Nowadays Architecture Trends, from Monolith to Microservices and Serverless
summary
A real-world migration of a huge EAR Core Banking Application to a distributed stack shows concrete tactics and tradeoffs: introduce a queue and Apache Camel request–reply to separate UI and back end, split the monolith into Spring Boot modules, add logging with correlation IDs and ELK, then experiment with Spring Cloud Functions to AWS Lambda and Oracle Cloud. Security uses Auth0 for OAuth2/JWT/JWS. Observed cold-starts ~0.65 ms with ~139 MB memory and cost ~262k requests per dollar; reactive Reactor/WebFlux handles very high throughput but operations and CI/CD costs rise.
AI-written summary · about
chapters (4)
related talks
42mEnterprise Serverless Adoption. An Experience Report
18mSoftware Architectures Gone Wild
51mMicroservices and Microfrontends: Scaling, Learning, Growing
01h 02mServerless Architecture Patterns
53m"Serverless" Is What J2EE Was Meant To Be
55mArchitectural Patterns for Rapid, Reliable, Frequent and Sustainable Development
45mPatterns & practices for developing enterprise-scale serverless applications
45mThe Evolution From Monolith to Microservices to Serverless and Beyond