云计算环境下利用工作流引擎的并行策略性能评估研究

The Research of Performance Evaluation of Parallel Strategies by Using Workflow Engine in Cloud Computing Environment

针对很少有方法对云中系统发育基因组学分析工作流程的并行性进行评估的问题,提出了一种适用于真实云环境中SciPhylomics执行的性能评估工作流程。呈现了SciCumulus云工作流引擎,在亚马逊EC2云上,使用两种并行执行方法(SciCumulus和Hadoop)实施该工作流程。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,但此类实验仍然适合在云中执行。所评估的工作流程呈现了几组数据密集型工作流程的许多特征,实验结果表明,这些云执行结果可以扩展到其他实验类型。

There is no previous approach that evaluates the performance of parallel execution of phylogenomic tree produced by each analysis executed. To solve this problem, a performance evaluation for SciPhylomics executions in a real cloud environment is pro-posed. The SciCumulus workflow engine is explained. The workflow is executed by using two parallel execution approaches （Sci-Cumulus and Hadoop） at the Amazon EC2 cloud. The experiment results demonstrate that this class of bioinformatics experiment is suitable to be executed in the cloud despite its need for high performance capabilities. The evaluated workflow shows many features of several data intensive workflows, which present that these cloud execution results can be extended to other classes of experiments.