Performance Modeling Of Computer Systems Research Articles

APEM — Approximate Performance Evaluation for Multi-Core Computers

Mean Value Analysis (MVA) has long been a standard approach for performance analysis of computer systems. While the exact load-dependent MVA algorithm is an efficient technique for computer system performance modeling, it fails to address multi-core computer systems with Dynamic Frequency Scaling (DFS). In addition, the load-dependent MVA algorithm suffers from numerical difficulties under heavy load conditions. The goal of our paper is to find an efficient and robust method which is easy to use in practice and is also accurate for performance prediction for multi-core platforms. The proposed method, called Approximate Performance Evaluation for Multi-core computers (APEM), uses a flow-equivalent performance model designed specifically to address multi-core computer systems and identify the influence on the CPU demand of the effect of DFS. We adopt an approximation technique to estimate resource demands to parametrize MVA algorithms. To validate the application of our method, we investigate three case studies with extended TPC-W benchmark kits, showing that our method achieves better accuracy compared with other commonly used MVA algorithms. We compare the three different performance models, and we also extend our approach to multi-class models.

Interval parameters for capturing uncertainties in an EJB performance model

Exact as well as approximate analytical solutions for quantitative performance models of computer systems are usually obtained by performing a series of arithmetical operations on the input parameters of the model. However, especially during early phases of system design and implementation, not all the parameter values are usually known exactly. In related research contributions, intervals have been proposed as a means to capture parameter uncertainties. Furthermore, methods to adapt existing solution algorithms to parameter intervals have been discussed. In this paper we present the adaptation of an existing performance model to parameter intervals. The approximate solution of a queueing network modelling an Enterprise JavaBeans server implementation is adapted to interval arithmetic in order to represent the uncertainty in some of the parameters of the model. A new interval splitting method is applied to obtain reasonable tight performance measure intervals. Monotonicity properties of intermediate computation results are exploited to achieve a more efficient interval solution. In addition, parts of the original solution algorithm are modified to increase the efficiency of the corresponding interval arithmetical solution.

Compositional Development of Performance Models in Poems

Performance models are software systems in which the components implement abstractions of the behavior of a total system. This paper describes a capability for semiautomatic development of performance models of computer systems spanning applications, operating systems, and hardware by composition from a library of components. Compositional development of performance models is a domain-specific instance of the general problem of software component reuse or design reuse. The concepts enabling compositional development of performance models in POEMS are encapsulation of analysis-level objects with associative interfaces and hierarchical dynamic data flow graphs as a structuring model. Objects with associative interfaces will be called compositional objects. Compositional objects and hierarchical dynamic data flow graphs provide a framework for the development of performance models that incorporate multiple modes of evaluation, span multiple semantic domains, span multiple levels of abstraction, and parallel implementation. Algorithms for composition through associative interfaces with automatic generation of parallel executables for the performance models will be defined.

3.1.4 A Hierarchical Method for Analyzing Candidate Solutions and Selecting the Best Option

ABSTRACTIn this paper the authors describe a case study which used a Systems Engineering Capability Maturity Model (SE‐CMM)‐compliant process for Analyzing Candidate Solutions to evaluate computer system performance modeling tools and recommend one for company‐wide use. As part of the paper, the authors describe the background and objectives of the effort and the details of the technical approach employed. While evaluating competing vendors or commercial‐off‐the‐shelf (COTS) products may be the most common use, trade studies are also useful for analyzing requirements trade‐offs, system or network architecture alternatives, make‐buy decisions, schedule changes, and other decisions faced by engineers and managers.

Performance modeling of computer systems

The analysis of computer systems behavior serves for analyzing of performance respectively performance prediction. The methodology is represented by the modeling in the unity of model forma tion, analysis/synthesis and valuation. Analytical, simulative and empirical behavior models are applied. An important method is the analysis on base the queueing theory especially closed queueing networks. To support use of analysis methods the program system BNETD and a program packet for performance evaluation are developed. The application is demonstrated on example of multiprogramming.

Stationary deterministic flows in discrete systems 1

We consider a deterministic system whose state space is the n-dimensional first orthant. It may be considered as a network of (deterministic) queues, a Karp-Miller vector addition system, a Petri net, a complex computer system, etc. Weak assumptions are then made concerning the asymptotic or limiting behaviour of the instants at which events are observed across a cut in the system: these instants may be considered as ‘arrival’ or ‘departure’ instants. Thus, like in operational analysis, we deal with deterministic and observable properties and we need no stochastic assumptions or restrictions (such as independence, identical distributions, etc.). We consider however asymptotic or stationary properties, as in conventional queuing analysis. Under our assumptions a set of standard theorems are proved: concerning arrival and departure instant measures, concerning ‘birth and death’ type equations, and concerning Little's formula. Our intention is to set the framework for a new approach to performance modelling of computer systems in a context close to that used in actual measurements, but taking into account infinite time behaviour in order to take advantage of the useful mathematical properties of asymptotic results.

Cross-entropy minimization given fully decomposable subset and aggregate constraints (Corresp.)

The principle of maximum entropy and the principle of minimum cross-entropy (minimum directed divergence, minimum discrimination information) have been applied recently to problems in queuing theory and computer-system performance modeling. These information-theoretic principles estimate probability distributions based on information in the form of known expected values. In the case of queuing theory and computer-system modeling, the known expected values arise from rate balance equations. This correspondence concerns situations in which the system state probabilities decompose into disjoint subsets and in which the known expected values are either expectations conditional on a specific subset or expectations involving aggregate subset probabilities. New properties of minimum cross-entropy distributions are derived and an efficient method of computing these distributions is derived. Computational examples are included. In the case of queuing theory and computer-system modeling, the disjoint subsets correspond to internal device states, and the aggregate probabilities correspond to overall device states. The results here apply when one has both rate balance equations for device equilibrium involving internal device state probabilities, as well as rate balance equations for system equilibrium involving aggregate device state probabilities.

Stationary deterministic flows in discrete systems

We consider a deterministic system whose state space is the n-dimensional first orthant. It may be considered as a network of (deterministic) queues, a Karp-Miller vector addition system, a Petrinet, a complex computer system, etc. Weak assumptions are then made concerning the asymptotic or limiting behaviour of the instants at which events are observed accross a cut in the system : these instants may be considered as "arrival" or "departure" instants. Thus, like in operational analysis, we deal with deterministic and observable properties and we need no stochastic assumptions or restrictions (such as independence, identical distributions, etc.). We consider however asymptotic or stationary properties, as in conventional queueing analysis. Under our assumptions a set of standard theorems are proved: concerning arrival and departure instant measures, concerning, "birth and death" type equations, and concerning Little's formula. Our intention is to set the framework for a new approach to performance modelling of computer systems in a context close to that used in actual measurements, but taking into account infinite time behaviour in order to take advantage of the useful mathematical properties of asymptotic results.

Computer performance modelling: an overview

An overview of the state-of-the-art in the performance modelling of computer systems is presented in this paper. The various forms of modelling procedures used in the past are reviewed with a greater emphasis given to new approaches which include queueing networks for mutliple-resource system analysis. Aspects of accuracy and computational efficiency of the algorithms involved are also discussed.

Queueing Models of Computer Systems

Benchmarking is older. Simulation is more popular. But analytical queueing models may offer the most cost-effective technique for computer system performance modeling.