Scientific Workflow Execution with Multiple Objectives in Multisite Clouds

Ji Liu (1, 2) , Esther Pacitti (2, 3) , Patrick Valduriez (3, 2) , Daniel de Oliveira (4) , Marta Mattoso (5)

1 MSR - INRIA - Microsoft Research - Inria Joint Centre
2 ZENITH - Scientific Data Management
3 IBC - Institut de Biologie Computationnelle
4 Fluminense Federal University [Niterói]
5 COPPE-UFRJ - Instituto Alberto Luiz Coimbra de Pós-Graduação e Pesquisa de Engenharia

In this short paper (see [4] for the extended version), we propose a general solution based on multi-objective scheduling to execute SWfs in a multisite cloud with the following main contributions: the design of a multi-objective cost model, SSVP VM provisioning approach, ActGreedy scheduling algorithm and an extensive experimental evaluation.

Domaines: Base de données [cs.DB]

Ji Liu, Esther Pacitti, Patrick Valduriez, Daniel de Oliveira, Marta Mattoso. Scientific Workflow Execution with Multiple Objectives in Multisite Clouds. BDA: Gestion de Données — Principes, Technologies et Applications, LIAS / ISAE-ENSMA, Poitiers, Nov 2016, Poitiers, France. ⟨lirmm-01409092⟩