Abstract
For parallel computations requiring massive data Input/Output, one of the goals is to maximize the use of the underling storage topology, in particular exploit the benefit of using local disks. This paper presents a mechanism to distribute independent data records from multiple files to multiple processing nodes and vice-versa. An allocation problem is solved using the Max-Flow algorithm. We give timing results using various read/distribute protocols. One application is to redistributing tasks (regions) for restarting parallel numerical integration runs.
Chapter PDF
Similar content being viewed by others
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
References
ParInt web site, http://www.cs.wmich.edu/parint
MPI web site, http://www-unix.mcs.anl.gov/mpi/index.html
Cormen, T.H., Leiserson, C.E., Rivest, R.L.: Introduction to Algorithms. MIT Press, Cambridge (1994)
Cucos, L.: Load Sharing Strategies in Distributed Environments. PhD thesis, Western Michigan University (December 2003)
Zanny, R., de Doncker, E., Kaugars, K., Cucos, L.: ParInt1.2 User’s Manual, Available at http://www.cs.wmich.edu/parint
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2005 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Cucos, L., de Doncker, E. (2005). Parallel Files Distribution. In: Sunderam, V.S., van Albada, G.D., Sloot, P.M.A., Dongarra, J. (eds) Computational Science – ICCS 2005. ICCS 2005. Lecture Notes in Computer Science, vol 3516. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11428862_155
Download citation
DOI: https://doi.org/10.1007/11428862_155
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-26044-8
Online ISBN: 978-3-540-32118-7
eBook Packages: Computer ScienceComputer Science (R0)