A lightweight performance measurement framework for a Virtual Machine on cloud

Due to advancement in technology and surge in usage of computing, many new areas are being developed and used. Cloud Computing is one among them and there is a significant growth in this area. Cloud computing provides different resources dynamically and economically over the internet. There are mainly three service models in the cloud computing stack - Infrastructure-as-a-Service (IaaS), Software-as-a-Service (SaaS), and Platform-as-a-Service (PaaS), each of which specifies different levels of capabilities provided to the consumer. Though there are numerous advantages and benefits that cloud computing offers, it also has many problems and challenges like reliability of service, performance issues, security etc. There is a lot of research work being done in most of these areas, but comparatively very less in performance profiling of cloud workloads which in turn provides reliability of service. In most of the applications running on cloud, performance and behavior of applications are essentially different from the traditional computing. In this paper, we have discussed about deploying a virtual machine on cloud and provide a framework to efficiently measure performance of applications running on Virtual Machine (VM). It monitors the variations in resource utilizations that helps in maintaining the health of the systems. In the developed framework, we have tested for few applications with varying CPU utilization levels. The results gives an overview for the service providers to decide the amount of resources to be allocated to the VM and provides service assurance. It is proved that performance monitoring will reduce the risk of failure and any sudden interruption in cloud service.