Integrating logical and physical file models in the MPI-IO implementation for "Clusterfile"

• Published in: Sixth IEEE International Symposium on Cluster Computing and the Grid (CCGRID'06) Vol. 1; pp. 8 pp. - 462
• Main Authors: Isaila, F., Singh, D., Carretero, J., Garcia, F., Szeder, G., Moschny, T.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 2006
• Subjects: Application software; Computer architecture; Computer science; Design optimization; File systems; Grid computing; Libraries; Physics computing; Processor scheduling; Programming profession
• ISBN: 9780769525853; 0769525857
• DOI: 10.1109/CCGRID.2006.62

This paper presents the design and implementation of the MPI-IO interface for the Clusterfile parallel file system. The approach offers the opportunity of achieving a high correlation between the file access patterns of parallel applications and the physical file distribution. First, any physical file distribution can be expressed by means of MPI data types. Second, mechanisms such as views and collective I/O operations are portably implemented inside the file system, unifying the I/O scheduling strategies of the MPI-IO library and the file system. The experimental section demonstrates performance benefits of more than one order of magnitude.