Name
Abstract | ||
|---|---|---|
In the past few decades, large-scale distributed computing systems such as grids have been widely used to serve a growing number of applications in a time-shared manner. In such environment, resources should be strictly assigned to achieve high performance. Hence, resource monitoring and usage prediction are required for the scheduling. Among these resources, CPU load has a significant effect on the performance. So prediction of CPU load plays an important role in the scheduling. In recent years, some research has been carried out in the field of CPU load prediction. Many prediction models were developed, such as Network Weather Service, the most popular performance prediction system. However, most of them adopt one-step-ahead or short-term prediction strategies, which cannot meet the requirement of the applications with much longer execution time. In this paper, we present a new long-term prediction model applying Fourier transform to exploit the periods of the CPU waves and using tendency-based methods to predict the variation. |
Year | DOI | Venue |
|---|---|---|
2011 | 10.1109/DASC.2011.28 | DASC |
Keywords | Field | DocType |
fourier transforms,high performance,long-term cpu load prediction,large-scale distributed computing system,new long-term prediction model,tendency-based method,fourier transform,tendency-based methods,cpu wave,network weather service,popular performance prediction system,long-term cpu load,long-term prediction,usage prediction,cpu load,resource allocation,short-term prediction strategy,resource monitoring,distributed computing,cpu load prediction,cpu waves,prediction model,distributed processing,time sharing,computer model,central processing unit,predictive models,time series analysis,meteorology,computational modeling | Time series,Central processing unit,Long-term prediction,Fair-share scheduling,Scheduling (computing),Computer science,Exploit,Real-time computing,Resource allocation,Performance prediction,Distributed computing | Conference |
ISBN | Citations | PageRank |
978-1-4673-0006-3 | 0 | 0.34 |
References | Authors | |
0 | 4 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Jianhuang Liang | 1 | 0 | 0.34 |
| Jian Cao | 2 | 0 | 0.68 |
| Jun-Gang Wang | 3 | 4 | 1.88 |
| Yuxia Xu | 4 | 0 | 0.34 |