臺灣博碩士論文加值系統

Teamster-G為一個基於開放網格服務架構的系統。對服務導向的系統而言，維護系統的QoS是一項重要的議題。以往的相關研究經常採用預訂資源保留的方法來維護系統的QoS，卻因其所提出的方法並沒有將資源未來的可用度納入考量，導致系統的工作服務品質不佳。有鑑於此，本論文為Teamster-G的Workflow工作設計一套基於未來資源可用度的預訂資源保留機制。此預訂資源保留機制會依照Worflow的期限，有效地為每個Workflow工作設定一個合適的執行時間長度，而Resource Broker則會依照每個工作的起始時間與執行時間長度以及資源未來可用度的預測為每個工作預留適當的計算資源。為了讓Workflow即使在資源可用度變化超乎預期時仍能順利地在Deadline之前完成，預測器會使用比預測值還高的負載狀態來評估每一個計算資源的未來可用度。本論文所提出的機制與其效能已經在Teamster-G上實現與評估。

Teamster-G is a grid system based on Open Grid Service Architecture. QoS is a very important issue for a service-oriented grid system. Although many related studies have proposed methods for the QoS problem of Grids, all of these methods do not take future resource availability into account. As a result, the QoS of user jobs is effectively maintained due to dynamic resource availability. To overcome this drawback, this paper is aimed at design and implementation of a resource reservation scheme based on future resource availability for the workflow jobs on Teamster-G. When this scheme is applied, each job of a workflow will be assigned a start time, and an end time based on the deadline of this workflow, and then the resource broker of Grids will reserve a proper resource for each of these jobs based on the period between its start time and end time, and the prediction of a long termed resource availability predictor. In order to maintain the QoS of the jobs, the predictor will forecast the future availability of each resource by a higher load value than that it predicted for this resource. The proposed scheme has been implemented on Teamster-G, and the effectiveness of this scheme has been evaluated in this study.

目錄…........................................................................................................................... iv
圖目錄............................................................................................................................ v
表目錄........................................................................................................................ viii
第1 章 緒論............................................................................................................. - 1 -
1.1 研究動機與目的........................................................................................ - 1 -
1.2 論文架構.................................................................................................... - 2 -
第2 章 相關研究與背景......................................................................................... - 3 -
2.1 相關研究.................................................................................................... - 3 -
2.2 GT4 ............................................................................................................. - 6 -
2.2.1 Globus Toolkit 4 規範 .............................................................................. - 7 -
2.2.2 Globus Web Service 的標準規格 ............................................................ - 7 -
第3 章 Teamster-G 系統架構 ............................................................................... - 12 -
3.1 資源預訂排程器(Reservation Scheduler) ............................................... - 13 -
3.2 未來資源可用度預測器.......................................................................... - 22 -
3.3 Resource Broker、Agent Server 與Reservation Server .......................... - 26 -
第4 章 系統實作................................................................................................... - 29 -
4.1 實現平台Platform................................................................................... - 29 -
4.2 Grid Resource Broker or Schedulers ........................................................ - 31 -
4.2.2 Resource Broker..................................................................................... - 34 -
4.2.3 Resource Monitor................................................................................... - 42 -
4.3 Local Server .............................................................................................. - 42 -
4.3.1 Reservation Server ................................................................................. - 43 -
4.3.2 Agent Server........................................................................................... - 43 -
第5 章 效能評估................................................................................................... - 45 -
5.1 實驗環境與方式....................................................................................... - 45 -
5.2 實驗評估的結果...................................................................................... - 48 -
第6 章 結論與未來工作....................................................................................... - 88 -
參考文獻................................................................................................................. - 89 -

