TY - GEN
T1 - Noncontiguous I/O accesses through MPI-IO
AU - Ching, Avery
AU - Choudhary, Alok
AU - Coloma, Kenin
AU - Liao, Wei Keng
AU - Ross, Robert
AU - Gropp, William
PY - 2003
Y1 - 2003
N2 - I/O performance remains a weakness of parallel computing systems today. While this weakness is partly attributed to rapid advances in other system components, I/O interfaces available to programmers and the I/O methods supported by file systems have traditionally not matched efficiently with the types of I/O operations that scientific applications perform, particularly noncontiguous accesses. The MPI-IO interface allows for rich descriptions of the I/O patterns desired for scientific applications and implementations such as ROMIO have taken advantage of this ability while remaining limited by underlying file system methods. A method of noncontiguous data access, list I/O, was recently implemented in the Parallel Virtual File System (PVFS). We implement support for this interface in the ROMIO MPI-IO implementation. Through a suite of noncontiguous I/O tests we compared ROMIO list I/O to current methods of ROMIO noncontiguous access and found that the list I/O interface provides performance benefits in many noncontiguous cases.
AB - I/O performance remains a weakness of parallel computing systems today. While this weakness is partly attributed to rapid advances in other system components, I/O interfaces available to programmers and the I/O methods supported by file systems have traditionally not matched efficiently with the types of I/O operations that scientific applications perform, particularly noncontiguous accesses. The MPI-IO interface allows for rich descriptions of the I/O patterns desired for scientific applications and implementations such as ROMIO have taken advantage of this ability while remaining limited by underlying file system methods. A method of noncontiguous data access, list I/O, was recently implemented in the Parallel Virtual File System (PVFS). We implement support for this interface in the ROMIO MPI-IO implementation. Through a suite of noncontiguous I/O tests we compared ROMIO list I/O to current methods of ROMIO noncontiguous access and found that the list I/O interface provides performance benefits in many noncontiguous cases.
UR - http://www.scopus.com/inward/record.url?scp=84887562601&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84887562601&partnerID=8YFLogxK
U2 - 10.1109/CCGRID.2003.1199358
DO - 10.1109/CCGRID.2003.1199358
M3 - Conference contribution
AN - SCOPUS:84887562601
SN - 0769519199
SN - 9780769519197
T3 - Proceedings - CCGrid 2003: 3rd IEEE/ACM International Symposium on Cluster Computing and the Grid
SP - 104
EP - 111
BT - Proceedings - CCGrid 2003
T2 - 3rd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2003
Y2 - 12 May 2003 through 15 May 2003
ER -