基于Hadoop平台的分布式重删存储系统

针对数据中心存在大量数据冗余的问题,特别是备份数据造成的存储容量浪费,提出一种基于Hadoop平台的分布式重复数据删除解决方案。该方案通过检测并消除特定数据集内的冗余数据,来显著降低数据存储容量,优化存储空间利用率。利用Hadoop大数据处理平台下的分布式文件系统(HDFS)和非关系型数据库HBase两种数据管理模式,设计并实现一种可扩展分布式重删存储系统。其中,MapReduce并行编程框架实现分布式并行重删处理,HDFS负责重删后的数据存储,在HBase数据库中构建索引表,实现高效数据块索引查询。最后,利用虚拟机镜像文件数据集对系统进行了测试,基于Hadoop平台的分布式重删系统能在保证高重删率的同时,具有高吞吐率和良好的可扩展性。     

Hadoop系统map阶段的排序性能分析与优化

MapReduce计算框架已被广泛用于大规模数据分析的应用。虽然它具有弹性的可扩展性和细粒度的容错系统,然而性能却并不令人满意。MapReduce可以通过分配更多的计算节点来实现更好的性能,但是,这种做法并不符合成本效益。用户渴望MapReduce在提供弹性的可扩展性和细密度容错的同时,可以具有更高的计算效率。该文提出了一种动态优化map阶段排序性能的方法,并进行了测试,测试结果表明,该方法能够提升MapReduce的基准测试性能。     

IBFET: Index-based features extraction technique for scalable code clone detection at file level granularity

Many techniques have been developed over the years to detect code clones in different software systems to maintain security measures. These techniques often require the source code to compare the subject system against a very large data set of big code. This paper presents index-based features extraction technique (IBFET) to detect code clones at a very large-scale level to billions of LOC at file level granularity. We performed preprocessing, indexing, and clone detection for more than 324 billion of LOC using a Hadoop distributed environment, which is quite faster and more efficient as compared to existing distributed indexing and clone detection techniques; meanwhile, it detects all three types of clones efficiently. The MapReduce rule of divide and conquer is used for a count and retrieve the similar features between different systems. We evaluated the execution time, scalability, precision, and recall of IBFET by using a well-known clone detection data set IJaDataset and BigCloneBench; furthermore, we compared the results with other state-of-the-art tools. Our approach is faster, flexible, scalable, and provides accurate results with high authenticity and can be implemented at a large-scale level.     

基于MapReduce的H-mine算法*

频繁模式挖掘是一种非常有效的从数据中获取知识的方法,但是随着大数据时代的来临,现有算法及其计算环境的运算速度、内外存容量面临严峻挑战。针对以上问题,本文紧密结合MapReduce模型提供的高效分布式编程和运行框架,在深入分析H-mine频繁模式挖掘算法的基础上,通过对H-mine算法频繁模式挖掘过程的并行化改进,提出了一种新颖的基于MapReduce模型的H-mine算法（简称：MRH-mine）。MRH-mine算法实现了对H-mine算法在分布式运行环境下的改造,实验表明该算法在面对数据大规模增长的情况下,具有良好的性能和扩展性。     

A partitioning framework for Cassandra NoSQL database using Rendezvous hashing

Due to the gradual expansion in data volume used in social networks and cloud computing, the term “Big data” has appeared with its challenges to store the immense datasets. Many tools and algorithms appeared to handle the challenges of storing big data. NoSQL databases, such as Cassandra and MongoDB, are designed with a novel data management system that can handle and process huge volumes of data. Partitioning data in NoSQL databases is considered one of the critical challenges in database design. In this paper, a MapReduce Rendezvous Hashing-Based Virtual Hierarchies (MR-RHVH) framework is proposed for scalable partitioning of Cassandra NoSQL database. The MapReduce framework is used to implement MR-RHVH on Cassandra to enhance its performance in highly distributed environments. MR-RHVH distributes the nodes to rendezvous regions based on a proposed Adopted Virtual Hierarchies strategy. Each region is responsible for a set of nodes. In addition, a proposed bloom filter evaluator is used to ensure the accurate allocation of keys to nodes in each region. Moreover, a number of experiments were performed to evaluate the performance of MR-RHVH framework, using YCSB for database benchmarking. The results show high scalability rate and less time consuming for MR-RHVH framework over different recent systems.     

