Marina Ribaudo

GreatSPN 1.7: graphical editor and analyzer for timed and stochastic Petri nets

This paper describes the GreatSPN 1.7 package for the modeling, validation, and performance evaluation of distributed systems using Generalized Stochastic Petri Nets and their colored extension. The tool provides a friendly framework to experiment with timed Petri net based modeling techniques. It implements efficient analysis algorithms to allow its use on “real” applications, not only toy examples. Developed in a university for non-profit purposes, it is distributed free of charge to other universities for educational and research purposes. An overview of the complete architecture of the package is given together with examples of its application. Then the various analysis and simulation modules are described.

PEPA Nets: A Structured Performance Modelling Formalism

In this paper we describe a formalism which uses the stochastic process algebra PEPA as the inscription language for labelled stochastic Petri nets. Viewed in another way, the net is used to provide a structure for combining related PEPA systems. The combined modelling language naturally represents such applications as mobile code systems where the PEPA terms are used to model the program code which moves between network hosts (the places in the net). We describe the implementation of a tool to support this modelling formalism and apply this to model a peer-to-peer filestore.

An efficient algorithm for aggregating PEPA models

Performance Evaluation Process Algebra (PEPA) is a formal language for performance modeling based on process algebra. It has previously been shown that, by using the process algebra apparatus, compact performance models can be derived which retain the essential behavioral characteristics of the modeled system. However, no efficient algorithm for this derivation was given. We present an efficient algorithm which recognizes and takes advantage of symmetries within the model and avoids unnecessary computation. The algorithm is illustrated by a multiprocessor example.

PEPA nets: a structured performance modelling formalism

In this paper we describe a formalism which uses the stochastic process algebra PEPA as the inscription language for labelled stochastic Petri nets. Viewed in another way, the net is used to provide a structure for linking related PEPA systems. The combined modelling language naturally represents such applications as mobile code systems where the PEPA terms are used to model the program code which moves between network hosts (the places in the net). We describe, the implementation of a tool to support this modelling formalism and apply this to model a hierarchical cellular network.

A comparison of performance evaluation process algebra and generalized stochastic Petri nets

Generalized stochastic Petri nets (GSPN) and performance evaluation process algebra (PEPA) can both be used to study qualitative and quantitative behaviour of systems in a single environment. The paper presents a comparison of the two formalisms in terms of the facilities that they provide to the modeller; considering both the definition and the analysis of the performance model. Our goal is to provide a better understanding of both formalisms; and to prepare a fertile ground for exchanging ideas and techniques between the two. To illustrate similarities and differences, we make the different issues more concrete by means of an example modelling resource contention.

Software performance modelling using PEPA nets

Modelling and analysing distributed and mobile software systems is a challenging task. PEPA nets---coloured stochastic Petri nets---are a recently introduced modelling formalism which clearly capture important features such as location, synchronisation and message passing. In this paper we describe PEPA nets and the newly-developed platform support for software performance modelling using them. Crucial to this support is the compilation from PEPA nets into Hillstons PEPA stochastic process algebra in order to access the software tools which support the PEPA algebra. In addition to derivation of steady state performance measures, this suite of tools allows properties of the system to be verified using model-checking. We show the application of PEPA nets in the modelling and analysis of a secure Web service.

Stochastic Petri net semantics for stochastic process algebras

We define a stochastic Petri net (SPN) semantics for stochastic process algebras (SPAs), a recently introduced formalism that offers a novel approach for performance modeling. The proposed semantics is evaluated in terms of three criteria: concurrency and retrievability, as defined by Olderog (1987, 1991) for untimed net semantics, and Markov equivalence for the stochastic aspects.

Exploiting Modal Logic to Express Performance Measures

Stochastic process algebras such as PEPA provide ample support for the component-based construction of models. Tools compute the numerical solution of these models; however, the stochastic process algebra methodology has lacked support for the specification and calculation of complex performance measures. In this paper we present a stochastic modal logic which can aid the construction of a reward structure over the model. We discuss its relationship to the underlying theory of PEPA. We also present a performance specification language which supports high level reasoning about PEPA models, and allows queries about their equilibrium behaviour. The meaning of the specification language has its foundations in the stochastic modal logic. We describe the implementation of the logic within the PEPA Workbench and a case study is presented to illustrate the approach.

On the computation of performance characteristics of concurrent programs using GSPNs

Abstract Petri net based models are proposed in this paper as the vehicles for the characterization of the processing and communication parameters of concurrent programs. A technique for the construction of a GSPN model as a direct translation of a concurrent program is used and a method is presented that allows to build, from a GSPN model, graph representations that are useful for preliminary assessment of the performance of the programs and for solving the problem of allocating the processes of a concurrent program to the processors of a given architecture. The concept of phase in the execution of a concurrent program is introduced and used to construct communication graphs that are simpler than those obtained considering the execution of a program as a whole. An example of the application of the technique is presented together with experimental results obtained from direct measurements of a real program. The agreement between the performance indices computed with the models and the results of direct measurements shows the practical relevance of this technique.

On the Aggregation Techniques in Stochastic Petri Nets and Stochastic Process Algebras

Stochastic Petri Nets and Stochastic Process Algebras can both be used for the qualitative and quantitative analysis of parallel and distributed systems. However, sometimes the complexity of the system can make such analysis infeasible, due to the large size of the state space underlying the model. To overcome this problem different aggregation techniques have been proposed in both formalisms. Two of them are compared in this paper.