公有云中基于系统工作流的并行策略性能评估

摘要

The performance of parallel execution of phylogenomic tree is studied.A performance evaluation for SciPhylomics exe-cutions in a real cloud environment is proposed.Firstly,the Hadoop,a MapReduce model implementation is introduced.Then, the SciCumulus workflow engine is explained.Finally,the workflow is executed using two parallel execution approaches (SciCu-mulus and Hadoop)at the Amazon EC2 cloud.The experiment results demonstrate that the bioinformatics experiment is suitable to be executed in the cloud despite its need for high performance capabilities.Many features of the evaluated workflow are same as other data intensive workflows.Thus,proposed method could be used to analyze other experiments.%分析系统发育基因组工作流程并行处理性能，提出了一种适用于云计算平台中SciPhylomics执行的性能评估工作流程。首先，介绍了映射化简模型的应用实现Hadoop；然后，呈现了SciCumulus云工作流程引擎；最后，在亚马逊EC2云上使用两种并行执行方法（SciCumulus和Hadoop）实施工作流程。实验结果表明，尽管系统发育基因组学实验对计算环境要求严格，但实验仍然适合在云中执行。此外，所评估的工作流程呈现出几组数据密集型工作流程的许多特征，本方法可以扩展到其他实验类型。