Project Description

This project aims to develop a scalable cloud-based web application that performs image recognition in a hybrid cloud environment using AWS and OpenStack. The application dynamically manages resources by scaling up or down the number of processing instances based on the volume of incoming user requests. Key AWS services utilized include EC2 for computation, SQS for message queuing, and S3 for storage, while OpenStack is used to create a private cloud infrastructure. The web app is designed with three main components: a web tier for handling user requests, a controller to manage auto-scaling, and an app tier that processes image recognition tasks. Auto-scaling ensures that the system can handle varying loads efficiently by launching or terminating instances as needed.

The project demonstrates how cloud infrastructure can be optimized to manage workloads dynamically, ensuring cost efficiency and resource availability. The team also implemented a caching mechanism to reduce response times and enhanced the auto-scaling logic to overcome issues like inconsistent queue reporting and unnecessary polling of SQS queues. This application provides a practical solution for handling high-traffic workloads in a cloud environment, showcasing key concepts of hybrid cloud computing, image processing, and resource optimization.