Forecast-Microservices

By Melek Elloumi, Mohamed Ben Salah and Mohamed Karim Mellouli, GL4

Demo:

• This is a demo of a forecast-planning application to study a microservice-based architecture.

• The demo is composed of 3 microservices made with Flask in python.

  • Forecast UI for web interface
  • Forecast Backend for fetching data from mongodb database
  • Security for authentification

• We chose flask because it's a microframework, it's lightweight and quick to setup api routes, and it's based on python which opens up a lot of libraries for forecasting.

  Execution

  

Architecture:

• This is part of a software architecture lab

• The application is devided into 3 main applications, forecast and planning applications are hosted and provide 2 different services and a headless application that runs every week to generate forecasts and plans in the databases.

• This solution have non-functional requirements: security, maintability, flexibility and performance.

• The microservice architecture is best suited for this type of solution:

  • We can reuse planning and forecast applications for other solutions that require these modules.
  • It makes increasing the scalabilty easy to manage which improve performances
  • We can apply non-functional implementations like security for any microservice.

• This image show the architectural structure of the project. The demo contains only 1 use case of the forecast application alongside security.