Migration Performance of Cloud Applications- A Quantitative Analysis

Abstract

Performance of a cloud data centre must abide the Service Level Agreement parameters and must provide negotiated Quality of Service values. One of the key areas in cloud computing, where the possibility of performance tuning is the maximum, is live virtual machine migration. Migration entails time and traffic which are crucial factors as far as the QoS is concerned. This paper focuses on quantitative analysis of live migration within a cloud data centre with the aim of understanding the factors which are responsible for cloud's efficiency. Various key parameters, such as, virtual machine size, network bandwidth available and dirty rate of a cloud application are discussed in detail and given the comparisons also, to give a clear view of their role in live migration's performance. The analysis presented in this paper gives a proper platform for considering future enhancements and/or modifications in the existing migration technology.