臺灣博碩士論文加值系統

大量且集中的電腦，完整的網路環境，充分的電力系統，足夠的空調環境等，是構成高速運算叢集系統的必要原件與環境，這樣的環境條件，一般而言通常在營利的研發單位或專屬的研究機構中才能擁有，要建製一個這樣的叢集環境在一般的組織單位下，似乎沒有一個地方能夠輕易的滿足這樣的條件，除了學校的電腦教室，於是我們開始思考如何應用電腦教室教學之餘的閒置資源，配合無碟工作站的技術，來建立一套可以提供高效能運算所需環境的基礎平台，並且能以自動或手動的方式將電腦教室的資源在「教學」與「高效能計算所需環境」間作切換, 其中我們還以LAM/MPI與PVM平行運算程式以矩陣相乘的模式進行運算，來證明我們所建出來高效能運算的平台是可以正常工作，並以此實驗做出該平台效能量測的數據，另外我們還將該平台安裝生物計算軟體FASTA來執行基因序列的比對，以應證該平台可以實際的應用，此外我們還在該系統平台上建置一些系統與網路效能的監測軟體如 ntop、Ganglia，來做為未來系統效能調教的依據，最後我們在不更動學校電腦教室現有網路的環境下，以虛擬專屬網路-VPN 的方式，將多個電腦教室的資源串接起來，甚至能跨到不同的網段，把不同的電腦教室，以叢集的型態相連，形成一個大型高速運算的叢集環境，提供更多的運算資源。

The objective of this thesis is to design and implement a high-performance computing environment with idle Lab desktops on campus. A high-performance computing cluster system can only be built within an environment with large number of concentrated desktops, an intact network environment, abundant power system, sufficient air conditioning, etc.; those generally are possessed by the commercial research and development department, some special research institution, as well as labs on campus. So that we here start with tying desktops in the Lab in with the technology of diskless slave nodes to develop a high-performance computing platform which can be automatically or manually switched between “teaching” and “high-performance computing”. First of all, the efficiency of this platform is demonstrated by simulation results of the matrix multiply with LAM/MPI and PVM. The biological algorithm program, FASTA, is executed smoothly in the cluster architecture as well. In addition, the system and network monitoring software, ntop and ganglia, are installed in the PC clusters for the future need of system performance tuning. Last, desktops on different locations or even on different network are clustered via VPN without changing existing configuration of the network to form a large-scale high-performance computing environment to provide more computing power.

摘要 i
Abstract ii
Acknowledgements iii
Table of Contents iv
List of Tables vi
List of Figures vii
Chapter 1 Introduction 1
1.1 Motivation 1
1.2 Contribution 2
1.3 Thesis Organization 3
Chapter 2 4
Background 4
2.1 Preboot eXecution Environment 4
2.2 NIS and NIS + 4
2.3 Network File System 5
2.4 Cluster Computing 6
2.5 Diskless PC Cluster 7
2.6 Virtual Private Network 7
Chapter 3 9
Implementation 9
3.1 The Lab 9
3.2 Clustering Environment 12
3.3 Monitor Tool - Ganglia 14
3.4 Switch Function 14
3.5 Cross Different Lab via VPN 18
Chapter 4 Experimental Results 22
4.1 Diskless Network Performance 22
4.2 The Results for FASTA 26
4.3 Linking Clusters via VPN 27
4.4 Bioinformatics Performance Evaluation 28
Chapter 5 Conclusions and Future Work 31
References 34

[1]Hamilton Bioinformatics Portal, http://academics.hamilton.edu/biology/home/bioinformatics/bioserver.html
[2]B. Wilkinson and M. Allen, Parallel Programming: Techniques and Applications Using Networked Workstations and Parallel Computers, 2nd Edition, Prentice Hall PTR, NJ, 2005.
[3]C. T. Yang, S. S. Tseng, M. C. Hsiao, and S. H. Kao, “A Portable parallelizing compiler with loop partitioning,” Proceedings of the NSC ROC(A), Vol. 23, No. 6, 19999, pp. 751-765.
[4]Resources for DHCP, http://www.dhcp.org/
[5]DRBL, http://drbl.sourceforge.net/
[6]E. Choi, “Performance test and analysis for an adaptive load balancing mechanism on distributed server cluster systems,” Future Generation Computer Systems, Vol. 20, No. 2, 2004, pp. 237-247.
[7]G. R. Watson, M. J. Sottile, Ronald G. Minnich, Sung-Eun Choi, Erik A. Hendriks, “Pink: A 1024-node Single-System Image Linux Cluster,” Proceedings of High Performance Computing and Grid in Asia Pacific Region, Seventh International Conference on (HPCAsia’04), 2004, pp. 454-461.
[8]J. R. V. Behren, S. Czerwinski, A. D. Joseph, E. A. Brewer, J. Kubiatowicz, “The Design of a High-Performance Clustered, Distributed E-mail System”, Parallel Processing, 2000. Proceedings. 2000 International Workshops on, 21-24 Aug. 2000, pp. 151-158.
[9]LAM/MPI Parallel Computing, http://www.lam-mpi.org
[10]K. Limaye, B. Leangsuksun, V. K. Munganuru, Z. Greenwood, S. L. Scott, R. Libby, K. Chanchio, “Grid aware HA-OSCAR,” Proceedings of High Performance Computing Systems and Applications, 2005. HPCS 2005. 19th International Symposium (HPCS'05), 2005, pp. 333-339.
[11]M. L. Massie, B. N. Chun, D. E. Culler, “The ganglia distributed monitoring system: design, implementation, and experience,” Proceedings of Parallel Computing, Vol. 30, Issue: 7, Jul. 2004, pp. 817-840.
[12]Linux NFS, http://nfs.sourceforge.net/
[13]ntop.org, http://www.ntop.org/
[14]Network Booting of a diskless Linux device with PXE, http://www.linuxdevices.com/articles/AT5834950453.html
[15]R. Buyya, D. Abramson, and S. Venugopal, “The Grid Economy,” PROCEEDINGS OF THE IEEE, Vol. 93, No. 3, Mar. 2005, pp. 698-714
[16]R. Buyya, High Performance Cluster Computing: System and Architectures, Vol. 1, Prentice Hall PTR, NJ, 1999.
[17]R. Buyya, High Performance Cluster Computing: Programming and Applications, Vol. 2, Prentice Hall PTR, NJ, 1999.
[18]S. M. Diesburg, D. Joiner, P. A. Gray, “High Performance Computing Environments Without the Fuss: The Bootable Cluster CD,” Proceedings of the 19th IEEE International Parallel and Distributed Processing Symposium (IPDPS’05) Workshop 13, Vol. 14, pp. 252.
[19]TFTP, http://www.faqs.org/rfcs/rfc1350.html,
[20]T. L. Sterling, J. Salmon, D. J. Backer, D. F. Savarese, How to Build a Beowulf: A Guide to the Implementation and Application of PC Clusters, 2nd Printing, MIT Press, Cambridge, Massachusetts, USA, 1999.
[21]Virtual Private Network Daemon, http://vpnd.dotsrc.org