A P2P Approach to Many Tasks Computing for Scientific Workflows
Eduardo Ogasawara (Federal University of Rio de Janeiro)
Jonas Dias (Federal University of Rio de Janeiro)
Daniel Oliveira (Federal University of Rio de Janeiro)
Carla Rodrigues (Federal University of Rio de Janeiro)
Carlos Pivotto (Federal University of Rio de Janeiro)
Rafael Antas (Federal University of Rio de Janeiro)
Vanessa Braganholo (Federal University of Rio de Janeiro)
Patrick Valduriez (INRIA & Laboratoire d'Informatique, de Robotique et de Microélectronique de Montpellier)
Marta Mattoso (Federal University of Rio de Janeiro)
Abstract:
Scientific Workflow Management Systems (SWfMS) are being used intensively to support large scale in silico experiments. In order to reduce execution time, current SWfMS have exploited workflow parallelization under the arising Many Tasks Computing (MTC) paradigm in homogeneous computing environments, such as multiprocessors, clusters and grids with centralized control. Although successful, this solution no longer applies to heterogeneous computing environments, such as hybrid clouds, which may combine users’ own computing resources with multiple edge clouds. A promising approach to address this challenge is Peer-to-Peer (P2P) which relies on decentralized control to deal with scalability and dynamic behavior of resources. In this paper, we propose a new P2P approach to apply MTC in scientific workflows. Through the results of simulation experiments, we show that our approach is promising.
Keywords:
Parallel and Distributed Computing