Performance evaluation of client-server systems

Ibe, Oliver C.; Choi, Hoon; Trivedi, Kishor S.

doi:10.1109/71.250101

Cited by 45 publications

(15 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Traffic characterizations of NFS workloads are presented in [13], whereas the generation of synthetic workloads for the performance analysis of distributed file servers is discussed in [6]. Using stochastic reward nets for the modelling of ethernet-based client server systems considered in [14]. Comparatively little attention has been paid to the development of techniques for performance improvement of these systems.…”

Section: Introductionmentioning

confidence: 99%

Performance of Scheduling Strategies for Client–Server Systems

Graham¹,

Majumdar

1999

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

Performance of Scheduling Strategies for Client–Server Systems

Graham¹,

Majumdar

1999

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

“…The firing of transition sk (1 ≤ k ≤ N ) means that the server sends a reply to station k. A token in place PSW means that the service is completed. In ( [16]), an approximation of this model is described to reduce the state space size. This is done by considering a tagged client and lumping the remaining clients into one super-client.…”

Section: Problem Descriptionmentioning

confidence: 99%

“…This is a common feature in distributed computing. The analysis of such systems becomes difficult by various kinds of dependencies in the system ( [16]). Generally Markov chain is able to capture the dependencies but a hand construction of such a Markov chain is often tedious and infeasible.…”

Section: Client/server Examplementioning

confidence: 99%

“…We consider here two real life examples: a client/server system and a cable modem termination system (CMTS) for the purpose of illustration. Client/server system is a system where a server station receives requests from its client stations, processes the requests and replies to the requesting stations ( [16]). We compare the solution methods while varying the number of stations connected to the server.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Simulation versus Analytic-Numeric Methods: Illustrative Examples

Tuffin¹,

Choudhary²,

Hirel³

et al. 2007

Proceedings of the 2nd International ICST Conference on Performance Evaluation Methodologies and Tools

Self Cite

View full text Add to dashboard Cite

Performance along with dependability analysis is a tremendous challenge in the design or improvement of modern complex systems. Two different classes of solution methods are generally used: analytic-numeric methods and simulation methods. As most of the literature explains, the choice between them depends more on the analyst's background than on the system itself. In this paper, we illustrate the advantages and drawbacks of each method on real problems and compare the results. Finally we conclude the paper providing some hints to choose a solution method depending on the model. We use SPNP, a Petri net analysis package, and CSIM 19, as simulation package to model and evaluate systems.

show abstract

“…But if multiple alternatives are acceptable, a quantitative method is required so that design trade-off decisions can be made. Literature exists on the performance modeling of client-server [3] and producer-consumer [1] systems. For messaging services, the recent work [8] on performance models for the configurable delivery and discard policies found in the CORBA Notification Service is relevant.…”

Section: Introductionmentioning

confidence: 99%