Monolith vs. Microservices
summary
Real-world Scala teams running systems that process millions of transactions discover microservices fail or succeed based on concrete choices: HTTP/JSON causes nested-request latency, while gRPC over JSON/HTTP cuts RPC cost; Kafka+Avro helps async patterns but raises QA and local-run pain; duplicated encoders/decoders and a shared common library anti-pattern reintroduce coupling; six public companies notably revert to monoliths; recommend refined types in IDL, strict contract testing, observability, retry/circuit-breakers, and careful SBT/module management.
AI-written summary · about
chapters (11)
- 0:00Intro
- 5:30Monolith vs microservices
- 7:02Small business capabilities
- 11:48Common dependencies
- 18:14Microservices vs Monolith
- 18:46Summary
- 20:03Requirements
- 25:01Cost
- 28:26Threat
- 28:58Communication
- 29:56Protocols
related talks
47mMicroservices? Monoliths? An Annoying Discussion!
50mEmbracing Microservices
01h 03mMicroservices vs The Distributed Monolith
43mMicroMonolith - Top anti-patterns of distributed systems
49mChallenges in Implementing Microservices
39mConsidering Migrating a Monolith to Microservices?
42mQ&A - Monolith to Microservices
19mMicroservices vs. Monolith for Scalability and Efficiency