Abstract
Recent studies have demonstrated that significant I/O operations are performed by a number of different classes of parallel applications. Appropriate I/O management strategies are required however for harnessing the power of parallel I/O. This paper focuses on two I/O management issues that affect system performance in multiprogrammed parallel environments. Characterization of the I/O behavior of parallel applications in terms of four different models is discussed first, followed by an investigation of the performance of a number of different data distribution strategies. Using computer simulations, this research shows I/O characteristics of applications and data distribution have an important effect on system performance. Applications which can simultaneously do computation and I/O, plus strategies that can incorporate centralized I/O management are found to be beneficial for a multiprogrammed parallel environment.
Preview
Unable to display preview. Download preview PDF.
References
P.F. Corbett, D.G. Feitelson, J.-P. Prost, S.J. Baylor, “Parallel Access to Files in the Vesta File System”, Proc. Supercomputing '93 Conf., 1993, pp. 472–481.
R. Cypher, P. Messina, “Architectural Requirements of Parallel Scientific Applications with Explicit Communication”, Proc. International Symp. on Computer Architecture, 1993.
J.M. Del Rosario, R. Bordawekar, A. Choudhary, “Improved Parallel I/O via a Two-Phase Run-Time Access Strategy”, ACM Computer Architecture News, Vol. 21, No. 5, December 1993, pp. 31–38.
J.M. Del Rosario, A. Choudhary, “High Performance I/O for Massively Parallel Computers”, IEEE Computer, March 1994, pp. 59–68.
R. Jain, K. Somalwar, J. Werth, J.C. Browne, “Scheduling Parallel I/O Operations”, ACM Computer Architecture News, Vol. 21, No. 5, December 1993, pp. 47–54.
M.Y. Kim, “Synchronized Disk Interleaving”, IEEE Trans on Computers, vol 35, November 1986, pp. 972–988.
D. Kotz, N. Nieuwejaar, “Dynamic File Access Characteristics of a Production Parallel Scientific Workload”, Proc of Supercomputing '94, November 1994, pp. 640–649.
D. Kotz, N. Nieuwejaar, et al.,“File-Access Characteristics of Parallel Scientific Workloads”, Tech Report PCS-TR95-263, Dept. of Math. and Comp. Science, Dartmouth College, Hanover, U.S.A, 1995.
P. Kwong, “Management of Parallel I/O in Multiprogrammed Parallel Systems”, M. Eng Thesis, Carleton University, Ottawa, Canada. To be available in the Fall of 1996.
S. Leuttenegger and M. Vernon, “The Performance of Multiprogrammed Multiprocessor Scheduling Policies”, Proc. 1990 ACM SIGMETRICS Conf. on Measurement and Modeling of Computer Systems, Boulder (CO), May 1990, pp. 226–236. College, Hanover, U.S.A.
M. Livny, “Multi-Disk Management Algorithms”, Proc. 1988 ACM SIGMETRICS Conf. on Measurement and Modeling of Computer Systems, May 1987, pp. 69–77.
S. Majumdar, D.L. eager, and R.B. Bunt, “Scheduling in Multiprogrammed Parallel Systems”, Proc 1988 ACM SIGMETRICS Conf on Measurement and Modeling of Computer Systems, Santa Fe, New Mexico, May 1988, pp. 104–113.
S. Majumdar, D.L. Eager and R.B. Bunt, “Characterization of Programs for Scheduling in Multiprogrammed Parallel Systems”, Performance Evaluation, Vol. 13 (1991) Elsevier North-Holland.
S. Majumdar, Y.M. Leung, “Characterization of Applications with I/O for Processor Scheduling in Multiprogrammed Parallel Systems”, Proc. Sixth IEEE Symp. on Parallel and Distributed Processing, Dallas, 1994, pp. 298–307.
S. Majumdar, F. Shad,“Characterization and Management of I/O on Multiprogrammed Parallel Systems”, Proc. seventh IEEE symposium in Parallel & Distributed Processing, San Antonio, Oct. 1995.
D.A. Patterson, G. Gibson, R.H. Katz, “A Case for Redundant Arrays of Inexpensive Disks (RAID)”, Proc. ACM SIGMOD Conference, June 1988, pp. 102–116.
Y.N. Patt, “The I/O Subsystem: A Candidate for Improvement”, IEEE Computer, March 1994, pp. 15–16.
A.L.N. Reddy, P. Banerjee, “An Evaluation of Multiple-Disk I/O Systems”, IEEE Trans. on Computers, Vol. 38, No. 12, December 1989, pp. 1680–1690.
K.C. Sevcik, “Application Scheduling and Processor Allocation in Multiprogrammed Parallel Systems”, Performance Evaluation, 1994.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1996 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Kwong, P., Majumdar, S. (1996). Study of data distribution strategies for parallel I/O management. In: Böszörményi, L. (eds) Parallel Computation. ACPC 1996. Lecture Notes in Computer Science, vol 1127. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-61695-0_2
Download citation
DOI: https://doi.org/10.1007/3-540-61695-0_2
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-61695-5
Online ISBN: 978-3-540-70645-8
eBook Packages: Springer Book Archive