A communication-reduced and computation-balanced framework for fast graph computation

Abstract

The bulk synchronous parallel (BSP) model is very user friendly for coding and debugging parallel graph algorithms. However, existing BSP-based distributed graphprocessing frameworks, such as Pregel, GPS and Giraph, routinely suffer from high communication costs. These high communication costs mainly stem from the fine-grained message-passing communication model. In order to address this problem, we propose a new computation model with low communication costs, called LCC-BSP. We use this model to design and implement a high-performance distributed graphprocessing framework called LCC-Graph. This framework eliminates high communication costs in existing distributed graph-processing frameworks. Moreover, LCC-Graph also balances the computation workloads among all compute nodes by optimizing graph partitioning, significantly reducing the computation time for each superstep. Evaluation of LCC-Graph on a 32-node cluster, driven by real-world graph datasets, shows that it significantly outperforms existing distributed graph-processing frameworks in terms of runtime, particularly when the system is supported by a highbandwidth network. For example, LCC-Graph achieves an order of magnitude performance improvement over GPS and GraphLab.

Keywords

graph computation / communication decrease / computation balance

References

[1] Valiant L G. A bridging model for parallel computation. Communications of the ACM, 1990, 33(8): 103–111

[2] Malewicz G, Austern M H, Bik A J C. Pregel: a system for large-scale graph processing. In: Proceedings of ACM International Conference on Management of Data. 2010, 135–146

[3] Salihoglu S, Widom J. GPS: a graph processing system. In: Proceedings of ACM International Conference on Scientific and Statistical Database Management. 2013, 22–32

[4] Ching A, Edunov S, Kabiljo K, Logothetis D, Muthukrishnan S. One trillion edges: graph processing at facebook-scale. Proceedings of the VIDB Endowment, 2015, 8(12): 1804–1815

[5] Wang G, Xie W, Demers A J, Gehrke J. Asynchronous large-scale graph processing made easy. in: Proceedings of International Conference on Innovation Database Research. 2013, 135–146

[6] Simmhan Y, Kumbhare A, Wickramaarachchi C, Nagarkar S, Ravi S, Raghavendra C, Prasanna V. Goffish: a sub-graph centric framework for large-scale graph analytics. In: Proceedings of European Conference on Parallel Processing. 2014, 451–462

[7] Kyrola A, Blelloch G E, Guestrin G. Graphchi: large-scale graph computation on just a PC. In: Proceedings of Usenix Conference on Operating Systems Design and Implementation. 2012, 31–46

[8] Cheng Y L, Wang F, Jiang H, Hua Y, Feng D, Wang X N. DD-graph: a highly cost-effective distributed disk-based graph-processing framework. In: Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing. 2016, 259–262

[9] Cheng Y L, Wang F, Jiang H, Hua Y, Feng D, Wang X N. LCC-graph: a high-performance graph-processing framework with low communication costs. In: Proceedings of IEEE/ACM International Symposium on Quality of Service. 2016, 91–100

[10] Cheng Y L, Jiang H, Wang F, Hua Y, Feng D. BlitzG: exploiting highbandwidth networks for fast graph processing. In: Proceedings of IEEE International Conference on Computer Communications. 2017, 2340–2348

[11] Page L. The pagerank citation ranking : bringing order to the web. Stanford Digital Libraries Working Paper, 1998, 9(1): 1–14

[12] Stefano L D, Bulgarelli A. A simple and efficient connected components labeling algorithm. In: Proceedings of International Conference on Image Analysis and Processing. 1999, 322–327

[13] Fortunato S. Community detection in graphs. Physics Reports, 2010, 486(3): 75–174

[14] Kothari R, Jain V. Learning from labeled and unlabeled data. In: Proceedings of IEEE International Joint Conference on Neural Networks. 2002, 2803–2808

[15] Lee M, Kim E J, Yousif M. Security enhancement in infiniband architecture. In: Proceedings of IEEE International Parallel and Distributed Processing Symposium. 2005, 105–114

[16] Karypis G, Kumar V. A fast and high quality multilevel scheme for partitioning irregular graphs. Journal of Scientific Computing, 1998, 20(1): 359–392

[17] Kernighan B W, Lin S. An efficient heuristic procedure for partitioning graphs. Journal of Bell System Technical, 1970, 49(2): 291–307

[18] Bui T N, Moon B R. Genetic algorithm and graph partitioning. IEEE Transactions on Computers, 1996, 45(7): 841–855