[1] Tyng-Yeu Liang, Chun-Yi Wu, Jyh-Biau Chang, Ce-Kuen Shieh, “Teamster-G : A Grid-enabled Software DSM System”, In of the Proceeding of The 2005 International Workshop on Distributed Shared Memory on Clusters at the IEEE International Symposium on Cluster Computing and the Grid (CCGrid 2005), pp. 905-912, Cardiff, UK, May 9-12, 2005.
[2] J. B. Chang, Ce-Kuen Shieh, Tyng-Yeu Liang, “A Transparent Distributed Shared memory for Clustered Symmetric Multiprocessors”, Journal of Supercomputing, vol. 37, issue 2, pp. 145-160, August 2006.
[3] Cristiana A., Alan L. C., Sandhya D., Pete K., Honghui L., Ramakrishnan R., Weimin Y., Willy Z., “TreadMarks: Shared Memory Computing on Networks of Workstations”, IEEE Computer, pp. 18-28, 1996.
[4] Speight E. and Bennett J.K., “Brazos: A third generation DSM system”, Proceedings of the 1997 USENIX Windows/NT Workshop, pp. 95-106, 1997.
[5] IVY K Li,“A shared virtual memory system for parallel computing”, In Proceedings of the 1988 International Conference on Parallel Processing (ICPP'88), pp. 94-101, 1988.
[6] Carter J.B., Bennett J.K. and Zwaenepoel W., “Implementation and Performance of Munin”, In Proceedings of 13th ACM Symposium on Operating System Principles, pp. 152-164, 1991.
[7] Tyng-Yeu Liang, We-Chen Li, Jin-An Huang, “Developing A Workflow Management System for Teamster-G”, Proceedings of National Computer Symposium, pp. 7-16, Nov. 27-28, 2009.
[8] Tyng-Yeu L., I-Han W., Sheng-Yuan C., “A Long-term Resource Availability Predictor Using Frequent Workload Patterns”, The 5th Workshop on Grid Technologies and Applications (WoGTA’08), pp. 125-130, 2008.
[9] Jia Yu, Rajkumar Buyya and Chen Khong Tham, “Cost-based Scheduling of Scientific Workflow Applications on Utility Grids”, The First International Conference on e-Science and Grid Computing, pp. 140-147, 1-1 July 2005.
[10] TORQUE Resource Manager. http://www.clusterresources.com/pages/products/torque-resourcemanager.php.

[11] Hidemoto Nakada, Atsuko Takefusa, Katsuhiko Ookubo, Makoto Kishimoto Tomohiro Kudoh, Yoshio Tanaka, Satoshi Sekiguchi, “Design and Implementation of a Local Scheduling System with Advance Reservation for Co-allocation on the Grid”, The Sixth International Conference on Computer and Information Technology at IEEE. pp. 65-65, Sept 2006.
[12] Globus project. http://www.globus.org/toolkit/docs/4.2/4.2.1/
[13] I. Foster. Globus toolkit version 4: Software for serviceoriented systems. In IFIP International Conference on Network and Parallel Computing, Springer-Verlag LNCS 3779, pp. 2-13, 2005.
[14] OASIS Web Services Resource Framework.
http://www.oasis-open.org/committees/tc_home.php?wg_abbrev=wsrf
[15] Foster I., Kesselman C., “The Grid2: Blueprint for a New Computing Infrastructure”, Morgan Kaufmann Publishers, 2004.
[16] Foster I., Kesselman C., “The Anatomy of The Grid: Enabling Scalable Virtual Organizations”, International Journal of High Performance Computing Applications, pp. 200-222, 2001.
[17] Foster I., Kesselman C., Nick J.M., Tuecke S., “Grid Services for Distributed System Integration”, IEEE Computer, pp. 37-46, 2002.
[18] Tyng-Yeu L., I-Han W., Sheng-Yuan C., “A Long-term Resource Availability Predictor Using Frequent Workload Patterns”, The 5th Workshop on Grid Technologies and Applications (WoGTA’08), pp. 125-130, 2008.
[19] Jonas X. Y., “Liferay Portal Enterprise Intranets”, Packt Publishing, 2008.
[20] Huang W., Huang C. L., and Wu C. H., “The Development of a Computational Grid Portal”, The 6th IEEE International Symposium on Cluster Computing and the Grid, pp. 56, 2006.
[21] Liferay portal enterprise, http://www.liferay.com, accessed 2005.
[22] Matthew L. M., Brent N. C., David E. C., “The Ganglia Distributed Monitoring System: Design, Implementation, and Experience”, Parallel Computing, pp. 817-840, 2004.
[23] Foster I, Kesselman C (eds.), “The Grid: Blueprint for a Future Computing Infrastructure”, Morgan Kaufmann: San Mateo, CA, 1999.
[24] Oram A (ed.), “Peer-to-Peer: Harnessing the Power of Disruptive Technologies”, O’Reilly, 2001.
[25] Buyya R (ed.), “High Performance Cluster Computing: Architectures and Systems”, vol. 1, Prentice-Hall: Englewood Cliffs, NJ, 1999.
[26] Condor High Throughput Computing. http://www.cs.wisc.edu/condor/

[27] Rajkumar Buyya, Manzur Murshed, “GridSim: a toolkit for the modeling and simulation of distributed rsource management and scheduling for Grid computing”, pp. 1175-1220, Concurrency and Computation: Practice and Experience, Wiley, May 2002.