Modelling and analysis of distributed software using GSPNs

4Citations
Citations of this article
4Readers
Mendeley users who have this article in their library.
Get full text

Abstract

This chapter discusses the role that Generalized Stochastic Petri Nets (GSPN) can play in the static analysis of distributed software. The material is organized along two main lines: The need and the advantages of studying both qualitative and quantitative aspects of a program, and the need for doing it in an automatic manner. The role of performance evaluation in the analysis of distributed software is illustrated through a small example, classical in the qualitative approach (the dining philosophers). Although small this example allows to point out the need and the requirements of automatic translation and to discuss the main hypothesis behind program performance evaluation through GSPN models. A procedure for the automatic generation of GSPN models starting from a distributed program written in a CSP-like language, and for the definition of program performance indices in terms of GSPN ones is then given and illustrated by means of a realistic example.

Cite

CITATION STYLE

APA

Donatelli, S., & Franceschinis, G. (1996). Modelling and analysis of distributed software using GSPNs. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 1492, 438–477. https://doi.org/10.1007/3-540-65307-4_54

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free