An efficient synchronous indexing technique for full-text retrieval in distributed databases

Abstract

Despite the fact that distributed databases (DDB) have become a requirement of users and real-world applications, these DDBs are still too far from the performance of centralized DBs that have shown efficiency and robustness with their native search engines. Indeed, DDB couldn’t benefit from textual search engines because of the timing issue in the distributed context and keeping data integrity among the distributed architecture. As a contribution in this research field, we propose in this paper, a new technology for indexing DDB based on the Lucene indexing engine. We prove that this technology allows, in addition to textual search, 1) the transparency of distributed architecture’s scaling [8] and external search indexes, 2) the optimization of the exchange protocol data between nodes, 3) the synchronous processing of insertion, deletion or failure of a node of the distributed topology.