The next phase of deploying an application for production use is ensuring that you're fully utilizing the system resources available to you, and ensuring that the service can scale to meet traffic requirements.

V8 is a single thread JavaScript engine. This is good enough for the Chrome browser but it means a Node-based server on that shiny new 16 core server will have 1 CPU core going flat out, and 15 CPU cores sitting idle. Your manager may want an explanation for this.

A single thread process will only use one core. That's a fact of life. Using multiple cores in a single process requires multi-threaded software. But Node's no threads design paradigm, while keeping the programming model simple, also means that Node does not make use of multiple cores. What are you to do? Or more important, how are you to keep your manager happy?

The general technique is to share out incoming requests to several processes, or threads, depending on application server...