A cluster heartbeat algorithm with ARMA-based prediction

Author

Chaoqun Xu ; Yi Zhuang ; Wei Zhu

Author_Institution

Sch. of Comput. Sci. & Technol., Nanjing Univ. of Aeronaut. & Astronaut., Nanjing, China

fYear

2013

fDate

12-13 Oct. 2013

Firstpage

218

Lastpage

221

Abstract

The high availability of the cluster system for supporting an overall platform seems more and more important with the increasing development of cloud computing. In this paper, a cluster heart beat algorithm with ARMA-based prediction (ARMA-HB) is proposed for the high availability of heterogeneous cluster. It records the historical data about the arrival time of the heartbeat packets from other nodes, and then predicts the arrival time of the next heartbeat packet from corresponding node in ARMA. The experiment shows that, compared with DHB algorithm, ARMA-HB can predict more accurately.

Keywords

autoregressive moving average processes; cloud computing; workstation clusters; ARMA-HB; ARMA-based prediction; DHB algorithm; auto-regressive and moving average model; cloud computing; cluster heartbeat algorithm; cluster system; heartbeat packets arrival time; heterogeneous cluster; historical data; Biomedical monitoring; Clustering algorithms; Delays; Heart beat; Monitoring; Prediction algorithms; Predictive models; ARMA; Heartbeat detection; cluster; heterogeneous;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Science and Network Technology (ICCSNT), 2013 3rd International Conference on

Conference_Location

Dalian

Type

conf

DOI

10.1109/ICCSNT.2013.6967099

Filename

6967099