Ask the architects
summary
Panel lists concrete failures and fixes: extracting a first service takes months when teams try to change UI, data schema and add features at once; delaying cross‑cutting concerns (auth, tracing) doubles work. Uses: Prometheus for metrics, EFK or centralized logs, Apache Kafka for events, Netflix OSS + Spring Boot for circuit breaking, Kubernetes for orchestration. Notes: team of 13 struggled with event payload versioning, recommends consumer‑driven contracts and CI/CD, warns about 17–18 overlay network choices and cautious adoption of service mesh.
AI-written summary · about
chapters (10)
- 0:00Intro
- 1:00Common pitfalls
- 4:33Logging
- 6:50Load Balancing
- 8:09Versioning
- 11:15Service Mesh
- 12:54Service Mesh comments
- 14:30Microservices comments
- 16:53Opensource vs delegated infrastructure
- 21:43Microservices maturity
related talks
42mPanel: Architecture
50mJourneys To Cloud Native Architecture: Sun, Sea & Emergencies
27mArchitecting for Change Debate
53mHow do we build high‑load microservices
47mMicroservice Panel
39mArchitectural design patterns applied to Distributed Systems
42mMicroservices are the answer - but what is the question?
37mPanel Discussion: Ask Us Anything: Microservices and Service Mesh