Optimizing load balancing for edge applications

Abstract

Hybrid clouds have gained popularity in recent years due to their ability to run applications across the on-premises data-center/edge and public clouds. The presence of edge cloud has significantly decreased latency and help provide greater reliability for latency sensitive applications such as Autonomous navigation (using cars, drones etc). Running application workloads on the edge has further helped reduce resource consumption like bandwidth and organizational costs by not utilizing cloud resources. However, poor load balancing decisions within the edge cloud can overwhelm resources, causing substantial latency and possible outages for users. This compromises on user experience and decreases revenue. In this work I am working on improving the existing load balancing strategies used within the existing edge cloud frameworks.