NDC Oslo 2021
• 47mThe Rise of Reactive MicroservicesDavid Leitner · Craft Conference 2022
• 30mReacting to the future of application architectureGrace Jansen · Voxxed Days Microservices 2019
• 54mReacting to an Event-Driven WorldGrace Jansen · Conf42 Enterprise Software 2021
• 50mAddressing Recent EventsViktor Klang · Scala Italy 2018
• 48mIntegrating microservices by accepting the data dichotomyDamien Gasparina · Voxxed Days Microservices 2018
• 52mBuilding monolith applications that could scale to microservices..David Gomez G. · Devoxx Ukraine 2021
• 49mMicroservices guided by DDD and reactive systems paradigmDinkar Gupta · Riga Dev Days 2019