Workload prediction in load balancing and resource management system

Cloud computing is becoming the primary source of computing power, which has the advantage of high availability, high flexibility, low cost, dynamic resource sharing. Workload balancing is necessary in cloud computing, aiming at using resources at the most balanced. However, virtualization technology used in cloud computing will generate a lot of virtual resources, which will easily cause workload imbalance in cloud environment, making some computing nodes overburdened while some are unoccupied. So we propose a workload balancing and resource management system, by workload prediction and resource adjustments, workload in each virtual machine can be better balanced. In this study, we focus on the workload prediction module. Experimental results demonstrate that our prediction module can get relatively accurate prediction results, which can make big contribution to the workload balancing and resource management system. © 2013 Asian Network for Scientific Information.