A scalable content-addressable network

Abstract

Hash tables -which map "keys" onto "values" -are an essential building block in modern software systems. We believe a similar functionality would be equally valuable to large distributed systems. In this paper, we introduce the concept of a Content-Addressable Network (CAN) as a distributed infrastructure that provides hash table-like functionality on Internet-like scales. The CAN is scalable, fault-tolerant and completely self-organizing, and we demonstrate its scalability, robustness and low-latency properties through simulation.

Content-based publish-subscribe over structured P2P networks

“…Distributed Hash Tables (DHTs [7], [8], [9], [11]) have been adopted to create peer-to-peer data networks. In a DHT each node has a unique identifier (nodeID) selected from a very large address space.…”

“…A popular class of them is the "structured" p2p systems. The most prominent of these systems are built using a Distributed Hash Table (DHT [7], [8], [9]), which is a mechanism that provides scalable resource lookup/routing.…”

An Optimal Broadcast Algorithm for Content-Addressable Networks

“…A CAN [3] is a structured P2P network based on a d -dimensional Cartesian coordinate space labeled D. This space is dynamically partitioned among all peers in the system such that each node is responsible for storing data, in the form of (key, value) pairs, in a sub-zone of D. To store a (k, v) pair, the key k is deterministically mapped onto a point in D and the value v is stored by the node responsible for the zone comprising this point. The search for the value corresponding to a key k is achieved by applying the same deterministic function on k to find the node responsible for storing the corresponding value.…”

“…Indeed, considering our definition of a CAN (each node is responsible for a hyperrectangle zone), the intersection between an hyperrectangle to be covered and a CAN remains a CAN, thus our algorithm is still valid to multicast on a range of coordinates, or to cover only a certain number of dimensions. An alternative definition of CAN [3] keeps track of the history of joining nodes, which forms a tree. Using this tree as a spanning tree has two disadvantages: first, this would limit the contribution to a subset of all possible CAN.…”

iCluster: A Self-organizing Overlay Network for P2P Information Retrieval

“…Distributed Hash Tables (DHTs) [2,3] and Semantic Overlay Networks (SONs) [4,5] are common solutions to the problem of fast information search in p2p networks. DHTs provide fast lookup mechanisms facilitating information search over the network assuming that each peer is connected to other peers and is responsible for a part of the distributed index.…”