Architektury Aplikacji: The last mile of microservices
summary
Argues that the “last mile” needs concrete choices: use an API gateway (Kong, Traefik, AWS API Gateway, Spring Cloud Gateway, Netflix Zuul) to handle routing, auth and edge logic at the cost of an extra network hop and latency; use GraphQL when clients must shape responses to avoid dozens of REST calls (he cites front ends sometimes needing ~50 requests per view); or embed a server-side UI (Vaadin) in the gateway to hide complexity — it ships a ~300 KB JS engine and forces server sessions but dramatically simplifies client code. Also notes organizations must weigh architecture tradeoffs (e.g., Amazon’s 11s deployment pace as a design goal).
AI-written summary · about
chapters (12)
- 0:00Introduction to microservices
- 1:19Big ball of mud problems
- 4:32Scaling dimensions explained
- 7:47Testing and legacy issues
- 9:33Breaking down monoliths
- 12:05Client access to microservices
- 14:36API gateway pattern
- 17:51Back-ends for front-ends
- 20:59Graph-based APIs
- 22:48Server-side UI approach
- 25:40Vaadin server-side demo
- 30:39Q&A and closing remarks
related talks
32mAdopting a Modern API Strategy for Hybrid, Multi-Cloud Architectures
18mSoftware Architecture Building Better Applications with Architecture Best Practices
23mMicroservices Architectures: Agility vs Complexity
56mArchitecting Cloud Native Microservices
27mMaster API Gateways: Boost Your Microservices Strategy
26mBuilding Scalable Microservice APIs From The Beginning
27mManaging the Complexity of Microservices Deployments - Prithpal Bhogill, Google & Kenny Bastini
23mClient/Server Architecture: Past, Present, & Future