Large scale graph processing systems: survey and an experimental evaluation

被引:50
作者
Batarfi, Omar [1 ,5 ]
El Shawi, Radwa [2 ]
Fayoumi, Ayman G. [1 ,5 ]
Nouri, Reza [3 ]
Beheshti, Seyed-Mehdi-Reza [3 ,6 ]
Barnawi, Ahmed [1 ,5 ]
Sakr, Sherif [3 ,4 ,7 ]
机构
[1] King Abdulaziz Univ, Jeddah 21413, Saudi Arabia
[2] Princess Nourah Bint Abdulrahman Univ, Coll Comp Sci & Informat Technol, Riyadh, Saudi Arabia
[3] Univ New S Wales, Sydney, NSW, Australia
[4] King Saud Bin Abdulaziz Univ Hlth Sci, Dept Hlth Informat, Riyadh, Saudi Arabia
[5] King Abdulaziz Univ, Fac Comp & Informat Technol, Jeddah 21413, Saudi Arabia
[6] Univ New S Wales, Sch Comp Sci & Engn CSE, Serv Oriented Comp Grp, Sydney, NSW, Australia
[7] Univ New S Wales, Comp Sci, Sydney, NSW, Australia
来源
CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS | 2015年 / 18卷 / 03期
关键词
Big graph; Graph processing; Experimental evaluation; FRAMEWORK; ANALYTICS; MAPREDUCE;
D O I
10.1007/s10586-015-0472-6
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Graph is a fundamental data structure that captures relationships between different data entities. In practice, graphs are widely used for modeling complicated data in different application domains such as social networks, protein networks, transportation networks, bibliographical networks, knowledge bases and many more. Currently, graphs with millions and billions of nodes and edges have become very common. In principle, graph analytics is an important big data discovery technique. Therefore, with the increasing abundance of large graphs, designing scalable systems for processing and analyzing large scale graphs has become one of the most timely problems facing the big data research community. In general, scalable processing of big graphs is a challenging task due to their size and the inherent irregular structure of graph computations. Thus, in recent years, we have witnessed an unprecedented interest in building big graph processing systems that attempted to tackle these challenges. In this article, we provide a comprehensive survey over the state-of-the-art of large scale graph processing platforms. In addition, we present an extensive experimental study of five popular systems in this domain, namely, GraphChi, Apache Giraph, GPS, GraphLab and GraphX. In particular, we report and analyze the performance characteristics of these systems using five common graph processing algorithms and seven large graph datasets. Finally, we identify a set of the current open research challenges and discuss some promising directions for future research in the domain of large scale graph processing.
引用
收藏
页码:1189 / 1213
页数:25
相关论文
共 47 条
[1]  
Abouzeid A., 2009, PVLDB, V2, P922, DOI 10.14778/1687627.1687731
[2]   The Stratosphere platform for big data analytics [J].
Alexandrov, Alexander ;
Bergmann, Rico ;
Ewen, Stephan ;
Freytag, Johann-Christoph ;
Hueske, Fabian ;
Heise, Arvid ;
Kao, Odej ;
Leich, Marcus ;
Leser, Ulf ;
Markl, Volker ;
Naumann, Felix ;
Peters, Mathias ;
Rheinlaender, Astrid ;
Sax, Matthias J. ;
Schelter, Sebastian ;
Hoeger, Mareike ;
Tzoumas, Kostas ;
Warneke, Daniel .
VLDB JOURNAL, 2014, 23 (06) :939-964
[3]  
[Anonymous], 2010, P 19 ACM INT S HIGH, DOI DOI 10.1145/1851476.1851593
[4]  
[Anonymous], 2010, P HOTCLOUD
[5]  
[Anonymous], 2012, PROC 10 USENIX C OPE
[6]  
Barnawi A., 2014, P TPC TECHN C TPCTC
[7]  
Borkar V, 2011, PROC INT CONF DATA, P1151, DOI 10.1109/ICDE.2011.5767921
[8]   Pregelix: Big(ger) Graph Analytics on A Dataflow Engine [J].
Bu, Yingyi ;
Borkar, Vinayak ;
Jia, Jianfeng ;
Carey, Michael J. ;
Condie, Tyson .
PROCEEDINGS OF THE VLDB ENDOWMENT, 2014, 8 (02) :161-172
[9]   The HaLoop approach to large-scale iterative data analysis [J].
Bu, Yingyi ;
Howe, Bill ;
Balazinska, Magdalena ;
Ernst, Michael D. .
VLDB JOURNAL, 2012, 21 (02) :169-190
[10]  
Chattopadhyay B, 2011, PROC VLDB ENDOW, V4, P1318