Skip to content

The aim of this repo is to implement a full server-side environment that manage entities with a potentially vast number of properties available for each.

License

Notifications You must be signed in to change notification settings

rfreydi/data-management-server

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 

Repository files navigation

data-management-server

The aim of this repo is to implement a full server-side environment that manage entities with a potentially vast number of properties available for each. To get a good subject I'll try to create an application that manage lifecycle in production factory, we will need to manage history, export, flexible entities and also real-time system, administration of the entities through an interface and more... The front part will be split into another repository, so we will implement a REST API to be able to switch from multiple front-end (Angular, React, VueJS).

Code Style

The code style of this project is based on the StandardJS code, but with some changes :

  • Semi-columns are mandatory
  • No space before functions declaration parentheses
  • No trailing comma

Technology used

Here is the current choice for the technology used. If you want to know why these choice as been made, look at the R&D history.

  • NodeJS
  • MySQL

NB: For NodeJS, it's the default choice since I love JavaScript and don't really like PHP syntax. The second reason is it need less skill to build backend and frontend if they're based on the same language. So even if I'm more comfortable with PHP 7 because I never used NodeJS before, I'll stick on JavaScript and acquire skill in NodeJS.

R&D

Here will be listed all sources that as been seem/read to make technical decision and what as been chosen/extracted from sources.

Subject to define later : Data Access Layer choice

Next subject: Message-based Interface Layer choice (GraphQL Implementation)

Possible sources :

2018-06-25: Message-based Interface Layer choice (RESTful API Framework)

I was looking for building a RESTful API, but was wondering which framework was better to do so. LoopBack 4 was pretty interesting but as for now it's on developer release and there is too much documentation missing for starting with it. I looked for ExpressJS and Restify, Restify looking better because it doesn't include useless feature for my case but as cons it look like it perform less than ExpressJS... So I decided to start with all of them (ExpressJS, Restify, LoopBack 3, LoopBack 4) side by side, and since I'm doing layering they should all access to the same business, forcing me to decouple my business from the interface at the same time.

But a friend said to me that there was something better than RESful API nowaday, I searched a bit and found GraphQL, Which look to fit my project perfectly, I'll look on it and it ecosystem (ArangoDB ?) before continue on RESTful API

Sources:

2018-05-27: Business Layer choice

The Business Layer is the core of the application, it'll be accessed through the Message-based Interface and will update send work to the Data Access Layer. To get a full business layer, it will be split into 3 parts: business workflow, business rules and business entities. They will communicate through a Chain of Responsibility. The business workflow will follow the Saga pattern.

Sources :

2018-04-15: Layers Architecture

My current knowledge on database access is through an Access Layer like DAO which I already used in a project and ORM which I didn't use it yet. I searched for the best way to access the database, globally and for the JavaScript/NodeJS. I found that it's better to build the Business Layer first and then exploit it to see if it's reliable before starting to create the database itself. Beside the choice of the Data Access Layer, it showed that we need to create an abstraction level between Business Layer and Data Access Layer to reduce dependencies. We will implement the architecture as follow using a top-down interaction approach :

  • Message-based Interface Layer (RESTful API)
  • Business Layer
  • Interface Layer (Abstract interface, Common design type, Dependency inversion)
  • Data Access Layer (ORM, DAO, ...)
  • Database (MySQL)

Sources:

2018-04-01: Data Model choice

I searched for an easy way to manage flexible properties and come across the Entity-Attribute-Value model (also named as vertical database model or open schema). After reading some article, I found one who spoke about the implementation of JSON Data Type to PostgreSQL and that it's perform better than EAV model. Perfect, it confirm my intention to use JSON Data Type for flexible properties. Beside this article, I also read one who showed the downside of using an EAV model

Sources:

2018-04-01: Database choice

My first though was to go for a NoSQL database, since we need a flexible way to add properties (so our code is generic as much as possible). But since we need to follow th ACID principles, NoSQL database seem to not be a good choice. Beside this specification MySQL released JSON Data Type in the v5.7.8, and will implement roles management in the v8, which seem to fit what I want to achieve: managing entities with flexible properties (and fixed properties) with access restriction on them. For the choice between MySQL or other RDBMS like PostgreSQL, it just a preference since I have a lot of knowledge on MySQL, I'll stick on it.

Sources:

About

The aim of this repo is to implement a full server-side environment that manage entities with a potentially vast number of properties available for each.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published