Journal of Intelligent & Fuzzy Systems 31 (2016) 1731–1743 DOI:10.3233/JIFS-152094 IOS Press 1731 A bio-inspired scheduler for minimizing makespan and flowtime of computational mechanics applications on federated clouds Elina Pacini a,b,d,∗ , Cristian Mateos c,d , Carlos Garc´ ıa Garino a,e , Claudio Careglio a,e and An´ ıbal Mirasso e a ITIC Research Institute – UNCuyo University, Mendoza, Argentina b Facultad de Ciencias Exactas y Naturales – UNCuyo University, Mendoza, Argentina c ISISTAN Research Institute – UNICEN University, Tandil, Buenos Aires, Argentina d Consejo Nacional de Investigaciones Cient´ ıficas y T´ ecnicas, Argentina e Facultad de Ingenier´ ıa – UNCuyo University, Mendoza, Argentina Abstract. Computational Mechanics (CM) concerns the use of computational methods to study phenomena under the principles of mechanics. A representative CM application is parameter sweep experiments (PSEs), which involves the execution of many CPU-intensive jobs and thus computing environments such as Clouds must be used. We focus on federated Clouds, where PSEs are processed via virtual machines (VM) that are lauched in hosts belonging to different datacenters, minimizing both the makespan and flowtime. Scheduling is performed at three levels: a) broker, where datacenters are selected based on their network latencies via three policies, b) infrastructure, where two bio-inspired schedulers based on Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO) for VM-host mapping in a datacenter are implemented, and c)VM, where jobs are assigned into the preallocated VMs based on job priorities. Simulated experiments performed with job data from two real PSEs show that our scheduling approach allows for a more agile job handling while reducing PSE makespan and flowtime. Keywords: Cloud computing, computational mechanics, scheduling, ant colony optimization, particle swarm optimization 1. Introduction CM involves the use of computational approaches to characterize and simulate physical events and engi- neering systems governed by the laws of mechanics. PSEs are CM simulations that require performing repeated analyses, where certain input parameters are varied among those defining the problem of inter- est. PSE users need a computing environment that delivers large amounts of computational power over ∗ Corresponding author. Elina Pacini. Tel./Fax: +54 (261) 4291000; E-mail: epacini@uncu.edu.ar. a long period of time, such as Clouds [4]. However, since in single-datacenter Clouds resource availabil- ity might be limited, the option of obtaining extra resources from an arrangement of Cloud providers –or federating Clouds– is an appealing solution [4]. For efficiently executing PSEs in federated Clouds it is necessary to properly manage physical resources from geographically distributed datacenters. There- fore, job scheduling should be performed at three levels [24]. At the broker level, scheduling strategies are used for selecting datacenters considering fac- tors such as network interconnections or monetary cost of allocating VMs on hosts. At the infrastructure 1064-1246/16/$35.00 © 2016 – IOS Press and the authors. All rights reserved