A Bio-inspired Scheduler for Minimizing Makespan and Flowtime of Computational Mechanics Applications on Federated Clouds [JCR]

ELINA PACINI; CRISTIAN MATEOS; CARLOS GARCÍA GARINO; CLAUDIO CAREGLIO; ANIBAL MIRASSO

JOURNAL OF INTELLIGENT AND FUZZY SYSTEMS

IOS PRESS

Lugar: Amsterdam; Año: 2016

1064-1246

/* Layout-provided Styles */div.standard {text-align: left;}Computational Mechanics (CM) is a discipline concerned with the use of computational methods to study phenomena governed by the principles of mechanics. A representative application of this discipline is parameter sweep experiments (PSEs), which involve the execution of many CPU-intensive jobs and thus computing environments such as Clouds must be used. Particularly, this work focuses on the federated Cloud model, where custom virtual machines (VM) are launched in appropriate hosts belonging to different resource providers (datacenters) to execute PSEs, minimizing both the makespan and flowtime. Concretely, scheduling is performed at three levels. First, at the broker level, datacenters are selected based on their network latencies via three policies ?Lowest-Latency-Time-First, First-Latency-Time-First, and Latency-Time-In-Round?. Second, at the infrastructure level, two bio-inspired schedulers based on Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO) for mapping VMs to appropriate hosts in a datacenter are implemented. Finally, at the VM level, jobs are assigned for execution into the preallocated VMs through a job-priority based policy. Simulated experiments performed with job data from two real PSEs show that the combination of policies at the broker level with ACO and PSO, plus a job-priority policy, allows for a more agile job handling while reducing makespan and flowtime of the PSEs compared to using these policies combined with Genetic Algorithms and Best Effort.