is an interesting presentation by Jimmy Bogard about a big and well known computer factory failed to build a microservice architecture. After the first attempt, which took 18 months to build with hundreds of developers, it took 9,5 minutes before the first page was presented on their brand new website
With the help of Jimmy they decided they would change one main service – the search engine to a completely new design:
Instead of making 200 API-calls they reversed the arrows and collected and stored all the needed data close to the service and sure, it meant duplication of data, but the response time went down to only 1-2 sec.
This success for the search service meant sadly that they took the same model and applied it all across all divisions.
The driving force behind this architecture was that the only way to be promoted as a manager was to grow and keep adding people to your own department. The solution architecture went from being small and focused to become many ”big balls of mud”.
Within Microservice architecture you often come across examples of Conways law:
Which interestingly is summaries in Jimmy law reagarding architecture:
Is the opposite true? If you start with a healthy and good organization with the right incentives – then the micro service architecture will follow.