DISTRIBUTED TRIGGER PROCESSING IN PEER DATABASES

Irfan Ahmed Makki (University of Ottawa)

Abstract

Hyperion supports data sharing and querying in a peer-to-peer data management environment, that is, environment based on the peer-to-peer paradigm where each peer manages its own data and shares relevant data with acquainted peers. To support data sharing, the use of mapping tables has been proposed. These mapping tales use pairs of corresponding data values that reside in different peers. In terms of querying, a framework has been proposed in which users pose queries only with respect to their local peer. Then, a translation mechanism has been provided that uses the mapping tables to translate a locally expressed query to a set of queries over the acquainted peers. In addition to querying the local and remote peers, distributed triggers are to be established in order to coordinate the different operations between the related peers in order to maintain data and mapping consistencies as well as form the overall functionality of the P2P auction system. We have used a peer-to-peer action environment as a motivating scenario, where all the participating peers take part in a public action, like eBay, by sharing the relevant data over acquainted peers. A Motor Bike dataset has been selected for used with the online auction system where each peer will have a set of Motor Bikes that will be put up for auction. At the same time, a network design has been selected for experimentation purposes, where specific peers share their data with a set of acquainted peers in order to facilitate an effective peer-to-peer auctioning environment.