How to deal with more than 40,000 requests per minute. Over the years, we at Rabobank built a vast and complex microservices architecture. The entry point to this landscape, our edge service, stands in front of nearly all of our microservices’ consumer facing traffic.
As we grew, we foresaw that the current edge service wouldn’t hold up to future customer traffic demands in a sustainable way. So we started looking for better solution which would involve the replacement of the edge service.
Throughout this session we will take you on a journey into our struggles and learnings with performance engineering, critical zero-downtime deployments, technical migrations from Zuul to Spring Cloud Gateway, Spring Boot 1.5 to 2.1, imperative to reactive, and much more. During this talk we’d like to show you the importance of lifecycle management taking one step at a time.
Using bleeding edge software sounds awesome but it’s called bleeding for a reason.