Abstract | ||
---|---|---|
Data stream applications grow rapidly in recent years. This paper focuses on analytically modeling the performance of middleware-based data stream processing center, which is usually structured in a multi-tier architecture and deployed on HPC clusters. We present a closed queueing network to model data stream center. The Mean-Value Analysis (MVA) algorithm is employed to derive the throughput of the system and help identify the system performance bottleneck. The model is validated through experiments on a concrete network monitoring data processing system hosted on a SMP cluster. Experimental results show that the proposed model well captures the system performance. |
Year | DOI | Venue |
---|---|---|
2006 | 10.1007/11610113_92 | APWeb |
Keywords | Field | DocType |
data stream application,mean-value analysis,system performance,model data stream center,middleware-based data stream processing,closed queueing network model,hpc cluster,concrete network monitoring data,multi-tier data stream processing,system performance bottleneck,closed queueing network,middleware,mean value analysis,data processing,network monitoring | Middleware,Data modeling,Computer science,Data stream,Data processing system,Multiprocessing,Layered queueing network,Queueing theory,Network monitoring,Distributed computing | Conference |
Volume | ISSN | ISBN |
3841 | 0302-9743 | 3-540-31142-4 |
Citations | PageRank | References |
1 | 0.36 | 6 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Yufeng Wang | 1 | 15 | 2.53 |
Wang Huaimin | 2 | 1025 | 121.31 |
Yan Jia | 3 | 1 | 0.36 |
Bixin Liu | 4 | 9 | 3.04 |