An Effective Lookup Strategy for Recursive and Iterative Lookup on Hierarchical DHT
Tomonori Funahashi, Yoshitaka Nakamura, Yoh Shiraishi, Osamu Takahashi · Institutional Repositories DataBase (IRDB) · 2012
Recursive and iterative lookups on the perfor- mance of distributed hash table (DHT) are deteriorated by churn when nodes leave the network. When churn occurs infrequently, recursive lookup outperforms iterative lookup, but back when churn occurs frequently, the opposite is the case. Therefore, optimal lookup needs recursive and itera- tive lookups to be separated by the frequency of churn. We propose a lookup strategy that separates recursive and itera- tive lookups by the churn rate. However, a common DHT makes it difficult establish the neighboring churn rate. Hier- archical DHT takes into consideration the reliability of nodes to ascertain the churn rate, but it uses only a lookup strategy, recursive or iterative in the DHT. We believe that each lookup strategy should be used that match the churn rate in hierarchical DHT. Therefore, we compared our lookup strategy with both recursive and iterative lookup on hierarchical DHT.