TY - GEN
T1 - Towards a high performance implementation of MPI-IO on the lustre file system
AU - Dickens, Phillip
AU - Logan, Jeremy
PY - 2008
Y1 - 2008
N2 - Lustre is becoming an increasingly important file system for large-scale computing clusters. The problem is that many data-intensive applications use MPI-IO for their I/O requirements, and it has been well documented that MPI-IO performs poorly in a Lustre file system environment. However, the reasons for such poor performance are not currently well understood. We believe that the primary reason for poor performance is that the assumptions underpinning most of the parallel I/O optimizations implemented in MPI-IO do not hold in a Lustre environment. Perhaps the most important assumption that appears to be incorrect is that optimal performance is obtained by performing large, contiguous I/O operations. Our research suggests that this is often the worst approach to take in a Lustre file system. In fact, we found that the best performance is sometimes achieved when each process performs a series of smaller, non-contiguous I/O requests. In this paper, we provide experimental results showing that such assumptions do not apply in Lustre, and explore new approaches that appear to provide significantly better performance.
AB - Lustre is becoming an increasingly important file system for large-scale computing clusters. The problem is that many data-intensive applications use MPI-IO for their I/O requirements, and it has been well documented that MPI-IO performs poorly in a Lustre file system environment. However, the reasons for such poor performance are not currently well understood. We believe that the primary reason for poor performance is that the assumptions underpinning most of the parallel I/O optimizations implemented in MPI-IO do not hold in a Lustre environment. Perhaps the most important assumption that appears to be incorrect is that optimal performance is obtained by performing large, contiguous I/O operations. Our research suggests that this is often the worst approach to take in a Lustre file system. In fact, we found that the best performance is sometimes achieved when each process performs a series of smaller, non-contiguous I/O requests. In this paper, we provide experimental results showing that such assumptions do not apply in Lustre, and explore new approaches that appear to provide significantly better performance.
KW - Collective I/O
KW - Lustre
KW - MPI-IO
KW - Parallel File Systems
KW - Parallel I/O
UR - http://www.scopus.com/inward/record.url?scp=58049111206&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-88871-0_61
DO - 10.1007/978-3-540-88871-0_61
M3 - Conference contribution
AN - SCOPUS:58049111206
SN - 3540888705
SN - 9783540888703
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 870
EP - 885
BT - On the Move to Meaningful Internet Systems
T2 - OTM 2008 Confederated International Conferences CoopIS, DOA, GADA, IS, and ODBASE 2008
Y2 - 9 November 2008 through 14 November 2008
ER -