10 years of microservices at FINN.no - and we still haven’t slain that dragon!
summary
FINN.no runs a few hundred services while the original monolith continues to handle core revenue, targeting 99.9% uptime and hitting 99.97% year‑to‑date. The stack uses Kubernetes on a 60‑node private cloud, 500 microservices (about 180 engineers), Kafka with 1332 Kafka topics and Kafka Streams, and micro‑frontends (Podium/Troika). Legacy migration realities include a Sybase off‑Solaris move optimized from a 48‑hour myth down to a 6→4 hour migration window.
AI-written summary · about
chapters (38)
- 0:00Intro
- 1:13Hennings background
- 1:40Quiz
- 3:22Deployments
- 4:35Marketplace
- 6:51Outline
- 7:29What were we looking for
- 7:58Uptime
- 9:01User experience
- 9:12Conclusions
- 11:16Design
- 11:49Legacy
- 12:19The problem
- 16:01Getting to the time line
- 18:27Thrift
- 19:40Reboot
- 19:50Product Catalog
- 21:49Sam Newman
- 21:59Hystrix
- 23:43Netflix
- 23:53Deathstar
- 24:22Eric Evans Fred George
- 27:42Cloud
- 28:32New Business Center
- 37:20The graph
- 38:20Migration off Solaris
- 38:45What we learned
- 41:04Why we havent slain the monolithic dragon
- 41:42What would you have done differently
- 43:30Now we have a direction
- 43:59The dragon will go to sleep
- 44:12Start with where you are
- 45:35Kafka streams
- 47:32How many developers are managed
- 49:23Grand redesign syndrome
- 52:05What are we doing now
- 55:02Do we have a strategy
- 56:35How do we discover microservices
related talks
38mMaintaining highly distributed services with Erlang: 20y experiences in fintech
18mMicroservices Out In the Wild
51mMicroservices and Microfrontends: Scaling, Learning, Growing
44mRoute to microservices the epic voyage of a fintech leader
26mFighting the Good Fight at the Hot Gates of Micro-Services
25mFrom Millions of Customers to Petabytes of Data: A Journey to Microservices
29mMoving from A and B to ~150 microservices, the journey and learnings
28mMicroservices – Boyan Dimitrov