Balbo, G., & Serazzi, G. (1997). Asymptotic analysis of multiclass closed queueing networks: Multiple bottlenecks. Perform. Eval. 30(3), 115--152.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....inaccuracies have been reported as an effect of ignoring such variabilities. Existence of multiple This work was supported by the Austrian Science Foundation (FWF research grant No. S05303 MAT) BNs for systems modeled with multiclass closed queueing networks is studied by Balbo and Serazzi [1]. Uncertainty analysis in the context of performability modeling of computer systems using Markov reward models is considered by Haverkort and Meeuwissen [7] This paper is organized as follows: in Section 2, WLC for systems with variabilities and uncertainties in workload is discussed. ....

Balbo, G. and Serazzi, G., Asymptotic Analysis of Multiclass Closed Queueing Networks: Multiple Bottlenecks. Performance Evaluation, 1997, in print.

....is the unique system bottleneck. Because of possibly overlapping intervals, in the case of single class QNMs with intervals as input parameters, several queueing centers may be identified as potential bottleneck devices. Non uniqueness of bottleneck devices is also observed in multiclass QNMs [2]. For the representation of multiple potential bottlenecks, we define the set B of potential bottlenecks and the set B 1 of potential unique bottlenecks, respectively: B = ae k j k K q ; d k Kq max j=1 d j oe B 1 = k j k K q ; d k Kq max j=1 j 6=k d j ) For every ....

G. Balbo and G. Serazzi. Asymptotic Analysis of Multiclass Closed Queueing Networks: Multiple Bottlenecks. Performance Evaluation, 1997. In print.

....for single class QNMs have been adapted to handle histogrambased workload characterization [16] Large inaccuracies have been reported as an effect of ignoring such variabilities. Existence of multiple BNs for systems modeled with multiclass closed queueing networks is studied by Balbo and Serazzi [1]. Other related works include the use of interval arithmetic for computing performance bounds for closed queueing networks [19] and for concurrent software using rendezvous style communication [17] As opposed to sensitivity analysis for QNMs (see for example [4] or [23] where the impact of ....

....(i.e. models characterized by VU list parameters) In different VU list components, we may identify different stations to be the j ary BN. Additionally, parameter intervals in models with uncertainties may be overlapping, hindering the identification of a unique BN device for every BN degree. In [1], Balbo and Serazzi observe similar non uniqueness of BNs for different job mixes in multiclass closed QNMs. An approach to generalize the identification of BNs to systems with uncertainties is the so called set of potential BNs proposed in [14] Using the simplifying assumption that the parameter ....

G. Balbo and G. Serazzi. Asymptotic analysis of multiclass closed queueing networks: Multiple bottlenecks. Performance Evaluation, 30(3):115--152, September 1997.

No context found.

Balbo, G., & Serazzi, G. (1997). Asymptotic analysis of multiclass closed queueing networks: Multiple bottlenecks. Perform. Eval. 30(3), 115--152.

No context found.

Balbo, G., & Serazzi, G. (1997). Asymptotic analysis of multiclass closed queueing networks: Multiple bottlenecks. Perform. Eval. 30(3), 115--152.

No context found.

Balbo, G., Serazzi, G., "Asymptotic analysis of multiclass closed queueing networks: Multiple bottleneck", Performance Evaluation 30, 1997, pp. 115 -- 152.

No context found.

Balbo, G., Serazzi, G., "Asymptotic analysis of multiclass closed queueing networks: Common bottleneck", Performance Evaluation 26, 1996, pp. 51 -- 72.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC