Hierarchical Queue-Based Task Scheduling
The concepts of Cloud Computing provide users with infrastructure, platform and software as service, which make those services more accessible for people via Internet. To better analyze the performance of Cloud Computing provisioning policies as well as resources allocation strategies, a toolkit named CloudSim was proposed. With CloudSim, the Cloud Computing environment can be easily constructed by modeling and simulating cloud computing components, such as datacenter, host, and virtual machine. A good scheduling strategy is the key to achieve load balancing among different machines as well as to improve the utilization of basic resources. Recently, the existing scheduling algorithms may work well in some presumptive cases in a single machine; however they are unable to make the best decision for the unforeseen future. In real world scenario, there would be numbers of tasks as well as several virtual machines working in parallel. Based on the concepts of multi-queue, this paper presents a new scheduling algorithm to schedule tasks with CloudSim by taking into account several parameters, including the machines’ capacity, the priority of tasks and the history log.
Journal of Advances in Computer Networks
Digital Object Identifier (DOI)