Live Migration of Stateful Processes across Edge Servers

Abstract

Edge Computing facilitates low-latency for real- time applications. It can be achieved by installing computing infrastructure at the network edge or bringing cloud benefits closer to the user. In Edge Computing, computing resources are not more than one hop away from the user equipment. The mobility of a user in such a case can cause problems. If a user continues to move, maintaining low-latency becomes challenging. So, service is migrated between edge servers as the user moves. This migration can lead to a time duration for which the service is not available, called Downtime. While migrating, to leverage low- latency, Downtime must be kept to a minimum. In this paper, we introduced an approach to migrate service so that the minimum amount of data is transferred during Downtime. We have also discussed some existing techniques for migrating service and compared them based on how they reduce data to be transferred to lessen the Downtime. Our approach transfers lesser data that causes less Downtime than one of the existing techniques.