Affiliation:
1. Communication University of China
Abstract
Considering common mode failure (CMF) in the rendering cluster systems, the availability of rendering cluster systems with the increase of cluster’s number was studied. Firstly, based on availability of system with one cluster node, system with two cluster nodes was modeled with continuous time markov chain (CTMC) model. Then, the CTMC model was extended to the case of system with three cluster nodes. Furthermore, by solving these three CTMC models, availability for different cases were numerically deduced. Additionally, during one year the unavailable time for different cases was calculated and analysis by comparison was conducted. Finally, conclusions on different cases’ advantages and disadvantages are derived thereby, which offers theoretical foundations for establishing rendering cluster systems.
Publisher
Trans Tech Publications, Ltd.
Reference6 articles.
1. Information on http: /www. teragrid. org.
2. S. L. Gooding, L. Arns, P. Smith, et al.: Implementation of a Distributed Rendering Environment for the TeraGrid, in Proc. of IEEE Challenges of Large Applications in Distributed Environments (CLADE). Paris, France, (2006), pp.13-21.
3. X. Liu: Research of Dependability Evaluation of Cluster System, Master thesis, Graduate School of Chinese Academy of Sciences, Beijing. Dec. (2003). (In Chinese).
4. C. Leangsuksun, L. Shen, T. Liu, et al.: Dependability Prediction of High Availability OSCAR Cluster Server, In Proc. of 2003 International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA '03), Las Vegas, Nevada, USA, June 23-26. (2003).
5. M. D. David: QoS and Traffic Management in IP and ATM Networks. McGraw-Hill Professional Publishing: New York (2000).
Cited by
2 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献