Elasticity in cloud computing: State of the art and new challenges

Elasticity is undoubtedly one of the most striking features of cloud computing, appearing as a differential of such kind of distributed system in comparison with others like grid computing and peer-to-peer. Concerning both the paradigms of on-demmand computing and pay-as-you-go, elasticity offers the ability to dynamically increase or decrease the number of instances or computing nodes, as well as to reconfigure CPU, memory and network bandwidth data for a cloud service. Besides the obvious benefits of cost and performance for the user, the cloud provider also has the advantage of providing a better use of resources for the subscribers. In this context, this paper presents the state-of-art in the area of cloud elasticity, emphasizing since the standard approach which uses Web transactions to research initiatives on high performance computing. Furthermore, the article discusses about metrics for elasticity activation, its level of activity (SaaS, PaaS or IaaS), as well as the user interface (without user intervention, command line, graphical tool or application programming interface). For purposes of experimentation, we present a case study considering the elasticity topic applied to high performance applications. Finally, the article discusses the challenges and research opportunities in the area, both in the deployment of private and public clouds.