DevOps workflow verification and duration prediction using non‐Markovian stochastic Petri nets

Abstract

In this paper, we provide a non‐Markovian Stochastic Petri Net (SPN) model for DevOps workflow specification, and we determine how business processes are carried out. After describing our model semantics, we show how general properties related to liveness and safety can be checked. After that, we provide several extensions on SPNs (SPN) the notation and expressivity to check some specific properties related to actors' (Developers and Operators) availability, interactions between the actors, and execution failures detection linked to the DevOps steps. Next, we validate the proposed model relevance with MATLAB simulation through a specific DevOps case study. Finally, we propose a truncated density function to anticipate the delays related to the DevOps business process overall steps.

Production Chain Modeling based on Learning Flow Stochastic Petri Nets

“…This is achieved via the functions of the model, which are based on the laws of probability. We note that this simulation (see figure3 and figure4) is similar to that presented in [27,28]. The objective of this simulation is to evaluate the execution time and the delay.…”

Understanding DevOps critical success factors and organizational practices

“…According to Claps et al [10] continuous integration process and support has an impact on DevOps success. Furthermore, automatic testing technique improvement [27], test automation [27], Verification of soundness [4], automation tools [4] may affect the success of DevOps.…”

FMS-Workflow Modeling Based on P-Timed Stochastic Petri Net

“…To achieve this, they must tend to totally or partially automate their processes because this method is apt to show the efficiency and tendency towards this objective [1] [5]. Several research works propose methodologies for modeling and automating business processes [4] [6] [7] [8].…”

“…This is related to the reasons that business analysts strive to formalize simple approaches to modeling and simulating their processes over a long period of time [18]. These approaches' disadvantage, according to [8] [18] [19], comes down to the need for operational expertise (manipulation and simulation). Furthermore, the representative model may deviate from the actual simulated due to the significant increase in the number of tests (time and simulation cost ratio) [8] [18].…”