# TeamCodeRing Team Members Nethra C Sashikar Neha Bisht Prashath Kumar Murali Parth Patel

Weather Prediction Application

This is a weather prediction application based on a micro-service architecture. There are 4 services:

1. Anemometer - The API gateway , a Ruby on Rails application
2. Rain Gauge - The Data Ingestor , a Python Flask application
3. Weather Ballon - The Storm Predictor , a Spark application
4. Satellite - The registry , a Node JS application (Under Construction)

The data is stored in Mysql Database and Redis is used for persistence storage. The registry solely revolves around the Redis and every application writes identifyable keys to redis as and when their work is done.

Application Workflow

The UI for all is present in the anemometer app. But the actual processing takes place in the microservices. The app takes in a station and a time for processing data and passes it on to the data ingestor application which forms the URL and fetches data. This is then passed on the storm prediction and clustering app which will predict the storm based on the data. The result will then be posted to redis from which the gateway can pick up and present it to the user. In every application, the step completion is recorded in the redis with a unique identifier. The registry simply pulls in the keys, segregates based on the identifiers and will be able to present the status. Currently the data ingestor and API gateway are fully connected and functional. The storm predictor is available but not connected. The redis has not been used as the registry has not been built. The app can form the data url with User authentication based on Authlogic.