[19] Roy A, Mihailovic I, Zwaenepoel W. X-stream: edge-centric graph processing using streaming partitions. In: Proceedings of the 24th ACM Symposium on Operating Systems Principles. 2013, 472–488

[20] Nishimura J, Ugander J. Restreaming graph partitioning: simple versatile algorithms for advanced balancing. In: Proceedings of ACM International Conference on Knowledge Discovery and Data Mining. 2013, 1106–1114

[21] Low Y, Bickson D, Gonzalez J E, Guestrin C, Kyrola A. Distributed graphlab: a framework for machine learning and data mining in the cloud. Proceedings of the VLDB Endowment, 2012, 5(8): 716–727

[22] Gonzalez J E, Low Y, Haijie G, Danny B, Carlos G. Powergraph: distributed graph-parallel computation on natural graphs. In: Proceedings of Usenix Conference on Operating Systems Design and Implementation. 2012, 17–30

[23] Backstrom L, Huttenlocher D, Kleinberg J, Lan X. Group formation in large social networks: membership, growth, and evolution. In: Proceedings of the 12th ACM International Conference on Knowledge Discovery and Data Mining. 2006, 44–54

[24] Yan D, Cheng J, Xing K, Lu Y, Ng W, Bu Y G. Pregel algorithms for graph connectivity problems with performance guarantees. Proceedings of the VLDB Endowment, 2014, 7(14): 1821–1832

[25] Han M, Daudjee K. Giraph unchained: barrierless asynchronous parallel execution in pregel-like graph processing systems. Proceedings of the VLDB Endowment, 2015, 8(9): 950–961

[26] Yang S, Yan X, Zong B, Khan A. Towards effective partition management for large graphs. In: Proceedings of ACM Conference on Management of Data. 2012: 517–528

[27] Xin R S, Crankshaw D, Dave A, Gonzalez J E, Franklin M J, Stoica I. Graphx: unifying data-parallel and graph-parallel analytics, Computer Science, 2014, 8(3): 125–137

[28] Da Y, Yingyi B, Yuanyuan T, Deshpande A. Big graph analytics platforms. Foundations and Trends in Databases, 2017, 7(2): 180–195

[29] Lu Y, Cheng J, Yan D, Wu H. Large-scale distributed graph computing systems: an experimental evaluation. Proceedings of the VLDB Endowment, 2014, 8(3): 281–292

[30] Zhou C, Gao J, Sun B, Yu J X. Mocgraph: scalable distributed graph processing using message online computing. Proceedings of the VLDB Endowment, 2014, 8(4): 377–388

[31] Yan D, Cheng J, Lu Y, Ng Y. Blogel: a block-centric framework for distributed computation on real-world graphs. Proceedings of the VLDB Endowment, 2014, 7(14): 1981–1992

[32] Yan D, Cheng J, Ozsu M T, Lu Y. A general-purpose query-centric framework for querying big graphs. Proceedings of the VLDB Endowment, 2016, 9(7): 564–575

[33] Devine K D, Boman E G, Heaphy R T. New challenges in dynamic load balancing. Applied Numerical Mathematics, 2005, 52(2): 133–152

[34] Cheng J, Liu Q, Li Z, Fan W, Lui J C S. VENUS: vertex-centric streamlined graph computation on a single PC. In: Proceedings of IEEE International Conference on Data Engineering. 2015, 1131–1142

[35] Malicevic J, Roy A, Zwaenepoel W. Scale-up graph processing in the cloud:challenges and solutions. In: Proceedings of International Workshop on Cloud Data and Platforms. 2014, 1–6

[36] Pearce R, Gokhale M, Amato N M. Scaling techniques for massive scale-free graphs in distributed (external) memory. In: Proceedings of International Symposium on Parallel and Distributed Processing. 2013, 825–836

[37] Roy A, Bindschaedler L, Malicevic J, Zwaenepoel W. Chaos: scale-out graph processing from secondary storage. In: Proceedings of Symposium on Operating Systems Principles. 2015, 410–424

[38] Zhu X, Han W, Chen W. GridGraph: large-scale graph processing on a single machine using 2-level hierarchical partitioning. In: Proceedings of Usenix Conference on Usenix Technical Conference. 2015, 375–386

[39] Xu N, Chen L, Cui B. LogGP: a log-based dynamic graph partitioning method. Proceedings of the VLDB Endowment, 2014, 7(14): 1917–1928

[40] Ming W, Fan Y, Jilong X, Xiao W, Miao Y. GRAM: scaling graph computation to the trillions. In: Proceedings of ACM Symposium on Cloud Computing. 2015, 408–421

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature