Equal peers in peer-to-peer (P2P) networks are the drawbacks of system in term of bandwidth, scalability and efficiency. The super-peer model is based on heterogeneity and different characteristics of peers in P2P networks. The P2P networks and large- scale distributed systems based on P2P networks use the super-peer model to design the query processing mechanism. This chapter first reviews the query processing methods in P2P networks, in which the authors classify theses query processing approaches in Unstructured and Structured mechanisms. Furthermore, the query processing techniques in distributed systems based on P2P networks are discussed. Afterward, authors concentrate on super-peer model to process the query of peers in P2P networks. Authors present the query processing methods in P2P-based distributed systems using the super node. Finally, the chapter provides some examples of each of the presented query processing techniques, and then illustrates the properties of each of them in terms of scalability and performance issues.
TopIntroduction
P2P networks are distributed systems in which resources are shared by direct exchange between autonomous nodes. The shared resources contain documents, storage capacity, bandwidth, and CPU cycles. Each peer links to a small subset of other peers, so logical overlay networks are formed on top of the physical one (usually the Internet). Existing P2P systems have already had advantages of scalability, load balancing, self-organization, adaptation and fault tolerance (Xiao, Zhuang, & Liu, 2005; Hawa, As-Sayid-Ahmad, & Khalaf, 2013; Torkestani, 2012). In pure P2P systems, all peers play equal roles and take same responsibilities regardless of their capabilities. Moreover, in query flooding process, any peer could be a query sender and the query receiver (Navimpour & Milani, 2015). However, as the size of P2P network grows, weak peers will seriously limit the scalability of P2P systems and become bottlenecks.
The peers in P2P networks usually have different characteristics with respect to their capabilities, e.g. available bandwidth, storage space or processing power (Nejdl et al., 2003; Cholvi, Felber, & Biersack, 2004). A discussion about the different capabilities of peers lead to an efficient network topology in P2P networks (Yang & Garcia-Molina, 2002). There is a small subset of peers with high capacity, called super-peers that takes over specific responsibilities for peer aggregation, query routing, query processing and possibly mediation (Cao, Li, & Liu, 2008). In this topology, the symmetry of pure P2P systems is broken by assigning additional responsibilities to super nodes (Garbacki, Epema, & van Steen, 2007). Consequently, super-peers make the decentralized networks more efficient by exploiting heterogeneity of peers and distributing load to machines that can handle the burden.
The super-peer network comprises the two-layer architecture (Awan, Ferreira, Jagannathan, & Grama, 2005). In the low layer, the super-peer is a node that acts as a centralized server to a subset of connected clients. When the clients need resource/file, submit queries to their super-peer and receive results from it (Yang & Garcia-m, 2003; Tan, Lü, & Lin, 2012). In the higher-layer, the super-peers are connected to each other and form the overlay network. The super-peers route messages over this overlay network, and submit and answer queries on behalf of their clients (Mastroianni, Talia, & Verta, 2005b). The super-peers are equal in terms of search, and all the client peers are equal in terms of download. Consequently, the client peers with low capacity are shielded from massive query traffic, which improves the scalability of the system. The peers use resources of network more efficiently because the super-peers provide the efficient and reliable query processing function.