LOBS: Load Balancing for Similarity Peer-to-Peer Structures

Investor logo

Warning

This publication doesn't include Faculty of Education. It includes Faculty of Informatics. Official publication website can be found on muni.cz.
Authors

NOVÁK David ZEZULA Pavel

Year of publication 2007
Type Article in Proceedings
Conference Databases Information Systems and Peer-to-Peer Computing 2007
MU Faculty or unit

Faculty of Informatics

Citation
Web Extended pre-proceedings version of the paper
Field Informatics
Keywords peer-to-peer; load-balancing; similarity search
Description The concept of peer-to-peer structures has recently been applied on the problem of large-scale similarity search. This resulted in systems where the computational load of the peers is of a high importance. Since no current load-balancing technique is designed for structures of this kind, we propose LOBS - a general system for load-balancing in peer-to-peer structures with time-consuming searching. LOBS is based on the following principles: measuring the computational load, separation of the logical and the physical level of the system, and detailed analysis of the load source to exploit either data relocation or replication. This work contains results of experiments we conducted using a prototype implementation of LOBS. In these trials, we used a real-life dataset and we varied the number of peers and the distribution of the query traffic in the system. The results show that LOBS is able to cope with any query-distribution and that it improves both the utilization of resources and the performance of the query processing. The costs of balancing are reasonable and are very small if there is time to adapt to a query-traffic. The behaviour of LOBS is independent of the network size.
Related projects:

You are running an old browser version. We recommend updating your browser to its latest version.