Vassilis Mertsiotakis

Compositional performance modelling with the TIPPtool

Stochastic process algebras have been proposed as compositional specification formalisms for performance models. In this paper, we describe a tool which aims at realising all beneficial aspects of compositional performance modelling, the TIPPtool. It incorporates methods for compositional specification as well as solution, based on state-of-the-art techniques, and wrapped in a user-friendly graphical front end. Apart from highlighting the general benefits of the tool, we also discuss some lessons learned during development and application of the TIPPtool. A non-trivial model of a real life communication system serves as a case study to illustrate benefits and limitations.

Stochastic process algebras—between LOTOS and Markov chains

Abstract This paper introduces stochastic process algebras as an approach for the structured design and analysis of both the functional behavior and performance characteristics of parallel and distributed systems. This is achieved by integrating stochastic modelling and analysis into process algebras like CCS or LOTOS. We demonstrate how notions of equivalent behavior – substantial for process algebras – stand in a new light for stochastic process algebras. In particular we discuss the impact of stochastic versions of such equivalences on compositional performance analysis by means of a running example. In addition we discuss analysis techniques, tool support and include a survey of related work as well as recent trends in this area.

A Construction and Analysis Tool Based on the Stochastic Process Algebra TIPP

There are many ways to incorporate a notion of time into process algebras in order to integrate functional design and performance analysis. One major research strand, stochastic process algebras, concentrates on the annotation of actions with exponentially distributed random variables. This paper presents a tool for the functional analysis and performance evaluation of complex systems based on the stochastic process algebra paradigm. The TIPP-tool provides facilities for model specification, reachability analysis, as well as several numerical algorithms for the solution of the underlying Markov chain and the computation of performance measures.

Stochastic process algebras: integrating qualitative and quantitative modelling

Stochastic process algebras (SPA) have emerged from work over the last five years investigating the use of process algebras for performance modelling. Like other stochastic extensions of existing system description techniques, they offer exciting possibilities for integrating qualitative and quantitative analysis. The use of such languages facilitates the early consideration of the temporal properties of an emerging design Unlike existing approaches, SPA exploit the compositionality, known from classical process algebras. The algebraic framework of this compositionality has advantages for both model construction and model solution.

Automatic scalability analysis of parallel programs based on modeling techniques

When implementing parallel programs for parallel computer systems the performance scalability of these programs should be tested and analyzed on different computer configurations and problem sizes. Since a complete scalability analysis is too time consuming and is limited to only existing systems, extensions of modeling approaches can be considered for analyzing the behavior of parallel programs under different problem and system scenarios. In this paper, a method for automatic scalability analysis using modeling is presented. Initially, we identify the important problems that arise when attempting to apply modeling techniques to scalability analysis. Based on this study, we define the Parallelization Description Language (PDL) that is used to describe parallel execution attributes of a generic program workload. Based on a parallelization description, stochastic models like graph models or Petri net models can be automatically generated from a generic model to analyze performance for scaled parallel systems as well as scaled input data. The complexity of the graph models produced depends significantly on the type of parallel computation described. We present several computation classes where tractable graph models can be generated and then compare the results of these automatically scaled models with their exact solutions using the modeling tool PEPP.

Exploiting stochastic process algebra achievements for generalized stochastic Petri nets

Constructing large generalized stochastic Petri nets (GSPN) by hierarchical composition of smaller components is a promising way to cope with the complexity of the design process for models of real hardware and software systems. The composition of nets is inspired by process algebraic operators. A solid theoretical framework of such operators relies on equivalences that are substitutive with respect to the operators. Practically important, such equivalences allow compositional reduction techniques, where components may be replaced by smaller but equivalent nets without affecting significant properties of the whole model. However substitutive equivalence notions for GSPN have not been published. In this paper we adopt operators and equivalences originally developed in the context of stochastic process algebras to GSPN. The equivalences are indeed substitutive with respect to two composition operators, parallel composition and hiding. This bears the potential to exploit hierarchies in the model definition to obtain performance indices of truly large composite GSPN by stepwise compositional reduction. We illustrate the effect of composition as well as compositional reduction by means of a running example. A case study of a workstation cluster highlights the potential of compositional reduction.

A Stochastic Process Algebra Based Modelling Tool

The incorporation of time into classical process algebras aiming at the integration of functional design and performance analysis has become very popular recently. There are many ways to include time in process algebras. Current research in this area concentrates mainly on the annotation of actions with exponentially distributed random variables. This allows us to make use of a large repertoire of analysis algorithms. This paper presents some first results of ongoing work which aims at providing a tool for the efficient performance evaluation and functional analysis of computer and communication systems based on the stochastic process algebra paradigm. It provides facilities for model creation, reachability analysis, as well as several numerical algorithms for the solution of the underlying Markov chain and the computation of characteristic performance measures.

Stochastic Process Algebras

We introduce Stochastic Process Algebras as a novel approach for the structured design and analysis of both the functional behaviour and performability (i.e performance and dependability) characteristics of parallel and distributed systems. This is achieved by integrating stochastic modelling and analysis into the powerful and well investigated formal description techniques of process algebras.

Compositional Performance Modelling with TIPPtool

Stochastic Process Algebras have been proposed as compositional specification formalisms for performance models. In this paper, we describe a tool which aims at realising all beneficial aspects of compositional performance modelling, the TIPPtool. It incorporates methods for compositional specification as well as solution, based on state-of-the-art-techniques, and wrapped in a user-friendly graphical front end.

Stochastic modeling of scaled parallel programs

Testing the performance scalability of parallel programs can be a time consuming task, involving many performance runs for different computer configurations, processor numbers, and problem sizes. Ideally, scalability issues would be addressed during parallel program design, but tools are not presently available that allow program developers to study the impact of algorithmic choices under different problem and system scenarios. Hence, scalability analysis is often reserved to existing (and available) parallel machines as well as implemented algorithms. In this paper we propose techniques for analyzing scaled parallel programs using stochastic modeling approaches. Although allowing more generality and flexibility in analysis, stochastic modeling of large parallel