Parallel and Distributed Processing Symposium, International (2008)
Miami, FL, USA
Apr. 14, 2008 to Apr. 18, 2008
Seetharami Seelam , IBM Thomas J. Watson Research Center Yorktown Heights, New York 10598 USA
I-Hsin Chung , IBM Thomas J. Watson Research Center Yorktown Heights, New York 10598 USA
Ding-Yong Hong , Department of Computer Science National Tsing Hua University HsinChu 300, Taiwan
Hui-Fang Wen , IBM Thomas J. Watson Research Center Yorktown Heights, New York 10598 USA
Hao Yu , IBM Thomas J. Watson Research Center Yorktown Heights, New York 10598 USA
On todays massively parallel processing (MPP) supercomputers, it is increasingly important to understand I/O performance of an application both to guide scalable application development and to tune its performance. These two critical steps are often enabled by performance analysis tools to obtain performance data on thousands of processors in an MPP system. To this end, we present the design, implementation, and early experiences of an application level I/O tracing library and the corresponding tool for analyzing and optimizing I/O performance on Blue Gene (BG) MPP systems. This effort was a part of IBM HPC Toolkit for BG systems. To our knowledge, this is the first comprehensive application-level I/O monitoring, playback, and optimizing tool available on BG systems. The preliminary experiments on popular NPB BTIO benchmark show that the tool is much useful on facilitating detailed I/O performance analysis.
Hui-Fang Wen, I-Hsin Chung, S. Seelam, Ding-Yong Hong and Hao Yu, "Early experiences in application level I/O tracing on blue gene systems," 2008 IEEE International Parallel & Distributed Processing Symposium(IPDPS), Miami, FL, 2008, pp. 1-8.