基于列存储的MapReduce分布式Hash连接算法

大数据具有规模大、深度大、宽度大、处理时间短、硬件系统普通化、软件系统开源化的特点。传统关系型数据库在对大数据进行操作时存在系统性能严重下降、计算效率提升有限以及可扩展性差等问题,因此引入MapReduce并行计算模型,提出一种大数据上基于列存储的MapReduce分布式Hash连接算法。首先,设计面向大数据的分布式计算模型,在设计的分片聚集并行连接的基础上,利用Hash连接以及动态探测方法优化了数据并行连接处理效率;然后,针对该算法开发了基于Hadoop的原型系统。通过实验证明,在大数据分析处理中,所提算法在执行时间和负载能力上都有很好的性能表现,也能提供良好的可扩展性。     

MapReduce and Its Applications,Challenges, and Architecture: a Comprehensive Review and Directions for Future Research

Profound attention to MapReduce framework has been caught by many different areas. It is presently a practical model for data-intensive applications due to its simple interface of programming, high scalability, and ability to withstand the subjection to flaws. Also, it is capable of processing a high proportion of data in distributed computing environments (DCE). MapReduce, on numerous occasions, has proved to be applicable to a wide range of domains. However, despite the significance of the techniques, applications, and mechanisms of MapReduce, there is no comprehensive study at hand in this regard. Thus, this paper not only analyzes the MapReduce applications and implementations in general, but it also provides a discussion of the differences between varied implementations of MapReduce as well as some guidelines for planning future research.     

Adapting scientific computing problems to clouds using MapReduce

Cloud computing, with its promise of virtually infinite resources, seems to suit well in solving resource greedy scientific computing problems. To study this, we established a scientific computing cloud (SciCloud) project and environment on our internal clusters. The main goal of the project is to study the scope of establishing private clouds at the universities. With these clouds, students and researchers can efficiently use the already existing resources of university computer networks, in solving computationally intensive scientific, mathematical, and academic problems. However, to be able to run the scientific computing applications on the cloud infrastructure, the applications must be reduced to frameworks that can successfully exploit the cloud resources, like the MapReduce framework. This paper summarizes the challenges associated with reducing iterative algorithms to the MapReduce model. Algorithms used by scientific computing are divided into different classes by how they can be adapted to the MapReduce model; examples from each such class are reduced to the MapReduce model and their performance is measured and analyzed. The study mainly focuses on the Hadoop MapReduce framework but also compares it to an alternative MapReduce framework called Twister, which is specifically designed for iterative algorithms. The analysis shows that Hadoop MapReduce has significant trouble with iterative problems while it suits well for embarrassingly parallel problems, and that Twister can handle iterative problems much more efficiently. This work shows how to adapt algorithms from each class into the MapReduce model, what affects the efficiency and scalability of algorithms in each class and allows us to judge which framework is more efficient for each of them, by mapping the advantages and disadvantages of the two frameworks. This study is of significant importance for scientific computing as it often uses complex iterative methods to solve critical problems and adapting such methods to cloud computing frameworks is not a trivial task.     

基于最小生成树的大规模数据分类模型及其MapReduce实现

数据的快速增长,为我们提供了更多的信息,然而,也对传统信息获取技术提出了挑战。这篇论文提出了MCMM算法,它是基于MapReduce的大规模数据分类模型的最小生成树（MST）的算法。它可以看做是介于传统的KNN方法和基于聚类分类方法之间的模型,旨在克服这两种方法的不足并能处理大规模的数据。在这一模型中,训练集作为有权重的无向完全图来处理。顶点是对象,两点之间边的权重是对象间的距离。这一距离,不同于欧几里得距离,它是一个特定的距离度量。这样,可以找到图中最小生成树集,其中,图中每棵树代表一个类。为了降低时间复杂度,提取了每棵树中最具代表性的点来代表该树。这些压缩了的点集,可以通过计算无标签对象和它们之间的距离,来进行分类。MCMM模型基于MapReduce实现并且部署在Hadoop平台。该模型可扩展处理大规模的数据,是因为Hadoop支持数据密集分布应用,并且这些应用可以和数以千计的节点和数据一起运作。另外,MapReduce 和Hadoop能在由商品机组成的集群上很好的运行。MCMM模型使用云平台并且通过使用MapReduce 和Hadoop进行云计算是有益处的。实验采用的数据集包括从UCI数据库得到的真实数据和一些模拟数据,实验使用了4000个集群。实验表明,MCMM模型在精确度和扩展性上优于KNN和其他一些经常使用的基础分类方法。     

