An Evaluation of EpiChord in OverSim

摘要

EpiChord is a Distributed Hash Table (DHT) algorithm which supports data storage/retrieval in large scale distributed systems. It removes the typical O( logn)-state-per-node restriction imposed by the majority of other DHT topologies by employing a reactive routing state maintenance strategy that amortizes network maintenance costs into lookup queries. Under ideal condition, EpiChord's lookup performance can approach O(1) hops - with maintenance costs comparable to traditional multi-hop DHTs. This paper presents an implementation of EpiChord in OverSim, and validates the performance of our model against the performance reported in the original EpiChord paper. We also present some adjustments to the algorithm to remove a discrepancy and then compare our modified results with the original ones. Finally, we present additional results showing the EpiChord algorithm is stable over time and performs well for larger networks.