【摘 要】
：
Current popular systems, Hadoop and Spark, cannot achieve satisfied performance because of the ine?cient overlapping of computation and communication when runni
论文部分内容阅读
Current popular systems, Hadoop and Spark, cannot achieve satisfied performance because of the ine?cient overlapping of computation and communication when running iterative big data applications. The pipeline of computing, data movement, and data management plays a key role for current distributed data computing systems. In this paper, we first analyze the overhead of shu?e operation in Hadoop and Spark when running PageRank workload, and then propose an event-driven pipeline and in-memory shu?e design with better overlapping of computation and communication as DataMPI-Iteration, an MPI-based library, for iterative big data computing. Our performance evaluation shows DataMPI-Iteration can achieve 9X～21X speedup over Apache Hadoop, and 2X～3X speedup over Apache Spark for PageRank and K-means.
其他文献
我们发现应用全自动生化分析仪检测血清甘油三酯(TG)时,若设置二氧化碳(CO2)位于TG之前检测,可使血清TG测定结果明显升高.为进一步探讨CO2试剂对TG检测的影响,我们进行了如下
按照吉林省农业委员会、吉林省财政厅《关于印发〈吉林省2011年农村劳动力培训阳光工程项目实施方案〉的通知》精神,桦甸市农村经济管理总站在省阳光办公室和省农村经济管理
该文从挂篮荷载计算、施工流程、支座及临时固结施工、挂篮安装及试验、合拢段施工、模板制作安装、钢筋安装、混凝土的浇筑及养生、测量监控等方面人手,介绍了S226海滨大桥
We present Fatman, an enterprise-scale archival storage based on volunteer contribution resources from under-utilized web servers, usually deployed on thousands
黑鱼君在我国黄、渤海及东海水域为常见种，其肉质鲜美，营养丰富，是深受广大消费者喜爱的名贵海产鱼。该鱼抗病力强，生长较快，在我国北方沿海可自然越冬，是增养殖的优良品种之一。一
在省农委组织的“干名处长进干村”活动中,省农业环境保护与农村能源管理总站帮扶对象是东丰县那丹伯镇石顶子村.该村位于全镇西侧,共有6个自然屯,农户226户,人口813人,党员2
患者男,13岁。主因呕血3次(共约400mL)入院。查体:体温36.7℃,脉搏98次/min,血压86/54mm Hg,神志清楚,面色苍白,腹平坦,左上腹见纵形长约6cm切口愈合瘢痕。血常规:白细胞9.9
目的 探讨多层面螺旋CT仿真内窥镜 (multi -slicehelicalCTvirtualendoscopy ,MSCTVE)诊断大肠肿瘤的检查方法和临床应用价值。方法 2 6例怀疑有大肠肿瘤的患者行结肠、直
该文从挂篮荷载计算、施工流程、支座及临时固结施工、挂篮安装及试验、合拢段施工、模板制作安装、钢筋安装、混凝土的浇筑及养生、测量监控等方面人手,介绍了S226海滨大桥
本文讨论了迟熟 α-淀粉酶 (L ate m atureα- amylase,L MA)的概念及其与小麦发芽籽粒中其它 α-淀粉酶同工酶的区别 ,综述了其遗传特性、表达的差异性、影响因素和鉴定方法