Abstract | ||
---|---|---|
In this paper we study the impact of different types of constraints on the maximum throughput that a system can handle. In particular, we focus on constraints limiting the use of resources and/or the allowed response time. The problem is made even more difficult by the pronounced diversity in resource requirements of the different applications in execution, i.e., by the multiclass characteristic of the workloads. The proposed approach allows to determine the maximum load of the different classes, while still satisfying the considered performance objectives. An experimental validation of the described technique through the study of a realistic e-commerce application is presented. |
Year | DOI | Venue |
---|---|---|
2014 | 10.1007/978-3-319-08219-6_17 | ANALYTICAL AND STOCHASTIC MODELLING TECHNIQUES AND APPLICATIONS |
Field | DocType | Volume |
Admission control,Computer science,Response time,Throughput maximization,Throughput,Limiting,Distributed computing | Conference | 8499 |
ISSN | Citations | PageRank |
0302-9743 | 0 | 0.34 |
References | Authors | |
17 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Davide Cerotti | 1 | 159 | 14.74 |
Marco Gribaudo | 2 | 491 | 59.46 |
Ingolf Krüger | 3 | 768 | 63.23 |
Pietro Piazzolla | 4 | 96 | 11.72 |
Filippo Seracini | 5 | 12 | 1.92 |
Giuseppe Serazzi | 6 | 909 | 123.23 |