A Predictive Method for Workload Forecasting in the Cloud Environment
Cloud computing provides powerful computing capabilities, and supplies users with a flexible pay mechanism, which makes the cloud more convenient. People are getting more and more usage of the cloud environment due to a steady increase of data. In order t
- PDF / 1,855,561 Bytes
- 9 Pages / 439.37 x 666.142 pts Page_size
- 85 Downloads / 221 Views
Abstract Cloud computing provides powerful computing capabilities, and supplies users with a flexible pay mechanism, which makes the cloud more convenient. People are getting more and more usage of the cloud environment due to a steady increase of data. In order to improve the performance and energy saving of the cloud computing, the efficiency of resource allocation has become an important issue. In this study, a neural network model with learning algorithm is applied to predict the workload of the cloud server. The resource manager deployed on the cloud server provides the service of managing the jobs with a resource allocation algorithm. With this prediction mechanism, cloud service providers can forecast the following time workload of cloud servers in advance. The experimental results show that resources can be allocated efficiently and become load balanced by proposed mechanism. Therefore, the cloud server can avoid the problem of inadequate resources. Keywords Cloud computing algorithm
Predictive workload Neural network Learning
Y.-C. Chang (&) Department of Computer Science and Information Engineering, National Taitung University, Taitung, Taiwan, Republic of China e-mail: [email protected] R.-S. Chang F.-W. Chuang Department of Computer Science and Information Engineering, National Dong Hwa University, Hualien, Taiwan, Republic of China e-mail: [email protected] F.-W. Chuang e-mail: [email protected]
Y.-M. Huang et al. (eds.), Advanced Technologies, Embedded and Multimedia for Human-centric Computing, Lecture Notes in Electrical Engineering 260, DOI: 10.1007/978-94-007-7262-5_65, Springer Science+Business Media Dordrecht 2014
577
578
Y.-C. Chang et al.
Introduction Numbers of physical and virtual resources provide a powerful computing ability in the cloud computing environment. When a user requests to execute the particular application, the resource manager of cloud computing environment will create a new virtual machine, and assign jobs to the host which operating under servers. Virtual resources can be added or removed at any time in the cloud computing environment. This characteristic makes the cloud platform more flexible and efficient. However, this feature also brings some issues, such as resource allocation imbalance and energy consumption in equable. On the other hand, the workload of the cloud environment may change frequently in a short period due to the large number of jobs or when only a small amount of new jobs is received from users. Furthermore, when the cloud server receives a large number of new jobs, it also results in the same situation in a short period of time. This varying workload may result a situation where there are not enough resources to be allocated or cause a resource imbalance. Therefore, a prediction method for cloud environments to allow the cloud provider to avoid this situation is needed. In this study, a prediction method to forecasting the workload of cloud environments based on a neural network is proposed. The workload information from the
Data Loading...
