Sets, lists and trees: Distributed data structures on distributed hash tables

Abstract

Distributed hash tables and unstructured peer-to-peer overlay networks allow to store, search and retrieve single data elements. While this is useful for simple applications such as file sharing, for future sophisticated applications more sophisticated distributed data structures should be supported. In order to build, for example, architectures for distributed online social networks or distributed computing graphs, more advances data structures such as sets, lists of sorted trees should be supported by the peer-to-peer storage overlay with convenient access to the entries. Taking, for example, the set of albums in a distributed online social network. This set consists of a set of albums consist of sets of images, which themselves have each a list of comments. Using distributed data structures for sets, lists or trees in a distributed hash table allow to easily build sophisticated applications. In this paper, we present and evaluate a concept for such distributed data structures in peer-to-peer networks. In the evaluation, we show that all elements of each distributed data structure is successfully stored and retrieved and that the approach comes with low overhead and delay.