Abstract | ||
---|---|---|
Many large-scale production parallel programs often run for a very long time and require data checkpoint periodically to save the state of the computation for program restart and/or tracing the progress. Such a write-only pattern has become a dominant part of an application’s I/O workload and implies the importance of its optimization. Existing approaches for write-behind data buffering at both file system and MPI I/O levels have been proposed, but challenges still exist for efficient design to maintain data consistency among distributed buffers. To address this problem, we propose a buffering scheme that coordinates the compute processes to achieve the consistency control. Different from other earlier work, our design can be applied to files opened in read-write mode and handle the patterns with mixed MPI collective and independent I/O calls. Performance evaluation using BTIO and FLASH IO benchmarks is presented, which shows a significant improvement over the method without buffering. |
Year | DOI | Venue |
---|---|---|
2005 | 10.1007/11557265_17 | PVM/MPI |
Keywords | Field | DocType |
efficient design,flash io benchmarks,write-behind data,mixed mpi,o level,cooperative write-behind data,dominant part,consistency control,mpi i,o workload,data consistency,buffering scheme | File system,Virtual machine,Workload,Computer science,Parallel computing,Input/output,Tracing,Message passing,Computation,Distributed computing,Data consistency | Conference |
Volume | ISSN | ISBN |
3666 | 0302-9743 | 3-540-29009-5 |
Citations | PageRank | References |
15 | 0.77 | 6 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Wei-keng Liao | 1 | 1095 | 87.98 |
Kenin Coloma | 2 | 137 | 9.36 |
Alok Choudhary | 3 | 322 | 31.06 |
Lee Ward | 4 | 185 | 9.81 |