Title
Optimizing Inter-server Communications by Exploiting Overlapping Communities in Online Social Networks.
Abstract
As the rapid growth of online social networks (OSNs), inter-server communications are becoming an obstacle to scaling the storage systems of OSNs. To address the problem, network partitioning and data replication are two commonly used approaches. In this paper, we exploit the combination of both approaches simultaneously and propose a data placement scheme based on overlapping communities detection. The principle behind the proposed scheme is to co-locate frequently interactive users together as long as it brings positive traffic reduction and satisfies load constraint. We conduct trace-driven experiments and the results show that our scheme significantly reduces the inter-server communications as well as preserving good load balancing.
Year
DOI
Venue
2016
10.1007/978-3-319-49583-5_18
ALGORITHMS AND ARCHITECTURES FOR PARALLEL PROCESSING, ICA3PP 2016
Keywords
Field
DocType
Inter-server communications,Online social networks,Data placement,Network partitioning,Data replication
Obstacle,Replication (computing),Social network,Load balancing (computing),Computer science,Exploit,Traffic reduction,Inter-server,Distributed computing
Conference
Volume
ISSN
Citations 
10048
0302-9743
1
PageRank 
References 
Authors
0.36
10
4
Name
Order
Citations
PageRank
Jing-Ya Zhou16416.35
Jianxi Fan271860.15
Baolei Cheng3252.76
Juncheng Jia462.45