Learn how we fixed a nasty bottleneck in a large Node.js Microservices infrastructure, and how we scaled it so it could serve 1000s of requests per second.