Distributed data structures improvement for collective retrieval time
Read Al-Aaridhi, Ahmet Yuksektepe, Tobias Amft, Kálmán Graffi · Wireless Personal Multimedia Communications · 2016
In peer-to-peer overlay networks, distributed hash tables have been characterized as self-organizing, scalable, and robust. Up to now, distributed hash tables allow to store, search and retrieve only single data elements, which make the creation of more complex applications, such as distributed online social networks, very cumbersome. Sophisticated applications require the support for various and advanced data structures such as sets, lists and trees. Ideally, the DHT should support these advanced Distributed Data Structures. In previous work, we presented a new approach for Distributed Data Structure on distributed hash tables, which comes with low overhead and delay. In this paper, we address the issue of the time-consuming retrieval of deep data structures and present therefore a Cache Structures, which helps to accelerate the retrieval process by magnitudes. Simulations show that the Cache Structure gives us new functionality and an alternative way to retrieve the complete Distributed Data Structure much faster.