PinS: Peer-to-Peer Interrogation and Indexing System

  • Authors:
  • Maria-Del-Pilar Villamil;Claudia Roncancio;Cyril Labbe

  • Affiliations:
  • LSR-IMAG Laboratory;LSR-IMAG Laboratory;LSR-IMAG Laboratory

  • Venue:
  • IDEAS '04 Proceedings of the International Database Engineering and Applications Symposium
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

Distributed Hash Table (DHT) P2P systems provide flexible management of large scale distributed systems. They also provide efficient object localization given its key. However, they donýt provide high level query languages to formulate such location queries. Recent efforts have been made to improve such querying capabilities. Nevertheless, most of the proposed solutions are based on special hashing functions or on an additional level of peers. This paper presents PinS, a P2P interrogation and indexation middleware for DHT based P2P systems. It improves data sharing in such systems by supporting declarative queries and some facilities on data management without using particular hash functions or other requirements. Location queries may be conjunctions or disjunctions of conditions including comparison terms, and users may specify some evaluation constraints. Comprehensive or partial answers are provided. PinS supports pre-defined and customized attributes to describe objects and allows sharing objects with restricted access. Additionally, PinS enables several query execution strategies and proposes the use of indexes to improve comparison queries support. Our performance analysis shows the scalability of the system. In particular, the complexity of the evaluation of conjunctive queries is independent of the number of objects registered in the system.