A Fair Load Balancing Algorithm for Hypercube-Based DHT Networks

Load balance is an important problem in the DHT-based P2P networks. In recent years, many solutions have been proposed to address this problem. However, these solutions have some limitations in our opinion. They either make some unrealistic assumptions about the network, or have high communication or maintenance overhead. In this paper, we present a distributed load balancing algorithm for the hypercube-based DHT networks. Our algorithm is based on the concept of fairness and uses the fairness index as the fairness metric. The purpose of our algorithm is to distribute the query load fairly to nodes. Each node periodically monitors the fairness index of current load distribution by using only local computation and it tries to achieve a fairer load distribution by dynamically adjusting its indegree according to its experienced load and the fairness index. The results of our experiments show that our algorithm has low overhead and it can achieve good load balance without unrealistic assumptions about the network.

[1]  Cheng-Zhong Xu,et al.  Locality-aware randomized load balancing algorithms for DHT networks , 2005, 2005 International Conference on Parallel Processing (ICPP'05).

[2]  Raj Jain,et al.  A Quantitative Measure Of Fairness And Discrimination For Resource Allocation In Shared Computer Systems , 1998, ArXiv.

[3]  Richard M. Karp,et al.  Load balancing in dynamic structured P2P systems , 2004, IEEE INFOCOM 2004.

[4]  Robert Morris,et al.  Chord: A scalable peer-to-peer lookup service for internet applications , 2001, SIGCOMM 2001.

[5]  Cheng-Zhong Xu,et al.  Elastic Routing Table with Provable Performance for Congestion Control in DHT Networks , 2006, 26th IEEE International Conference on Distributed Computing Systems (ICDCS'06).

[6]  David R. Karger,et al.  Chord: a scalable peer-to-peer lookup protocol for internet applications , 2003, TNET.

[7]  Cheng-Zhong Xu,et al.  Hash-based proximity clustering for load balancing in heterogeneous DHT networks , 2006, Proceedings 20th IEEE International Parallel & Distributed Processing Symposium.

[8]  Ben Y. Zhao,et al.  Tapestry: a resilient global-scale overlay for service deployment , 2004, IEEE Journal on Selected Areas in Communications.

[9]  Ben Y. Zhao,et al.  An Infrastructure for Fault-tolerant Wide-area Location and Routing , 2001 .

[10]  Stefan Saroiu,et al.  A Measurement Study of Peer-to-Peer File Sharing Systems , 2001 .

[11]  Vana Kalogeraki,et al.  A Fair Resource Allocation Algorithm for Peer-to-Peer Overlays , 2006, Proceedings IEEE INFOCOM 2006. 25TH IEEE International Conference on Computer Communications.

[12]  Yiming Hu,et al.  Efficient, proximity-aware load balancing for DHT-based P2P systems , 2005, IEEE Transactions on Parallel and Distributed Systems.

[13]  Mark Handley,et al.  A scalable content-addressable network , 2001, SIGCOMM 2001.

[14]  Ion Stoica,et al.  Peer-to-Peer Systems II , 2003, Lecture Notes in Computer Science.

[15]  Antony I. T. Rowstron,et al.  Pastry: Scalable, Decentralized Object Location, and Routing for Large-Scale Peer-to-Peer Systems , 2001, Middleware.

[16]  Richard M. Karp,et al.  Load Balancing in Structured P2P Systems , 2003, IPTPS.

[17]  Krishna P. Gummadi,et al.  Measurement, modeling, and analysis of a peer-to-peer file-sharing workload , 2003, SOSP '03.

[18]  David R. Karger,et al.  Wide-area cooperative storage with CFS , 2001, SOSP.

[19]  David R. Karger,et al.  Consistent hashing and random trees: distributed caching protocols for relieving hot spots on the World Wide Web , 1997, STOC '97.