Low traffic overlay networks with large routing tables

Abstract

The routing tables of Distributed Hash Tables (DHTs) can vary from size O(1) to O(n). Currently, what is lacking is an analytic framework to suggest the optimal routing table size for a given workload. This paper (1) compares DHTs with O(1) to O(n) routing tables and identifies some good design points; and (2) proposes protocols to realize the potential of those good design points.We use total traffic as the uniform metric to compare heterogeneous DHTs and emphasize the balance between maintenance cost and lookup cost. Assuming a node on average processes 1,000 or more lookups during its entire lifetime, our analysis shows that large routing tables actually lead to both low traffic and low lookup hops. These good design points translate into one-hop routing for systems of medium size and two-hop routing for large systems.Existing one-hop or two-hop protocols are based on a hierarchy. We instead demonstrate that it is possible to achieve completely decentralized one-hop or two-hop routing, i.e., without giving up being peer-to-peer. We propose 1h-Calot for one-hop routing and 2h-Calot for two-hop routing. Assuming a moderate lookup rate, compared with DHTs that use O(log n) routing tables, 1h-Calot and 2h-Calot save traffic by up to 70% while resolving lookups in one or two hops as opposed to O(log n) hops.

An analytical study on optimizing the lookup performance of distributed hash table systems under churn

“…The value of l is different in various DHTs, for example. [2], O(1) in Onehop Overlay [34], etc. We denote the average onehop routing latency by , and the average round-trip time by RTT = 2 .…”

Improving the Video Data Lookup Efficiency in Peer-to-Peer Live Streaming Network

“…For the fast DHT lookup with one hop, some classic One-hop DHTs have been proposed such as D1HT [1], 1h-Calot [2] and OneHop [3]. A comprehensive comparison between these One-hop DHTs can be found in [7] , which shows that D1HT consistently has the smallest overhead and better performance in load balancing than 1h-Calot and OneHop.…”

Survey on hierarchical routing schemes in “flat” distributed hash tables

“…Many popular DHTs, such as Chord [11] and Pastry [13], provide |T u | = O(log N) and |s → + d| = O(log N), or in other words, O(log N) state and O(log N) routing. 6 Better tradeoffs are possible, e.g., O(1) state and O(log N) routing in Viceroy [17] and Ulysses [5], or O( √ N) state and O(1) routing in Kelips [27] and 2h-Calot [69]. However, they have more complicated topologies or large routing tables, hence the maintenance cost becomes an issue [7].…”

“…It leads to systems with large routing tables, up to complete membership information. Extreme examples are OneHop [30], 1h-Calot [69], and D1HT [70] where every node must keep a complete routing table; work [112] makes their comparison. Note that SmartBoa [68] also utilizes large sizes for high-bandwidth nodes (see Section 4.3).…”

“…The principle evolves to DHTs with large routing tables or with high replication. The former family includes EpiChord [66], Accordion [67], SmartBoa [68], OneHop [30], 1h-Calot [69], and D1HT [70]. The latter family is represented by DHash [71], Beehive [72], and Yarqs [73].…”