Search papers, labs, and topics across Lattice.
This paper addresses the computational challenges in performance inference for multiclass closed queueing networks, which are critical for managing enterprise applications. By reformulating the normalizing constant of equilibrium state probabilities as a multidimensional integral, the authors provide explicit expressions and develop efficient evaluation methods using cubature rules and asymptotic expansions. The proposed techniques significantly enhance accuracy and efficiency in likelihood computation, particularly for optimization-based inference in large models.
Transforming the computation of normalizing constants into manageable multidimensional integrals could revolutionize performance analysis in complex enterprise systems.
Recent years have seen a rapid growth of interest in exploiting monitoring data collected from enterprise applications for automated management and performance analysis. In spite of this trend, even simple performance inference problems involving queueing theoretic formulas often incur computational bottlenecks, for example upon computing likelihoods in models of batch systems. Motivated by this issue, we revisit the solution of multiclass closed queueing networks, which are popular models used to describe batch and distributed applications with parallelism constraints. We first prove that the normalizing constant of the equilibrium state probabilities of a closed model can be reformulated exactly as a multidimensional integral over the unit simplex. This gives as a by-product novel explicit expressions for the multiclass normalizing constant. We then derive a method based on cubature rules to efficiently evaluate the proposed integral form in small and medium-sized models. For large models, we propose novel asymptotic expansions and Monte Carlo sampling methods to efficiently and accurately approximate normalizing constants and likelihoods. We illustrate the resulting accuracy gains in problems involving optimization-based inference.
