Building a microservice system of any significant size comes with challenges. One of the key challenges is managing a number of discrete processes in development. Tools like Fuge can help us to manage this complexity and accelerate our development experience.
Under the hood Fuge reads its configuration file to determine what processes it needs to manage. Using standard Node interfaces (such as process.stdin and process.stdout) along with a variety of modules (such as chokidar, chalk and ps-tree, and much more) Fuge provides an interactive execution environment for those processes and watches our code for changes, automatically reloading a service when a change occurs.
Not only is this highly useful when developing systems with a significant number of microservices, its also requires no service specific configuration on a developers part. This is modus operandi that Fuge embraces: enhanced developer experience of microservice (or other types of multiprocess) systems.
Fuge can also manage Docker containers locally for us, alongside the Node processes. This is the subject of a a recipe later in this chapter: Using Containerized Infrastructure.
It should be noted that Fuge is a development tool, something that is used locally. Fuge should not be used for running microservices in a production environment.
Take a look at Chapter 11, Deploying Systems for information on deploying distributed Node systems.