Abstract | ||
---|---|---|
The processing power and memory capacity of indepen- dent and heterogeneousparallel machines can be combined to form a single parallel system that is more powerful than any of its constituents. However, achieving satisfactory application performance on such a metacomputer is hard because the high latency of inter-machine communication as well as differences in hardware of constituent machines may introduce various types of wait states. In our earlier work, we have demonstrated that automatic pattern search in event traces can identify the sources of wait states in par- allel applications running on a single computer. In this article, we describe how this approach can be extended to metacomputing environments with special emphasis on performance problems related to inter-machine communi- cation. In addition, we demonstrate the benefits of our so- lution using a real-world multi-physics application. |
Year | DOI | Venue |
---|---|---|
2007 | 10.1109/IPDPS.2007.370238 | IPDPS |
Keywords | Field | DocType |
meta- computing,grid computing,event tracing,performance tools,application software,parallel systems,hardware,concurrent computing,pattern search,weather forecasting,computer science | Grid computing,Latency (engineering),Computer science,Parallel computing,Event tracing,Metacomputing,Pattern search,Distributed computing | Conference |
Citations | PageRank | References |
23 | 1.14 | 13 |
Authors | ||
6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Daniel Becker | 1 | 23 | 2.15 |
Felix Wolf | 2 | 1264 | 107.97 |
Wolfgang Frings | 3 | 154 | 18.07 |
Markus Geimer | 4 | 35 | 2.12 |
Brian J. N. Wylie | 5 | 554 | 45.86 |
Bernd Mohr | 6 | 1019 | 89.99 |