Paradigm and performance analysis of distributed frequent itemset mining algorithms based on Mapreduce

FrequentItemsetMining (FIM) is one of the most important data mining tasks and is the foundation of many data mining tasks. In Big Data era, centralized FIM algorithms cannot meet the needs of FIM for big data in terms of time and space, so Distributed Frequent Itemset Mining (DFIM) algorithms have been designed to meet the above challenges. In this paper, LocalGlobal and RedistributionMining which are two main paradigms of DFIM algorithm are discussed; Two algorithms of these paradigms on MapReduce named LG and RM are proposed while MapReduce is a popular distributed computing model, and also the related work is discussed. The experimental results show that the RM algorithm has better performance in terms of computation and scalability of sites, and can be used as the basis for designing the DFIM algorithm based on MapReduce. This paper also discusses the main ideas of improving the DFIM algorithms based on MapReduce.     

Distributed media indexing based on MPI and MapReduce

Web-scale digital assets comprise millions or billions of documents. Due to such increase, sequential algorithms cannot cope with this data, and parallel and distributed computing become the solution of choice. MapReduce is a programming model proposed by Google for scalable data processing. MapReduce is mainly applicable for data intensive algorithms. In contrast, the message passing interface (MPI) is suitable for high performance algorithms. This paper proposes an adapted structure of the MapReduce programming model using MPI for multimedia indexing. Experimental results are done on various multimedia applications to validate our model. The experiments indicate that our proposed model achieves good speedup compared to the original sequential versions, Hadoop and the earlier versions of MapReduce using MPI.     

Cloud computing-based parallel genetic algorithm for gene selection in cancer classification

Cancer classification is one of the main steps during patient healing process. This fact enforces modern clinical researchers to use advanced bioinformatics methods for cancer classification. Cancer classification is usually performed using gene expression data gained in microarray experiment and advanced machine learning methods. Microarray experiment generates huge amount of data, and its processing via machine learning methods represents a big challenge. In this study, two-step classification paradigm which merges genetic algorithm feature selection and machine learning classifiers is utilized. Genetic algorithm is built in MapReduce programming spirit which makes this algorithm highly scalable for Hadoop cluster. In order to improve the performance of the proposed algorithm, it is extended into a parallel algorithm which process on microarray data in distributed manner using the Hadoop MapReduce framework. In this paper, the algorithm was tested on eleven GEMS data sets (9 tumors, 11 tumors, 14 tumors, brain tumor 1, lung cancer, brain tumor 2, leukemia 1, DLBCL, leukemia 2, SRBCT, and prostate tumor) and its accuracy reached 100% for less than 25 selected features. The proposed cloud computing-based MapReduce parallel genetic algorithm performed well on gene expression data. In addition, the scalability of the suggested algorithm is unlimited because of underlying Hadoop MapReduce platform. The presented results indicate that the proposed method can be effectively implemented for real-world microarray data in the cloud environment. In addition, the Hadoop MapReduce framework demonstrates substantial decrease in the computation time.

     

分布式系统Hadoop 平台的视频转码

在研究了目前主流的视频转码方案基础上,提出了一种分布式转码系统。系统采用HDFS（Hadoop Distributed File System）进行视频存储,利用MapReduce思想和FFMPEG进行分布式转码。详细讨论了视频分布式存储时的分段策略,以及分段大小对存取时间的影响。同时,定义了视频存储和转换的元数据格式。提出了基于MapReduce编程框架的分布式转码方案,即Mapper端进行转码和Reducer端进行视频合并。实验数据显示了转码时间随视频分段大小和转码机器数量不同而变化的趋势。结果表明,相比单机转码,提出的系统在采用8台机器并行转码时,可以节约80％左右的时间。     

MapReduce环境下的并行Dwarf立方构建

针对数据密集型应用,提出了一种基于MapReduce框架的并行Dwarf数据立方构建算法.算法将传统Dwarf立方等价分割为多个独立的子Dwarf立方,采用MapReduce架构,实现了Dwarf立方的并行构建、查询和更新.实验证明,并行Dwarf算法一方面结合了MapReduce框架的并行性和高可扩展性,另一方面结合...     

