An Approach to Evaluate Applications Running on Web-Based Remote Virtual Machines in Cloud Computing

Abstract

The exploitation of remote Virtual Machines (VM) over the Cloud gives new opportunities to both business and private clients for extending their processing and storage infrastructures. Typically, a remote VM can be accessed through consolidated technologies based on SSH or Remote Desktop Viewer. Regarding Remote Desktop Viewer, several solutions are emerging on the market to access remote VMs via web. However, evaluating performance of desktop applications running on remote VMs is quite hard, because CPU clocks of VMs are typically misaligned from physical devices. In addition, the overhead due to network latency can considerably degrade performance in terms of application responsivity. In this paper, we propose a new methodology that can help scientists and software developers to evaluate performance of applications running on Remote Web-Based VMs (RWVMs) over the Cloud by means of estimating the LAtency Gap (LAG). We use such a methodology to analyze the behavior of a Cloud IaaS considering two different applications running on RWVMs, which can be widely adopted for IoT purposes.