PlanetP: using gossiping to build content addressable peer-to-peer information sharing communities

Cuenca-Acuna, Francisco Matias; Peery, Christopher; Martin, Richard P.; Nguyen, Thu D.

doi:10.1109/hpdc.2003.1210033

Cited by 191 publications

(164 citation statements)

References 15 publications

Supporting

Mentioning

163

Contrasting

Unclassified

Order By: Relevance

“…For each related cluster, the virtual peer computes the partial cosine similarity based on the retrieved scores. Then (lines [23][24][25][26][27][28][29][30][31], for the µ clusters with the highest partial cosine similarity, it sends its Bloom filter to the peer that holds the cluster's centroid (the cache peer), and retrieves the expected overlap size (cf. Eqn.…”

Section: Joining Peers In the Clustering-enhanced Algorithmmentioning

confidence: 99%

See 1 more Smart Citation

PCIR: Combining DHTs and peer clusters for efficient full-text P2P indexing

Papapetrou¹,

Siberski²,

Nejdl³

2010

Computer Networks

View full text Add to dashboard Cite

Distributed hash tables (DHTs) are very efficient for querying based on key lookups. However, building huge term indexes, as required for IR-style keyword search, poses a scalability challenge for plain DHTs. Due to the large sizes of document term vocabularies, peers joining the network cause huge amounts of key inserts and, consequently, a large number of index maintenance messages. Thus, the key to exploiting DHTs for distributed information retrieval is to reduce index maintenance costs. Various approaches in this direction have been pursued, including the use of hybrid infrastructures, or changing the granularity of the inverted index to peer level. We show that indexing costs can be significantly reduced further by letting peers form groups in a self-organized fashion. Instead of each individual peer submitting index information separately, all peers of a group cooperate to publish the index updates to the DHT in batches. Our evaluation shows that this approach reduces index maintenance cost by an order of magnitude, while still keeping a complete and correct term index for query processing.

show abstract

Section: Joining Peers In the Clustering-enhanced Algorithmmentioning

confidence: 99%

“…The first proposals focused on distributed IR in unstructured networks, using approximated system-wide information. PlanetP [28] is one of these systems. It uses gossiping to distribute peer content summaries, represented by Bloom filters, to all participating peers.…”

Section: P2p Information Retrievalmentioning

confidence: 99%

PCIR: Combining DHTs and peer clusters for efficient full-text P2P indexing

Papapetrou¹,

Siberski²,

Nejdl³

2010

Computer Networks

View full text Add to dashboard Cite

show abstract

“…PlantP [14] is a content addressable content publish/subscribe service for unstructured P2P, which use gossiping to build content-addressable communities. Zhang et al [15] proposed an enhanced clustering cache replacement scheme for Freenet by forcing the routing tables to resemble neighbour relationships in a small-world acquaintance graph.…”

Section: Small World Overlay In Peer-to-peer Networkmentioning

confidence: 99%

Fault-tolerant peer-to-peer search on small-world networks

Liu¹,

Antonopoulos²,

Mackin³

2007

Future Generation Computer Systems

View full text Add to dashboard Cite

This paper presents small world architecture for P2P networks (SWAN) for content discovery in multi-group P2P systems. A semi-structured P2P algorithm of SWAN is utilized to create and find long-range shortcuts toward remote peer groups. In SWAN, not every peer node needs to be connected to remote groups, but every peer node can easily find which peer nodes have external connections to a specific peer group. From our analysis and simulation, SWAN has the advantages of both structured and unstructured P2P networks and can achieve good performance in both stable and dynamic environments.Keywords: Peer-to-peer; Small world; Information search IntroductionMost early Internet applications are distributed using client/server architecture that has a series of drawbacks, such as performance bottlenecks and low fault tolerance. On the contrary, peer-to-peer (P2P) architecture does not rely on centralized servers to provide access to services and offers an appealing alternative to the client/server model, especially for large-scale distributed applications. This paper presents small world architecture for P2P networks (SWAN) for resource discovery and discovery in multi-group P2P systems. A semi-structured P2P algorithm of SWAN is used to create and discover long-range shortcuts between different peer groups, which can satisfy the following requirements of design:(1) Not every peer node needs to connect to other peer groups;(2) Each peer node needs to know or can easily find which peer nodes have external connections to which peer groups; 3 (3) External links to other peer groups need to be distributed within the peer group and cannot be centralized in one or a few peer nodes.The rest of paper is organized as follows: Section 2 discusses related work. Section 3 presents the SWAN algorithms. The small-world properties of SWAN are analysed in Section 4. The performance evaluation and simulations are given in Section 5 and Section 6, respectively. We conclude the work in Section 7. Related Work Resource Discovery in Peer-to-Peer NetworksExisting P2P solutions for resource discovery can be generally classified into two categories: structured and unstructured P2P systems. Distributed hash tables (DHTs) have become the dominant methodology for resource discovery in structured P2Pnetworks [5]. The observed problem of most DHTs is that the cost of maintaining a consistent distributed index is too high in dynamic environments [6]. Some structured P2P protocols (e.g. Kademlia [7]) are beginning to seek ways to save the cost of maintaining a consistent index. In contrast, unstructured P2P systems (e.g. Gnutella) do not control data placement and are more resilient in dynamic environments, but current unstructured P2P search techniques tend to either require high search overhead or generate massive network traffic.In contrast, SWAN uses a semi-structured P2P search method combining the techniques of both structured and unstructured search methods. SWAN does not strictly rely on DHTs. It can find the requested data inside and ...

show abstract

“…Allowing nodes to maintain summaries of remote peers' content enables them to guide queries to peers more likely to provide relevant answers. Bloom filters [2] are a probabilistic data structure for content summarization that has been used in peer-topeer environments [6,11,14,19]. In Planet-P [6] gossiping is used for propagating content indices across peers.…”

Section: Introductionmentioning

confidence: 99%

“…Bloom filters [2] are a probabilistic data structure for content summarization that has been used in peer-topeer environments [6,11,14,19]. In Planet-P [6] gossiping is used for propagating content indices across peers. In OceanStore [19] attenuated Bloom filters are propagated to improve searching in a structured peer-to-peer network.…”

Section: Introductionmentioning

confidence: 99%

Efficient Data Dissemination in Overlays

Repantis

Kalogeraki

2009

2009 Software Technologies for Future Dependable Distributed Systems

View full text Add to dashboard Cite

In this paper we propose adaptive data dissemination algorithms for intelligently routing search queries in a peerto-peer network. In our mechanism nodes build content synopses of their data and adaptively disseminate them to the most appropriate nodes. Based on the content synopses, a routing mechanism is being built to forward the queries to those nodes that have a high probability of providing the desired results. Our simulation results show that our approach is highly scalable and significantly improves resources usage by saving both bandwidth and processing power.

show abstract

PlanetP: using gossiping to build content addressable peer-to-peer information sharing communities

Cited by 191 publications

References 15 publications

PCIR: Combining DHTs and peer clusters for efficient full-text P2P indexing

PCIR: Combining DHTs and peer clusters for efficient full-text P2P indexing

Fault-tolerant peer-to-peer search on small-world networks

Efficient Data Dissemination in Overlays

Contact Info

Product

Resources

About