A survey and classification of the workload forecasting methods in cloud computing
- PDF / 3,873,267 Bytes
- 26 Pages / 595.276 x 790.866 pts Page_size
- 71 Downloads / 307 Views
(0123456789().,-volV)(0123456789(). ,- volV)
A survey and classification of the workload forecasting methods in cloud computing Mohammad Masdari1 • Afsane Khoshnevis1 Received: 8 July 2019 / Revised: 1 September 2019 / Accepted: 23 October 2019 Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract Workload prediction is one of the important parts of proactive resource management and auto-scaling in cloud computing. Accurate prediction of workload in cloud computing is of high importance for improving cloud performance, mitigate energy consumptions, meeting the required quality of service (QoS) level, predicting the energy consumption of data centers (DCs), and improving the cloud service providers’ scalability. However, in cloud computing context workload prediction is a challenging issue and various schemes using machine learning, data mining, and mathematical methods to deal with this issue. This scheme presents an extensive literature review of the workload prediction schemes proposed in the literature to improve resource management in the cloud DCs. It first provides the required knowledge regarding the workload prediction context and presents a taxonomy of the workload prediction schemes according to their applied prediction algorithm. Moreover, the main contributions of these schemes are illustrated and their major advantages and limitation are specified. At last, the open research opportunities in the workload prediction field are focused and the concluding remarks are presented. Keywords SVM ANN SVR Deep learning Collaborative filtering Ensemble
1 Introduction Cloud computing is a promising technology aimed to bring various visualized resources, software, and platforms as services to its customers based on the pay-for-use model [1]. To provide high-performance cloud services for endusers, conducting resource management in cloud DCs is of high importance [2, 3] and it can decrease the energy consumption costs as well as CO2 emissions [4, 5]. At general, resource management scheme can be classified as reactive and proactive categories which in the first case, when the workload increases/decreases to a predefined specific threshold, resource management will be conducted [6]. But, regarding the boot time of the VMs, the reactive method cannot deal with the sudden burst of the workload & Afsane Khoshnevis [email protected] Mohammad Masdari [email protected] 1
Department of Computer Engineering, Urmia Branch, Islamic Azad University, Urmia, Iran
[7] and may result in service level agreement (SLA) violations. On the other hand, proactive methods solve this problem by predicting the future workload of DC by recognizing the possible resource usage patterns and provisioning the required resource. Consequently, by effective prediction, the performance degradation can be deterred and idle resources can be reduced to further improve the profit. However, conducting proactive resource management is not a trivial process and variable workload of the cloud-hosted service
Data Loading...
