Modeling and simulation of task-to-task relationships in cloud computing environment

• Grant number: 19/06281-8
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Effective date (Start): June 01, 2019
• Effective date (End): May 31, 2020
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computer Systems
• Principal Investigator: Aleardo Manacero Junior
• Grantee: Luís Vinícius Omar Baldissera
• Host Institution: Instituto de Biociências, Letras e Ciências Exatas (IBILCE). Universidade Estadual Paulista (UNESP). Campus de São José do Rio Preto. São José do Rio Preto , SP, Brazil

Abstract

The analysis of computing systems, when performed through simulation, implies in the specification of several aspects of the analyzed system, such as its hardware and the running applications. Hardware modeling usually involves specific parameters such as processor's speed, storage's capacity or communication capacity. Application modeling, however, may involve different abstraction levels, from simple blackboxes with given computing loads up to including dependency relationships between tasks. Although one can simulate relationships, the literature shows that most tools model tasks as blackboxes since modeling relationships implies in a remarkable growth in the model's complexity. In this context, the specification of a simpler approach to model task relationships is quite relevant, mostly if the goal tool works with a graphical interface for modeling/simulation. Therefore, in this research project the goal is the implementation of a graphical approach to model task dependencies in a cloud computing environment. The work of the scholarship's recipient will focus in the definition of which relationships should be modeled, finishing with their implementation in the iSPD (iconic Simulator of Parallel and Distributed Systems), which is a simulator developed at the UNESP's Parallel and Distributed Systems Lab. -- GSPD -- with grants from FAPESP (2008/09312-7 e2012/15127-3). Results from this work will enable the validation of the proposed modeling approach as well as its application to evaluate how task dependencies affect the performance of cloud systems.