Improving Many-Task computing in scientific workflows using P2P techniques

Abstract

Large-scale scientific experiments are usually supported by scientific workflows that may demand high performance computing infrastructure. Within a given experiment, the same workflow may be explored with different sets of parameters. However, the parallelization of the workflow instances is hard to be accomplished mainly due to the heterogeneity of its activities. Many-Task computing paradigm seems to be a candidate approach to support workflow activity parallelism. However, scheduling a huge amount of workflow activities on large clusters may be susceptible to resource failures and overloading. In this paper, we propose Heracles, an approach to apply consolidated P2P techniques to improve Many-Task computing of workflow activities on large clusters. We present a fault tolerance mechanism, a dynamic resource management and a hierarchical organization of computing nodes to handle workflow instances execution properly. We have evaluated Heracles by executing experimental analysis regarding the benefits of P2P techniques on the workflow execution time.

Towards a Cost Model for Scheduling Scientific Workflows Activities in Cloud Environments

“…Dias et al [12] proposes Heracles, an approach that uses Peer-to-Peer (P2P) techniques to manage distributed scientific workflows on huge clusters based on QoS criteria such as deadlines. However, different from the approach proposed in this paper, Heracles is focused on clusters and does not address cloud configuration issues.…”

Desenvolvimento de um ambiente de computação voluntária baseado em computação ponto-a-ponto

“…32 Tabela A computação voluntária (CV) tenta aproveitar o recurso ocioso "doando-o" para um determinado projeto, istoé, o responsável pelo projeto transmite (geralmente via internet) para outro computador (doador) uma atividade para que este a processe. Esse paradigma tem o poder de fornecer muito processamento (ANDERSON, 2004), mas quando se trata de experimentos científicos existem muitos fatores que podem tornar esta solução pouco eficiente, como por exemplo, processamento de longa duração (DETHIER et al, 2008), grande volume de dados (DUAN et al, 2012), heterogenia de sistemas (DORNEMANN et al, 2012) e instabilidade dos computadores voluntários (DIAS et al, 2010).…”