Intro

We all want our systems to be responsive, scalable, fault tolerant, maintainable and for the last, but not least, easy and fun to develop. With this goals in mind i decided to build a micro-services framework for nodejs using and architecture freely inspired on actors model. I present you Studio

Studio makes easy to create code without ANY dependency between your services, so you can deploy all in a single machine or just easily change to each one in a different machine or anything in between. It also enables operations timeouts, zero-downtime reload, let-it-crash approach (stop to be afraid of exceptions, Studio handles it to you), plugins and makes it nearly impossible to falls in a callback hell. Supports any web framework (we have examples with express) and helps you with flow-control using bluebird.

Studio encourages you to use the best practices of nodejs, it helps you to write simple, clean and completely decoupled code. And makes it very easy and fun.

First of all, everything in a Studio-based application is a service.

So if you're used to build SOA or micro-services all your services (and possible layers, as DAOs for instance) are going to be declared as a STATELESS SINGLETON services. Services have an unique identifier and communicate (always) asynchronously through message passing. The benefits of this approach is that it is really easy to take just some of your servers to different servers and make a better use of it. Also, your services have the free benefit of deep copying the parameters before the message is delivered (so one service can't mess with the objects of another service) increasing your code security.

And this is it... this is all you need to create reactive applications.

Why

Now you might be wondering why systems created with Studio can be called a reactive system. As stated by the reactive manifesto, reactive systems are those who follow 4 principles:

• Responsive : > Responsive systems focus on providing rapid and consistent response times, establishing reliable upper bounds so they deliver a consistent quality of service.

Using Studio you just add a thin layer over your functions without compromising the responsiveness while giving you the power to interact with your application in runtime as in aspect-oriented programming

• Resilient : > The system stays responsive in the face of failure. This applies not only to highly-available, mission critical systems any system that is not resilient will be unresponsive after a failure.

This is critical for thoses using nodejs, Studio enforces you to use the best practices to avoid your process or any of workers to crash. And as all your services are written with async flow in mind it also makes easy to add redundance

• Elastic : > The system stays responsive under varying workload. This is critical for Studio. All service calls are async so you never release zalgo, also every service call receives a copy of the parameters so a service cant mess with other service code. And for the last but not least using Studio plugins you can have measures of your code in realtime as using the timer plugin you can check the time needed to execute every single service call in your application and you can even send it easily for a statsd/grafana metrics dashboard. So this way you have an application ready to scale horizontally and also with the metrics to help you to decide when to do this.

• Message driven : > Reactive Systems rely on asynchronous message-passing to establish a boundary between components that ensures loose coupling, isolation, location transparency, and provides the means to delegate errors as messages.

All service calls in Studio are async, even if youre doing some sync code, Studio will make it run async. Also all call goes through the Studio router which enforces a deep clone of the parameters for security reasons, and all services are COMPLETELY DECOUPLED and isolated from each other

So the main reason to use Studio is because it makes it to reason about your code and make it scalable as hell.