Transactional model and performance analysis for business processes and e-Science applications

Abstract

In modern science and business process studies the main challenge is not anymore in data acquisition but mainly in data processing, analysis, storage and transfer. To treat these challenges, workow and business process systems have received increased attention from research and industry communities. Despite of the signicant eorts and software tools developed, the workows and business processes still oer a scenario with several research challenges, mainly related to the scalability and reliability of their execution in the new collaborative and high-performance computing platforms. In this pro ject, we are focusing in two important issues (called axes) belonging to this scenario: transaction processing and prediction of performance related to dierent phases of the workow life cycle. The rst axis is related to transaction processing aiming to grant a exible and reliable environment for the modeling and execution of workows. The second research axis is related to prediction of performance of workows by means of formal modeling, in order to support a better provisioning of resources and a better scheduling of tasks in such type of applications. To validate the results obtained in these two research axes, we will use two real-world case studies that are important examples of both workows and e-Science applications. In addition to the academic results provided in this project, we also hope to increase the research capability of the DATA group through two important research axes and also the validation of academic results based on two important case studies. (AU)