Font Size: a A A

The Research Of Resource Scheduling Technology In Cloud Computing

Posted on:2016-01-21Degree:MasterType:Thesis
Country:ChinaCandidate:L ZhangFull Text:PDF
GTID:2308330464464992Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Since the concept of cloud computing, cloud computing has attracted extensive attention, and has been developing rapidly. Cloud computing based on computer resource virtualization, by virtue of its "on-demand-pay, flexibility-of-use, low-pay-high-efficiency" properties can be used in a variety of scales for personal, researching, and business applications and so on.It is very attractive to users.The scheduling technic of cloud computing resources is key to ensuring the stable and efficient operation of cloud computing, and also is an important modules in the study of cloud computing.It has been the research topic in recent years.With development in the study of cloud computing, the emergence of new technology led to a more comprehensive, more complex cloud computing services. At the same time, users’ demands for the quality of service also rise sharply. In this situation, the existing scheduling must keep pace with the time, adopting a more comprehensive and more detailed scheduling algorithm or policy.In-depth research and analysis of traditional scheduling algorithm, we researched two scheduling algorithms:scheduling oriented to topology of virtual machines resource and scheduling oriented to response time in cloud computing with algorithm RTWSQPSO(Response Time Workflow Scheduling with QPSO).Algorithm of scheduling oriented to topology of virtual machines resource. At first, the model of cloud computing tasks and the model of VMs’ topology were built. Then a VM-Task fitness evaluation function was given, integrated with processing performance of VM and topological distance between VMs. After evaluation of function, the VM with excellent comprehensive performance were selected and assigned with the task. In comparison experiments with HEFT and DCP, the results showed that proposed algorithm had good adaptability for different types or different scales of tasks, the average makespan of proposed algorithm was smaller than that of HEFT and DCP when considering the topology of VMs.Algorithm of scheduling oriented to response time in cloud computing with algorithm RTWSQPSO. At first, the model of cloud computing workflow and the model of VM resources were built. Then the workflow scheduling response time constraint was introduced, as well as the influence of renting total cost on response time. Based on QPSO, introduced the "task-number-particle-coordinate dimension", "serial-number-particle-coordinate of virtual machine resources" of encoding and decoding scheme, and applied QPSO to cloud computing workflow scheduling, and RTWSQPSO algorithm was also put forward. Then, set meeting the response-time demand as the main target, as well as reducing total cost of rent for secondary objective solution strategy, and introduced fitness function fitness. Then realized the combination of RTWSQPSO particle evaluation with cloud computing workflow scheduling scheme.The results of experiment showed that RTWSQPSO was of feasibility, with strong optimization capability,at the same time, could meet response-time demand well and reduce scheduling total rent cost effect. In a word the algorithm RTWSQPSO was good, and had a better scheduling performance.
Keywords/Search Tags:cloud computing, vinual machine, topology, response time, scheduling algorithm
PDF Full Text Request
Related items