基于HBase的遥感数据分布式存储与查询方法研究

遥感影像的存储与查询是地理信息处理中重要的内容,在海量遥感影像的实时处理中发挥着重要作用。针对传统的遥感影像处理中存在单节点故障、扩展性低和处理效率低等问题,提出了一种基于HBase的遥感数据分布式存储与查询方案。该方法首先采用均匀网格对遥感影像进行划分,并根据划分结果设计了一种基于网格ID和Hilbert曲线相结合的索引方案。然后,通过利用HBase的过滤机制设计了过滤列族,达到了在查询时筛选数据的目的。另外,采用MapReduce的并行处理方法对影像数据进行并行写入和查询。实验结果表明,与MySQL和MapFile相比,该方法可以有效地提高数据的写入和查询速度,且具有较好的可扩展性。     

基于分布式数据仓库的分类分析研究

针对GAC-RDB分类算法只能应用于单机版数据仓库的局限性,为了能够更方便、快捷地在云计算平台上开展数据挖掘工作,基于分布式数据仓库HBase,结合GAC-RDB分类算法的实现机理,制定适合分布式平台的运行策略,使用原生HiveQL语言提出了一种分布式GAC-RDB分类算法。实验显示,随着集群中节点的不断增加,算法的运行时间稳步下降。结果表明,在保证算法准确率的前提下,分布式数据仓库能够有效提高GACRDB分类算法的扩展性和运行效率,相对于MapReduce框架,HiveQL语言降低了对数据挖掘从业人员的技术要求,更大程度地减少了算法的开发时间,为挖掘海量数据提供了新的解决方案。     

Kassandra: A framework for distributed simulation of heterogeneous cooperating objects

To address the heterogeneity and scalability issues of simulating Cooperating Objects (COs) systems, we propose Kassandra, a conceptual framework for enabling distributed COs simulation by integrating existing simulation tools. Moreover, Kassandra exploits the communication middleware used by real-world COs as underlying communication mechanism for integrating Kassandra-enabled simulation tools. In this way, real-world COs can be included with simulated objects in a seamless way to perform more accurate system performance evaluation. Moreover, such a hardware-in-the-loop approach is not limited to pre-deployment performance analysis, and can offer possibilities to analyse performance at different phases of CO applications. The concept of Kassandra has been carried out in the EU PLANET project. In this paper, we introduce the Kassandra framework components and show their interactions at different phases for node deployments in PLANET use cases. The result demonstrates the applicability of Kassandra to facilitate the development of CO applications.     

A survey of large-scale analytical query processing in MapReduce

Enterprises today acquire vast volumes of data from different sources and leverage this information by means of data analysis to support effective decision-making and provide new functionality and services. The key requirement of data analytics is scalability, simply due to the immense volume of data that need to be extracted, processed, and analyzed in a timely fashion. Arguably the most popular framework for contemporary large-scale data analytics is MapReduce, mainly due to its salient features that include scalability, fault-tolerance, ease of programming, and flexibility. However, despite its merits, MapReduce has evident performance limitations in miscellaneous analytical tasks, and this has given rise to a significant body of research that aim at improving its efficiency, while maintaining its desirable properties. This survey aims to review the state of the art in improving the performance of parallel query processing using MapReduce. A set of the most significant weaknesses and limitations of MapReduce is discussed at a high level, along with solving techniques. A taxonomy is presented for categorizing existing research on MapReduce improvements according to the specific problem they target. Based on the proposed taxonomy, a classification of existing research is provided focusing on the optimization objective. Concluding, we outline interesting directions for future parallel data processing systems.     

Hybrid address spaces: A methodology for implementing scalable high-level programming models on non-coherent many-core architectures

This paper introduces hybrid address spaces as a fundamental design methodology for implementing scalable runtime systems on many-core architectures without hardware support for cache coherence. We use hybrid address spaces for an implementation of MapReduce, a programming model for large-scale data processing, and the implementation of a remote memory access (RMA) model. Both implementations are available on the Intel SCC and are portable to similar architectures. We present the design and implementation of HyMR, a MapReduce runtime system whereby different stages and the synchronization operations between them alternate between a distributed memory address space and a shared memory address space, to improve performance and scalability. We compare HyMR to a reference implementation and we find that HyMR improves performance by a factor of 1.71× over a set of representative MapReduce benchmarks. We also compare HyMR with Phoenix++, a state-of-art implementation for systems with hardware-managed cache coherence in terms of scalability and sustained to peak data processing bandwidth, where HyMR demonstrates improvements of a factor of 3.1× and 3.2× respectively. We further evaluate our hybrid remote memory access (HyRMA) programming model and assess its performance to be superior of that of message passing.