We present a framework for assessing operational reliability and performance of Web Services. The framework, named WS-REPAS, is hybrid in that it combines a modelling approach, based on Discrete Time Markov Chains (DTMC), with monitoring and in vivo testing of the service under assessment. Through the passive observation of the software in operation, field data are gathered and used to continuously update values of the parameters of the service DTMC model; changes in the service provisioning, or in the way it is used, trigger an active strategy, which executes proper testing sessions, ultimately yielding faithful estimates of the current service reliability and/or performance. We illustrate the framework and the automated support it provides. We show how it works describing experiments with a Web service publicly available in the Github repository.
A Hybrid Framework for Web Services Reliability and Performance Assessment / Guerriero, Antonio; Mirandola, Raffaela; Pietrantuono, Roberto; Russo, Stefano. - (2019), pp. 185-192. (Intervento presentato al convegno 1st International Workshop of Governing Adaptive and Unplanned Systems of Systems (GAUSS) tenutosi a Berlin, GERMANY nel 28/10/2019) [10.1109/ISSREW.2019.00070].
A Hybrid Framework for Web Services Reliability and Performance Assessment
GUERRIERO, ANTONIO;Roberto Pietrantuono;Stefano Russo
2019
Abstract
We present a framework for assessing operational reliability and performance of Web Services. The framework, named WS-REPAS, is hybrid in that it combines a modelling approach, based on Discrete Time Markov Chains (DTMC), with monitoring and in vivo testing of the service under assessment. Through the passive observation of the software in operation, field data are gathered and used to continuously update values of the parameters of the service DTMC model; changes in the service provisioning, or in the way it is used, trigger an active strategy, which executes proper testing sessions, ultimately yielding faithful estimates of the current service reliability and/or performance. We illustrate the framework and the automated support it provides. We show how it works describing experiments with a Web service publicly available in the Github repository.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.