This paper proposes a novel, hypergraph partitioning based strategy to schedule multiple data analysis tasks with batch-shared I/O behavior. This strategy formulates the sharing of files among tasks as a hypergraph to minimize the I/O overheads due to transferring of the same set of files multiple times and employs a dynamic scheme for file transfers to reduce contention on the storage system. We experimentally evaluate the proposed approach using application emulators from two application domains; analysis of remotely-sensed data and biomedical imaging.
%0 Conference Paper
%1 khanna2005hypergraph
%A Khanna, G.
%A Vydyanathan, N.
%A Kurc, T.
%A Catalyurek, U.
%A Wyckoff, P.
%A Saltz, J.
%A Sadayappan, P.
%B Cluster Computing and the Grid, 2005. CCGrid 2005. IEEE International Symposium on
%D 2005
%K batch hypergraph partition scheduling
%P 792--799
%R 10.1109/CCGRID.2005.1558643
%T A hypergraph partitioning based approach for scheduling of tasks with batch-shared I/O
%V 2
%X This paper proposes a novel, hypergraph partitioning based strategy to schedule multiple data analysis tasks with batch-shared I/O behavior. This strategy formulates the sharing of files among tasks as a hypergraph to minimize the I/O overheads due to transferring of the same set of files multiple times and employs a dynamic scheme for file transfers to reduce contention on the storage system. We experimentally evaluate the proposed approach using application emulators from two application domains; analysis of remotely-sensed data and biomedical imaging.
@inproceedings{khanna2005hypergraph,
abstract = {This paper proposes a novel, hypergraph partitioning based strategy to schedule multiple data analysis tasks with batch-shared I/O behavior. This strategy formulates the sharing of files among tasks as a hypergraph to minimize the I/O overheads due to transferring of the same set of files multiple times and employs a dynamic scheme for file transfers to reduce contention on the storage system. We experimentally evaluate the proposed approach using application emulators from two application domains; analysis of remotely-sensed data and biomedical imaging.},
added-at = {2012-12-11T02:25:44.000+0100},
author = {Khanna, G. and Vydyanathan, N. and Kurc, T. and Catalyurek, U. and Wyckoff, P. and Saltz, J. and Sadayappan, P.},
biburl = {https://www.bibsonomy.org/bibtex/2754a8c0b19aa402e974951424ff100b9/ytyoun},
booktitle = {Cluster Computing and the Grid, 2005. CCGrid 2005. IEEE International Symposium on},
doi = {10.1109/CCGRID.2005.1558643},
interhash = {d9a9862a3d627046c148b2c8bd8b7812},
intrahash = {754a8c0b19aa402e974951424ff100b9},
keywords = {batch hypergraph partition scheduling},
organization = {IEEE},
pages = {792--799},
timestamp = {2012-12-11T02:26:18.000+0100},
title = {A hypergraph partitioning based approach for scheduling of tasks with batch-shared I/O},
volume = 2,
year = 2005
}