Name
Abstract | ||
|---|---|---|
MapReduce has been very successful in implementing large-scale data-intensive applications. Because of its simple programming model, MapReduce has also begun being utilized as a programming tool for more general distributed and parallel HPC applications. However, its applicability is often limited due to relatively inefficient runtime performance and hence insufficient support for flexible workflows. In particular, the performance problem is not negligible in iterative MapReduce applications. We implemented new MapReduce framework SSS based on distributed key-value store, that supports flexible workflows. Mappers and reducers read key-values only from its local storage enjoying high throughput and low latency. We evaluated SSS comparing with Hadoop using synthetic benchmark and iterative application. The result showed that SSS is faster than Hadoop, except for simple sequential read from disks. |
Year | DOI | Venue |
|---|---|---|
2011 | 10.1145/2148600.2148642 | SC Companion |
Keywords | Field | DocType |
simple sequential,performance problem,iterative application,high throughput,key-value store,simple programming model,flexible workflows,mapreduce framework,iterative mapreduce application,new mapreduce framework,programming tool,inefficient runtime performance,low latency,key value store,parallel computer,parallel computing,programming model | Programming paradigm,Computer science,Parallel computing,Associative array,SSS*,Latency (engineering),Throughput,Workflow,Distributed computing | Conference |
Citations | PageRank | References |
1 | 0.37 | 1 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Hidemoto Nakada | 1 | 956 | 118.87 |
| Hirotaka Ogawa | 2 | 196 | 23.58 |
| Tomohiro Kudoh | 3 | 344 | 50.92 |