Implementation of a framework for a DHT-based Distributed Location Service

Distributed Hash Tables (DHTs) are structured peer-to-peer systems in which a number of peer nodes collectively cooperate to provide a key/value pair information storage and retrieval service. All DHTs are characterized by desirable features such as scalability, fault tolerance, and self-organization. Each DHT node is responsible for maintaining a subset of the stored information, which depends on the specific DHT algorithm. Many Internet-based applications strongly rely on a Location Service (LS), such as DNS, in order to map a URI to one or more IP addresses (and port numbers) that specify where the resource identified by the URI can actually be accessed. However, Location Services typically introduce centralization points into the architectures they are used in and therefore expose the overall system to possible failures. Because of their nature, DHT appear to be a perfect mean for setting up a Distributed Location Service (DLS). In this paper we present a Java-based framework that can be used for building a DLS independent from the specific DHT algorithm and communication protocol adopted.